diff --git "a/sft/hyperrouter/trainer_state.json" "b/sft/hyperrouter/trainer_state.json" new file mode 100644--- /dev/null +++ "b/sft/hyperrouter/trainer_state.json" @@ -0,0 +1,74887 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999398785546805, + "eval_steps": 500, + "global_step": 8316, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00012024289063909097, + "flos": 24895993628400.0, + "grad_norm": 39.864779989280635, + "learning_rate": 0.0, + "loss": 1.8565, + "num_input_tokens_seen": 20375, + "step": 1 + }, + { + "epoch": 0.00024048578127818193, + "flos": 30620581122240.0, + "grad_norm": 54.92542440522194, + "learning_rate": 5.021476677069823e-07, + "loss": 1.8897, + "num_input_tokens_seen": 39035, + "step": 2 + }, + { + "epoch": 0.0003607286719172729, + "flos": 18999150600000.0, + "grad_norm": 39.93986874332933, + "learning_rate": 7.958852231401551e-07, + "loss": 1.6146, + "num_input_tokens_seen": 57600, + "step": 3 + }, + { + "epoch": 0.00048097156255636386, + "flos": 19288748899200.0, + "grad_norm": 37.48539917577492, + "learning_rate": 1.0042953354139647e-06, + "loss": 1.6442, + "num_input_tokens_seen": 76465, + "step": 4 + }, + { + "epoch": 0.0006012144531954548, + "flos": 13971521020800.0, + "grad_norm": 55.01484840406156, + "learning_rate": 1.1659507774310057e-06, + "loss": 1.9349, + "num_input_tokens_seen": 94350, + "step": 5 + }, + { + "epoch": 0.0007214573438345458, + "flos": 23115903379680.0, + "grad_norm": 45.517488912230206, + "learning_rate": 1.2980328908471373e-06, + "loss": 1.611, + "num_input_tokens_seen": 114595, + "step": 6 + }, + { + "epoch": 0.0008417002344736367, + "flos": 67566737500320.0, + "grad_norm": 4.57081056146968, + "learning_rate": 1.4097067265369432e-06, + "loss": 0.8146, + "num_input_tokens_seen": 179590, + "step": 7 + }, + { + "epoch": 0.0009619431251127277, + "flos": 21250766937120.0, + "grad_norm": 41.122534773660746, + "learning_rate": 1.506443003120947e-06, + "loss": 1.5845, + "num_input_tokens_seen": 195090, + "step": 8 + }, + { + "epoch": 0.0010821860157518186, + "flos": 23297989532160.0, + "grad_norm": 17.402762428324955, + "learning_rate": 1.5917704462803102e-06, + "loss": 1.4765, + "num_input_tokens_seen": 211635, + "step": 9 + }, + { + "epoch": 0.0012024289063909096, + "flos": 16985683880640.0, + "grad_norm": 14.760614229153033, + "learning_rate": 1.6680984451379884e-06, + "loss": 1.5282, + "num_input_tokens_seen": 224705, + "step": 10 + }, + { + "epoch": 0.0013226717970300007, + "flos": 21257754852000.0, + "grad_norm": 14.74182109043859, + "learning_rate": 1.7371455188905097e-06, + "loss": 1.3263, + "num_input_tokens_seen": 244635, + "step": 11 + }, + { + "epoch": 0.0014429146876690916, + "flos": 27198937453440.0, + "grad_norm": 13.006070406992292, + "learning_rate": 1.8001805585541196e-06, + "loss": 1.2521, + "num_input_tokens_seen": 265765, + "step": 12 + }, + { + "epoch": 0.0015631575783081825, + "flos": 19035025152000.0, + "grad_norm": 6.70701105895694, + "learning_rate": 1.8581671739548328e-06, + "loss": 1.2932, + "num_input_tokens_seen": 283500, + "step": 13 + }, + { + "epoch": 0.0016834004689472734, + "flos": 48069920786400.0, + "grad_norm": 7.738848138309248, + "learning_rate": 1.9118543942439254e-06, + "loss": 1.1336, + "num_input_tokens_seen": 305685, + "step": 14 + }, + { + "epoch": 0.0018036433595863645, + "flos": 34919717412480.0, + "grad_norm": 5.588484688853832, + "learning_rate": 1.961836000571161e-06, + "loss": 1.1296, + "num_input_tokens_seen": 327340, + "step": 15 + }, + { + "epoch": 0.0019238862502254555, + "flos": 59682759360480.0, + "grad_norm": 3.7751046677788347, + "learning_rate": 2.0085906708279293e-06, + "loss": 0.6466, + "num_input_tokens_seen": 382710, + "step": 16 + }, + { + "epoch": 0.0020441291408645466, + "flos": 20784880968000.0, + "grad_norm": 4.361612818882017, + "learning_rate": 2.0525099325728135e-06, + "loss": 1.1604, + "num_input_tokens_seen": 400890, + "step": 17 + }, + { + "epoch": 0.0021643720315036373, + "flos": 63765607585440.0, + "grad_norm": 3.5238151939940523, + "learning_rate": 2.0939181139872922e-06, + "loss": 0.7215, + "num_input_tokens_seen": 462605, + "step": 18 + }, + { + "epoch": 0.0022846149221427284, + "flos": 31240038304320.0, + "grad_norm": 5.674765278671672, + "learning_rate": 2.1330868934640175e-06, + "loss": 1.0151, + "num_input_tokens_seen": 483280, + "step": 19 + }, + { + "epoch": 0.002404857812781819, + "flos": 51010730508960.0, + "grad_norm": 3.543961996636809, + "learning_rate": 2.170246112844971e-06, + "loss": 0.7645, + "num_input_tokens_seen": 537620, + "step": 20 + }, + { + "epoch": 0.0025251007034209102, + "flos": 15792940092000.0, + "grad_norm": 4.374386117645538, + "learning_rate": 2.2055919496770983e-06, + "loss": 1.016, + "num_input_tokens_seen": 555760, + "step": 21 + }, + { + "epoch": 0.0026453435940600014, + "flos": 37797332809440.0, + "grad_norm": 4.002523341590351, + "learning_rate": 2.2392931865974923e-06, + "loss": 0.8953, + "num_input_tokens_seen": 578450, + "step": 22 + }, + { + "epoch": 0.002765586484699092, + "flos": 21111803439840.0, + "grad_norm": 4.569145484015754, + "learning_rate": 2.271496085962064e-06, + "loss": 1.0187, + "num_input_tokens_seen": 596145, + "step": 23 + }, + { + "epoch": 0.002885829375338183, + "flos": 20637888802560.0, + "grad_norm": 3.7219036418283555, + "learning_rate": 2.3023282262611022e-06, + "loss": 1.0266, + "num_input_tokens_seen": 614920, + "step": 24 + }, + { + "epoch": 0.003006072265977274, + "flos": 34773766000320.0, + "grad_norm": 3.4859106674084956, + "learning_rate": 2.3319015548620114e-06, + "loss": 0.9247, + "num_input_tokens_seen": 636060, + "step": 25 + }, + { + "epoch": 0.003126315156616365, + "flos": 24386938150080.0, + "grad_norm": 2.6532252424194445, + "learning_rate": 2.3603148416618152e-06, + "loss": 0.9283, + "num_input_tokens_seen": 655575, + "step": 26 + }, + { + "epoch": 0.003246558047255456, + "flos": 23588517075360.0, + "grad_norm": 2.4475933321340038, + "learning_rate": 2.3876556694204647e-06, + "loss": 1.0089, + "num_input_tokens_seen": 675730, + "step": 27 + }, + { + "epoch": 0.003366800937894547, + "flos": 17794884185760.0, + "grad_norm": 3.3811908032775886, + "learning_rate": 2.414002061950908e-06, + "loss": 0.907, + "num_input_tokens_seen": 694605, + "step": 28 + }, + { + "epoch": 0.003487043828533638, + "flos": 24391175502720.0, + "grad_norm": 3.634793436721916, + "learning_rate": 2.4394238264681557e-06, + "loss": 0.9977, + "num_input_tokens_seen": 714340, + "step": 29 + }, + { + "epoch": 0.003607286719172729, + "flos": 26102852241600.0, + "grad_norm": 2.282746020425468, + "learning_rate": 2.4639836682781433e-06, + "loss": 0.9958, + "num_input_tokens_seen": 734470, + "step": 30 + }, + { + "epoch": 0.00372752960981182, + "flos": 20563649324640.0, + "grad_norm": 2.5489686189128804, + "learning_rate": 2.487738122623307e-06, + "loss": 1.0016, + "num_input_tokens_seen": 753380, + "step": 31 + }, + { + "epoch": 0.003847772500450911, + "flos": 22642211760960.0, + "grad_norm": 5.103734440570763, + "learning_rate": 2.510738338534912e-06, + "loss": 0.9888, + "num_input_tokens_seen": 772105, + "step": 32 + }, + { + "epoch": 0.003968015391090002, + "flos": 17942248048800.0, + "grad_norm": 2.64984309613665, + "learning_rate": 2.5330307420306648e-06, + "loss": 1.0251, + "num_input_tokens_seen": 788955, + "step": 33 + }, + { + "epoch": 0.004088258281729093, + "flos": 27264813735360.0, + "grad_norm": 2.726250485818532, + "learning_rate": 2.554657600279796e-06, + "loss": 0.8832, + "num_input_tokens_seen": 810230, + "step": 34 + }, + { + "epoch": 0.004208501172368184, + "flos": 23225348353920.0, + "grad_norm": 2.5277414377509118, + "learning_rate": 2.5756575039679493e-06, + "loss": 1.0358, + "num_input_tokens_seen": 829780, + "step": 35 + }, + { + "epoch": 0.0043287440630072746, + "flos": 17287548200640.0, + "grad_norm": 2.4731806339140556, + "learning_rate": 2.5960657816942747e-06, + "loss": 0.9524, + "num_input_tokens_seen": 848695, + "step": 36 + }, + { + "epoch": 0.004448986953646365, + "flos": 53016726106560.0, + "grad_norm": 1.3541602599163598, + "learning_rate": 2.6159148575788668e-06, + "loss": 0.6097, + "num_input_tokens_seen": 906730, + "step": 37 + }, + { + "epoch": 0.004569229844285457, + "flos": 13425225393600.0, + "grad_norm": 4.320213385716047, + "learning_rate": 2.635234561171e-06, + "loss": 0.9881, + "num_input_tokens_seen": 925125, + "step": 38 + }, + { + "epoch": 0.0046894727349245475, + "flos": 16186370731680.0, + "grad_norm": 2.594247609399402, + "learning_rate": 2.6540523970949877e-06, + "loss": 0.9426, + "num_input_tokens_seen": 939970, + "step": 39 + }, + { + "epoch": 0.004809715625563638, + "flos": 23880011032320.0, + "grad_norm": 2.913243401082827, + "learning_rate": 2.6723937805519533e-06, + "loss": 0.9285, + "num_input_tokens_seen": 957470, + "step": 40 + }, + { + "epoch": 0.00492995851620273, + "flos": 20743245103200.0, + "grad_norm": 2.4524682982220543, + "learning_rate": 2.690282243737839e-06, + "loss": 0.9326, + "num_input_tokens_seen": 976405, + "step": 41 + }, + { + "epoch": 0.0050502014068418205, + "flos": 20309628219360.0, + "grad_norm": 3.6256707984886978, + "learning_rate": 2.7077396173840807e-06, + "loss": 0.9955, + "num_input_tokens_seen": 994690, + "step": 42 + }, + { + "epoch": 0.005170444297480911, + "flos": 25957644224640.0, + "grad_norm": 16.008396806380922, + "learning_rate": 2.7247861909342594e-06, + "loss": 0.9282, + "num_input_tokens_seen": 1015615, + "step": 43 + }, + { + "epoch": 0.005290687188120003, + "flos": 20923955974560.0, + "grad_norm": 3.4158460205269363, + "learning_rate": 2.7414408543044743e-06, + "loss": 0.8297, + "num_input_tokens_seen": 1031255, + "step": 44 + }, + { + "epoch": 0.005410930078759093, + "flos": 15828926153280.0, + "grad_norm": 4.442113707604093, + "learning_rate": 2.7577212237113157e-06, + "loss": 0.7965, + "num_input_tokens_seen": 1048295, + "step": 45 + }, + { + "epoch": 0.005531172969398184, + "flos": 21074850964800.0, + "grad_norm": 2.183217898319022, + "learning_rate": 2.7736437536690466e-06, + "loss": 1.0448, + "num_input_tokens_seen": 1067925, + "step": 46 + }, + { + "epoch": 0.005651415860037276, + "flos": 20814957037440.0, + "grad_norm": 2.3192591475774815, + "learning_rate": 2.789223836941131e-06, + "loss": 1.0791, + "num_input_tokens_seen": 1088060, + "step": 47 + }, + { + "epoch": 0.005771658750676366, + "flos": 13241466601920.0, + "grad_norm": 2.3932528544773257, + "learning_rate": 2.8044758939680847e-06, + "loss": 1.0891, + "num_input_tokens_seen": 1104130, + "step": 48 + }, + { + "epoch": 0.005891901641315457, + "flos": 24390432107520.0, + "grad_norm": 3.205401992939471, + "learning_rate": 2.8194134530738863e-06, + "loss": 1.025, + "num_input_tokens_seen": 1122900, + "step": 49 + }, + { + "epoch": 0.006012144531954548, + "flos": 23043225031680.0, + "grad_norm": 4.593601146543992, + "learning_rate": 2.834049222568994e-06, + "loss": 0.9048, + "num_input_tokens_seen": 1140250, + "step": 50 + }, + { + "epoch": 0.006132387422593639, + "flos": 22496594876640.0, + "grad_norm": 2.257532333362146, + "learning_rate": 2.848395155712969e-06, + "loss": 0.926, + "num_input_tokens_seen": 1160470, + "step": 51 + }, + { + "epoch": 0.00625263031323273, + "flos": 27588428098560.0, + "grad_norm": 2.578236525464216, + "learning_rate": 2.8624625093687977e-06, + "loss": 0.9793, + "num_input_tokens_seen": 1177605, + "step": 52 + }, + { + "epoch": 0.006372873203871821, + "flos": 23077575623520.0, + "grad_norm": 2.4771338925817763, + "learning_rate": 2.876261897070029e-06, + "loss": 0.8902, + "num_input_tokens_seen": 1197735, + "step": 53 + }, + { + "epoch": 0.006493116094510912, + "flos": 22823331499680.0, + "grad_norm": 4.504466065869985, + "learning_rate": 2.889803337127447e-06, + "loss": 0.9259, + "num_input_tokens_seen": 1216335, + "step": 54 + }, + { + "epoch": 0.006613358985150003, + "flos": 23038578811680.0, + "grad_norm": 3.684941972459413, + "learning_rate": 2.903096296321516e-06, + "loss": 0.8474, + "num_input_tokens_seen": 1234480, + "step": 55 + }, + { + "epoch": 0.006733601875789094, + "flos": 26499479480640.0, + "grad_norm": 2.3838807759971345, + "learning_rate": 2.9161497296578907e-06, + "loss": 0.9171, + "num_input_tokens_seen": 1253870, + "step": 56 + }, + { + "epoch": 0.006853844766428185, + "flos": 15501185946720.0, + "grad_norm": 2.3590648523062003, + "learning_rate": 2.928972116604173e-06, + "loss": 0.8607, + "num_input_tokens_seen": 1270145, + "step": 57 + }, + { + "epoch": 0.006974087657067276, + "flos": 24209089350240.0, + "grad_norm": 2.615583697030526, + "learning_rate": 2.9415714941751377e-06, + "loss": 1.0221, + "num_input_tokens_seen": 1291365, + "step": 58 + }, + { + "epoch": 0.007094330547706367, + "flos": 25735483337280.0, + "grad_norm": 2.3681743314854673, + "learning_rate": 2.9539554871897396e-06, + "loss": 0.9367, + "num_input_tokens_seen": 1311535, + "step": 59 + }, + { + "epoch": 0.007214573438345458, + "flos": 21288388467840.0, + "grad_norm": 2.5450074070970707, + "learning_rate": 2.9661313359851253e-06, + "loss": 0.9759, + "num_input_tokens_seen": 1329420, + "step": 60 + }, + { + "epoch": 0.007334816328984549, + "flos": 24901410729120.0, + "grad_norm": 2.455070355994072, + "learning_rate": 2.978105921839922e-06, + "loss": 0.9412, + "num_input_tokens_seen": 1349965, + "step": 61 + }, + { + "epoch": 0.00745505921962364, + "flos": 18484083304800.0, + "grad_norm": 3.3408054684886537, + "learning_rate": 2.9898857903302893e-06, + "loss": 0.7222, + "num_input_tokens_seen": 1368915, + "step": 62 + }, + { + "epoch": 0.007575302110262731, + "flos": 18451331012640.0, + "grad_norm": 3.320080414682894, + "learning_rate": 3.001477172817253e-06, + "loss": 0.8805, + "num_input_tokens_seen": 1386805, + "step": 63 + }, + { + "epoch": 0.007695545000901822, + "flos": 24937433960160.0, + "grad_norm": 2.8628393979823836, + "learning_rate": 3.012886006241894e-06, + "loss": 0.9646, + "num_input_tokens_seen": 1406190, + "step": 64 + }, + { + "epoch": 0.007815787891540913, + "flos": 21293778083040.0, + "grad_norm": 2.144457317252166, + "learning_rate": 3.0241179513858383e-06, + "loss": 0.8835, + "num_input_tokens_seen": 1425500, + "step": 65 + }, + { + "epoch": 0.007936030782180003, + "flos": 21544825607520.0, + "grad_norm": 2.660301081318, + "learning_rate": 3.035178409737647e-06, + "loss": 0.8792, + "num_input_tokens_seen": 1442950, + "step": 66 + }, + { + "epoch": 0.008056273672819095, + "flos": 20092113552000.0, + "grad_norm": 3.2222741842647564, + "learning_rate": 3.046072539090907e-06, + "loss": 0.889, + "num_input_tokens_seen": 1460915, + "step": 67 + }, + { + "epoch": 0.008176516563458186, + "flos": 18307052239680.0, + "grad_norm": 2.428875557505249, + "learning_rate": 3.056805267986779e-06, + "loss": 1.0512, + "num_input_tokens_seen": 1478385, + "step": 68 + }, + { + "epoch": 0.008296759454097276, + "flos": 21840854275200.0, + "grad_norm": 4.141399267968318, + "learning_rate": 3.0673813091022194e-06, + "loss": 0.9529, + "num_input_tokens_seen": 1497605, + "step": 69 + }, + { + "epoch": 0.008417002344736368, + "flos": 63318418005600.0, + "grad_norm": 1.2340046290769882, + "learning_rate": 3.0778051716749317e-06, + "loss": 0.62, + "num_input_tokens_seen": 1561150, + "step": 70 + }, + { + "epoch": 0.008537245235375458, + "flos": 22933631378400.0, + "grad_norm": 4.635067472985261, + "learning_rate": 3.0880811730470094e-06, + "loss": 0.9054, + "num_input_tokens_seen": 1580605, + "step": 71 + }, + { + "epoch": 0.008657488126014549, + "flos": 61896265226400.0, + "grad_norm": 1.1582921526241505, + "learning_rate": 3.098213449401257e-06, + "loss": 0.5852, + "num_input_tokens_seen": 1647535, + "step": 72 + }, + { + "epoch": 0.00877773101665364, + "flos": 30253175048160.0, + "grad_norm": 2.7295049404330873, + "learning_rate": 3.1082059657570015e-06, + "loss": 0.9905, + "num_input_tokens_seen": 1666770, + "step": 73 + }, + { + "epoch": 0.00889797390729273, + "flos": 23480447382240.0, + "grad_norm": 11.912383334538239, + "learning_rate": 3.1180625252858496e-06, + "loss": 0.9662, + "num_input_tokens_seen": 1685200, + "step": 74 + }, + { + "epoch": 0.009018216797931822, + "flos": 23042853334080.0, + "grad_norm": 3.30170105527415, + "learning_rate": 3.1277867780021663e-06, + "loss": 0.8037, + "num_input_tokens_seen": 1701835, + "step": 75 + }, + { + "epoch": 0.009138459688570914, + "flos": 15896103376800.0, + "grad_norm": 1.9763340271392924, + "learning_rate": 3.1373822288779824e-06, + "loss": 0.9563, + "num_input_tokens_seen": 1718415, + "step": 76 + }, + { + "epoch": 0.009258702579210003, + "flos": 26977817319360.0, + "grad_norm": 2.135476261951054, + "learning_rate": 3.1468522454274533e-06, + "loss": 0.7942, + "num_input_tokens_seen": 1738770, + "step": 77 + }, + { + "epoch": 0.009378945469849095, + "flos": 26864506690080.0, + "grad_norm": 2.1073405309327127, + "learning_rate": 3.15620006480197e-06, + "loss": 0.9196, + "num_input_tokens_seen": 1758040, + "step": 78 + }, + { + "epoch": 0.009499188360488187, + "flos": 35642672407200.0, + "grad_norm": 3.210623547132452, + "learning_rate": 3.1654288004333087e-06, + "loss": 0.7508, + "num_input_tokens_seen": 1776705, + "step": 79 + }, + { + "epoch": 0.009619431251127276, + "flos": 21472370278080.0, + "grad_norm": 4.359627227596258, + "learning_rate": 3.1745414482589353e-06, + "loss": 0.7621, + "num_input_tokens_seen": 1795915, + "step": 80 + }, + { + "epoch": 0.009739674141766368, + "flos": 17396324119200.0, + "grad_norm": 4.793496398981727, + "learning_rate": 3.1835408925606204e-06, + "loss": 0.8711, + "num_input_tokens_seen": 1814055, + "step": 81 + }, + { + "epoch": 0.00985991703240546, + "flos": 27487643678400.0, + "grad_norm": 2.9065152939335497, + "learning_rate": 3.1924299114448214e-06, + "loss": 0.8925, + "num_input_tokens_seen": 1834535, + "step": 82 + }, + { + "epoch": 0.00998015992304455, + "flos": 13788579963840.0, + "grad_norm": 2.357289485169021, + "learning_rate": 3.2012111819909055e-06, + "loss": 0.8362, + "num_input_tokens_seen": 1851865, + "step": 83 + }, + { + "epoch": 0.010100402813683641, + "flos": 20162524544640.0, + "grad_norm": 2.910960819726355, + "learning_rate": 3.2098872850910627e-06, + "loss": 0.9498, + "num_input_tokens_seen": 1868540, + "step": 84 + }, + { + "epoch": 0.010220645704322733, + "flos": 17177211152160.0, + "grad_norm": 2.987388327837757, + "learning_rate": 3.2184607100038194e-06, + "loss": 0.8936, + "num_input_tokens_seen": 1887180, + "step": 85 + }, + { + "epoch": 0.010340888594961822, + "flos": 21439766664960.0, + "grad_norm": 2.8806201900177997, + "learning_rate": 3.2269338586412414e-06, + "loss": 0.9316, + "num_input_tokens_seen": 1904765, + "step": 86 + }, + { + "epoch": 0.010461131485600914, + "flos": 22969951967520.0, + "grad_norm": 2.30318036989815, + "learning_rate": 3.2353090496083106e-06, + "loss": 0.9671, + "num_input_tokens_seen": 1922600, + "step": 87 + }, + { + "epoch": 0.010581374376240005, + "flos": 33498605386560.0, + "grad_norm": 3.2274099350114853, + "learning_rate": 3.2435885220114572e-06, + "loss": 0.8145, + "num_input_tokens_seen": 1943950, + "step": 88 + }, + { + "epoch": 0.010701617266879095, + "flos": 21732041186880.0, + "grad_norm": 2.19679660792912, + "learning_rate": 3.2517744390519113e-06, + "loss": 0.9425, + "num_input_tokens_seen": 1962815, + "step": 89 + }, + { + "epoch": 0.010821860157518187, + "flos": 19032980815200.0, + "grad_norm": 2.6673406305585927, + "learning_rate": 3.259868891418298e-06, + "loss": 0.7551, + "num_input_tokens_seen": 1980580, + "step": 90 + }, + { + "epoch": 0.010942103048157278, + "flos": 25410939730080.0, + "grad_norm": 2.8838907515194157, + "learning_rate": 3.2678739004917757e-06, + "loss": 0.851, + "num_input_tokens_seen": 2000315, + "step": 91 + }, + { + "epoch": 0.011062345938796368, + "flos": 27454631197920.0, + "grad_norm": 1.7561146314044571, + "learning_rate": 3.275791421376029e-06, + "loss": 0.9231, + "num_input_tokens_seen": 2023760, + "step": 92 + }, + { + "epoch": 0.01118258882943546, + "flos": 16048447987680.0, + "grad_norm": 2.290658440027056, + "learning_rate": 3.2836233457634622e-06, + "loss": 0.963, + "num_input_tokens_seen": 2041895, + "step": 93 + }, + { + "epoch": 0.011302831720074551, + "flos": 20639152574400.0, + "grad_norm": 2.5830019974237475, + "learning_rate": 3.2913715046481135e-06, + "loss": 0.8535, + "num_input_tokens_seen": 2061640, + "step": 94 + }, + { + "epoch": 0.011423074610713641, + "flos": 13053581966880.0, + "grad_norm": 3.033484353159221, + "learning_rate": 3.299037670895023e-06, + "loss": 0.8883, + "num_input_tokens_seen": 2078255, + "step": 95 + }, + { + "epoch": 0.011543317501352733, + "flos": 30292060350720.0, + "grad_norm": 3.8026037130062535, + "learning_rate": 3.3066235616750667e-06, + "loss": 0.8033, + "num_input_tokens_seen": 2099490, + "step": 96 + }, + { + "epoch": 0.011663560391991824, + "flos": 15500962928160.0, + "grad_norm": 2.297642170987127, + "learning_rate": 3.3141308407736276e-06, + "loss": 0.9245, + "num_input_tokens_seen": 2116125, + "step": 97 + }, + { + "epoch": 0.011783803282630914, + "flos": 19873781149920.0, + "grad_norm": 3.3273088054340176, + "learning_rate": 3.321561120780869e-06, + "loss": 0.8669, + "num_input_tokens_seen": 2134835, + "step": 98 + }, + { + "epoch": 0.011904046173270006, + "flos": 22308561562560.0, + "grad_norm": 2.313197675056483, + "learning_rate": 3.3289159651708192e-06, + "loss": 1.0142, + "num_input_tokens_seen": 2152410, + "step": 99 + }, + { + "epoch": 0.012024289063909096, + "flos": 19072906871040.0, + "grad_norm": 2.5278636745441947, + "learning_rate": 3.3361968902759768e-06, + "loss": 0.9787, + "num_input_tokens_seen": 2172090, + "step": 100 + }, + { + "epoch": 0.012144531954548187, + "flos": 14990950720320.0, + "grad_norm": 2.333075565904783, + "learning_rate": 3.343405367163663e-06, + "loss": 0.9395, + "num_input_tokens_seen": 2189020, + "step": 101 + }, + { + "epoch": 0.012264774845187279, + "flos": 15100618713120.0, + "grad_norm": 3.1065045626619248, + "learning_rate": 3.350542823419951e-06, + "loss": 0.8146, + "num_input_tokens_seen": 2205620, + "step": 102 + }, + { + "epoch": 0.012385017735826368, + "flos": 13929364779360.0, + "grad_norm": 4.218481210693676, + "learning_rate": 3.3576106448465615e-06, + "loss": 0.8768, + "num_input_tokens_seen": 2219000, + "step": 103 + }, + { + "epoch": 0.01250526062646546, + "flos": 23589483489120.0, + "grad_norm": 2.214153633045815, + "learning_rate": 3.3646101770757797e-06, + "loss": 0.8814, + "num_input_tokens_seen": 2237790, + "step": 104 + }, + { + "epoch": 0.012625503517104552, + "flos": 34590193057440.0, + "grad_norm": 1.8411760628973486, + "learning_rate": 3.371542727108104e-06, + "loss": 0.8561, + "num_input_tokens_seen": 2259965, + "step": 105 + }, + { + "epoch": 0.012745746407743641, + "flos": 17796556824960.0, + "grad_norm": 3.106730347609652, + "learning_rate": 3.3784095647770114e-06, + "loss": 0.9026, + "num_input_tokens_seen": 2278610, + "step": 106 + }, + { + "epoch": 0.012865989298382733, + "flos": 20565693661440.0, + "grad_norm": 4.428099224739733, + "learning_rate": 3.3852119241449547e-06, + "loss": 0.8879, + "num_input_tokens_seen": 2297730, + "step": 107 + }, + { + "epoch": 0.012986232189021825, + "flos": 23914621812480.0, + "grad_norm": 2.672471810772982, + "learning_rate": 3.3919510048344295e-06, + "loss": 0.9639, + "num_input_tokens_seen": 2315740, + "step": 108 + }, + { + "epoch": 0.013106475079660914, + "flos": 23691234323040.0, + "grad_norm": 2.2115071240249344, + "learning_rate": 3.3986279732976907e-06, + "loss": 0.8651, + "num_input_tokens_seen": 2334215, + "step": 109 + }, + { + "epoch": 0.013226717970300006, + "flos": 21072211911840.0, + "grad_norm": 2.252179150185444, + "learning_rate": 3.4052439640284983e-06, + "loss": 0.9539, + "num_input_tokens_seen": 2353130, + "step": 110 + }, + { + "epoch": 0.013346960860939098, + "flos": 24828323513760.0, + "grad_norm": 2.888906266901942, + "learning_rate": 3.4118000807190217e-06, + "loss": 0.8148, + "num_input_tokens_seen": 2374010, + "step": 111 + }, + { + "epoch": 0.013467203751578187, + "flos": 28141451452320.0, + "grad_norm": 1.714205125492575, + "learning_rate": 3.4182973973648723e-06, + "loss": 0.7599, + "num_input_tokens_seen": 2395220, + "step": 112 + }, + { + "epoch": 0.013587446642217279, + "flos": 18889296758400.0, + "grad_norm": 3.145197789411897, + "learning_rate": 3.424736959321014e-06, + "loss": 0.9523, + "num_input_tokens_seen": 2413025, + "step": 113 + }, + { + "epoch": 0.01370768953285637, + "flos": 23954324849760.0, + "grad_norm": 2.3097968927902737, + "learning_rate": 3.431119784311155e-06, + "loss": 0.8851, + "num_input_tokens_seen": 2432700, + "step": 114 + }, + { + "epoch": 0.01382793242349546, + "flos": 39146101015200.0, + "grad_norm": 1.8454819642629732, + "learning_rate": 3.43744686339307e-06, + "loss": 0.7795, + "num_input_tokens_seen": 2455020, + "step": 115 + }, + { + "epoch": 0.013948175314134552, + "flos": 41293847842080.0, + "grad_norm": 3.5459569339449932, + "learning_rate": 3.44371916188212e-06, + "loss": 0.9083, + "num_input_tokens_seen": 2475775, + "step": 116 + }, + { + "epoch": 0.014068418204773643, + "flos": 22421388984960.0, + "grad_norm": 2.411799385445391, + "learning_rate": 3.449937620235143e-06, + "loss": 0.8609, + "num_input_tokens_seen": 2496370, + "step": 117 + }, + { + "epoch": 0.014188661095412733, + "flos": 23771086434720.0, + "grad_norm": 1.8296705706144634, + "learning_rate": 3.456103154896722e-06, + "loss": 0.8924, + "num_input_tokens_seen": 2517645, + "step": 118 + }, + { + "epoch": 0.014308903986051825, + "flos": 23626435964160.0, + "grad_norm": 2.5459763515360008, + "learning_rate": 3.462216659109757e-06, + "loss": 0.9228, + "num_input_tokens_seen": 2537825, + "step": 119 + }, + { + "epoch": 0.014429146876690916, + "flos": 20638520688480.0, + "grad_norm": 2.736668785394862, + "learning_rate": 3.4682790036921077e-06, + "loss": 0.8532, + "num_input_tokens_seen": 2556485, + "step": 120 + }, + { + "epoch": 0.014549389767330006, + "flos": 20200331924160.0, + "grad_norm": 1.9778197138813425, + "learning_rate": 3.4742910377810193e-06, + "loss": 0.8308, + "num_input_tokens_seen": 2573945, + "step": 121 + }, + { + "epoch": 0.014669632657969098, + "flos": 17978568637920.0, + "grad_norm": 3.465053891730288, + "learning_rate": 3.4802535895469042e-06, + "loss": 0.8885, + "num_input_tokens_seen": 2592695, + "step": 122 + }, + { + "epoch": 0.01478987554860819, + "flos": 22709091626400.0, + "grad_norm": 2.657959511361692, + "learning_rate": 3.4861674668779934e-06, + "loss": 0.8965, + "num_input_tokens_seen": 2610925, + "step": 123 + }, + { + "epoch": 0.01491011843924728, + "flos": 17174051722560.0, + "grad_norm": 2.0609594429241973, + "learning_rate": 3.492033458037272e-06, + "loss": 0.8431, + "num_input_tokens_seen": 2629495, + "step": 124 + }, + { + "epoch": 0.01503036132988637, + "flos": 17648672585280.0, + "grad_norm": 2.647119965189984, + "learning_rate": 3.497852332293018e-06, + "loss": 0.8722, + "num_input_tokens_seen": 2645070, + "step": 125 + }, + { + "epoch": 0.015150604220525462, + "flos": 18851303530080.0, + "grad_norm": 1.9662831913207361, + "learning_rate": 3.5036248405242356e-06, + "loss": 0.966, + "num_input_tokens_seen": 2663825, + "step": 126 + }, + { + "epoch": 0.015270847111164552, + "flos": 39363355494240.0, + "grad_norm": 2.049725127981984, + "learning_rate": 3.509351715802146e-06, + "loss": 0.8282, + "num_input_tokens_seen": 2684710, + "step": 127 + }, + { + "epoch": 0.015391090001803644, + "flos": 43699332750240.0, + "grad_norm": 2.727103105963436, + "learning_rate": 3.5150336739488763e-06, + "loss": 0.7845, + "num_input_tokens_seen": 2706995, + "step": 128 + }, + { + "epoch": 0.015511332892442733, + "flos": 18889408267680.0, + "grad_norm": 2.2862064393713117, + "learning_rate": 3.5206714140744143e-06, + "loss": 0.8422, + "num_input_tokens_seen": 2727050, + "step": 129 + }, + { + "epoch": 0.015631575783081827, + "flos": 24500174439840.0, + "grad_norm": 4.441283426456939, + "learning_rate": 3.5262656190928208e-06, + "loss": 0.8751, + "num_input_tokens_seen": 2745350, + "step": 130 + }, + { + "epoch": 0.015751818673720917, + "flos": 62239510956480.0, + "grad_norm": 1.0297192859519182, + "learning_rate": 3.5318169562186737e-06, + "loss": 0.7143, + "num_input_tokens_seen": 2814195, + "step": 131 + }, + { + "epoch": 0.015872061564360006, + "flos": 23844396668640.0, + "grad_norm": 2.185775076753605, + "learning_rate": 3.5373260774446292e-06, + "loss": 0.8221, + "num_input_tokens_seen": 2834645, + "step": 132 + }, + { + "epoch": 0.0159923044549991, + "flos": 23334495970080.0, + "grad_norm": 3.694053257920231, + "learning_rate": 3.542793620000961e-06, + "loss": 0.9039, + "num_input_tokens_seen": 2854120, + "step": 133 + }, + { + "epoch": 0.01611254734563819, + "flos": 17832914583840.0, + "grad_norm": 3.2960934548785867, + "learning_rate": 3.5482202067978894e-06, + "loss": 0.8668, + "num_input_tokens_seen": 2871330, + "step": 134 + }, + { + "epoch": 0.01623279023627728, + "flos": 20924699369760.0, + "grad_norm": 2.5357502621990267, + "learning_rate": 3.553606446851471e-06, + "loss": 0.7626, + "num_input_tokens_seen": 2888070, + "step": 135 + }, + { + "epoch": 0.016353033126916373, + "flos": 15720113064960.0, + "grad_norm": 1.817661368593131, + "learning_rate": 3.5589529356937613e-06, + "loss": 0.832, + "num_input_tokens_seen": 2906465, + "step": 136 + }, + { + "epoch": 0.016473276017555463, + "flos": 18780520839840.0, + "grad_norm": 1.8707691779177673, + "learning_rate": 3.5642602557679627e-06, + "loss": 0.7689, + "num_input_tokens_seen": 2924915, + "step": 137 + }, + { + "epoch": 0.016593518908194552, + "flos": 24317790929280.0, + "grad_norm": 2.7093756406226714, + "learning_rate": 3.569528976809202e-06, + "loss": 0.8407, + "num_input_tokens_seen": 2942130, + "step": 138 + }, + { + "epoch": 0.016713761798833646, + "flos": 22314211366080.0, + "grad_norm": 1.9750280040252493, + "learning_rate": 3.5747596562115522e-06, + "loss": 0.9008, + "num_input_tokens_seen": 2962745, + "step": 139 + }, + { + "epoch": 0.016834004689472735, + "flos": 17796482485440.0, + "grad_norm": 4.226071482568799, + "learning_rate": 3.5799528393819138e-06, + "loss": 0.91, + "num_input_tokens_seen": 2981825, + "step": 140 + }, + { + "epoch": 0.016954247580111825, + "flos": 20490153241920.0, + "grad_norm": 5.712382456031337, + "learning_rate": 3.585109060081286e-06, + "loss": 0.8823, + "num_input_tokens_seen": 3001625, + "step": 141 + }, + { + "epoch": 0.017074490470750915, + "flos": 22057253849760.0, + "grad_norm": 4.618071031026637, + "learning_rate": 3.590228840753992e-06, + "loss": 0.7869, + "num_input_tokens_seen": 3022055, + "step": 142 + }, + { + "epoch": 0.01719473336139001, + "flos": 15974580207360.0, + "grad_norm": 2.203245551957212, + "learning_rate": 3.5953126928453423e-06, + "loss": 0.8714, + "num_input_tokens_seen": 3039605, + "step": 143 + }, + { + "epoch": 0.017314976252029098, + "flos": 22459902589920.0, + "grad_norm": 2.203461425201234, + "learning_rate": 3.600361117108239e-06, + "loss": 0.8071, + "num_input_tokens_seen": 3059405, + "step": 144 + }, + { + "epoch": 0.017435219142668188, + "flos": 21986619838560.0, + "grad_norm": 2.4798083289741784, + "learning_rate": 3.6053746038991616e-06, + "loss": 0.9732, + "num_input_tokens_seen": 3078490, + "step": 145 + }, + { + "epoch": 0.01755546203330728, + "flos": 72136627021440.0, + "grad_norm": 1.0783347326640071, + "learning_rate": 3.6103536334639843e-06, + "loss": 0.5839, + "num_input_tokens_seen": 3131755, + "step": 146 + }, + { + "epoch": 0.01767570492394637, + "flos": 25300751360640.0, + "grad_norm": 2.3671900179506515, + "learning_rate": 3.615298676214041e-06, + "loss": 0.8548, + "num_input_tokens_seen": 3152035, + "step": 147 + }, + { + "epoch": 0.01779594781458546, + "flos": 20420039607360.0, + "grad_norm": 2.0879116360612584, + "learning_rate": 3.6202101929928317e-06, + "loss": 0.8934, + "num_input_tokens_seen": 3170625, + "step": 148 + }, + { + "epoch": 0.017916190705224554, + "flos": 16230571309920.0, + "grad_norm": 6.2540651452526825, + "learning_rate": 3.6250886353337413e-06, + "loss": 0.8856, + "num_input_tokens_seen": 3188155, + "step": 149 + }, + { + "epoch": 0.018036433595863644, + "flos": 23297208967200.0, + "grad_norm": 2.1669763441568195, + "learning_rate": 3.6299344457091488e-06, + "loss": 0.8681, + "num_input_tokens_seen": 3209015, + "step": 150 + }, + { + "epoch": 0.018156676486502734, + "flos": 18561556551840.0, + "grad_norm": 4.509355357042066, + "learning_rate": 3.634748057771256e-06, + "loss": 0.9401, + "num_input_tokens_seen": 3224955, + "step": 151 + }, + { + "epoch": 0.018276919377141827, + "flos": 25412315011200.0, + "grad_norm": 1.644431503061687, + "learning_rate": 3.639529896584965e-06, + "loss": 0.8596, + "num_input_tokens_seen": 3246330, + "step": 152 + }, + { + "epoch": 0.018397162267780917, + "flos": 20018728978560.0, + "grad_norm": 4.370634680428648, + "learning_rate": 3.6442803788531233e-06, + "loss": 0.8893, + "num_input_tokens_seen": 3264290, + "step": 153 + }, + { + "epoch": 0.018517405158420007, + "flos": 27525822755520.0, + "grad_norm": 2.955854496700415, + "learning_rate": 3.6489999131344357e-06, + "loss": 0.9609, + "num_input_tokens_seen": 3287065, + "step": 154 + }, + { + "epoch": 0.0186376480490591, + "flos": 19326481939200.0, + "grad_norm": 2.0926507370046683, + "learning_rate": 3.653688900054313e-06, + "loss": 0.9058, + "num_input_tokens_seen": 3305595, + "step": 155 + }, + { + "epoch": 0.01875789093969819, + "flos": 26649593905920.0, + "grad_norm": 2.5100054793279045, + "learning_rate": 3.6583477325089526e-06, + "loss": 0.7579, + "num_input_tokens_seen": 3326135, + "step": 156 + }, + { + "epoch": 0.01887813383033728, + "flos": 24318162626880.0, + "grad_norm": 3.375182535392241, + "learning_rate": 3.6629767958628916e-06, + "loss": 1.0433, + "num_input_tokens_seen": 3343510, + "step": 157 + }, + { + "epoch": 0.018998376720976373, + "flos": 14626815585120.0, + "grad_norm": 4.042320486976953, + "learning_rate": 3.667576468140291e-06, + "loss": 0.856, + "num_input_tokens_seen": 3361325, + "step": 158 + }, + { + "epoch": 0.019118619611615463, + "flos": 29262855004320.0, + "grad_norm": 2.695461191692284, + "learning_rate": 3.672147120210184e-06, + "loss": 0.8876, + "num_input_tokens_seen": 3377925, + "step": 159 + }, + { + "epoch": 0.019238862502254553, + "flos": 20856927430080.0, + "grad_norm": 2.0888020138339645, + "learning_rate": 3.6766891159659177e-06, + "loss": 0.8634, + "num_input_tokens_seen": 3396335, + "step": 160 + }, + { + "epoch": 0.019359105392893646, + "flos": 21330173011680.0, + "grad_norm": 2.9998901664973276, + "learning_rate": 3.6812028124990075e-06, + "loss": 0.8777, + "num_input_tokens_seen": 3413605, + "step": 161 + }, + { + "epoch": 0.019479348283532736, + "flos": 16260424360800.0, + "grad_norm": 2.9564653507694443, + "learning_rate": 3.6856885602676016e-06, + "loss": 0.8164, + "num_input_tokens_seen": 3429640, + "step": 162 + }, + { + "epoch": 0.019599591174171826, + "flos": 22059707053920.0, + "grad_norm": 2.5286706097558103, + "learning_rate": 3.6901467032597733e-06, + "loss": 0.9438, + "num_input_tokens_seen": 3448125, + "step": 163 + }, + { + "epoch": 0.01971983406481092, + "flos": 19581729646560.0, + "grad_norm": 2.6276594949473036, + "learning_rate": 3.694577579151804e-06, + "loss": 0.8738, + "num_input_tokens_seen": 3466535, + "step": 164 + }, + { + "epoch": 0.01984007695545001, + "flos": 19072163475840.0, + "grad_norm": 2.5697202608918897, + "learning_rate": 3.6989815194616703e-06, + "loss": 0.7404, + "num_input_tokens_seen": 3483730, + "step": 165 + }, + { + "epoch": 0.0199603198460891, + "flos": 20818599673920.0, + "grad_norm": 2.521421968026123, + "learning_rate": 3.703358849697888e-06, + "loss": 0.7991, + "num_input_tokens_seen": 3503640, + "step": 166 + }, + { + "epoch": 0.020080562736728192, + "flos": 21839478994080.0, + "grad_norm": 1.8613950511192343, + "learning_rate": 3.7077098895038803e-06, + "loss": 0.8245, + "num_input_tokens_seen": 3523010, + "step": 167 + }, + { + "epoch": 0.020200805627367282, + "flos": 21657801708960.0, + "grad_norm": 2.263690833209008, + "learning_rate": 3.712034952798045e-06, + "loss": 0.9677, + "num_input_tokens_seen": 3541125, + "step": 168 + }, + { + "epoch": 0.02032104851800637, + "flos": 33494665392000.0, + "grad_norm": 3.8607268031640665, + "learning_rate": 3.7163343479096656e-06, + "loss": 0.845, + "num_input_tokens_seen": 3562700, + "step": 169 + }, + { + "epoch": 0.020441291408645465, + "flos": 31641460442400.0, + "grad_norm": 2.3731621105958447, + "learning_rate": 3.720608377710802e-06, + "loss": 0.8297, + "num_input_tokens_seen": 3582790, + "step": 170 + }, + { + "epoch": 0.020561534299284555, + "flos": 20856890260320.0, + "grad_norm": 2.732510176043745, + "learning_rate": 3.7248573397443277e-06, + "loss": 0.8661, + "num_input_tokens_seen": 3601715, + "step": 171 + }, + { + "epoch": 0.020681777189923645, + "flos": 20966669762400.0, + "grad_norm": 2.261491464319173, + "learning_rate": 3.729081526348224e-06, + "loss": 0.9759, + "num_input_tokens_seen": 3620245, + "step": 172 + }, + { + "epoch": 0.020802020080562738, + "flos": 28217029041600.0, + "grad_norm": 2.379042046841213, + "learning_rate": 3.7332812247762777e-06, + "loss": 0.8502, + "num_input_tokens_seen": 3641545, + "step": 173 + }, + { + "epoch": 0.020922262971201828, + "flos": 19653181392480.0, + "grad_norm": 2.6276682054159552, + "learning_rate": 3.737456717315293e-06, + "loss": 0.9548, + "num_input_tokens_seen": 3660510, + "step": 174 + }, + { + "epoch": 0.021042505861840918, + "flos": 15644052268800.0, + "grad_norm": 2.859508573820411, + "learning_rate": 3.7416082813989552e-06, + "loss": 0.9068, + "num_input_tokens_seen": 3677505, + "step": 175 + }, + { + "epoch": 0.02116274875248001, + "flos": 21111840609600.0, + "grad_norm": 2.690728665613299, + "learning_rate": 3.745736189718439e-06, + "loss": 0.8961, + "num_input_tokens_seen": 3696760, + "step": 176 + }, + { + "epoch": 0.0212829916431191, + "flos": 24680327764800.0, + "grad_norm": 4.0170670231989, + "learning_rate": 3.749840710329894e-06, + "loss": 0.7231, + "num_input_tokens_seen": 3717465, + "step": 177 + }, + { + "epoch": 0.02140323453375819, + "flos": 16621548745440.0, + "grad_norm": 3.834512543018463, + "learning_rate": 3.7539221067588938e-06, + "loss": 0.98, + "num_input_tokens_seen": 3731440, + "step": 178 + }, + { + "epoch": 0.021523477424397284, + "flos": 20270259709920.0, + "grad_norm": 3.892263657212638, + "learning_rate": 3.757980638101964e-06, + "loss": 0.9352, + "num_input_tokens_seen": 3744935, + "step": 179 + }, + { + "epoch": 0.021643720315036374, + "flos": 26066605992000.0, + "grad_norm": 2.3800452000469146, + "learning_rate": 3.7620165591252806e-06, + "loss": 0.8977, + "num_input_tokens_seen": 3763035, + "step": 180 + }, + { + "epoch": 0.021763963205675464, + "flos": 24752076868800.0, + "grad_norm": 2.011235862478353, + "learning_rate": 3.766030120360636e-06, + "loss": 0.9441, + "num_input_tokens_seen": 3782665, + "step": 181 + }, + { + "epoch": 0.021884206096314557, + "flos": 25776635995200.0, + "grad_norm": 2.667781259140009, + "learning_rate": 3.7700215681987578e-06, + "loss": 0.903, + "num_input_tokens_seen": 3802435, + "step": 182 + }, + { + "epoch": 0.022004448986953647, + "flos": 20054045984160.0, + "grad_norm": 2.0546376037341045, + "learning_rate": 3.7739911449800767e-06, + "loss": 0.8219, + "num_input_tokens_seen": 3822490, + "step": 183 + }, + { + "epoch": 0.022124691877592736, + "flos": 20451230769600.0, + "grad_norm": 2.036927256210866, + "learning_rate": 3.7779390890830114e-06, + "loss": 0.8072, + "num_input_tokens_seen": 3841140, + "step": 184 + }, + { + "epoch": 0.02224493476823183, + "flos": 23553125730240.0, + "grad_norm": 2.131552411323836, + "learning_rate": 3.7818656350098723e-06, + "loss": 0.8586, + "num_input_tokens_seen": 3862090, + "step": 185 + }, + { + "epoch": 0.02236517765887092, + "flos": 16885940213760.0, + "grad_norm": 3.5709366868387815, + "learning_rate": 3.7857710134704447e-06, + "loss": 0.7742, + "num_input_tokens_seen": 3881025, + "step": 186 + }, + { + "epoch": 0.02248542054951001, + "flos": 43445460324000.0, + "grad_norm": 2.2957531531352875, + "learning_rate": 3.7896554514633234e-06, + "loss": 0.7898, + "num_input_tokens_seen": 3905310, + "step": 187 + }, + { + "epoch": 0.022605663440149103, + "flos": 23334235781760.0, + "grad_norm": 2.0941117386977255, + "learning_rate": 3.7935191723550955e-06, + "loss": 0.8449, + "num_input_tokens_seen": 3924265, + "step": 188 + }, + { + "epoch": 0.022725906330788193, + "flos": 28978163113440.0, + "grad_norm": 4.493336300117879, + "learning_rate": 3.797362395957408e-06, + "loss": 0.8851, + "num_input_tokens_seen": 3944830, + "step": 189 + }, + { + "epoch": 0.022846149221427282, + "flos": 24461028948960.0, + "grad_norm": 6.384862712916619, + "learning_rate": 3.8011853386020055e-06, + "loss": 0.7847, + "num_input_tokens_seen": 3965735, + "step": 190 + }, + { + "epoch": 0.022966392112066376, + "flos": 15501557644320.0, + "grad_norm": 2.999906856362475, + "learning_rate": 3.804988213213804e-06, + "loss": 0.8978, + "num_input_tokens_seen": 3983975, + "step": 191 + }, + { + "epoch": 0.023086635002705466, + "flos": 55571061668160.0, + "grad_norm": 1.0195628194196373, + "learning_rate": 3.808771229382049e-06, + "loss": 0.6311, + "num_input_tokens_seen": 4043440, + "step": 192 + }, + { + "epoch": 0.023206877893344555, + "flos": 19285775318400.0, + "grad_norm": 2.0405971554408913, + "learning_rate": 3.8125345934296324e-06, + "loss": 0.8426, + "num_input_tokens_seen": 4061750, + "step": 193 + }, + { + "epoch": 0.02332712078398365, + "flos": 23039879753280.0, + "grad_norm": 2.1855844012082186, + "learning_rate": 3.81627850848061e-06, + "loss": 0.8784, + "num_input_tokens_seen": 4082345, + "step": 194 + }, + { + "epoch": 0.02344736367462274, + "flos": 24391026823680.0, + "grad_norm": 2.6646995342130975, + "learning_rate": 3.820003174525994e-06, + "loss": 0.8627, + "num_input_tokens_seen": 4101770, + "step": 195 + }, + { + "epoch": 0.02356760656526183, + "flos": 21548988620640.0, + "grad_norm": 2.3974315012194927, + "learning_rate": 3.823708788487851e-06, + "loss": 0.8268, + "num_input_tokens_seen": 4118770, + "step": 196 + }, + { + "epoch": 0.02368784945590092, + "flos": 25157587680480.0, + "grad_norm": 2.2369232585228005, + "learning_rate": 3.827395544281781e-06, + "loss": 0.8452, + "num_input_tokens_seen": 4141110, + "step": 197 + }, + { + "epoch": 0.02380809234654001, + "flos": 27522923514240.0, + "grad_norm": 2.3053495846196315, + "learning_rate": 3.831063632877802e-06, + "loss": 0.7917, + "num_input_tokens_seen": 4161430, + "step": 198 + }, + { + "epoch": 0.0239283352371791, + "flos": 18233816345280.0, + "grad_norm": 3.568592416788194, + "learning_rate": 3.834713242359712e-06, + "loss": 0.7596, + "num_input_tokens_seen": 4179260, + "step": 199 + }, + { + "epoch": 0.02404857812781819, + "flos": 21364486433760.0, + "grad_norm": 2.166332060090044, + "learning_rate": 3.838344557982959e-06, + "loss": 0.8711, + "num_input_tokens_seen": 4200640, + "step": 200 + }, + { + "epoch": 0.024168821018457284, + "flos": 16630618166880.0, + "grad_norm": 3.2002611709350886, + "learning_rate": 3.841957762231063e-06, + "loss": 0.8468, + "num_input_tokens_seen": 4218170, + "step": 201 + }, + { + "epoch": 0.024289063909096374, + "flos": 22788274682400.0, + "grad_norm": 2.3503796209959296, + "learning_rate": 3.8455530348706454e-06, + "loss": 0.8766, + "num_input_tokens_seen": 4237770, + "step": 202 + }, + { + "epoch": 0.024409306799735464, + "flos": 17723358100320.0, + "grad_norm": 2.0563388469018395, + "learning_rate": 3.849130553005099e-06, + "loss": 0.773, + "num_input_tokens_seen": 4255985, + "step": 203 + }, + { + "epoch": 0.024529549690374557, + "flos": 21585346379520.0, + "grad_norm": 2.165962679519443, + "learning_rate": 3.852690491126933e-06, + "loss": 0.8376, + "num_input_tokens_seen": 4276035, + "step": 204 + }, + { + "epoch": 0.024649792581013647, + "flos": 25514883579840.0, + "grad_norm": 3.0189393167781082, + "learning_rate": 3.856233021168845e-06, + "loss": 0.9134, + "num_input_tokens_seen": 4295730, + "step": 205 + }, + { + "epoch": 0.024770035471652737, + "flos": 34445951454240.0, + "grad_norm": 2.9405864976557505, + "learning_rate": 3.859758312553544e-06, + "loss": 0.9143, + "num_input_tokens_seen": 4317950, + "step": 206 + }, + { + "epoch": 0.02489027836229183, + "flos": 21474674803200.0, + "grad_norm": 2.248456213411311, + "learning_rate": 3.8632665322423735e-06, + "loss": 0.9175, + "num_input_tokens_seen": 4337605, + "step": 207 + }, + { + "epoch": 0.02501052125293092, + "flos": 23185719656160.0, + "grad_norm": 2.5611671144656327, + "learning_rate": 3.866757844782762e-06, + "loss": 0.8573, + "num_input_tokens_seen": 4358110, + "step": 208 + }, + { + "epoch": 0.02513076414357001, + "flos": 26350851845760.0, + "grad_norm": 2.6392083515632914, + "learning_rate": 3.870232412354527e-06, + "loss": 0.913, + "num_input_tokens_seen": 4374955, + "step": 209 + }, + { + "epoch": 0.025251007034209103, + "flos": 13571288315040.0, + "grad_norm": 2.2225235562529737, + "learning_rate": 3.873690394815086e-06, + "loss": 0.9268, + "num_input_tokens_seen": 4391535, + "step": 210 + }, + { + "epoch": 0.025371249924848193, + "flos": 15027940365120.0, + "grad_norm": 2.5802171021426576, + "learning_rate": 3.877131949743587e-06, + "loss": 0.9134, + "num_input_tokens_seen": 4408400, + "step": 211 + }, + { + "epoch": 0.025491492815487283, + "flos": 25517150935200.0, + "grad_norm": 7.224349851296363, + "learning_rate": 3.880557232483993e-06, + "loss": 0.7796, + "num_input_tokens_seen": 4427840, + "step": 212 + }, + { + "epoch": 0.025611735706126376, + "flos": 20601308025120.0, + "grad_norm": 2.162281170836802, + "learning_rate": 3.883966396187164e-06, + "loss": 0.8674, + "num_input_tokens_seen": 4447110, + "step": 213 + }, + { + "epoch": 0.025731978596765466, + "flos": 19035136661280.0, + "grad_norm": 2.206786771753798, + "learning_rate": 3.887359591851937e-06, + "loss": 0.9015, + "num_input_tokens_seen": 4464715, + "step": 214 + }, + { + "epoch": 0.025852221487404556, + "flos": 22132162383360.0, + "grad_norm": 1.7624085422892395, + "learning_rate": 3.890736968365265e-06, + "loss": 0.9246, + "num_input_tokens_seen": 4485030, + "step": 215 + }, + { + "epoch": 0.02597246437804365, + "flos": 26503939851840.0, + "grad_norm": 2.024847180286854, + "learning_rate": 3.894098672541412e-06, + "loss": 0.8498, + "num_input_tokens_seen": 4505935, + "step": 216 + }, + { + "epoch": 0.02609270726868274, + "flos": 32624718231840.0, + "grad_norm": 1.985590224737501, + "learning_rate": 3.89744484916025e-06, + "loss": 0.7542, + "num_input_tokens_seen": 4527045, + "step": 217 + }, + { + "epoch": 0.02621295015932183, + "flos": 26205718168320.0, + "grad_norm": 2.467704938051833, + "learning_rate": 3.900775641004673e-06, + "loss": 0.8733, + "num_input_tokens_seen": 4546360, + "step": 218 + }, + { + "epoch": 0.026333193049960922, + "flos": 42859424489760.0, + "grad_norm": 3.0929268406560135, + "learning_rate": 3.904091188897156e-06, + "loss": 0.7402, + "num_input_tokens_seen": 4565495, + "step": 219 + }, + { + "epoch": 0.026453435940600012, + "flos": 17938159375200.0, + "grad_norm": 2.7826989503598374, + "learning_rate": 3.90739163173548e-06, + "loss": 0.819, + "num_input_tokens_seen": 4583330, + "step": 220 + }, + { + "epoch": 0.026573678831239102, + "flos": 18957105867840.0, + "grad_norm": 2.719763292924102, + "learning_rate": 3.910677106527646e-06, + "loss": 0.8825, + "num_input_tokens_seen": 4600520, + "step": 221 + }, + { + "epoch": 0.026693921721878195, + "flos": 29199469096320.0, + "grad_norm": 4.982118675849075, + "learning_rate": 3.913947748426004e-06, + "loss": 0.8423, + "num_input_tokens_seen": 4617340, + "step": 222 + }, + { + "epoch": 0.026814164612517285, + "flos": 14107808295360.0, + "grad_norm": 3.062009188189135, + "learning_rate": 3.9172036907606136e-06, + "loss": 0.7652, + "num_input_tokens_seen": 4630820, + "step": 223 + }, + { + "epoch": 0.026934407503156375, + "flos": 23478291536160.0, + "grad_norm": 3.3609379880542125, + "learning_rate": 3.920445065071855e-06, + "loss": 0.9509, + "num_input_tokens_seen": 4651985, + "step": 224 + }, + { + "epoch": 0.027054650393795468, + "flos": 28908830043840.0, + "grad_norm": 4.646422184916411, + "learning_rate": 3.923672001142322e-06, + "loss": 0.801, + "num_input_tokens_seen": 4672295, + "step": 225 + }, + { + "epoch": 0.027174893284434558, + "flos": 31386435753600.0, + "grad_norm": 1.9266701401389907, + "learning_rate": 3.926884627027996e-06, + "loss": 0.8438, + "num_input_tokens_seen": 4696065, + "step": 226 + }, + { + "epoch": 0.027295136175073648, + "flos": 22022977597440.0, + "grad_norm": 2.4948589870684965, + "learning_rate": 3.930083069088744e-06, + "loss": 0.7733, + "num_input_tokens_seen": 4716065, + "step": 227 + }, + { + "epoch": 0.02741537906571274, + "flos": 60714120552960.0, + "grad_norm": 0.9816261238821327, + "learning_rate": 3.933267452018137e-06, + "loss": 0.5917, + "num_input_tokens_seen": 4775860, + "step": 228 + }, + { + "epoch": 0.02753562195635183, + "flos": 24570994299840.0, + "grad_norm": 4.601025069090731, + "learning_rate": 3.936437898872622e-06, + "loss": 0.8438, + "num_input_tokens_seen": 4795835, + "step": 229 + }, + { + "epoch": 0.02765586484699092, + "flos": 34045198371840.0, + "grad_norm": 5.507243155413747, + "learning_rate": 3.9395945311000525e-06, + "loss": 0.7977, + "num_input_tokens_seen": 4817460, + "step": 230 + }, + { + "epoch": 0.027776107737630014, + "flos": 14808938907360.0, + "grad_norm": 3.5863882541965344, + "learning_rate": 3.942737468567608e-06, + "loss": 0.9112, + "num_input_tokens_seen": 4835475, + "step": 231 + }, + { + "epoch": 0.027896350628269104, + "flos": 47851105177440.0, + "grad_norm": 3.269861858611439, + "learning_rate": 3.9458668295891026e-06, + "loss": 0.8621, + "num_input_tokens_seen": 4857760, + "step": 232 + }, + { + "epoch": 0.028016593518908194, + "flos": 21653415677280.0, + "grad_norm": 2.615091059249935, + "learning_rate": 3.948982730951712e-06, + "loss": 0.8682, + "num_input_tokens_seen": 4875855, + "step": 233 + }, + { + "epoch": 0.028136836409547287, + "flos": 18413151935520.0, + "grad_norm": 4.695975004607295, + "learning_rate": 3.9520852879421254e-06, + "loss": 0.8234, + "num_input_tokens_seen": 4893200, + "step": 234 + }, + { + "epoch": 0.028257079300186377, + "flos": 31531569431040.0, + "grad_norm": 2.751416015191735, + "learning_rate": 3.955174614372137e-06, + "loss": 0.8155, + "num_input_tokens_seen": 4912965, + "step": 235 + }, + { + "epoch": 0.028377322190825467, + "flos": 23479703987040.0, + "grad_norm": 4.433359712985916, + "learning_rate": 3.9582508226037045e-06, + "loss": 0.8462, + "num_input_tokens_seen": 4933105, + "step": 236 + }, + { + "epoch": 0.02849756508146456, + "flos": 20450190016320.0, + "grad_norm": 3.089893596261571, + "learning_rate": 3.9613140235734636e-06, + "loss": 0.9407, + "num_input_tokens_seen": 4950085, + "step": 237 + }, + { + "epoch": 0.02861780797210365, + "flos": 14263200826560.0, + "grad_norm": 2.257041931655236, + "learning_rate": 3.96436432681674e-06, + "loss": 0.8119, + "num_input_tokens_seen": 4968075, + "step": 238 + }, + { + "epoch": 0.02873805086274274, + "flos": 25771432228800.0, + "grad_norm": 2.6013968016863975, + "learning_rate": 3.967401840491044e-06, + "loss": 0.8894, + "num_input_tokens_seen": 4987355, + "step": 239 + }, + { + "epoch": 0.028858293753381833, + "flos": 17279408023200.0, + "grad_norm": 7.541189638791276, + "learning_rate": 3.97042667139909e-06, + "loss": 0.8743, + "num_input_tokens_seen": 5004680, + "step": 240 + }, + { + "epoch": 0.028978536644020923, + "flos": 23005380482400.0, + "grad_norm": 2.248048359781764, + "learning_rate": 3.973438925011327e-06, + "loss": 0.8712, + "num_input_tokens_seen": 5022965, + "step": 241 + }, + { + "epoch": 0.029098779534660012, + "flos": 28289075503680.0, + "grad_norm": 2.814845738766254, + "learning_rate": 3.976438705488002e-06, + "loss": 0.9137, + "num_input_tokens_seen": 5042625, + "step": 242 + }, + { + "epoch": 0.029219022425299106, + "flos": 13862076046560.0, + "grad_norm": 3.300471942562175, + "learning_rate": 3.9794261157007744e-06, + "loss": 0.9289, + "num_input_tokens_seen": 5060380, + "step": 243 + }, + { + "epoch": 0.029339265315938196, + "flos": 19394662746240.0, + "grad_norm": 2.7122205177158345, + "learning_rate": 3.982401257253887e-06, + "loss": 0.8486, + "num_input_tokens_seen": 5078720, + "step": 244 + }, + { + "epoch": 0.029459508206577285, + "flos": 15647360377440.0, + "grad_norm": 3.987987720474423, + "learning_rate": 3.985364230504893e-06, + "loss": 0.8968, + "num_input_tokens_seen": 5096605, + "step": 245 + }, + { + "epoch": 0.02957975109721638, + "flos": 28180522603680.0, + "grad_norm": 2.140252965075998, + "learning_rate": 3.988315134584976e-06, + "loss": 0.8465, + "num_input_tokens_seen": 5116285, + "step": 246 + }, + { + "epoch": 0.02969999398785547, + "flos": 24281061472800.0, + "grad_norm": 3.242708389538745, + "learning_rate": 3.991254067418851e-06, + "loss": 0.8059, + "num_input_tokens_seen": 5136825, + "step": 247 + }, + { + "epoch": 0.02982023687849456, + "flos": 35028084463680.0, + "grad_norm": 2.195970212758699, + "learning_rate": 3.994181125744254e-06, + "loss": 0.8318, + "num_input_tokens_seen": 5158630, + "step": 248 + }, + { + "epoch": 0.02994047976913365, + "flos": 26139656037600.0, + "grad_norm": 2.140256863980849, + "learning_rate": 3.99709640513106e-06, + "loss": 0.7413, + "num_input_tokens_seen": 5179510, + "step": 249 + }, + { + "epoch": 0.03006072265977274, + "flos": 25588156644000.0, + "grad_norm": 2.375262697649045, + "learning_rate": 4e-06, + "loss": 0.8549, + "num_input_tokens_seen": 5199345, + "step": 250 + }, + { + "epoch": 0.03018096555041183, + "flos": 22095470096640.0, + "grad_norm": 3.5747494162923124, + "learning_rate": 3.999999848300794e-06, + "loss": 0.8844, + "num_input_tokens_seen": 5218330, + "step": 251 + }, + { + "epoch": 0.030301208441050925, + "flos": 30144770827200.0, + "grad_norm": 1.8026029098698877, + "learning_rate": 3.999999393203203e-06, + "loss": 0.8912, + "num_input_tokens_seen": 5240740, + "step": 252 + }, + { + "epoch": 0.030421451331690014, + "flos": 23587550661600.0, + "grad_norm": 4.013468478727138, + "learning_rate": 3.999998634707293e-06, + "loss": 0.8496, + "num_input_tokens_seen": 5260290, + "step": 253 + }, + { + "epoch": 0.030541694222329104, + "flos": 27888396760800.0, + "grad_norm": 2.6406390849984933, + "learning_rate": 3.999997572813182e-06, + "loss": 0.9663, + "num_input_tokens_seen": 5278100, + "step": 254 + }, + { + "epoch": 0.030661937112968194, + "flos": 18562002588960.0, + "grad_norm": 2.0276024566990714, + "learning_rate": 3.999996207521028e-06, + "loss": 0.8759, + "num_input_tokens_seen": 5296810, + "step": 255 + }, + { + "epoch": 0.030782180003607287, + "flos": 12950864719200.0, + "grad_norm": 3.1505672549582457, + "learning_rate": 3.999994538831039e-06, + "loss": 0.8232, + "num_input_tokens_seen": 5313395, + "step": 256 + }, + { + "epoch": 0.030902422894246377, + "flos": 23302152545280.0, + "grad_norm": 4.235415397727665, + "learning_rate": 3.99999256674347e-06, + "loss": 0.8587, + "num_input_tokens_seen": 5333545, + "step": 257 + }, + { + "epoch": 0.031022665784885467, + "flos": 55015287752160.0, + "grad_norm": 1.0031963168548128, + "learning_rate": 3.999990291258618e-06, + "loss": 0.5346, + "num_input_tokens_seen": 5392235, + "step": 258 + }, + { + "epoch": 0.03114290867552456, + "flos": 19290272859360.0, + "grad_norm": 2.603559947226818, + "learning_rate": 3.999987712376829e-06, + "loss": 0.8676, + "num_input_tokens_seen": 5410555, + "step": 259 + }, + { + "epoch": 0.031263151566163654, + "flos": 20929531438560.0, + "grad_norm": 2.3844106849009923, + "learning_rate": 3.999984830098494e-06, + "loss": 0.8224, + "num_input_tokens_seen": 5430135, + "step": 260 + }, + { + "epoch": 0.03138339445680274, + "flos": 14772060771840.0, + "grad_norm": 5.384798920825606, + "learning_rate": 3.999981644424051e-06, + "loss": 0.979, + "num_input_tokens_seen": 5444935, + "step": 261 + }, + { + "epoch": 0.03150363734744183, + "flos": 11639234837280.0, + "grad_norm": 3.035378702031444, + "learning_rate": 3.999978155353982e-06, + "loss": 0.8637, + "num_input_tokens_seen": 5462080, + "step": 262 + }, + { + "epoch": 0.03162388023808092, + "flos": 33679167578880.0, + "grad_norm": 3.420656030654604, + "learning_rate": 3.9999743628888186e-06, + "loss": 0.7999, + "num_input_tokens_seen": 5483870, + "step": 263 + }, + { + "epoch": 0.03174412312872001, + "flos": 20781163992000.0, + "grad_norm": 2.638893236747863, + "learning_rate": 3.999970267029133e-06, + "loss": 0.8952, + "num_input_tokens_seen": 5502830, + "step": 264 + }, + { + "epoch": 0.0318643660193591, + "flos": 23693836206240.0, + "grad_norm": 2.4257719128808324, + "learning_rate": 3.999965867775548e-06, + "loss": 0.8026, + "num_input_tokens_seen": 5523225, + "step": 265 + }, + { + "epoch": 0.0319846089099982, + "flos": 13898210786880.0, + "grad_norm": 2.8147627311436922, + "learning_rate": 3.9999611651287315e-06, + "loss": 0.8691, + "num_input_tokens_seen": 5541380, + "step": 266 + }, + { + "epoch": 0.03210485180063729, + "flos": 14732246225280.0, + "grad_norm": 4.919539996716855, + "learning_rate": 3.999956159089396e-06, + "loss": 0.7825, + "num_input_tokens_seen": 5558830, + "step": 267 + }, + { + "epoch": 0.03222509469127638, + "flos": 28871654550240.0, + "grad_norm": 2.6818525726511417, + "learning_rate": 3.999950849658302e-06, + "loss": 0.7943, + "num_input_tokens_seen": 5577750, + "step": 268 + }, + { + "epoch": 0.03234533758191547, + "flos": 16922335142400.0, + "grad_norm": 2.5673401873660597, + "learning_rate": 3.999945236836254e-06, + "loss": 0.8424, + "num_input_tokens_seen": 5596715, + "step": 269 + }, + { + "epoch": 0.03246558047255456, + "flos": 18962049445920.0, + "grad_norm": 3.747493033068634, + "learning_rate": 3.999939320624103e-06, + "loss": 0.948, + "num_input_tokens_seen": 5611865, + "step": 270 + }, + { + "epoch": 0.03258582336319365, + "flos": 23694245073600.0, + "grad_norm": 2.4994580635143873, + "learning_rate": 3.999933101022749e-06, + "loss": 0.8993, + "num_input_tokens_seen": 5632270, + "step": 271 + }, + { + "epoch": 0.032706066253832745, + "flos": 27631067546880.0, + "grad_norm": 2.096651259195374, + "learning_rate": 3.999926578033132e-06, + "loss": 0.8675, + "num_input_tokens_seen": 5652085, + "step": 272 + }, + { + "epoch": 0.032826309144471835, + "flos": 45558893728800.0, + "grad_norm": 2.9965043364769874, + "learning_rate": 3.999919751656244e-06, + "loss": 0.6325, + "num_input_tokens_seen": 5678985, + "step": 273 + }, + { + "epoch": 0.032946552035110925, + "flos": 25775409393120.0, + "grad_norm": 2.66016288229946, + "learning_rate": 3.9999126218931195e-06, + "loss": 0.7577, + "num_input_tokens_seen": 5697020, + "step": 274 + }, + { + "epoch": 0.033066794925750015, + "flos": 15100581543360.0, + "grad_norm": 2.8287067974060016, + "learning_rate": 3.99990518874484e-06, + "loss": 0.8968, + "num_input_tokens_seen": 5713460, + "step": 275 + }, + { + "epoch": 0.033187037816389105, + "flos": 22743776746080.0, + "grad_norm": 5.5400775361402035, + "learning_rate": 3.999897452212534e-06, + "loss": 0.9235, + "num_input_tokens_seen": 5730790, + "step": 276 + }, + { + "epoch": 0.033307280707028195, + "flos": 23298063871680.0, + "grad_norm": 2.320302430204896, + "learning_rate": 3.999889412297374e-06, + "loss": 1.0003, + "num_input_tokens_seen": 5750215, + "step": 277 + }, + { + "epoch": 0.03342752359766729, + "flos": 28798901862720.0, + "grad_norm": 2.3462625895646623, + "learning_rate": 3.999881069000581e-06, + "loss": 0.7909, + "num_input_tokens_seen": 5769945, + "step": 278 + }, + { + "epoch": 0.03354776648830638, + "flos": 19356706687680.0, + "grad_norm": 3.6381258920925026, + "learning_rate": 3.99987242232342e-06, + "loss": 0.8686, + "num_input_tokens_seen": 5784950, + "step": 279 + }, + { + "epoch": 0.03366800937894547, + "flos": 17833249111680.0, + "grad_norm": 2.9205697964251796, + "learning_rate": 3.9998634722672026e-06, + "loss": 0.7969, + "num_input_tokens_seen": 5805005, + "step": 280 + }, + { + "epoch": 0.03378825226958456, + "flos": 35899146716640.0, + "grad_norm": 2.272528502702059, + "learning_rate": 3.999854218833286e-06, + "loss": 0.7866, + "num_input_tokens_seen": 5825825, + "step": 281 + }, + { + "epoch": 0.03390849516022365, + "flos": 25665667060800.0, + "grad_norm": 2.70252437202975, + "learning_rate": 3.999844662023075e-06, + "loss": 0.8218, + "num_input_tokens_seen": 5845700, + "step": 282 + }, + { + "epoch": 0.03402873805086274, + "flos": 21254000706240.0, + "grad_norm": 2.1638859976391567, + "learning_rate": 3.999834801838018e-06, + "loss": 0.9208, + "num_input_tokens_seen": 5864680, + "step": 283 + }, + { + "epoch": 0.03414898094150183, + "flos": 22678458010560.0, + "grad_norm": 2.040376787731395, + "learning_rate": 3.9998246382796115e-06, + "loss": 0.74, + "num_input_tokens_seen": 5884260, + "step": 284 + }, + { + "epoch": 0.03426922383214093, + "flos": 18853050508800.0, + "grad_norm": 2.655298937906238, + "learning_rate": 3.999814171349399e-06, + "loss": 0.9095, + "num_input_tokens_seen": 5902120, + "step": 285 + }, + { + "epoch": 0.03438946672278002, + "flos": 34702054066080.0, + "grad_norm": 1.9746216302900788, + "learning_rate": 3.9998034010489655e-06, + "loss": 0.7366, + "num_input_tokens_seen": 5925810, + "step": 286 + }, + { + "epoch": 0.03450970961341911, + "flos": 22132459741440.0, + "grad_norm": 2.590419699491366, + "learning_rate": 3.999792327379946e-06, + "loss": 0.7593, + "num_input_tokens_seen": 5945185, + "step": 287 + }, + { + "epoch": 0.034629952504058197, + "flos": 21694717014240.0, + "grad_norm": 2.3774033482878836, + "learning_rate": 3.999780950344021e-06, + "loss": 0.9626, + "num_input_tokens_seen": 5963375, + "step": 288 + }, + { + "epoch": 0.034750195394697286, + "flos": 20019435204000.0, + "grad_norm": 2.4302063415580224, + "learning_rate": 3.999769269942916e-06, + "loss": 0.8285, + "num_input_tokens_seen": 5983415, + "step": 289 + }, + { + "epoch": 0.034870438285336376, + "flos": 27925795272960.0, + "grad_norm": 2.18968037524234, + "learning_rate": 3.999757286178402e-06, + "loss": 0.8127, + "num_input_tokens_seen": 6005850, + "step": 290 + }, + { + "epoch": 0.03499068117597547, + "flos": 21986062292160.0, + "grad_norm": 2.0806376246501608, + "learning_rate": 3.999744999052299e-06, + "loss": 0.9039, + "num_input_tokens_seen": 6027240, + "step": 291 + }, + { + "epoch": 0.03511092406661456, + "flos": 57159986658720.0, + "grad_norm": 0.9546996694934848, + "learning_rate": 3.9997324085664675e-06, + "loss": 0.6112, + "num_input_tokens_seen": 6087470, + "step": 292 + }, + { + "epoch": 0.03523116695725365, + "flos": 22895229282720.0, + "grad_norm": 2.3933804657562336, + "learning_rate": 3.999719514722821e-06, + "loss": 0.9196, + "num_input_tokens_seen": 6107600, + "step": 293 + }, + { + "epoch": 0.03535140984789274, + "flos": 36850098251040.0, + "grad_norm": 3.6029817842207805, + "learning_rate": 3.999706317523314e-06, + "loss": 0.7523, + "num_input_tokens_seen": 6126160, + "step": 294 + }, + { + "epoch": 0.03547165273853183, + "flos": 20420002437600.0, + "grad_norm": 2.3527544112273087, + "learning_rate": 3.999692816969948e-06, + "loss": 0.8647, + "num_input_tokens_seen": 6145695, + "step": 295 + }, + { + "epoch": 0.03559189562917092, + "flos": 69750804916320.0, + "grad_norm": 0.9947350304477397, + "learning_rate": 3.999679013064772e-06, + "loss": 0.6931, + "num_input_tokens_seen": 6212440, + "step": 296 + }, + { + "epoch": 0.03571213851981002, + "flos": 21620997912960.0, + "grad_norm": 3.482515805572111, + "learning_rate": 3.99966490580988e-06, + "loss": 0.8636, + "num_input_tokens_seen": 6229800, + "step": 297 + }, + { + "epoch": 0.03583238141044911, + "flos": 43882236637440.0, + "grad_norm": 2.570004059161235, + "learning_rate": 3.999650495207411e-06, + "loss": 0.661, + "num_input_tokens_seen": 6255825, + "step": 298 + }, + { + "epoch": 0.0359526243010882, + "flos": 18883684124640.0, + "grad_norm": 3.237226178582341, + "learning_rate": 3.999635781259553e-06, + "loss": 0.902, + "num_input_tokens_seen": 6271090, + "step": 299 + }, + { + "epoch": 0.03607286719172729, + "flos": 61580722434720.0, + "grad_norm": 0.9190827448332826, + "learning_rate": 3.999620763968535e-06, + "loss": 0.5227, + "num_input_tokens_seen": 6329965, + "step": 300 + }, + { + "epoch": 0.03619311008236638, + "flos": 27779174805120.0, + "grad_norm": 2.0601687115960328, + "learning_rate": 3.999605443336638e-06, + "loss": 0.8649, + "num_input_tokens_seen": 6352095, + "step": 301 + }, + { + "epoch": 0.03631335297300547, + "flos": 13601736082080.0, + "grad_norm": 2.758189664558529, + "learning_rate": 3.999589819366185e-06, + "loss": 0.8903, + "num_input_tokens_seen": 6365885, + "step": 302 + }, + { + "epoch": 0.036433595863644565, + "flos": 27591996395520.0, + "grad_norm": 2.17276048790873, + "learning_rate": 3.999573892059547e-06, + "loss": 0.8458, + "num_input_tokens_seen": 6385015, + "step": 303 + }, + { + "epoch": 0.036553838754283655, + "flos": 24536866726560.0, + "grad_norm": 2.38250800516679, + "learning_rate": 3.999557661419138e-06, + "loss": 0.8133, + "num_input_tokens_seen": 6405165, + "step": 304 + }, + { + "epoch": 0.036674081644922744, + "flos": 23370853728960.0, + "grad_norm": 3.6217713282758877, + "learning_rate": 3.9995411274474225e-06, + "loss": 0.8151, + "num_input_tokens_seen": 6424445, + "step": 305 + }, + { + "epoch": 0.036794324535561834, + "flos": 27450245166240.0, + "grad_norm": 2.118373749670307, + "learning_rate": 3.999524290146908e-06, + "loss": 0.8147, + "num_input_tokens_seen": 6444650, + "step": 306 + }, + { + "epoch": 0.036914567426200924, + "flos": 19435741064640.0, + "grad_norm": 4.976786322478531, + "learning_rate": 3.9995071495201485e-06, + "loss": 0.9257, + "num_input_tokens_seen": 6461755, + "step": 307 + }, + { + "epoch": 0.037034810316840014, + "flos": 22277444739840.0, + "grad_norm": 3.326791558861306, + "learning_rate": 3.999489705569744e-06, + "loss": 0.984, + "num_input_tokens_seen": 6479455, + "step": 308 + }, + { + "epoch": 0.03715505320747911, + "flos": 18561891079680.0, + "grad_norm": 2.352535110033261, + "learning_rate": 3.999471958298341e-06, + "loss": 0.8674, + "num_input_tokens_seen": 6498365, + "step": 309 + }, + { + "epoch": 0.0372752960981182, + "flos": 35903607087840.0, + "grad_norm": 2.169180096899666, + "learning_rate": 3.999453907708631e-06, + "loss": 0.7635, + "num_input_tokens_seen": 6520770, + "step": 310 + }, + { + "epoch": 0.03739553898875729, + "flos": 20784360591360.0, + "grad_norm": 2.5205985241026094, + "learning_rate": 3.999435553803353e-06, + "loss": 0.8123, + "num_input_tokens_seen": 6540170, + "step": 311 + }, + { + "epoch": 0.03751578187939638, + "flos": 20235165722880.0, + "grad_norm": 2.9484896333307478, + "learning_rate": 3.999416896585292e-06, + "loss": 0.8366, + "num_input_tokens_seen": 6557200, + "step": 312 + }, + { + "epoch": 0.03763602477003547, + "flos": 20638595028000.0, + "grad_norm": 3.4384167922506785, + "learning_rate": 3.9993979360572775e-06, + "loss": 0.8594, + "num_input_tokens_seen": 6577340, + "step": 313 + }, + { + "epoch": 0.03775626766067456, + "flos": 16667682151200.0, + "grad_norm": 2.9493401710828784, + "learning_rate": 3.999378672222185e-06, + "loss": 0.8284, + "num_input_tokens_seen": 6595125, + "step": 314 + }, + { + "epoch": 0.03787651055131366, + "flos": 21111617591040.0, + "grad_norm": 2.416946910087274, + "learning_rate": 3.9993591050829385e-06, + "loss": 0.8283, + "num_input_tokens_seen": 6612990, + "step": 315 + }, + { + "epoch": 0.037996753441952746, + "flos": 21986657008320.0, + "grad_norm": 1.9626688210347694, + "learning_rate": 3.999339234642506e-06, + "loss": 0.7931, + "num_input_tokens_seen": 6632740, + "step": 316 + }, + { + "epoch": 0.038116996332591836, + "flos": 27669655491360.0, + "grad_norm": 8.78745124119652, + "learning_rate": 3.9993190609038994e-06, + "loss": 0.8389, + "num_input_tokens_seen": 6651505, + "step": 317 + }, + { + "epoch": 0.038237239223230926, + "flos": 21147826670880.0, + "grad_norm": 3.2897522264303904, + "learning_rate": 3.999298583870182e-06, + "loss": 0.8322, + "num_input_tokens_seen": 6671090, + "step": 318 + }, + { + "epoch": 0.038357482113870016, + "flos": 25520013006720.0, + "grad_norm": 2.605005500480131, + "learning_rate": 3.999277803544458e-06, + "loss": 0.7731, + "num_input_tokens_seen": 6691245, + "step": 319 + }, + { + "epoch": 0.038477725004509106, + "flos": 59143079950560.0, + "grad_norm": 0.95942239799589, + "learning_rate": 3.999256719929882e-06, + "loss": 0.6234, + "num_input_tokens_seen": 6752520, + "step": 320 + }, + { + "epoch": 0.0385979678951482, + "flos": 67221044299680.0, + "grad_norm": 1.2562642490470857, + "learning_rate": 3.999235333029651e-06, + "loss": 0.67, + "num_input_tokens_seen": 6806460, + "step": 321 + }, + { + "epoch": 0.03871821078578729, + "flos": 22714221053280.0, + "grad_norm": 2.6047563622702046, + "learning_rate": 3.999213642847009e-06, + "loss": 0.8203, + "num_input_tokens_seen": 6827045, + "step": 322 + }, + { + "epoch": 0.03883845367642638, + "flos": 26242744982880.0, + "grad_norm": 1.9693111649279025, + "learning_rate": 3.999191649385247e-06, + "loss": 0.9121, + "num_input_tokens_seen": 6848220, + "step": 323 + }, + { + "epoch": 0.03895869656706547, + "flos": 56881539287520.0, + "grad_norm": 1.5245436345127434, + "learning_rate": 3.999169352647702e-06, + "loss": 0.5969, + "num_input_tokens_seen": 6909400, + "step": 324 + }, + { + "epoch": 0.03907893945770456, + "flos": 24827765967360.0, + "grad_norm": 2.1538990860502687, + "learning_rate": 3.999146752637755e-06, + "loss": 0.8284, + "num_input_tokens_seen": 6930445, + "step": 325 + }, + { + "epoch": 0.03919918234834365, + "flos": 18342443584800.0, + "grad_norm": 2.7054476599183843, + "learning_rate": 3.999123849358836e-06, + "loss": 0.8937, + "num_input_tokens_seen": 6948110, + "step": 326 + }, + { + "epoch": 0.03931942523898275, + "flos": 25189187710080.0, + "grad_norm": 2.5536840492131296, + "learning_rate": 3.999100642814418e-06, + "loss": 0.748, + "num_input_tokens_seen": 6968550, + "step": 327 + }, + { + "epoch": 0.03943966812962184, + "flos": 23224642128480.0, + "grad_norm": 2.467323663782101, + "learning_rate": 3.999077133008022e-06, + "loss": 0.8829, + "num_input_tokens_seen": 6987135, + "step": 328 + }, + { + "epoch": 0.03955991102026093, + "flos": 29126976597120.0, + "grad_norm": 2.6755262267271434, + "learning_rate": 3.9990533199432145e-06, + "loss": 0.906, + "num_input_tokens_seen": 7008630, + "step": 329 + }, + { + "epoch": 0.03968015391090002, + "flos": 17577406688160.0, + "grad_norm": 2.86144844296948, + "learning_rate": 3.999029203623608e-06, + "loss": 0.7592, + "num_input_tokens_seen": 7026350, + "step": 330 + }, + { + "epoch": 0.03980039680153911, + "flos": 21768175927200.0, + "grad_norm": 2.7068987837358502, + "learning_rate": 3.99900478405286e-06, + "loss": 0.8694, + "num_input_tokens_seen": 7045660, + "step": 331 + }, + { + "epoch": 0.0399206396921782, + "flos": 15173631588960.0, + "grad_norm": 2.9297954081689777, + "learning_rate": 3.998980061234676e-06, + "loss": 0.8198, + "num_input_tokens_seen": 7063575, + "step": 332 + }, + { + "epoch": 0.040040882582817294, + "flos": 14402313002880.0, + "grad_norm": 2.9156077688020767, + "learning_rate": 3.9989550351728055e-06, + "loss": 0.764, + "num_input_tokens_seen": 7080505, + "step": 333 + }, + { + "epoch": 0.040161125473456384, + "flos": 19253431893600.0, + "grad_norm": 2.7672812464820677, + "learning_rate": 3.998929705871046e-06, + "loss": 0.8458, + "num_input_tokens_seen": 7097860, + "step": 334 + }, + { + "epoch": 0.040281368364095474, + "flos": 17795739090240.0, + "grad_norm": 2.9504778264887963, + "learning_rate": 3.99890407333324e-06, + "loss": 0.8879, + "num_input_tokens_seen": 7116390, + "step": 335 + }, + { + "epoch": 0.040401611254734564, + "flos": 19545892264320.0, + "grad_norm": 2.8728432236307353, + "learning_rate": 3.998878137563275e-06, + "loss": 0.8719, + "num_input_tokens_seen": 7135940, + "step": 336 + }, + { + "epoch": 0.040521854145373654, + "flos": 22022977597440.0, + "grad_norm": 2.4313871565623675, + "learning_rate": 3.998851898565085e-06, + "loss": 0.8549, + "num_input_tokens_seen": 7155745, + "step": 337 + }, + { + "epoch": 0.04064209703601274, + "flos": 22642248930720.0, + "grad_norm": 2.11627504175087, + "learning_rate": 3.998825356342653e-06, + "loss": 0.8319, + "num_input_tokens_seen": 7175920, + "step": 338 + }, + { + "epoch": 0.04076233992665183, + "flos": 38526606663360.0, + "grad_norm": 3.300944632374711, + "learning_rate": 3.998798510900003e-06, + "loss": 0.7303, + "num_input_tokens_seen": 7198720, + "step": 339 + }, + { + "epoch": 0.04088258281729093, + "flos": 25848236420160.0, + "grad_norm": 2.6348900451013453, + "learning_rate": 3.998771362241207e-06, + "loss": 0.8376, + "num_input_tokens_seen": 7219125, + "step": 340 + }, + { + "epoch": 0.04100282570793002, + "flos": 19761548443680.0, + "grad_norm": 2.0074468494946864, + "learning_rate": 3.998743910370385e-06, + "loss": 0.8766, + "num_input_tokens_seen": 7237985, + "step": 341 + }, + { + "epoch": 0.04112306859856911, + "flos": 22532915465760.0, + "grad_norm": 2.2323716792940553, + "learning_rate": 3.998716155291702e-06, + "loss": 0.733, + "num_input_tokens_seen": 7255825, + "step": 342 + }, + { + "epoch": 0.0412433114892082, + "flos": 25004611183680.0, + "grad_norm": 1.936065019374821, + "learning_rate": 3.998688097009366e-06, + "loss": 0.9056, + "num_input_tokens_seen": 7276590, + "step": 343 + }, + { + "epoch": 0.04136355437984729, + "flos": 25155283155360.0, + "grad_norm": 2.585464664805726, + "learning_rate": 3.998659735527636e-06, + "loss": 0.8002, + "num_input_tokens_seen": 7295680, + "step": 344 + }, + { + "epoch": 0.04148379727048638, + "flos": 22933965906240.0, + "grad_norm": 2.0914141022481703, + "learning_rate": 3.998631070850813e-06, + "loss": 0.7786, + "num_input_tokens_seen": 7316300, + "step": 345 + }, + { + "epoch": 0.041604040161125476, + "flos": 14043604652640.0, + "grad_norm": 2.8278078354506055, + "learning_rate": 3.9986021029832455e-06, + "loss": 0.8305, + "num_input_tokens_seen": 7333615, + "step": 346 + }, + { + "epoch": 0.041724283051764566, + "flos": 12074821718400.0, + "grad_norm": 2.7832158031509486, + "learning_rate": 3.9985728319293285e-06, + "loss": 0.915, + "num_input_tokens_seen": 7347590, + "step": 347 + }, + { + "epoch": 0.041844525942403656, + "flos": 12367876805280.0, + "grad_norm": 20.578036760678316, + "learning_rate": 3.998543257693501e-06, + "loss": 0.8517, + "num_input_tokens_seen": 7364345, + "step": 348 + }, + { + "epoch": 0.041964768833042745, + "flos": 23735360561760.0, + "grad_norm": 2.1041109310611885, + "learning_rate": 3.998513380280251e-06, + "loss": 0.878, + "num_input_tokens_seen": 7384625, + "step": 349 + }, + { + "epoch": 0.042085011723681835, + "flos": 11858682332160.0, + "grad_norm": 3.209831399976013, + "learning_rate": 3.99848319969411e-06, + "loss": 0.9538, + "num_input_tokens_seen": 7402225, + "step": 350 + }, + { + "epoch": 0.042205254614320925, + "flos": 16850028492000.0, + "grad_norm": 2.343096934567901, + "learning_rate": 3.9984527159396564e-06, + "loss": 0.7948, + "num_input_tokens_seen": 7420865, + "step": 351 + }, + { + "epoch": 0.04232549750496002, + "flos": 25082530467840.0, + "grad_norm": 6.061717818621851, + "learning_rate": 3.9984219290215154e-06, + "loss": 0.8454, + "num_input_tokens_seen": 7441040, + "step": 352 + }, + { + "epoch": 0.04244574039559911, + "flos": 26686249022880.0, + "grad_norm": 1.7863950875730232, + "learning_rate": 3.998390838944356e-06, + "loss": 0.8916, + "num_input_tokens_seen": 7462545, + "step": 353 + }, + { + "epoch": 0.0425659832862382, + "flos": 20893396698240.0, + "grad_norm": 2.4437564859523704, + "learning_rate": 3.998359445712895e-06, + "loss": 0.903, + "num_input_tokens_seen": 7481650, + "step": 354 + }, + { + "epoch": 0.04268622617687729, + "flos": 23297766513600.0, + "grad_norm": 2.6549840266435085, + "learning_rate": 3.9983277493318955e-06, + "loss": 0.8104, + "num_input_tokens_seen": 7500945, + "step": 355 + }, + { + "epoch": 0.04280646906751638, + "flos": 25956752150400.0, + "grad_norm": 7.0673876028768134, + "learning_rate": 3.998295749806165e-06, + "loss": 0.8135, + "num_input_tokens_seen": 7522170, + "step": 356 + }, + { + "epoch": 0.04292671195815547, + "flos": 26868260835840.0, + "grad_norm": 3.883538345251728, + "learning_rate": 3.998263447140558e-06, + "loss": 0.8332, + "num_input_tokens_seen": 7542370, + "step": 357 + }, + { + "epoch": 0.04304695484879457, + "flos": 39400456648320.0, + "grad_norm": 2.174202200483581, + "learning_rate": 3.998230841339976e-06, + "loss": 0.8177, + "num_input_tokens_seen": 7564380, + "step": 358 + }, + { + "epoch": 0.04316719773943366, + "flos": 19618310424000.0, + "grad_norm": 2.37444258723972, + "learning_rate": 3.998197932409363e-06, + "loss": 0.8497, + "num_input_tokens_seen": 7582390, + "step": 359 + }, + { + "epoch": 0.04328744063007275, + "flos": 22420125213120.0, + "grad_norm": 2.9210869853575723, + "learning_rate": 3.9981647203537125e-06, + "loss": 0.8631, + "num_input_tokens_seen": 7599890, + "step": 360 + }, + { + "epoch": 0.04340768352071184, + "flos": 21252476746080.0, + "grad_norm": 2.1573096726011625, + "learning_rate": 3.998131205178063e-06, + "loss": 0.959, + "num_input_tokens_seen": 7618360, + "step": 361 + }, + { + "epoch": 0.04352792641135093, + "flos": 11567485733280.0, + "grad_norm": 3.3560578758332493, + "learning_rate": 3.998097386887498e-06, + "loss": 0.77, + "num_input_tokens_seen": 7635435, + "step": 362 + }, + { + "epoch": 0.04364816930199002, + "flos": 23589780847200.0, + "grad_norm": 1.9038775329552957, + "learning_rate": 3.998063265487148e-06, + "loss": 0.8487, + "num_input_tokens_seen": 7656485, + "step": 363 + }, + { + "epoch": 0.043768412192629114, + "flos": 14409375257280.0, + "grad_norm": 3.312768889256093, + "learning_rate": 3.99802884098219e-06, + "loss": 0.8126, + "num_input_tokens_seen": 7674595, + "step": 364 + }, + { + "epoch": 0.043888655083268203, + "flos": 26430964145760.0, + "grad_norm": 3.38742237635565, + "learning_rate": 3.997994113377845e-06, + "loss": 0.8231, + "num_input_tokens_seen": 7693495, + "step": 365 + }, + { + "epoch": 0.04400889797390729, + "flos": 27196224060960.0, + "grad_norm": 1.9433257593669269, + "learning_rate": 3.9979590826793815e-06, + "loss": 0.8325, + "num_input_tokens_seen": 7714685, + "step": 366 + }, + { + "epoch": 0.04412914086454638, + "flos": 20090812610400.0, + "grad_norm": 2.080896465528798, + "learning_rate": 3.997923748892113e-06, + "loss": 0.8097, + "num_input_tokens_seen": 7734005, + "step": 367 + }, + { + "epoch": 0.04424938375518547, + "flos": 22572432654240.0, + "grad_norm": 2.491003921383191, + "learning_rate": 3.9978881120214015e-06, + "loss": 0.8901, + "num_input_tokens_seen": 7755525, + "step": 368 + }, + { + "epoch": 0.04436962664582456, + "flos": 24097785888000.0, + "grad_norm": 1.989267520571856, + "learning_rate": 3.997852172072652e-06, + "loss": 0.7927, + "num_input_tokens_seen": 7776740, + "step": 369 + }, + { + "epoch": 0.04448986953646366, + "flos": 18196975379520.0, + "grad_norm": 3.092449338678401, + "learning_rate": 3.9978159290513155e-06, + "loss": 0.8969, + "num_input_tokens_seen": 7794820, + "step": 370 + }, + { + "epoch": 0.04461011242710275, + "flos": 30074694362400.0, + "grad_norm": 2.585978066855142, + "learning_rate": 3.997779382962892e-06, + "loss": 0.8028, + "num_input_tokens_seen": 7817705, + "step": 371 + }, + { + "epoch": 0.04473035531774184, + "flos": 29709815832000.0, + "grad_norm": 2.833518097032976, + "learning_rate": 3.997742533812924e-06, + "loss": 0.7385, + "num_input_tokens_seen": 7840970, + "step": 372 + }, + { + "epoch": 0.04485059820838093, + "flos": 13133025211200.0, + "grad_norm": 2.9719914458653602, + "learning_rate": 3.997705381607001e-06, + "loss": 0.9215, + "num_input_tokens_seen": 7857785, + "step": 373 + }, + { + "epoch": 0.04497084109902002, + "flos": 67996823256960.0, + "grad_norm": 0.9732415150474141, + "learning_rate": 3.997667926350761e-06, + "loss": 0.6028, + "num_input_tokens_seen": 7916115, + "step": 374 + }, + { + "epoch": 0.04509108398965911, + "flos": 64250561641440.0, + "grad_norm": 0.9012987435290116, + "learning_rate": 3.997630168049886e-06, + "loss": 0.5776, + "num_input_tokens_seen": 7974480, + "step": 375 + }, + { + "epoch": 0.045211326880298205, + "flos": 22239228492960.0, + "grad_norm": 2.0612463175299456, + "learning_rate": 3.997592106710101e-06, + "loss": 0.7758, + "num_input_tokens_seen": 7993940, + "step": 376 + }, + { + "epoch": 0.045331569770937295, + "flos": 32113182063840.0, + "grad_norm": 2.64609480239221, + "learning_rate": 3.997553742337182e-06, + "loss": 0.656, + "num_input_tokens_seen": 8013365, + "step": 377 + }, + { + "epoch": 0.045451812661576385, + "flos": 22131790685760.0, + "grad_norm": 1.9334508977968083, + "learning_rate": 3.997515074936949e-06, + "loss": 0.9145, + "num_input_tokens_seen": 8034240, + "step": 378 + }, + { + "epoch": 0.045572055552215475, + "flos": 16558460195520.0, + "grad_norm": 2.6992198099446183, + "learning_rate": 3.997476104515268e-06, + "loss": 0.8708, + "num_input_tokens_seen": 8052430, + "step": 379 + }, + { + "epoch": 0.045692298442854565, + "flos": 17578373101920.0, + "grad_norm": 2.374869694397706, + "learning_rate": 3.9974368310780485e-06, + "loss": 0.7737, + "num_input_tokens_seen": 8069605, + "step": 380 + }, + { + "epoch": 0.045812541333493655, + "flos": 26722755460800.0, + "grad_norm": 2.7302325450799283, + "learning_rate": 3.997397254631251e-06, + "loss": 0.7442, + "num_input_tokens_seen": 8090225, + "step": 381 + }, + { + "epoch": 0.04593278422413275, + "flos": 60164405307840.0, + "grad_norm": 1.2544308334568008, + "learning_rate": 3.997357375180878e-06, + "loss": 0.6003, + "num_input_tokens_seen": 8154505, + "step": 382 + }, + { + "epoch": 0.04605302711477184, + "flos": 21768138757440.0, + "grad_norm": 1.9511556233291527, + "learning_rate": 3.997317192732979e-06, + "loss": 0.7529, + "num_input_tokens_seen": 8173285, + "step": 383 + }, + { + "epoch": 0.04617327000541093, + "flos": 19431838239840.0, + "grad_norm": 2.2132620840592683, + "learning_rate": 3.99727670729365e-06, + "loss": 0.8245, + "num_input_tokens_seen": 8191845, + "step": 384 + }, + { + "epoch": 0.04629351289605002, + "flos": 25375102347840.0, + "grad_norm": 1.9852040879191535, + "learning_rate": 3.997235918869033e-06, + "loss": 0.7835, + "num_input_tokens_seen": 8212880, + "step": 385 + }, + { + "epoch": 0.04641375578668911, + "flos": 20528964204960.0, + "grad_norm": 2.116046893878367, + "learning_rate": 3.997194827465315e-06, + "loss": 0.8251, + "num_input_tokens_seen": 8231475, + "step": 386 + }, + { + "epoch": 0.0465339986773282, + "flos": 13170089195520.0, + "grad_norm": 3.427201258399654, + "learning_rate": 3.997153433088728e-06, + "loss": 0.915, + "num_input_tokens_seen": 8248600, + "step": 387 + }, + { + "epoch": 0.0466542415679673, + "flos": 25519938667200.0, + "grad_norm": 2.196828189964323, + "learning_rate": 3.997111735745554e-06, + "loss": 0.8126, + "num_input_tokens_seen": 8271570, + "step": 388 + }, + { + "epoch": 0.04677448445860639, + "flos": 22204989410400.0, + "grad_norm": 2.193803479012833, + "learning_rate": 3.997069735442118e-06, + "loss": 0.8258, + "num_input_tokens_seen": 8291265, + "step": 389 + }, + { + "epoch": 0.04689472734924548, + "flos": 28107472558080.0, + "grad_norm": 1.6962223687729039, + "learning_rate": 3.997027432184792e-06, + "loss": 0.8038, + "num_input_tokens_seen": 8315925, + "step": 390 + }, + { + "epoch": 0.04701497023988457, + "flos": 23115866209920.0, + "grad_norm": 6.052244655890561, + "learning_rate": 3.99698482597999e-06, + "loss": 0.8949, + "num_input_tokens_seen": 8333605, + "step": 391 + }, + { + "epoch": 0.04713521313052366, + "flos": 64735664376480.0, + "grad_norm": 0.8730675611669463, + "learning_rate": 3.99694191683418e-06, + "loss": 0.6388, + "num_input_tokens_seen": 8394405, + "step": 392 + }, + { + "epoch": 0.047255456021162746, + "flos": 18744906476160.0, + "grad_norm": 2.0077506326725003, + "learning_rate": 3.996898704753867e-06, + "loss": 0.8179, + "num_input_tokens_seen": 8412355, + "step": 393 + }, + { + "epoch": 0.04737569891180184, + "flos": 22022308541760.0, + "grad_norm": 2.632164034801691, + "learning_rate": 3.996855189745609e-06, + "loss": 0.8781, + "num_input_tokens_seen": 8431620, + "step": 394 + }, + { + "epoch": 0.04749594180244093, + "flos": 29015598795360.0, + "grad_norm": 2.2819552441887883, + "learning_rate": 3.996811371816007e-06, + "loss": 0.9283, + "num_input_tokens_seen": 8454045, + "step": 395 + }, + { + "epoch": 0.04761618469308002, + "flos": 35061617320800.0, + "grad_norm": 2.2295472601539554, + "learning_rate": 3.996767250971707e-06, + "loss": 0.7789, + "num_input_tokens_seen": 8476785, + "step": 396 + }, + { + "epoch": 0.04773642758371911, + "flos": 25594029466080.0, + "grad_norm": 8.829823450416624, + "learning_rate": 3.996722827219403e-06, + "loss": 0.8717, + "num_input_tokens_seen": 8498400, + "step": 397 + }, + { + "epoch": 0.0478566704743582, + "flos": 20602088590080.0, + "grad_norm": 2.557517627644128, + "learning_rate": 3.996678100565833e-06, + "loss": 0.832, + "num_input_tokens_seen": 8517455, + "step": 398 + }, + { + "epoch": 0.04797691336499729, + "flos": 18808626912000.0, + "grad_norm": 3.6423321433784683, + "learning_rate": 3.996633071017783e-06, + "loss": 0.8887, + "num_input_tokens_seen": 8534085, + "step": 399 + }, + { + "epoch": 0.04809715625563638, + "flos": 21069349840320.0, + "grad_norm": 2.498778521635963, + "learning_rate": 3.996587738582084e-06, + "loss": 0.8191, + "num_input_tokens_seen": 8550885, + "step": 400 + }, + { + "epoch": 0.04821739914627548, + "flos": 23771606811360.0, + "grad_norm": 3.0983301033494297, + "learning_rate": 3.9965421032656115e-06, + "loss": 0.8644, + "num_input_tokens_seen": 8570815, + "step": 401 + }, + { + "epoch": 0.04833764203691457, + "flos": 22168891839840.0, + "grad_norm": 4.929938792067329, + "learning_rate": 3.99649616507529e-06, + "loss": 0.9481, + "num_input_tokens_seen": 8589910, + "step": 402 + }, + { + "epoch": 0.04845788492755366, + "flos": 65810153957760.0, + "grad_norm": 0.8896910806898035, + "learning_rate": 3.996449924018088e-06, + "loss": 0.6316, + "num_input_tokens_seen": 8650370, + "step": 403 + }, + { + "epoch": 0.04857812781819275, + "flos": 19253840760960.0, + "grad_norm": 2.1449272193209916, + "learning_rate": 3.99640338010102e-06, + "loss": 0.7958, + "num_input_tokens_seen": 8669475, + "step": 404 + }, + { + "epoch": 0.04869837070883184, + "flos": 24028006781280.0, + "grad_norm": 2.2739018231539316, + "learning_rate": 3.996356533331146e-06, + "loss": 0.7866, + "num_input_tokens_seen": 8691345, + "step": 405 + }, + { + "epoch": 0.04881861359947093, + "flos": 25150934293440.0, + "grad_norm": 3.5418283463019447, + "learning_rate": 3.996309383715573e-06, + "loss": 0.6155, + "num_input_tokens_seen": 8710125, + "step": 406 + }, + { + "epoch": 0.048938856490110025, + "flos": 16339718926080.0, + "grad_norm": 3.917756733810483, + "learning_rate": 3.996261931261454e-06, + "loss": 0.7373, + "num_input_tokens_seen": 8728705, + "step": 407 + }, + { + "epoch": 0.049059099380749115, + "flos": 29852050268160.0, + "grad_norm": 1.888314782422157, + "learning_rate": 3.996214175975987e-06, + "loss": 0.8669, + "num_input_tokens_seen": 8749225, + "step": 408 + }, + { + "epoch": 0.049179342271388204, + "flos": 35867174989440.0, + "grad_norm": 3.144774319749473, + "learning_rate": 3.996166117866417e-06, + "loss": 0.7953, + "num_input_tokens_seen": 8771160, + "step": 409 + }, + { + "epoch": 0.049299585162027294, + "flos": 14590569335520.0, + "grad_norm": 2.3766498858094387, + "learning_rate": 3.996117756940035e-06, + "loss": 0.8658, + "num_input_tokens_seen": 8787845, + "step": 410 + }, + { + "epoch": 0.049419828052666384, + "flos": 19539870763200.0, + "grad_norm": 2.2879577678970966, + "learning_rate": 3.996069093204175e-06, + "loss": 0.9758, + "num_input_tokens_seen": 8803805, + "step": 411 + }, + { + "epoch": 0.049540070943305474, + "flos": 13640547045120.0, + "grad_norm": 5.881307584215136, + "learning_rate": 3.996020126666221e-06, + "loss": 0.8798, + "num_input_tokens_seen": 8820785, + "step": 412 + }, + { + "epoch": 0.04966031383394457, + "flos": 21800742370560.0, + "grad_norm": 2.71738636414452, + "learning_rate": 3.995970857333601e-06, + "loss": 0.8251, + "num_input_tokens_seen": 8841195, + "step": 413 + }, + { + "epoch": 0.04978055672458366, + "flos": 28577633049600.0, + "grad_norm": 1.974624045116592, + "learning_rate": 3.995921285213789e-06, + "loss": 0.7968, + "num_input_tokens_seen": 8861455, + "step": 414 + }, + { + "epoch": 0.04990079961522275, + "flos": 19799950539360.0, + "grad_norm": 2.925170345725813, + "learning_rate": 3.995871410314305e-06, + "loss": 0.8109, + "num_input_tokens_seen": 8880015, + "step": 415 + }, + { + "epoch": 0.05002104250586184, + "flos": 62645616484320.0, + "grad_norm": 0.9257821513626282, + "learning_rate": 3.995821232642714e-06, + "loss": 0.5964, + "num_input_tokens_seen": 8938420, + "step": 416 + }, + { + "epoch": 0.05014128539650093, + "flos": 27888471100320.0, + "grad_norm": 2.393853892967669, + "learning_rate": 3.995770752206629e-06, + "loss": 0.8251, + "num_input_tokens_seen": 8959495, + "step": 417 + }, + { + "epoch": 0.05026152828714002, + "flos": 17680161105600.0, + "grad_norm": 2.3372896803620122, + "learning_rate": 3.995719969013709e-06, + "loss": 0.9734, + "num_input_tokens_seen": 8976675, + "step": 418 + }, + { + "epoch": 0.05038177117777912, + "flos": 19105956521280.0, + "grad_norm": 4.42257156547059, + "learning_rate": 3.995668883071655e-06, + "loss": 0.8549, + "num_input_tokens_seen": 8992900, + "step": 419 + }, + { + "epoch": 0.050502014068418206, + "flos": 20638223330400.0, + "grad_norm": 4.709612152264331, + "learning_rate": 3.995617494388219e-06, + "loss": 0.9108, + "num_input_tokens_seen": 9011020, + "step": 420 + }, + { + "epoch": 0.050622256959057296, + "flos": 21330619048800.0, + "grad_norm": 2.5771817418035625, + "learning_rate": 3.995565802971196e-06, + "loss": 0.806, + "num_input_tokens_seen": 9030995, + "step": 421 + }, + { + "epoch": 0.050742499849696386, + "flos": 27633929618400.0, + "grad_norm": 2.496626103185835, + "learning_rate": 3.995513808828427e-06, + "loss": 0.679, + "num_input_tokens_seen": 9053790, + "step": 422 + }, + { + "epoch": 0.050862742740335476, + "flos": 19837720749120.0, + "grad_norm": 2.1852276580021046, + "learning_rate": 3.9954615119678e-06, + "loss": 0.768, + "num_input_tokens_seen": 9072345, + "step": 423 + }, + { + "epoch": 0.050982985630974566, + "flos": 22053351024960.0, + "grad_norm": 2.6202380265882343, + "learning_rate": 3.995408912397248e-06, + "loss": 0.8045, + "num_input_tokens_seen": 9090240, + "step": 424 + }, + { + "epoch": 0.05110322852161366, + "flos": 20711384885280.0, + "grad_norm": 3.0104471136927184, + "learning_rate": 3.99535601012475e-06, + "loss": 0.931, + "num_input_tokens_seen": 9105570, + "step": 425 + }, + { + "epoch": 0.05122347141225275, + "flos": 28506292812960.0, + "grad_norm": 2.6444264874351737, + "learning_rate": 3.995302805158333e-06, + "loss": 0.7533, + "num_input_tokens_seen": 9128225, + "step": 426 + }, + { + "epoch": 0.05134371430289184, + "flos": 19694705748000.0, + "grad_norm": 8.687684859909321, + "learning_rate": 3.9952492975060665e-06, + "loss": 0.8371, + "num_input_tokens_seen": 9148295, + "step": 427 + }, + { + "epoch": 0.05146395719353093, + "flos": 34409370676800.0, + "grad_norm": 2.9434768258438497, + "learning_rate": 3.995195487176067e-06, + "loss": 0.8497, + "num_input_tokens_seen": 9168525, + "step": 428 + }, + { + "epoch": 0.05158420008417002, + "flos": 21729141945600.0, + "grad_norm": 5.678558699190005, + "learning_rate": 3.995141374176499e-06, + "loss": 0.857, + "num_input_tokens_seen": 9186920, + "step": 429 + }, + { + "epoch": 0.05170444297480911, + "flos": 72450385664640.0, + "grad_norm": 0.869959516186247, + "learning_rate": 3.995086958515572e-06, + "loss": 0.6303, + "num_input_tokens_seen": 9244940, + "step": 430 + }, + { + "epoch": 0.05182468586544821, + "flos": 62327540415360.0, + "grad_norm": 0.861358572839104, + "learning_rate": 3.995032240201538e-06, + "loss": 0.5993, + "num_input_tokens_seen": 9307335, + "step": 431 + }, + { + "epoch": 0.0519449287560873, + "flos": 41167187535360.0, + "grad_norm": 0.9311383250821043, + "learning_rate": 3.9949772192427e-06, + "loss": 0.6308, + "num_input_tokens_seen": 9353960, + "step": 432 + }, + { + "epoch": 0.05206517164672639, + "flos": 17469708692640.0, + "grad_norm": 2.2329429892656343, + "learning_rate": 3.994921895647405e-06, + "loss": 0.7958, + "num_input_tokens_seen": 9372130, + "step": 433 + }, + { + "epoch": 0.05218541453736548, + "flos": 63911150130240.0, + "grad_norm": 0.8402414489192941, + "learning_rate": 3.994866269424043e-06, + "loss": 0.5532, + "num_input_tokens_seen": 9428500, + "step": 434 + }, + { + "epoch": 0.05230565742800457, + "flos": 19289789652480.0, + "grad_norm": 2.441882570722183, + "learning_rate": 3.9948103405810545e-06, + "loss": 0.7841, + "num_input_tokens_seen": 9447450, + "step": 435 + }, + { + "epoch": 0.05242590031864366, + "flos": 25262312095200.0, + "grad_norm": 2.625272010152247, + "learning_rate": 3.994754109126923e-06, + "loss": 0.8617, + "num_input_tokens_seen": 9468945, + "step": 436 + }, + { + "epoch": 0.052546143209282754, + "flos": 26174006629440.0, + "grad_norm": 5.804270178681287, + "learning_rate": 3.994697575070181e-06, + "loss": 0.9327, + "num_input_tokens_seen": 9488405, + "step": 437 + }, + { + "epoch": 0.052666386099921844, + "flos": 22126289561280.0, + "grad_norm": 2.765308254228769, + "learning_rate": 3.994640738419402e-06, + "loss": 0.9146, + "num_input_tokens_seen": 9507780, + "step": 438 + }, + { + "epoch": 0.052786628990560934, + "flos": 23847779116800.0, + "grad_norm": 3.1041892816799335, + "learning_rate": 3.9945835991832075e-06, + "loss": 0.8121, + "num_input_tokens_seen": 9529665, + "step": 439 + }, + { + "epoch": 0.052906871881200024, + "flos": 24569693358240.0, + "grad_norm": 2.389644446640517, + "learning_rate": 3.994526157370268e-06, + "loss": 0.9297, + "num_input_tokens_seen": 9548280, + "step": 440 + }, + { + "epoch": 0.053027114771839114, + "flos": 56381414235840.0, + "grad_norm": 0.8859330447991518, + "learning_rate": 3.994468412989296e-06, + "loss": 0.592, + "num_input_tokens_seen": 9609690, + "step": 441 + }, + { + "epoch": 0.053147357662478203, + "flos": 17286507447360.0, + "grad_norm": 2.551365269959042, + "learning_rate": 3.994410366049052e-06, + "loss": 0.9285, + "num_input_tokens_seen": 9627550, + "step": 442 + }, + { + "epoch": 0.0532676005531173, + "flos": 17140593204960.0, + "grad_norm": 2.724531318886383, + "learning_rate": 3.994352016558341e-06, + "loss": 0.8284, + "num_input_tokens_seen": 9644855, + "step": 443 + }, + { + "epoch": 0.05338784344375639, + "flos": 27780104049120.0, + "grad_norm": 2.44337293833041, + "learning_rate": 3.994293364526014e-06, + "loss": 0.7402, + "num_input_tokens_seen": 9665740, + "step": 444 + }, + { + "epoch": 0.05350808633439548, + "flos": 21477016498080.0, + "grad_norm": 8.526155508714877, + "learning_rate": 3.99423440996097e-06, + "loss": 0.8497, + "num_input_tokens_seen": 9685280, + "step": 445 + }, + { + "epoch": 0.05362832922503457, + "flos": 20055681453600.0, + "grad_norm": 3.52593648388324, + "learning_rate": 3.994175152872152e-06, + "loss": 0.8146, + "num_input_tokens_seen": 9704365, + "step": 446 + }, + { + "epoch": 0.05374857211567366, + "flos": 26099655642240.0, + "grad_norm": 2.4123507359634133, + "learning_rate": 3.994115593268548e-06, + "loss": 0.7843, + "num_input_tokens_seen": 9724145, + "step": 447 + }, + { + "epoch": 0.05386881500631275, + "flos": 27447271585440.0, + "grad_norm": 4.111066462872889, + "learning_rate": 3.994055731159195e-06, + "loss": 0.8209, + "num_input_tokens_seen": 9741615, + "step": 448 + }, + { + "epoch": 0.053989057896951846, + "flos": 23551304412000.0, + "grad_norm": 2.2646469636018276, + "learning_rate": 3.993995566553172e-06, + "loss": 0.8674, + "num_input_tokens_seen": 9760580, + "step": 449 + }, + { + "epoch": 0.054109300787590936, + "flos": 25192756007040.0, + "grad_norm": 1.8581480578221445, + "learning_rate": 3.993935099459607e-06, + "loss": 0.7714, + "num_input_tokens_seen": 9782195, + "step": 450 + }, + { + "epoch": 0.054229543678230026, + "flos": 23807407023840.0, + "grad_norm": 2.6617448407443196, + "learning_rate": 3.993874329887673e-06, + "loss": 0.7398, + "num_input_tokens_seen": 9800850, + "step": 451 + }, + { + "epoch": 0.054349786568869116, + "flos": 16297599854400.0, + "grad_norm": 3.205877544936781, + "learning_rate": 3.993813257846589e-06, + "loss": 0.8665, + "num_input_tokens_seen": 9817605, + "step": 452 + }, + { + "epoch": 0.054470029459508205, + "flos": 18634160560320.0, + "grad_norm": 3.833214440091447, + "learning_rate": 3.993751883345619e-06, + "loss": 0.9306, + "num_input_tokens_seen": 9835965, + "step": 453 + }, + { + "epoch": 0.054590272350147295, + "flos": 17760013217280.0, + "grad_norm": 4.231082776595413, + "learning_rate": 3.993690206394073e-06, + "loss": 0.8765, + "num_input_tokens_seen": 9852265, + "step": 454 + }, + { + "epoch": 0.054710515240786385, + "flos": 17760682272960.0, + "grad_norm": 2.7435659220822517, + "learning_rate": 3.993628227001307e-06, + "loss": 0.8768, + "num_input_tokens_seen": 9870465, + "step": 455 + }, + { + "epoch": 0.05483075813142548, + "flos": 48141186683520.0, + "grad_norm": 2.370214660677216, + "learning_rate": 3.993565945176726e-06, + "loss": 0.7141, + "num_input_tokens_seen": 9891490, + "step": 456 + }, + { + "epoch": 0.05495100102206457, + "flos": 19654184976000.0, + "grad_norm": 2.5351524882002137, + "learning_rate": 3.993503360929776e-06, + "loss": 0.8425, + "num_input_tokens_seen": 9910375, + "step": 457 + }, + { + "epoch": 0.05507124391270366, + "flos": 26322299736480.0, + "grad_norm": 1.9075654915782143, + "learning_rate": 3.99344047426995e-06, + "loss": 0.8107, + "num_input_tokens_seen": 9931635, + "step": 458 + }, + { + "epoch": 0.05519148680334275, + "flos": 22569013036320.0, + "grad_norm": 12.685673624511123, + "learning_rate": 3.993377285206789e-06, + "loss": 0.934, + "num_input_tokens_seen": 9951420, + "step": 459 + }, + { + "epoch": 0.05531172969398184, + "flos": 40494088656000.0, + "grad_norm": 1.7928638302890547, + "learning_rate": 3.99331379374988e-06, + "loss": 0.8645, + "num_input_tokens_seen": 9975025, + "step": 460 + }, + { + "epoch": 0.05543197258462093, + "flos": 23444349811680.0, + "grad_norm": 2.2474913917520922, + "learning_rate": 3.993249999908852e-06, + "loss": 0.8003, + "num_input_tokens_seen": 9995175, + "step": 461 + }, + { + "epoch": 0.05555221547526003, + "flos": 18598100159520.0, + "grad_norm": 2.27570966194825, + "learning_rate": 3.993185903693384e-06, + "loss": 0.8692, + "num_input_tokens_seen": 10011975, + "step": 462 + }, + { + "epoch": 0.05567245836589912, + "flos": 23553385918560.0, + "grad_norm": 3.138579352391353, + "learning_rate": 3.9931215051131995e-06, + "loss": 0.8237, + "num_input_tokens_seen": 10032620, + "step": 463 + }, + { + "epoch": 0.05579270125653821, + "flos": 27706905324480.0, + "grad_norm": 1.6816664282827267, + "learning_rate": 3.993056804178068e-06, + "loss": 0.8006, + "num_input_tokens_seen": 10054165, + "step": 464 + }, + { + "epoch": 0.0559129441471773, + "flos": 26975364115200.0, + "grad_norm": 2.665501896474175, + "learning_rate": 3.992991800897803e-06, + "loss": 0.8466, + "num_input_tokens_seen": 10073970, + "step": 465 + }, + { + "epoch": 0.05603318703781639, + "flos": 15208168029600.0, + "grad_norm": 2.7063012838460048, + "learning_rate": 3.9929264952822665e-06, + "loss": 0.9012, + "num_input_tokens_seen": 10091505, + "step": 466 + }, + { + "epoch": 0.05615342992845548, + "flos": 22234433593920.0, + "grad_norm": 2.4305270299297215, + "learning_rate": 3.992860887341366e-06, + "loss": 0.8884, + "num_input_tokens_seen": 10109675, + "step": 467 + }, + { + "epoch": 0.056273672819094574, + "flos": 23551192902720.0, + "grad_norm": 2.1594765889055165, + "learning_rate": 3.992794977085052e-06, + "loss": 0.808, + "num_input_tokens_seen": 10127675, + "step": 468 + }, + { + "epoch": 0.056393915709733664, + "flos": 19830175287840.0, + "grad_norm": 3.2350152160664023, + "learning_rate": 3.992728764523326e-06, + "loss": 0.8475, + "num_input_tokens_seen": 10146620, + "step": 469 + }, + { + "epoch": 0.05651415860037275, + "flos": 22132125213600.0, + "grad_norm": 2.0129460711165086, + "learning_rate": 3.99266224966623e-06, + "loss": 0.8055, + "num_input_tokens_seen": 10167535, + "step": 470 + }, + { + "epoch": 0.05663440149101184, + "flos": 19435815404160.0, + "grad_norm": 2.6646360971228944, + "learning_rate": 3.992595432523855e-06, + "loss": 0.8805, + "num_input_tokens_seen": 10184825, + "step": 471 + }, + { + "epoch": 0.05675464438165093, + "flos": 22637677050240.0, + "grad_norm": 2.2221787514673297, + "learning_rate": 3.992528313106338e-06, + "loss": 0.8593, + "num_input_tokens_seen": 10203025, + "step": 472 + }, + { + "epoch": 0.05687488727229002, + "flos": 16876870792320.0, + "grad_norm": 3.0547093539419667, + "learning_rate": 3.9924608914238595e-06, + "loss": 0.825, + "num_input_tokens_seen": 10218020, + "step": 473 + }, + { + "epoch": 0.05699513016292912, + "flos": 29127125276160.0, + "grad_norm": 3.357457949104483, + "learning_rate": 3.992393167486648e-06, + "loss": 0.8395, + "num_input_tokens_seen": 10237170, + "step": 474 + }, + { + "epoch": 0.05711537305356821, + "flos": 18889259588640.0, + "grad_norm": 2.575655314936586, + "learning_rate": 3.992325141304977e-06, + "loss": 0.8061, + "num_input_tokens_seen": 10255125, + "step": 475 + }, + { + "epoch": 0.0572356159442073, + "flos": 26720562444960.0, + "grad_norm": 2.720485966129349, + "learning_rate": 3.992256812889166e-06, + "loss": 0.8694, + "num_input_tokens_seen": 10271230, + "step": 476 + }, + { + "epoch": 0.05735585883484639, + "flos": 35065631654880.0, + "grad_norm": 3.0168089444942594, + "learning_rate": 3.992188182249582e-06, + "loss": 0.769, + "num_input_tokens_seen": 10293125, + "step": 477 + }, + { + "epoch": 0.05747610172548548, + "flos": 18707730982560.0, + "grad_norm": 2.5893089237018554, + "learning_rate": 3.992119249396633e-06, + "loss": 0.9056, + "num_input_tokens_seen": 10311810, + "step": 478 + }, + { + "epoch": 0.05759634461612457, + "flos": 27924643010400.0, + "grad_norm": 2.37351702518597, + "learning_rate": 3.992050014340778e-06, + "loss": 0.822, + "num_input_tokens_seen": 10332165, + "step": 479 + }, + { + "epoch": 0.057716587506763666, + "flos": 69192652135680.0, + "grad_norm": 0.8415675016722727, + "learning_rate": 3.99198047709252e-06, + "loss": 0.5502, + "num_input_tokens_seen": 10393685, + "step": 480 + }, + { + "epoch": 0.057836830397402755, + "flos": 24973382851680.0, + "grad_norm": 2.285516589845389, + "learning_rate": 3.991910637662408e-06, + "loss": 0.7884, + "num_input_tokens_seen": 10413975, + "step": 481 + }, + { + "epoch": 0.057957073288041845, + "flos": 25557262839840.0, + "grad_norm": 2.634496495532523, + "learning_rate": 3.9918404960610355e-06, + "loss": 0.8071, + "num_input_tokens_seen": 10433005, + "step": 482 + }, + { + "epoch": 0.058077316178680935, + "flos": 20747779813920.0, + "grad_norm": 2.852454811426977, + "learning_rate": 3.991770052299043e-06, + "loss": 0.7791, + "num_input_tokens_seen": 10451235, + "step": 483 + }, + { + "epoch": 0.058197559069320025, + "flos": 18889556946720.0, + "grad_norm": 2.4415990158274057, + "learning_rate": 3.991699306387118e-06, + "loss": 0.8765, + "num_input_tokens_seen": 10469705, + "step": 484 + }, + { + "epoch": 0.058317801959959115, + "flos": 24827728797600.0, + "grad_norm": 2.623873877797276, + "learning_rate": 3.991628258335991e-06, + "loss": 0.7812, + "num_input_tokens_seen": 10491910, + "step": 485 + }, + { + "epoch": 0.05843804485059821, + "flos": 23224716468000.0, + "grad_norm": 5.0254321847169, + "learning_rate": 3.991556908156442e-06, + "loss": 0.8738, + "num_input_tokens_seen": 10508435, + "step": 486 + }, + { + "epoch": 0.0585582877412373, + "flos": 23117501679360.0, + "grad_norm": 2.218028528819452, + "learning_rate": 3.9914852558592914e-06, + "loss": 0.8756, + "num_input_tokens_seen": 10529485, + "step": 487 + }, + { + "epoch": 0.05867853063187639, + "flos": 23472530223360.0, + "grad_norm": 3.7368031033855234, + "learning_rate": 3.991413301455413e-06, + "loss": 0.8081, + "num_input_tokens_seen": 10545935, + "step": 488 + }, + { + "epoch": 0.05879877352251548, + "flos": 29453415862080.0, + "grad_norm": 2.5075049285992366, + "learning_rate": 3.991341044955719e-06, + "loss": 0.7792, + "num_input_tokens_seen": 10565690, + "step": 489 + }, + { + "epoch": 0.05891901641315457, + "flos": 20128768668960.0, + "grad_norm": 2.3848813940342692, + "learning_rate": 3.991268486371172e-06, + "loss": 0.8164, + "num_input_tokens_seen": 10584245, + "step": 490 + }, + { + "epoch": 0.05903925930379366, + "flos": 24608913188640.0, + "grad_norm": 2.5007804606042274, + "learning_rate": 3.991195625712779e-06, + "loss": 0.8804, + "num_input_tokens_seen": 10601210, + "step": 491 + }, + { + "epoch": 0.05915950219443276, + "flos": 21220021812000.0, + "grad_norm": 2.3083374087668385, + "learning_rate": 3.991122462991592e-06, + "loss": 0.815, + "num_input_tokens_seen": 10620730, + "step": 492 + }, + { + "epoch": 0.05927974508507185, + "flos": 9888858644640.0, + "grad_norm": 3.373149223078093, + "learning_rate": 3.991048998218712e-06, + "loss": 0.8072, + "num_input_tokens_seen": 10634035, + "step": 493 + }, + { + "epoch": 0.05939998797571094, + "flos": 18233518987200.0, + "grad_norm": 3.4274252553813023, + "learning_rate": 3.990975231405281e-06, + "loss": 0.7714, + "num_input_tokens_seen": 10652485, + "step": 494 + }, + { + "epoch": 0.05952023086635003, + "flos": 28216062627840.0, + "grad_norm": 4.063663961484773, + "learning_rate": 3.990901162562491e-06, + "loss": 0.7905, + "num_input_tokens_seen": 10673175, + "step": 495 + }, + { + "epoch": 0.05964047375698912, + "flos": 14881394236800.0, + "grad_norm": 4.773469266145836, + "learning_rate": 3.9908267917015765e-06, + "loss": 0.905, + "num_input_tokens_seen": 10691235, + "step": 496 + }, + { + "epoch": 0.059760716647628206, + "flos": 23152446987360.0, + "grad_norm": 2.332652253704492, + "learning_rate": 3.990752118833821e-06, + "loss": 0.9265, + "num_input_tokens_seen": 10708675, + "step": 497 + }, + { + "epoch": 0.0598809595382673, + "flos": 22714444071840.0, + "grad_norm": 1.9720802599299108, + "learning_rate": 3.990677143970553e-06, + "loss": 0.7834, + "num_input_tokens_seen": 10729485, + "step": 498 + }, + { + "epoch": 0.06000120242890639, + "flos": 22095618775680.0, + "grad_norm": 2.815121594474172, + "learning_rate": 3.990601867123144e-06, + "loss": 0.8116, + "num_input_tokens_seen": 10749210, + "step": 499 + }, + { + "epoch": 0.06012144531954548, + "flos": 19144470126240.0, + "grad_norm": 2.614698997664678, + "learning_rate": 3.990526288303014e-06, + "loss": 0.8494, + "num_input_tokens_seen": 10768000, + "step": 500 + }, + { + "epoch": 0.06024168821018457, + "flos": 22750764660960.0, + "grad_norm": 22.034181570249054, + "learning_rate": 3.9904504075216295e-06, + "loss": 0.9051, + "num_input_tokens_seen": 10788760, + "step": 501 + }, + { + "epoch": 0.06036193110082366, + "flos": 18743977232160.0, + "grad_norm": 3.416757318798267, + "learning_rate": 3.990374224790501e-06, + "loss": 0.9373, + "num_input_tokens_seen": 10806965, + "step": 502 + }, + { + "epoch": 0.06048217399146275, + "flos": 17177136812640.0, + "grad_norm": 2.9193727927504836, + "learning_rate": 3.990297740121185e-06, + "loss": 0.7112, + "num_input_tokens_seen": 10824900, + "step": 503 + }, + { + "epoch": 0.06060241688210185, + "flos": 23989753364640.0, + "grad_norm": 2.9852865534608926, + "learning_rate": 3.990220953525284e-06, + "loss": 0.7825, + "num_input_tokens_seen": 10842010, + "step": 504 + }, + { + "epoch": 0.06072265977274094, + "flos": 14590532165760.0, + "grad_norm": 3.3340295700290548, + "learning_rate": 3.9901438650144465e-06, + "loss": 0.7413, + "num_input_tokens_seen": 10858260, + "step": 505 + }, + { + "epoch": 0.06084290266338003, + "flos": 20528964204960.0, + "grad_norm": 6.81085493191306, + "learning_rate": 3.990066474600367e-06, + "loss": 0.9237, + "num_input_tokens_seen": 10876230, + "step": 506 + }, + { + "epoch": 0.06096314555401912, + "flos": 22277333230560.0, + "grad_norm": 2.306735095771291, + "learning_rate": 3.989988782294786e-06, + "loss": 0.6798, + "num_input_tokens_seen": 10896320, + "step": 507 + }, + { + "epoch": 0.06108338844465821, + "flos": 19103986524000.0, + "grad_norm": 2.1553774187649313, + "learning_rate": 3.989910788109489e-06, + "loss": 0.9499, + "num_input_tokens_seen": 10912970, + "step": 508 + }, + { + "epoch": 0.0612036313352973, + "flos": 33535743710400.0, + "grad_norm": 2.5960379930604973, + "learning_rate": 3.989832492056307e-06, + "loss": 0.7552, + "num_input_tokens_seen": 10933995, + "step": 509 + }, + { + "epoch": 0.06132387422593639, + "flos": 27451137240480.0, + "grad_norm": 2.8719499414273315, + "learning_rate": 3.989753894147119e-06, + "loss": 0.8107, + "num_input_tokens_seen": 10954120, + "step": 510 + }, + { + "epoch": 0.061444117116575485, + "flos": 25847976231840.0, + "grad_norm": 2.3707765873004756, + "learning_rate": 3.989674994393846e-06, + "loss": 0.8024, + "num_input_tokens_seen": 10973595, + "step": 511 + }, + { + "epoch": 0.061564360007214575, + "flos": 28471124486400.0, + "grad_norm": 2.7575367905902435, + "learning_rate": 3.98959579280846e-06, + "loss": 0.9403, + "num_input_tokens_seen": 10991995, + "step": 512 + }, + { + "epoch": 0.061684602897853665, + "flos": 12077386431840.0, + "grad_norm": 3.026460147007282, + "learning_rate": 3.989516289402973e-06, + "loss": 0.8298, + "num_input_tokens_seen": 11007625, + "step": 513 + }, + { + "epoch": 0.061804845788492754, + "flos": 19504665266880.0, + "grad_norm": 2.785534847684515, + "learning_rate": 3.989436484189447e-06, + "loss": 0.8043, + "num_input_tokens_seen": 11025570, + "step": 514 + }, + { + "epoch": 0.061925088679131844, + "flos": 15319397152320.0, + "grad_norm": 3.4646074457451586, + "learning_rate": 3.9893563771799885e-06, + "loss": 0.8111, + "num_input_tokens_seen": 11042045, + "step": 515 + }, + { + "epoch": 0.062045331569770934, + "flos": 25882735691040.0, + "grad_norm": 4.080417266526761, + "learning_rate": 3.989275968386749e-06, + "loss": 0.8617, + "num_input_tokens_seen": 11059955, + "step": 516 + }, + { + "epoch": 0.06216557446041003, + "flos": 28070297064480.0, + "grad_norm": 2.4955424623814904, + "learning_rate": 3.989195257821926e-06, + "loss": 0.766, + "num_input_tokens_seen": 11078440, + "step": 517 + }, + { + "epoch": 0.06228581735104912, + "flos": 23444647169760.0, + "grad_norm": 2.441739332747395, + "learning_rate": 3.989114245497765e-06, + "loss": 0.8432, + "num_input_tokens_seen": 11098240, + "step": 518 + }, + { + "epoch": 0.06240606024168821, + "flos": 15173705928480.0, + "grad_norm": 2.785801880895048, + "learning_rate": 3.989032931426554e-06, + "loss": 0.9459, + "num_input_tokens_seen": 11115395, + "step": 519 + }, + { + "epoch": 0.06252630313232731, + "flos": 20602423117920.0, + "grad_norm": 2.601498348143913, + "learning_rate": 3.9889513156206295e-06, + "loss": 0.8649, + "num_input_tokens_seen": 11134835, + "step": 520 + }, + { + "epoch": 0.06264654602296639, + "flos": 20748448869600.0, + "grad_norm": 3.1314825340927777, + "learning_rate": 3.988869398092371e-06, + "loss": 0.7356, + "num_input_tokens_seen": 11154745, + "step": 521 + }, + { + "epoch": 0.06276678891360549, + "flos": 29563269703680.0, + "grad_norm": 2.7245860984108776, + "learning_rate": 3.988787178854206e-06, + "loss": 0.7832, + "num_input_tokens_seen": 11174280, + "step": 522 + }, + { + "epoch": 0.06288703180424457, + "flos": 22094652361920.0, + "grad_norm": 2.4488034420987463, + "learning_rate": 3.988704657918608e-06, + "loss": 0.8753, + "num_input_tokens_seen": 11193340, + "step": 523 + }, + { + "epoch": 0.06300727469488367, + "flos": 14955522205440.0, + "grad_norm": 2.9438429184035724, + "learning_rate": 3.988621835298094e-06, + "loss": 0.7998, + "num_input_tokens_seen": 11210555, + "step": 524 + }, + { + "epoch": 0.06312751758552275, + "flos": 24500285949120.0, + "grad_norm": 3.604629633868545, + "learning_rate": 3.988538711005229e-06, + "loss": 0.9196, + "num_input_tokens_seen": 11230010, + "step": 525 + }, + { + "epoch": 0.06324776047616185, + "flos": 21476310272640.0, + "grad_norm": 3.101550200699795, + "learning_rate": 3.988455285052622e-06, + "loss": 0.8814, + "num_input_tokens_seen": 11246190, + "step": 526 + }, + { + "epoch": 0.06336800336680094, + "flos": 21657021144000.0, + "grad_norm": 2.1079321553715076, + "learning_rate": 3.98837155745293e-06, + "loss": 0.8394, + "num_input_tokens_seen": 11264230, + "step": 527 + }, + { + "epoch": 0.06348824625744003, + "flos": 19473139576800.0, + "grad_norm": 2.827688531231292, + "learning_rate": 3.988287528218854e-06, + "loss": 0.7659, + "num_input_tokens_seen": 11283015, + "step": 528 + }, + { + "epoch": 0.06360848914807912, + "flos": 15459104044800.0, + "grad_norm": 2.716706132504615, + "learning_rate": 3.98820319736314e-06, + "loss": 0.9047, + "num_input_tokens_seen": 11299630, + "step": 529 + }, + { + "epoch": 0.0637287320387182, + "flos": 20563946682720.0, + "grad_norm": 3.159430082088147, + "learning_rate": 3.988118564898582e-06, + "loss": 0.8553, + "num_input_tokens_seen": 11319170, + "step": 530 + }, + { + "epoch": 0.0638489749293573, + "flos": 17387180358240.0, + "grad_norm": 3.142224108740231, + "learning_rate": 3.988033630838019e-06, + "loss": 0.8942, + "num_input_tokens_seen": 11333210, + "step": 531 + }, + { + "epoch": 0.0639692178199964, + "flos": 23773688317920.0, + "grad_norm": 9.429164413696865, + "learning_rate": 3.987948395194334e-06, + "loss": 0.8823, + "num_input_tokens_seen": 11355590, + "step": 532 + }, + { + "epoch": 0.06408946071063548, + "flos": 18450884975520.0, + "grad_norm": 3.2733531567762513, + "learning_rate": 3.987862857980458e-06, + "loss": 0.7696, + "num_input_tokens_seen": 11371535, + "step": 533 + }, + { + "epoch": 0.06420970360127458, + "flos": 27123285524640.0, + "grad_norm": 2.3596727802240345, + "learning_rate": 3.987777019209368e-06, + "loss": 0.7686, + "num_input_tokens_seen": 11392050, + "step": 534 + }, + { + "epoch": 0.06432994649191366, + "flos": 23625841248000.0, + "grad_norm": 2.1079303456838034, + "learning_rate": 3.987690878894084e-06, + "loss": 0.8082, + "num_input_tokens_seen": 11411765, + "step": 535 + }, + { + "epoch": 0.06445018938255276, + "flos": 23370482031360.0, + "grad_norm": 3.159336809027486, + "learning_rate": 3.987604437047673e-06, + "loss": 0.8547, + "num_input_tokens_seen": 11431565, + "step": 536 + }, + { + "epoch": 0.06457043227319184, + "flos": 19618719291360.0, + "grad_norm": 2.606968941085401, + "learning_rate": 3.987517693683251e-06, + "loss": 0.7775, + "num_input_tokens_seen": 11450140, + "step": 537 + }, + { + "epoch": 0.06469067516383094, + "flos": 16954492718400.0, + "grad_norm": 3.0506329176787714, + "learning_rate": 3.9874306488139745e-06, + "loss": 0.9611, + "num_input_tokens_seen": 11465400, + "step": 538 + }, + { + "epoch": 0.06481091805447003, + "flos": 23263155733440.0, + "grad_norm": 2.5547291230243587, + "learning_rate": 3.987343302453049e-06, + "loss": 0.8813, + "num_input_tokens_seen": 11486675, + "step": 539 + }, + { + "epoch": 0.06493116094510912, + "flos": 29130619233600.0, + "grad_norm": 1.7290345873045625, + "learning_rate": 3.987255654613724e-06, + "loss": 0.8272, + "num_input_tokens_seen": 11510440, + "step": 540 + }, + { + "epoch": 0.06505140383574821, + "flos": 19836754335360.0, + "grad_norm": 3.1884847877478135, + "learning_rate": 3.987167705309296e-06, + "loss": 0.7041, + "num_input_tokens_seen": 11529715, + "step": 541 + }, + { + "epoch": 0.0651716467263873, + "flos": 17898976714560.0, + "grad_norm": 2.202638612990346, + "learning_rate": 3.987079454553108e-06, + "loss": 0.9497, + "num_input_tokens_seen": 11547665, + "step": 542 + }, + { + "epoch": 0.0652918896170264, + "flos": 20812838361120.0, + "grad_norm": 3.2243136126476712, + "learning_rate": 3.986990902358546e-06, + "loss": 0.9115, + "num_input_tokens_seen": 11564605, + "step": 543 + }, + { + "epoch": 0.06541213250766549, + "flos": 21841300312320.0, + "grad_norm": 2.3850515383224113, + "learning_rate": 3.986902048739045e-06, + "loss": 0.9319, + "num_input_tokens_seen": 11584550, + "step": 544 + }, + { + "epoch": 0.06553237539830457, + "flos": 23077649963040.0, + "grad_norm": 3.654839137512587, + "learning_rate": 3.986812893708082e-06, + "loss": 0.8013, + "num_input_tokens_seen": 11600740, + "step": 545 + }, + { + "epoch": 0.06565261828894367, + "flos": 17898010300800.0, + "grad_norm": 2.552181029233907, + "learning_rate": 3.9867234372791826e-06, + "loss": 0.8125, + "num_input_tokens_seen": 11618695, + "step": 546 + }, + { + "epoch": 0.06577286117958275, + "flos": 22750467302880.0, + "grad_norm": 1.9552308593601333, + "learning_rate": 3.986633679465918e-06, + "loss": 0.8727, + "num_input_tokens_seen": 11638850, + "step": 547 + }, + { + "epoch": 0.06589310407022185, + "flos": 23662310516160.0, + "grad_norm": 2.700217739791581, + "learning_rate": 3.986543620281904e-06, + "loss": 0.8084, + "num_input_tokens_seen": 11658500, + "step": 548 + }, + { + "epoch": 0.06601334696086093, + "flos": 26826067424640.0, + "grad_norm": 1.985091553121337, + "learning_rate": 3.986453259740802e-06, + "loss": 0.9109, + "num_input_tokens_seen": 11676670, + "step": 549 + }, + { + "epoch": 0.06613358985150003, + "flos": 12550000127520.0, + "grad_norm": 2.531252825700101, + "learning_rate": 3.986362597856319e-06, + "loss": 0.788, + "num_input_tokens_seen": 11693170, + "step": 550 + }, + { + "epoch": 0.06625383274213913, + "flos": 18306420353760.0, + "grad_norm": 2.905509772425261, + "learning_rate": 3.986271634642211e-06, + "loss": 0.822, + "num_input_tokens_seen": 11710150, + "step": 551 + }, + { + "epoch": 0.06637407563277821, + "flos": 15353747744160.0, + "grad_norm": 2.8457775224060033, + "learning_rate": 3.986180370112274e-06, + "loss": 0.8173, + "num_input_tokens_seen": 11726670, + "step": 552 + }, + { + "epoch": 0.0664943185234173, + "flos": 23990905627200.0, + "grad_norm": 2.078262095123965, + "learning_rate": 3.986088804280354e-06, + "loss": 0.7453, + "num_input_tokens_seen": 11746400, + "step": 553 + }, + { + "epoch": 0.06661456141405639, + "flos": 20927524271520.0, + "grad_norm": 2.2219816501522422, + "learning_rate": 3.985996937160342e-06, + "loss": 0.9438, + "num_input_tokens_seen": 11765590, + "step": 554 + }, + { + "epoch": 0.06673480430469549, + "flos": 52148382979680.0, + "grad_norm": 2.299342484406293, + "learning_rate": 3.985904768766173e-06, + "loss": 0.6928, + "num_input_tokens_seen": 11788365, + "step": 555 + }, + { + "epoch": 0.06685504719533458, + "flos": 16194102041760.0, + "grad_norm": 4.28501715212653, + "learning_rate": 3.98581229911183e-06, + "loss": 0.7568, + "num_input_tokens_seen": 11807285, + "step": 556 + }, + { + "epoch": 0.06697529008597367, + "flos": 22459382213280.0, + "grad_norm": 2.062182430903342, + "learning_rate": 3.985719528211341e-06, + "loss": 0.9201, + "num_input_tokens_seen": 11826655, + "step": 557 + }, + { + "epoch": 0.06709553297661276, + "flos": 62598479495040.0, + "grad_norm": 0.8465701057596541, + "learning_rate": 3.985626456078777e-06, + "loss": 0.6296, + "num_input_tokens_seen": 11891310, + "step": 558 + }, + { + "epoch": 0.06721577586725185, + "flos": 11200823054400.0, + "grad_norm": 2.1903293295807105, + "learning_rate": 3.985533082728259e-06, + "loss": 0.865, + "num_input_tokens_seen": 11906965, + "step": 559 + }, + { + "epoch": 0.06733601875789094, + "flos": 25885077385920.0, + "grad_norm": 3.3520527372194766, + "learning_rate": 3.985439408173951e-06, + "loss": 0.7481, + "num_input_tokens_seen": 11927190, + "step": 560 + }, + { + "epoch": 0.06745626164853002, + "flos": 20783951724000.0, + "grad_norm": 2.2805626238600984, + "learning_rate": 3.9853454324300634e-06, + "loss": 0.7081, + "num_input_tokens_seen": 11946400, + "step": 561 + }, + { + "epoch": 0.06757650453916912, + "flos": 19801028462400.0, + "grad_norm": 2.313380614996704, + "learning_rate": 3.985251155510852e-06, + "loss": 0.7825, + "num_input_tokens_seen": 11965430, + "step": 562 + }, + { + "epoch": 0.06769674742980822, + "flos": 25702359347520.0, + "grad_norm": 2.1957423105019447, + "learning_rate": 3.98515657743062e-06, + "loss": 0.8016, + "num_input_tokens_seen": 11984895, + "step": 563 + }, + { + "epoch": 0.0678169903204473, + "flos": 13056221019840.0, + "grad_norm": 2.167912552938591, + "learning_rate": 3.985061698203711e-06, + "loss": 0.781, + "num_input_tokens_seen": 12002010, + "step": 564 + }, + { + "epoch": 0.0679372332110864, + "flos": 70764361793760.0, + "grad_norm": 0.8865889399622804, + "learning_rate": 3.984966517844523e-06, + "loss": 0.6379, + "num_input_tokens_seen": 12057255, + "step": 565 + }, + { + "epoch": 0.06805747610172548, + "flos": 28215876779040.0, + "grad_norm": 3.5720010516601928, + "learning_rate": 3.984871036367492e-06, + "loss": 0.8074, + "num_input_tokens_seen": 12077280, + "step": 566 + }, + { + "epoch": 0.06817771899236458, + "flos": 20091853363680.0, + "grad_norm": 2.084298851074874, + "learning_rate": 3.984775253787102e-06, + "loss": 0.8358, + "num_input_tokens_seen": 12095570, + "step": 567 + }, + { + "epoch": 0.06829796188300366, + "flos": 17905109724960.0, + "grad_norm": 3.090328135181321, + "learning_rate": 3.984679170117885e-06, + "loss": 0.8812, + "num_input_tokens_seen": 12111775, + "step": 568 + }, + { + "epoch": 0.06841820477364276, + "flos": 14481124361280.0, + "grad_norm": 2.6645909118146895, + "learning_rate": 3.984582785374415e-06, + "loss": 0.7836, + "num_input_tokens_seen": 12129215, + "step": 569 + }, + { + "epoch": 0.06853844766428185, + "flos": 21906916405920.0, + "grad_norm": 2.4409332549625122, + "learning_rate": 3.9844860995713155e-06, + "loss": 0.8062, + "num_input_tokens_seen": 12148755, + "step": 570 + }, + { + "epoch": 0.06865869055492094, + "flos": 16776792597600.0, + "grad_norm": 3.3117402842965094, + "learning_rate": 3.9843891127232524e-06, + "loss": 0.8253, + "num_input_tokens_seen": 12166290, + "step": 571 + }, + { + "epoch": 0.06877893344556003, + "flos": 19908837967200.0, + "grad_norm": 2.619597734335157, + "learning_rate": 3.984291824844938e-06, + "loss": 0.6698, + "num_input_tokens_seen": 12181385, + "step": 572 + }, + { + "epoch": 0.06889917633619912, + "flos": 23006458405440.0, + "grad_norm": 2.9499889050291226, + "learning_rate": 3.984194235951132e-06, + "loss": 0.8498, + "num_input_tokens_seen": 12197530, + "step": 573 + }, + { + "epoch": 0.06901941922683821, + "flos": 20930683701120.0, + "grad_norm": 3.6393025497203277, + "learning_rate": 3.9840963460566375e-06, + "loss": 0.8465, + "num_input_tokens_seen": 12216310, + "step": 574 + }, + { + "epoch": 0.06913966211747731, + "flos": 24786130102560.0, + "grad_norm": 2.6612429198837884, + "learning_rate": 3.983998155176305e-06, + "loss": 0.8979, + "num_input_tokens_seen": 12236670, + "step": 575 + }, + { + "epoch": 0.06925990500811639, + "flos": 58284766925280.0, + "grad_norm": 0.8138615131294794, + "learning_rate": 3.9838996633250305e-06, + "loss": 0.5705, + "num_input_tokens_seen": 12297185, + "step": 576 + }, + { + "epoch": 0.06938014789875549, + "flos": 12731974770720.0, + "grad_norm": 2.8757794194421638, + "learning_rate": 3.983800870517753e-06, + "loss": 0.8868, + "num_input_tokens_seen": 12313975, + "step": 577 + }, + { + "epoch": 0.06950039078939457, + "flos": 22787605626720.0, + "grad_norm": 5.928547424162758, + "learning_rate": 3.983701776769463e-06, + "loss": 0.7832, + "num_input_tokens_seen": 12331385, + "step": 578 + }, + { + "epoch": 0.06962063368003367, + "flos": 21909889986720.0, + "grad_norm": 6.54783226425661, + "learning_rate": 3.9836023820951885e-06, + "loss": 0.8619, + "num_input_tokens_seen": 12350600, + "step": 579 + }, + { + "epoch": 0.06974087657067275, + "flos": 20676551086560.0, + "grad_norm": 2.3695719075574493, + "learning_rate": 3.983502686510011e-06, + "loss": 0.6883, + "num_input_tokens_seen": 12371430, + "step": 580 + }, + { + "epoch": 0.06986111946131185, + "flos": 22605891171840.0, + "grad_norm": 2.3778148834902404, + "learning_rate": 3.9834026900290525e-06, + "loss": 0.7368, + "num_input_tokens_seen": 12390825, + "step": 581 + }, + { + "epoch": 0.06998136235195095, + "flos": 26904841613280.0, + "grad_norm": 2.186865360296546, + "learning_rate": 3.983302392667482e-06, + "loss": 1.0031, + "num_input_tokens_seen": 12411670, + "step": 582 + }, + { + "epoch": 0.07010160524259003, + "flos": 22460385796800.0, + "grad_norm": 2.267947565256402, + "learning_rate": 3.983201794440517e-06, + "loss": 0.9376, + "num_input_tokens_seen": 12431245, + "step": 583 + }, + { + "epoch": 0.07022184813322913, + "flos": 18306457523520.0, + "grad_norm": 1.8695131765688366, + "learning_rate": 3.9831008953634165e-06, + "loss": 0.6721, + "num_input_tokens_seen": 12450535, + "step": 584 + }, + { + "epoch": 0.07034209102386821, + "flos": 24640030011360.0, + "grad_norm": 2.348841515545254, + "learning_rate": 3.9829996954514864e-06, + "loss": 0.8123, + "num_input_tokens_seen": 12469675, + "step": 585 + }, + { + "epoch": 0.0704623339145073, + "flos": 25958684977920.0, + "grad_norm": 3.578626839040442, + "learning_rate": 3.982898194720079e-06, + "loss": 0.8412, + "num_input_tokens_seen": 12490405, + "step": 586 + }, + { + "epoch": 0.0705825768051464, + "flos": 25302052302240.0, + "grad_norm": 2.367575736360111, + "learning_rate": 3.982796393184592e-06, + "loss": 0.8247, + "num_input_tokens_seen": 12509485, + "step": 587 + }, + { + "epoch": 0.07070281969578548, + "flos": 65953020279840.0, + "grad_norm": 0.7933119608968094, + "learning_rate": 3.98269429086047e-06, + "loss": 0.626, + "num_input_tokens_seen": 12567325, + "step": 588 + }, + { + "epoch": 0.07082306258642458, + "flos": 23619522388800.0, + "grad_norm": 4.398555015498299, + "learning_rate": 3.982591887763199e-06, + "loss": 0.8618, + "num_input_tokens_seen": 12584785, + "step": 589 + }, + { + "epoch": 0.07094330547706366, + "flos": 13860403407360.0, + "grad_norm": 2.7375436788061633, + "learning_rate": 3.982489183908316e-06, + "loss": 0.8219, + "num_input_tokens_seen": 12601205, + "step": 590 + }, + { + "epoch": 0.07106354836770276, + "flos": 24609842432640.0, + "grad_norm": 2.1694362862737906, + "learning_rate": 3.982386179311399e-06, + "loss": 0.8482, + "num_input_tokens_seen": 12621725, + "step": 591 + }, + { + "epoch": 0.07118379125834184, + "flos": 16194250720800.0, + "grad_norm": 4.890682719807155, + "learning_rate": 3.982282873988075e-06, + "loss": 0.8771, + "num_input_tokens_seen": 12639600, + "step": 592 + }, + { + "epoch": 0.07130403414898094, + "flos": 19691918016000.0, + "grad_norm": 1.8187680220612275, + "learning_rate": 3.982179267954016e-06, + "loss": 0.8686, + "num_input_tokens_seen": 12660030, + "step": 593 + }, + { + "epoch": 0.07142427703962004, + "flos": 21840705596160.0, + "grad_norm": 62.16780859722538, + "learning_rate": 3.982075361224937e-06, + "loss": 0.9591, + "num_input_tokens_seen": 12678395, + "step": 594 + }, + { + "epoch": 0.07154451993025912, + "flos": 18270582971520.0, + "grad_norm": 2.175528328404489, + "learning_rate": 3.981971153816602e-06, + "loss": 0.8819, + "num_input_tokens_seen": 12696000, + "step": 595 + }, + { + "epoch": 0.07166476282089822, + "flos": 22128928614240.0, + "grad_norm": 1.9078330049029888, + "learning_rate": 3.981866645744819e-06, + "loss": 0.964, + "num_input_tokens_seen": 12716715, + "step": 596 + }, + { + "epoch": 0.0717850057115373, + "flos": 14116357340160.0, + "grad_norm": 2.89767898546537, + "learning_rate": 3.9817618370254416e-06, + "loss": 0.8115, + "num_input_tokens_seen": 12733370, + "step": 597 + }, + { + "epoch": 0.0719052486021764, + "flos": 30873301285920.0, + "grad_norm": 4.082282625334198, + "learning_rate": 3.9816567276743684e-06, + "loss": 0.8747, + "num_input_tokens_seen": 12753235, + "step": 598 + }, + { + "epoch": 0.0720254914928155, + "flos": 21256602589440.0, + "grad_norm": 2.493180606247745, + "learning_rate": 3.9815513177075466e-06, + "loss": 0.7749, + "num_input_tokens_seen": 12772020, + "step": 599 + }, + { + "epoch": 0.07214573438345458, + "flos": 27779286314400.0, + "grad_norm": 2.4167054162262183, + "learning_rate": 3.9814456071409646e-06, + "loss": 0.7037, + "num_input_tokens_seen": 12792555, + "step": 600 + }, + { + "epoch": 0.07226597727409367, + "flos": 25447074470400.0, + "grad_norm": 2.9589260385132308, + "learning_rate": 3.981339595990659e-06, + "loss": 0.8526, + "num_input_tokens_seen": 12811085, + "step": 601 + }, + { + "epoch": 0.07238622016473276, + "flos": 23480224363680.0, + "grad_norm": 2.822396731383838, + "learning_rate": 3.981233284272713e-06, + "loss": 0.8143, + "num_input_tokens_seen": 12830830, + "step": 602 + }, + { + "epoch": 0.07250646305537185, + "flos": 25417295759040.0, + "grad_norm": 1.7758552102875036, + "learning_rate": 3.981126672003253e-06, + "loss": 0.9008, + "num_input_tokens_seen": 12853505, + "step": 603 + }, + { + "epoch": 0.07262670594601094, + "flos": 27115851572640.0, + "grad_norm": 2.8149948922309136, + "learning_rate": 3.981019759198451e-06, + "loss": 0.7809, + "num_input_tokens_seen": 12872455, + "step": 604 + }, + { + "epoch": 0.07274694883665003, + "flos": 26613421995840.0, + "grad_norm": 2.486251314449586, + "learning_rate": 3.980912545874528e-06, + "loss": 0.8416, + "num_input_tokens_seen": 12892620, + "step": 605 + }, + { + "epoch": 0.07286719172728913, + "flos": 29819483824800.0, + "grad_norm": 3.9903691698965083, + "learning_rate": 3.980805032047746e-06, + "loss": 0.8557, + "num_input_tokens_seen": 12914090, + "step": 606 + }, + { + "epoch": 0.07298743461792821, + "flos": 17357290137600.0, + "grad_norm": 6.017530897351998, + "learning_rate": 3.980697217734415e-06, + "loss": 0.8112, + "num_input_tokens_seen": 12931830, + "step": 607 + }, + { + "epoch": 0.07310767750856731, + "flos": 19470351844800.0, + "grad_norm": 2.7675102403239364, + "learning_rate": 3.980589102950891e-06, + "loss": 0.9162, + "num_input_tokens_seen": 12949995, + "step": 608 + }, + { + "epoch": 0.07322792039920639, + "flos": 29126827918080.0, + "grad_norm": 2.703551330462907, + "learning_rate": 3.9804806877135755e-06, + "loss": 0.7589, + "num_input_tokens_seen": 12968040, + "step": 609 + }, + { + "epoch": 0.07334816328984549, + "flos": 23444275472160.0, + "grad_norm": 2.398524080882042, + "learning_rate": 3.980371972038915e-06, + "loss": 0.8606, + "num_input_tokens_seen": 12988530, + "step": 610 + }, + { + "epoch": 0.07346840618048459, + "flos": 22929505535040.0, + "grad_norm": 2.2013132207349484, + "learning_rate": 3.980262955943399e-06, + "loss": 0.8413, + "num_input_tokens_seen": 13008195, + "step": 611 + }, + { + "epoch": 0.07358864907112367, + "flos": 17648709755040.0, + "grad_norm": 5.283206554700827, + "learning_rate": 3.980153639443569e-06, + "loss": 0.8664, + "num_input_tokens_seen": 13024180, + "step": 612 + }, + { + "epoch": 0.07370889196176277, + "flos": 24062543221920.0, + "grad_norm": 2.1180809922766763, + "learning_rate": 3.980044022556005e-06, + "loss": 0.8015, + "num_input_tokens_seen": 13043865, + "step": 613 + }, + { + "epoch": 0.07382913485240185, + "flos": 25848868306080.0, + "grad_norm": 2.472632433639498, + "learning_rate": 3.9799341052973375e-06, + "loss": 0.7288, + "num_input_tokens_seen": 13063700, + "step": 614 + }, + { + "epoch": 0.07394937774304094, + "flos": 16849916982720.0, + "grad_norm": 3.1049469515435226, + "learning_rate": 3.979823887684241e-06, + "loss": 0.7539, + "num_input_tokens_seen": 13082640, + "step": 615 + }, + { + "epoch": 0.07406962063368003, + "flos": 20673912033600.0, + "grad_norm": 2.835564875226997, + "learning_rate": 3.979713369733434e-06, + "loss": 0.8506, + "num_input_tokens_seen": 13100505, + "step": 616 + }, + { + "epoch": 0.07418986352431912, + "flos": 21400286646240.0, + "grad_norm": 2.3386382243747827, + "learning_rate": 3.979602551461683e-06, + "loss": 0.8457, + "num_input_tokens_seen": 13121285, + "step": 617 + }, + { + "epoch": 0.07431010641495822, + "flos": 12004262046720.0, + "grad_norm": 3.9099590545227496, + "learning_rate": 3.979491432885799e-06, + "loss": 0.9168, + "num_input_tokens_seen": 13137550, + "step": 618 + }, + { + "epoch": 0.0744303493055973, + "flos": 20927598611040.0, + "grad_norm": 2.3743435805838353, + "learning_rate": 3.97938001402264e-06, + "loss": 0.8289, + "num_input_tokens_seen": 13156675, + "step": 619 + }, + { + "epoch": 0.0745505921962364, + "flos": 16230088103040.0, + "grad_norm": 3.5455833469461133, + "learning_rate": 3.979268294889105e-06, + "loss": 0.7946, + "num_input_tokens_seen": 13172225, + "step": 620 + }, + { + "epoch": 0.07467083508687548, + "flos": 50870843501280.0, + "grad_norm": 2.8677434148465926, + "learning_rate": 3.979156275502143e-06, + "loss": 0.7409, + "num_input_tokens_seen": 13195885, + "step": 621 + }, + { + "epoch": 0.07479107797751458, + "flos": 17504765509920.0, + "grad_norm": 3.252921038856335, + "learning_rate": 3.979043955878749e-06, + "loss": 0.9147, + "num_input_tokens_seen": 13213730, + "step": 622 + }, + { + "epoch": 0.07491132086815366, + "flos": 23440261138080.0, + "grad_norm": 2.376603538866112, + "learning_rate": 3.978931336035959e-06, + "loss": 0.8321, + "num_input_tokens_seen": 13232540, + "step": 623 + }, + { + "epoch": 0.07503156375879276, + "flos": 20128843008480.0, + "grad_norm": 2.85838726055852, + "learning_rate": 3.9788184159908595e-06, + "loss": 0.8228, + "num_input_tokens_seen": 13249670, + "step": 624 + }, + { + "epoch": 0.07515180664943186, + "flos": 15093890986560.0, + "grad_norm": 3.3647059525545115, + "learning_rate": 3.97870519576058e-06, + "loss": 0.8231, + "num_input_tokens_seen": 13266095, + "step": 625 + }, + { + "epoch": 0.07527204954007094, + "flos": 21257308814880.0, + "grad_norm": 2.8534093441449837, + "learning_rate": 3.978591675362295e-06, + "loss": 0.8132, + "num_input_tokens_seen": 13284810, + "step": 626 + }, + { + "epoch": 0.07539229243071004, + "flos": 21293740913280.0, + "grad_norm": 2.2044809916544796, + "learning_rate": 3.978477854813226e-06, + "loss": 0.8767, + "num_input_tokens_seen": 13304150, + "step": 627 + }, + { + "epoch": 0.07551253532134912, + "flos": 13024174953120.0, + "grad_norm": 4.176626635514627, + "learning_rate": 3.97836373413064e-06, + "loss": 0.8247, + "num_input_tokens_seen": 13322205, + "step": 628 + }, + { + "epoch": 0.07563277821198822, + "flos": 19181422601280.0, + "grad_norm": 3.0843177117590064, + "learning_rate": 3.978249313331848e-06, + "loss": 0.7437, + "num_input_tokens_seen": 13340435, + "step": 629 + }, + { + "epoch": 0.07575302110262731, + "flos": 19509385826400.0, + "grad_norm": 7.185813374203157, + "learning_rate": 3.978134592434208e-06, + "loss": 0.6204, + "num_input_tokens_seen": 13358185, + "step": 630 + }, + { + "epoch": 0.0758732639932664, + "flos": 67865107862880.0, + "grad_norm": 1.020617845069862, + "learning_rate": 3.978019571455123e-06, + "loss": 0.5938, + "num_input_tokens_seen": 13410130, + "step": 631 + }, + { + "epoch": 0.07599350688390549, + "flos": 18962160955200.0, + "grad_norm": 2.7326991586099054, + "learning_rate": 3.977904250412042e-06, + "loss": 0.8408, + "num_input_tokens_seen": 13429085, + "step": 632 + }, + { + "epoch": 0.07611374977454458, + "flos": 21038939243040.0, + "grad_norm": 2.2921962704064915, + "learning_rate": 3.97778862932246e-06, + "loss": 0.86, + "num_input_tokens_seen": 13446250, + "step": 633 + }, + { + "epoch": 0.07623399266518367, + "flos": 18488171978400.0, + "grad_norm": 2.5828320621919683, + "learning_rate": 3.9776727082039144e-06, + "loss": 0.9395, + "num_input_tokens_seen": 13463220, + "step": 634 + }, + { + "epoch": 0.07635423555582276, + "flos": 44599498925280.0, + "grad_norm": 0.7986601114213681, + "learning_rate": 3.977556487073991e-06, + "loss": 0.5549, + "num_input_tokens_seen": 13517775, + "step": 635 + }, + { + "epoch": 0.07647447844646185, + "flos": 21730034019840.0, + "grad_norm": 2.0397232563571985, + "learning_rate": 3.97743996595032e-06, + "loss": 0.8198, + "num_input_tokens_seen": 13537815, + "step": 636 + }, + { + "epoch": 0.07659472133710095, + "flos": 23772796243680.0, + "grad_norm": 1.7755216208161808, + "learning_rate": 3.9773231448505804e-06, + "loss": 0.8165, + "num_input_tokens_seen": 13559605, + "step": 637 + }, + { + "epoch": 0.07671496422774003, + "flos": 21439320627840.0, + "grad_norm": 2.1714630834088386, + "learning_rate": 3.977206023792491e-06, + "loss": 0.7801, + "num_input_tokens_seen": 13579495, + "step": 638 + }, + { + "epoch": 0.07683520711837913, + "flos": 16956797243520.0, + "grad_norm": 2.9709509470924114, + "learning_rate": 3.97708860279382e-06, + "loss": 0.8162, + "num_input_tokens_seen": 13597605, + "step": 639 + }, + { + "epoch": 0.07695545000901821, + "flos": 23443160379360.0, + "grad_norm": 2.156105036631007, + "learning_rate": 3.97697088187238e-06, + "loss": 0.7804, + "num_input_tokens_seen": 13618120, + "step": 640 + }, + { + "epoch": 0.07707569289965731, + "flos": 17609006717760.0, + "grad_norm": 2.4979558216766646, + "learning_rate": 3.976852861046029e-06, + "loss": 0.9195, + "num_input_tokens_seen": 13634735, + "step": 641 + }, + { + "epoch": 0.0771959357902964, + "flos": 25738754276160.0, + "grad_norm": 1.876835306302349, + "learning_rate": 3.97673454033267e-06, + "loss": 0.8014, + "num_input_tokens_seen": 13656835, + "step": 642 + }, + { + "epoch": 0.07731617868093549, + "flos": 19800396576480.0, + "grad_norm": 2.149333531189587, + "learning_rate": 3.976615919750254e-06, + "loss": 0.8227, + "num_input_tokens_seen": 13674535, + "step": 643 + }, + { + "epoch": 0.07743642157157458, + "flos": 21294224120160.0, + "grad_norm": 2.3078532572397026, + "learning_rate": 3.976496999316775e-06, + "loss": 0.8724, + "num_input_tokens_seen": 13693290, + "step": 644 + }, + { + "epoch": 0.07755666446221367, + "flos": 19940475166560.0, + "grad_norm": 2.407307281069349, + "learning_rate": 3.976377779050271e-06, + "loss": 0.8398, + "num_input_tokens_seen": 13711420, + "step": 645 + }, + { + "epoch": 0.07767690735285276, + "flos": 23589446319360.0, + "grad_norm": 2.9586630265894134, + "learning_rate": 3.976258258968831e-06, + "loss": 0.8441, + "num_input_tokens_seen": 13729965, + "step": 646 + }, + { + "epoch": 0.07779715024349185, + "flos": 22204469033760.0, + "grad_norm": 3.0335030588587624, + "learning_rate": 3.976138439090583e-06, + "loss": 0.7441, + "num_input_tokens_seen": 13748045, + "step": 647 + }, + { + "epoch": 0.07791739313413094, + "flos": 20925033897600.0, + "grad_norm": 2.547966658776814, + "learning_rate": 3.976018319433706e-06, + "loss": 0.8552, + "num_input_tokens_seen": 13765590, + "step": 648 + }, + { + "epoch": 0.07803763602477004, + "flos": 19284623055840.0, + "grad_norm": 2.5636543837134718, + "learning_rate": 3.9758979000164205e-06, + "loss": 0.9189, + "num_input_tokens_seen": 13782410, + "step": 649 + }, + { + "epoch": 0.07815787891540912, + "flos": 22678346501280.0, + "grad_norm": 2.069105264175383, + "learning_rate": 3.975777180856995e-06, + "loss": 0.7218, + "num_input_tokens_seen": 13801530, + "step": 650 + }, + { + "epoch": 0.07827812180604822, + "flos": 22678681029120.0, + "grad_norm": 2.616878388782801, + "learning_rate": 3.975656161973742e-06, + "loss": 0.8655, + "num_input_tokens_seen": 13820615, + "step": 651 + }, + { + "epoch": 0.0783983646966873, + "flos": 21694568335200.0, + "grad_norm": 3.2372970281772138, + "learning_rate": 3.9755348433850194e-06, + "loss": 0.8924, + "num_input_tokens_seen": 13835955, + "step": 652 + }, + { + "epoch": 0.0785186075873264, + "flos": 60553970292480.0, + "grad_norm": 0.9585729209670221, + "learning_rate": 3.975413225109232e-06, + "loss": 0.6355, + "num_input_tokens_seen": 13896505, + "step": 653 + }, + { + "epoch": 0.0786388504779655, + "flos": 23843727612960.0, + "grad_norm": 9.424819207923873, + "learning_rate": 3.975291307164829e-06, + "loss": 0.9402, + "num_input_tokens_seen": 13915150, + "step": 654 + }, + { + "epoch": 0.07875909336860458, + "flos": 15137236660320.0, + "grad_norm": 2.912660481366119, + "learning_rate": 3.975169089570306e-06, + "loss": 0.8522, + "num_input_tokens_seen": 13933525, + "step": 655 + }, + { + "epoch": 0.07887933625924368, + "flos": 22205100919680.0, + "grad_norm": 2.1001911439644068, + "learning_rate": 3.975046572344202e-06, + "loss": 0.9166, + "num_input_tokens_seen": 13949985, + "step": 656 + }, + { + "epoch": 0.07899957914988276, + "flos": 20748114341760.0, + "grad_norm": 2.1595373112007157, + "learning_rate": 3.974923755505103e-06, + "loss": 0.7112, + "num_input_tokens_seen": 13969215, + "step": 657 + }, + { + "epoch": 0.07911982204052186, + "flos": 22970174986080.0, + "grad_norm": 1.8839500441518782, + "learning_rate": 3.974800639071641e-06, + "loss": 0.9134, + "num_input_tokens_seen": 13989935, + "step": 658 + }, + { + "epoch": 0.07924006493116094, + "flos": 23078319018720.0, + "grad_norm": 2.558914560487465, + "learning_rate": 3.974677223062492e-06, + "loss": 1.0087, + "num_input_tokens_seen": 14007150, + "step": 659 + }, + { + "epoch": 0.07936030782180004, + "flos": 16448792202720.0, + "grad_norm": 3.5553844397482948, + "learning_rate": 3.974553507496378e-06, + "loss": 0.7443, + "num_input_tokens_seen": 14025725, + "step": 660 + }, + { + "epoch": 0.07948055071243913, + "flos": 23699374500480.0, + "grad_norm": 2.2140931164355013, + "learning_rate": 3.974429492392068e-06, + "loss": 0.8946, + "num_input_tokens_seen": 14045750, + "step": 661 + }, + { + "epoch": 0.07960079360307822, + "flos": 19545446227200.0, + "grad_norm": 2.3237594535352164, + "learning_rate": 3.974305177768373e-06, + "loss": 0.9119, + "num_input_tokens_seen": 14063960, + "step": 662 + }, + { + "epoch": 0.07972103649371731, + "flos": 23479778326560.0, + "grad_norm": 2.5142328426294944, + "learning_rate": 3.974180563644152e-06, + "loss": 0.8677, + "num_input_tokens_seen": 14082525, + "step": 663 + }, + { + "epoch": 0.0798412793843564, + "flos": 16703110666080.0, + "grad_norm": 3.8736514544438734, + "learning_rate": 3.97405565003831e-06, + "loss": 0.8954, + "num_input_tokens_seen": 14098610, + "step": 664 + }, + { + "epoch": 0.07996152227499549, + "flos": 18197421416640.0, + "grad_norm": 2.281042429220423, + "learning_rate": 3.973930436969794e-06, + "loss": 0.7835, + "num_input_tokens_seen": 14117065, + "step": 665 + }, + { + "epoch": 0.08008176516563459, + "flos": 20565507812640.0, + "grad_norm": 2.068775225252175, + "learning_rate": 3.973804924457602e-06, + "loss": 0.8568, + "num_input_tokens_seen": 14135145, + "step": 666 + }, + { + "epoch": 0.08020200805627367, + "flos": 31788638456640.0, + "grad_norm": 1.9728110073306926, + "learning_rate": 3.973679112520771e-06, + "loss": 0.8615, + "num_input_tokens_seen": 14156860, + "step": 667 + }, + { + "epoch": 0.08032225094691277, + "flos": 17758154729280.0, + "grad_norm": 2.5223958135890125, + "learning_rate": 3.973553001178389e-06, + "loss": 0.9924, + "num_input_tokens_seen": 14174365, + "step": 668 + }, + { + "epoch": 0.08044249383755185, + "flos": 24027300555840.0, + "grad_norm": 2.7320389714512348, + "learning_rate": 3.973426590449585e-06, + "loss": 0.7569, + "num_input_tokens_seen": 14192320, + "step": 669 + }, + { + "epoch": 0.08056273672819095, + "flos": 18197718774720.0, + "grad_norm": 2.228065492068081, + "learning_rate": 3.9732998803535364e-06, + "loss": 0.7496, + "num_input_tokens_seen": 14210380, + "step": 670 + }, + { + "epoch": 0.08068297961883003, + "flos": 19648609512000.0, + "grad_norm": 2.661987570581, + "learning_rate": 3.973172870909465e-06, + "loss": 0.8517, + "num_input_tokens_seen": 14225145, + "step": 671 + }, + { + "epoch": 0.08080322250946913, + "flos": 23115643191360.0, + "grad_norm": 2.9293204522675143, + "learning_rate": 3.973045562136638e-06, + "loss": 0.8113, + "num_input_tokens_seen": 14241960, + "step": 672 + }, + { + "epoch": 0.08092346540010822, + "flos": 21732598733280.0, + "grad_norm": 2.1329202589822147, + "learning_rate": 3.972917954054368e-06, + "loss": 0.9182, + "num_input_tokens_seen": 14260515, + "step": 673 + }, + { + "epoch": 0.08104370829074731, + "flos": 21002469974880.0, + "grad_norm": 3.3286636704621704, + "learning_rate": 3.972790046682013e-06, + "loss": 0.8224, + "num_input_tokens_seen": 14279190, + "step": 674 + }, + { + "epoch": 0.0811639511813864, + "flos": 20054529191040.0, + "grad_norm": 1.9362015834244308, + "learning_rate": 3.972661840038977e-06, + "loss": 0.7923, + "num_input_tokens_seen": 14299480, + "step": 675 + }, + { + "epoch": 0.08128419407202549, + "flos": 16812815828640.0, + "grad_norm": 2.914285032555423, + "learning_rate": 3.972533334144707e-06, + "loss": 0.8349, + "num_input_tokens_seen": 14316405, + "step": 676 + }, + { + "epoch": 0.08140443696266458, + "flos": 23735286222240.0, + "grad_norm": 2.1005281647026846, + "learning_rate": 3.972404529018699e-06, + "loss": 0.7856, + "num_input_tokens_seen": 14336265, + "step": 677 + }, + { + "epoch": 0.08152467985330367, + "flos": 24355263780960.0, + "grad_norm": 2.231908543958778, + "learning_rate": 3.972275424680493e-06, + "loss": 0.8556, + "num_input_tokens_seen": 14356375, + "step": 678 + }, + { + "epoch": 0.08164492274394276, + "flos": 19290235689600.0, + "grad_norm": 2.085134630316275, + "learning_rate": 3.972146021149673e-06, + "loss": 0.9193, + "num_input_tokens_seen": 14374650, + "step": 679 + }, + { + "epoch": 0.08176516563458186, + "flos": 14809496453760.0, + "grad_norm": 2.1246367873401084, + "learning_rate": 3.972016318445868e-06, + "loss": 0.7848, + "num_input_tokens_seen": 14392650, + "step": 680 + }, + { + "epoch": 0.08188540852522094, + "flos": 22570053789600.0, + "grad_norm": 2.5644745081339084, + "learning_rate": 3.971886316588757e-06, + "loss": 0.9227, + "num_input_tokens_seen": 14413155, + "step": 681 + }, + { + "epoch": 0.08200565141586004, + "flos": 19436001252960.0, + "grad_norm": 2.792406487056037, + "learning_rate": 3.9717560155980595e-06, + "loss": 0.7351, + "num_input_tokens_seen": 14428805, + "step": 682 + }, + { + "epoch": 0.08212589430649912, + "flos": 20565284794080.0, + "grad_norm": 2.253761855447047, + "learning_rate": 3.971625415493542e-06, + "loss": 0.9219, + "num_input_tokens_seen": 14447125, + "step": 683 + }, + { + "epoch": 0.08224613719713822, + "flos": 25916268548160.0, + "grad_norm": 2.764691470076387, + "learning_rate": 3.971494516295017e-06, + "loss": 0.8772, + "num_input_tokens_seen": 14466575, + "step": 684 + }, + { + "epoch": 0.08236638008777732, + "flos": 23734728675840.0, + "grad_norm": 2.1683133787878472, + "learning_rate": 3.971363318022341e-06, + "loss": 0.856, + "num_input_tokens_seen": 14487915, + "step": 685 + }, + { + "epoch": 0.0824866229784164, + "flos": 38743192086720.0, + "grad_norm": 1.8973560156995521, + "learning_rate": 3.971231820695417e-06, + "loss": 0.6858, + "num_input_tokens_seen": 14511530, + "step": 686 + }, + { + "epoch": 0.0826068658690555, + "flos": 23074564872960.0, + "grad_norm": 3.5030888142958534, + "learning_rate": 3.971100024334193e-06, + "loss": 0.8119, + "num_input_tokens_seen": 14529050, + "step": 687 + }, + { + "epoch": 0.08272710875969458, + "flos": 21106004957280.0, + "grad_norm": 2.3057267254429514, + "learning_rate": 3.970967928958663e-06, + "loss": 0.8667, + "num_input_tokens_seen": 14546165, + "step": 688 + }, + { + "epoch": 0.08284735165033368, + "flos": 19035880056480.0, + "grad_norm": 1.779507772130197, + "learning_rate": 3.970835534588865e-06, + "loss": 0.8381, + "num_input_tokens_seen": 14563660, + "step": 689 + }, + { + "epoch": 0.08296759454097276, + "flos": 16703705382240.0, + "grad_norm": 3.4602879588523714, + "learning_rate": 3.970702841244883e-06, + "loss": 0.8586, + "num_input_tokens_seen": 14581980, + "step": 690 + }, + { + "epoch": 0.08308783743161186, + "flos": 17979349202880.0, + "grad_norm": 2.084197500926885, + "learning_rate": 3.970569848946847e-06, + "loss": 0.8277, + "num_input_tokens_seen": 14601795, + "step": 691 + }, + { + "epoch": 0.08320808032225095, + "flos": 15058239453120.0, + "grad_norm": 3.6531780412019637, + "learning_rate": 3.970436557714932e-06, + "loss": 0.8292, + "num_input_tokens_seen": 14618315, + "step": 692 + }, + { + "epoch": 0.08332832321289003, + "flos": 22351126671360.0, + "grad_norm": 2.264447452103091, + "learning_rate": 3.970302967569358e-06, + "loss": 0.8663, + "num_input_tokens_seen": 14636905, + "step": 693 + }, + { + "epoch": 0.08344856610352913, + "flos": 24682000404000.0, + "grad_norm": 2.6270991515233706, + "learning_rate": 3.9701690785303896e-06, + "loss": 0.6861, + "num_input_tokens_seen": 14656100, + "step": 694 + }, + { + "epoch": 0.08356880899416821, + "flos": 25333726671360.0, + "grad_norm": 2.3059538901848073, + "learning_rate": 3.970034890618339e-06, + "loss": 0.8839, + "num_input_tokens_seen": 14675790, + "step": 695 + }, + { + "epoch": 0.08368905188480731, + "flos": 24318236966400.0, + "grad_norm": 2.3055645841654084, + "learning_rate": 3.969900403853562e-06, + "loss": 0.8856, + "num_input_tokens_seen": 14694830, + "step": 696 + }, + { + "epoch": 0.08380929477544641, + "flos": 18011543948640.0, + "grad_norm": 2.202394296662151, + "learning_rate": 3.96976561825646e-06, + "loss": 0.7814, + "num_input_tokens_seen": 14711920, + "step": 697 + }, + { + "epoch": 0.08392953766608549, + "flos": 26248617804960.0, + "grad_norm": 2.4925481361936974, + "learning_rate": 3.969630533847479e-06, + "loss": 0.871, + "num_input_tokens_seen": 14730880, + "step": 698 + }, + { + "epoch": 0.08404978055672459, + "flos": 22460014099200.0, + "grad_norm": 2.4171450788132782, + "learning_rate": 3.969495150647113e-06, + "loss": 0.8419, + "num_input_tokens_seen": 14749050, + "step": 699 + }, + { + "epoch": 0.08417002344736367, + "flos": 24791333868960.0, + "grad_norm": 1.8223086696500375, + "learning_rate": 3.969359468675899e-06, + "loss": 0.7642, + "num_input_tokens_seen": 14769180, + "step": 700 + }, + { + "epoch": 0.08429026633800277, + "flos": 16921851935520.0, + "grad_norm": 3.4781512455665484, + "learning_rate": 3.969223487954418e-06, + "loss": 0.8941, + "num_input_tokens_seen": 14786640, + "step": 701 + }, + { + "epoch": 0.08441050922864185, + "flos": 23807815891200.0, + "grad_norm": 3.2234365177569093, + "learning_rate": 3.969087208503301e-06, + "loss": 0.8295, + "num_input_tokens_seen": 14806720, + "step": 702 + }, + { + "epoch": 0.08453075211928095, + "flos": 25483655247840.0, + "grad_norm": 2.8656249400963563, + "learning_rate": 3.968950630343219e-06, + "loss": 0.8481, + "num_input_tokens_seen": 14823705, + "step": 703 + }, + { + "epoch": 0.08465099500992004, + "flos": 19504516587840.0, + "grad_norm": 2.6310144278211545, + "learning_rate": 3.968813753494892e-06, + "loss": 0.9331, + "num_input_tokens_seen": 14841865, + "step": 704 + }, + { + "epoch": 0.08477123790055913, + "flos": 29309174258880.0, + "grad_norm": 2.640066866605872, + "learning_rate": 3.968676577979084e-06, + "loss": 0.7546, + "num_input_tokens_seen": 14861415, + "step": 705 + }, + { + "epoch": 0.08489148079119822, + "flos": 18597988650240.0, + "grad_norm": 2.249829391792918, + "learning_rate": 3.968539103816605e-06, + "loss": 0.7834, + "num_input_tokens_seen": 14879215, + "step": 706 + }, + { + "epoch": 0.0850117236818373, + "flos": 23437622085120.0, + "grad_norm": 2.123088425665733, + "learning_rate": 3.9684013310283085e-06, + "loss": 0.8961, + "num_input_tokens_seen": 14897900, + "step": 707 + }, + { + "epoch": 0.0851319665724764, + "flos": 40566506815680.0, + "grad_norm": 2.2351514627564733, + "learning_rate": 3.9682632596350956e-06, + "loss": 0.6423, + "num_input_tokens_seen": 14919065, + "step": 708 + }, + { + "epoch": 0.0852522094631155, + "flos": 15856734867360.0, + "grad_norm": 2.1963813324743047, + "learning_rate": 3.968124889657911e-06, + "loss": 0.7825, + "num_input_tokens_seen": 14934165, + "step": 709 + }, + { + "epoch": 0.08537245235375458, + "flos": 14546220078240.0, + "grad_norm": 2.6826652686410575, + "learning_rate": 3.967986221117746e-06, + "loss": 0.9076, + "num_input_tokens_seen": 14950105, + "step": 710 + }, + { + "epoch": 0.08549269524439368, + "flos": 26431001315520.0, + "grad_norm": 2.10820630511358, + "learning_rate": 3.967847254035635e-06, + "loss": 0.8672, + "num_input_tokens_seen": 14969490, + "step": 711 + }, + { + "epoch": 0.08561293813503276, + "flos": 13571065296480.0, + "grad_norm": 3.595607244486528, + "learning_rate": 3.967707988432661e-06, + "loss": 0.8652, + "num_input_tokens_seen": 14987195, + "step": 712 + }, + { + "epoch": 0.08573318102567186, + "flos": 26904915952800.0, + "grad_norm": 2.5902861480532198, + "learning_rate": 3.967568424329949e-06, + "loss": 0.8794, + "num_input_tokens_seen": 15007620, + "step": 713 + }, + { + "epoch": 0.08585342391631094, + "flos": 67206914057280.0, + "grad_norm": 0.8278063754945647, + "learning_rate": 3.967428561748671e-06, + "loss": 0.555, + "num_input_tokens_seen": 15075590, + "step": 714 + }, + { + "epoch": 0.08597366680695004, + "flos": 22424548414560.0, + "grad_norm": 6.938990252470236, + "learning_rate": 3.967288400710045e-06, + "loss": 0.8775, + "num_input_tokens_seen": 15095055, + "step": 715 + }, + { + "epoch": 0.08609390969758914, + "flos": 23516842310880.0, + "grad_norm": 2.1744626225434134, + "learning_rate": 3.9671479412353335e-06, + "loss": 0.8848, + "num_input_tokens_seen": 15113520, + "step": 716 + }, + { + "epoch": 0.08621415258822822, + "flos": 25848793966560.0, + "grad_norm": 2.800234348844265, + "learning_rate": 3.967007183345843e-06, + "loss": 0.7418, + "num_input_tokens_seen": 15133375, + "step": 717 + }, + { + "epoch": 0.08633439547886732, + "flos": 13570396240800.0, + "grad_norm": 2.443557840972907, + "learning_rate": 3.966866127062927e-06, + "loss": 0.9008, + "num_input_tokens_seen": 15150500, + "step": 718 + }, + { + "epoch": 0.0864546383695064, + "flos": 57684941110080.0, + "grad_norm": 0.8990792014566589, + "learning_rate": 3.966724772407982e-06, + "loss": 0.6268, + "num_input_tokens_seen": 15208015, + "step": 719 + }, + { + "epoch": 0.0865748812601455, + "flos": 20018357280960.0, + "grad_norm": 2.1901287088066117, + "learning_rate": 3.966583119402454e-06, + "loss": 0.8833, + "num_input_tokens_seen": 15224780, + "step": 720 + }, + { + "epoch": 0.08669512415078459, + "flos": 35211731746080.0, + "grad_norm": 2.3683667157169657, + "learning_rate": 3.9664411680678305e-06, + "loss": 0.8178, + "num_input_tokens_seen": 15246535, + "step": 721 + }, + { + "epoch": 0.08681536704142367, + "flos": 65561225109600.0, + "grad_norm": 0.8436465376539235, + "learning_rate": 3.966298918425644e-06, + "loss": 0.6139, + "num_input_tokens_seen": 15304025, + "step": 722 + }, + { + "epoch": 0.08693560993206277, + "flos": 34480190536800.0, + "grad_norm": 2.4623933875537354, + "learning_rate": 3.966156370497476e-06, + "loss": 0.8246, + "num_input_tokens_seen": 15327635, + "step": 723 + }, + { + "epoch": 0.08705585282270185, + "flos": 23116349416800.0, + "grad_norm": 2.1365563925476847, + "learning_rate": 3.96601352430495e-06, + "loss": 0.891, + "num_input_tokens_seen": 15347405, + "step": 724 + }, + { + "epoch": 0.08717609571334095, + "flos": 29455274350080.0, + "grad_norm": 1.7682977886864153, + "learning_rate": 3.965870379869735e-06, + "loss": 0.835, + "num_input_tokens_seen": 15369450, + "step": 725 + }, + { + "epoch": 0.08729633860398003, + "flos": 20638000311840.0, + "grad_norm": 2.1787505511722536, + "learning_rate": 3.965726937213547e-06, + "loss": 0.8733, + "num_input_tokens_seen": 15388085, + "step": 726 + }, + { + "epoch": 0.08741658149461913, + "flos": 18343372828800.0, + "grad_norm": 2.2984626109957733, + "learning_rate": 3.965583196358144e-06, + "loss": 0.8119, + "num_input_tokens_seen": 15407120, + "step": 727 + }, + { + "epoch": 0.08753682438525823, + "flos": 18702861744000.0, + "grad_norm": 3.2869770751005136, + "learning_rate": 3.965439157325335e-06, + "loss": 0.7502, + "num_input_tokens_seen": 15424485, + "step": 728 + }, + { + "epoch": 0.08765706727589731, + "flos": 27736052149920.0, + "grad_norm": 4.718349818761538, + "learning_rate": 3.965294820136968e-06, + "loss": 0.7584, + "num_input_tokens_seen": 15446285, + "step": 729 + }, + { + "epoch": 0.08777731016653641, + "flos": 24354223027680.0, + "grad_norm": 2.3179191807768627, + "learning_rate": 3.965150184814938e-06, + "loss": 0.8709, + "num_input_tokens_seen": 15465770, + "step": 730 + }, + { + "epoch": 0.08789755305717549, + "flos": 21950485098240.0, + "grad_norm": 2.4059179498027072, + "learning_rate": 3.965005251381189e-06, + "loss": 0.7696, + "num_input_tokens_seen": 15483705, + "step": 731 + }, + { + "epoch": 0.08801779594781459, + "flos": 58277184294240.0, + "grad_norm": 0.9110653638661901, + "learning_rate": 3.964860019857705e-06, + "loss": 0.6462, + "num_input_tokens_seen": 15548620, + "step": 732 + }, + { + "epoch": 0.08813803883845367, + "flos": 23261483094240.0, + "grad_norm": 3.928716641999649, + "learning_rate": 3.964714490266518e-06, + "loss": 0.8416, + "num_input_tokens_seen": 15569265, + "step": 733 + }, + { + "epoch": 0.08825828172909277, + "flos": 63334369566240.0, + "grad_norm": 0.8868671242684812, + "learning_rate": 3.964568662629706e-06, + "loss": 0.646, + "num_input_tokens_seen": 15630570, + "step": 734 + }, + { + "epoch": 0.08837852461973186, + "flos": 26682569216640.0, + "grad_norm": 2.3775912754897366, + "learning_rate": 3.9644225369693895e-06, + "loss": 0.8437, + "num_input_tokens_seen": 15650870, + "step": 735 + }, + { + "epoch": 0.08849876751037095, + "flos": 27226114281600.0, + "grad_norm": 2.1653070830240098, + "learning_rate": 3.964276113307735e-06, + "loss": 0.8673, + "num_input_tokens_seen": 15670835, + "step": 736 + }, + { + "epoch": 0.08861901040101004, + "flos": 19800508085760.0, + "grad_norm": 2.709998768035496, + "learning_rate": 3.9641293916669574e-06, + "loss": 0.8091, + "num_input_tokens_seen": 15689795, + "step": 737 + }, + { + "epoch": 0.08873925329164913, + "flos": 23624837664480.0, + "grad_norm": 2.539249072922859, + "learning_rate": 3.9639823720693115e-06, + "loss": 0.8294, + "num_input_tokens_seen": 15711650, + "step": 738 + }, + { + "epoch": 0.08885949618228822, + "flos": 71729289157920.0, + "grad_norm": 0.8380977134615138, + "learning_rate": 3.963835054537102e-06, + "loss": 0.6, + "num_input_tokens_seen": 15780615, + "step": 739 + }, + { + "epoch": 0.08897973907292732, + "flos": 22314248535840.0, + "grad_norm": 2.3195429595106933, + "learning_rate": 3.963687439092676e-06, + "loss": 0.6128, + "num_input_tokens_seen": 15801300, + "step": 740 + }, + { + "epoch": 0.0890999819635664, + "flos": 21221508602400.0, + "grad_norm": 3.374587139732424, + "learning_rate": 3.963539525758427e-06, + "loss": 0.8042, + "num_input_tokens_seen": 15820860, + "step": 741 + }, + { + "epoch": 0.0892202248542055, + "flos": 25334432896800.0, + "grad_norm": 2.60429554938235, + "learning_rate": 3.9633913145567925e-06, + "loss": 0.678, + "num_input_tokens_seen": 15841350, + "step": 742 + }, + { + "epoch": 0.08934046774484458, + "flos": 24422738362560.0, + "grad_norm": 3.6439421771556835, + "learning_rate": 3.9632428055102575e-06, + "loss": 0.8159, + "num_input_tokens_seen": 15861360, + "step": 743 + }, + { + "epoch": 0.08946071063548368, + "flos": 35721297916800.0, + "grad_norm": 2.297765857477871, + "learning_rate": 3.9630939986413495e-06, + "loss": 0.6708, + "num_input_tokens_seen": 15883160, + "step": 744 + }, + { + "epoch": 0.08958095352612276, + "flos": 14336325211680.0, + "grad_norm": 1.8837738748415758, + "learning_rate": 3.962944893972643e-06, + "loss": 0.7821, + "num_input_tokens_seen": 15901610, + "step": 745 + }, + { + "epoch": 0.08970119641676186, + "flos": 17827710817440.0, + "grad_norm": 4.2165241162074, + "learning_rate": 3.962795491526756e-06, + "loss": 0.9103, + "num_input_tokens_seen": 15918770, + "step": 746 + }, + { + "epoch": 0.08982143930740095, + "flos": 20781907387200.0, + "grad_norm": 3.316582521173109, + "learning_rate": 3.962645791326354e-06, + "loss": 0.8926, + "num_input_tokens_seen": 15938025, + "step": 747 + }, + { + "epoch": 0.08994168219804004, + "flos": 24063249447360.0, + "grad_norm": 2.207072083762187, + "learning_rate": 3.962495793394146e-06, + "loss": 0.8334, + "num_input_tokens_seen": 15957215, + "step": 748 + }, + { + "epoch": 0.09006192508867913, + "flos": 57107528660160.0, + "grad_norm": 0.7460223658100666, + "learning_rate": 3.9623454977528864e-06, + "loss": 0.6121, + "num_input_tokens_seen": 16015870, + "step": 749 + }, + { + "epoch": 0.09018216797931822, + "flos": 20458367363520.0, + "grad_norm": 1.8806181343944681, + "learning_rate": 3.962194904425375e-06, + "loss": 0.8508, + "num_input_tokens_seen": 16036500, + "step": 750 + }, + { + "epoch": 0.09030241086995731, + "flos": 22605779662560.0, + "grad_norm": 2.840447321036168, + "learning_rate": 3.9620440134344566e-06, + "loss": 0.686, + "num_input_tokens_seen": 16054655, + "step": 751 + }, + { + "epoch": 0.09042265376059641, + "flos": 21840259559040.0, + "grad_norm": 3.5424847424702857, + "learning_rate": 3.9618928248030215e-06, + "loss": 0.8235, + "num_input_tokens_seen": 16074165, + "step": 752 + }, + { + "epoch": 0.0905428966512355, + "flos": 24281396000640.0, + "grad_norm": 2.9960456854236233, + "learning_rate": 3.961741338554005e-06, + "loss": 0.8303, + "num_input_tokens_seen": 16092505, + "step": 753 + }, + { + "epoch": 0.09066313954187459, + "flos": 35793455888160.0, + "grad_norm": 2.692975343984439, + "learning_rate": 3.9615895547103865e-06, + "loss": 0.755, + "num_input_tokens_seen": 16116150, + "step": 754 + }, + { + "epoch": 0.09078338243251367, + "flos": 29126865087840.0, + "grad_norm": 2.4821648513431103, + "learning_rate": 3.961437473295193e-06, + "loss": 0.7827, + "num_input_tokens_seen": 16136295, + "step": 755 + }, + { + "epoch": 0.09090362532315277, + "flos": 21876394299360.0, + "grad_norm": 2.9717228510564726, + "learning_rate": 3.961285094331495e-06, + "loss": 0.7229, + "num_input_tokens_seen": 16154210, + "step": 756 + }, + { + "epoch": 0.09102386821379185, + "flos": 27305297337600.0, + "grad_norm": 1.9497147962882249, + "learning_rate": 3.961132417842406e-06, + "loss": 0.8588, + "num_input_tokens_seen": 16173995, + "step": 757 + }, + { + "epoch": 0.09114411110443095, + "flos": 20783803044960.0, + "grad_norm": 2.8304482712243058, + "learning_rate": 3.960979443851089e-06, + "loss": 0.754, + "num_input_tokens_seen": 16191020, + "step": 758 + }, + { + "epoch": 0.09126435399507005, + "flos": 26108687893920.0, + "grad_norm": 2.1691953745918404, + "learning_rate": 3.96082617238075e-06, + "loss": 0.7864, + "num_input_tokens_seen": 16213645, + "step": 759 + }, + { + "epoch": 0.09138459688570913, + "flos": 24354000009120.0, + "grad_norm": 2.8199036531695865, + "learning_rate": 3.960672603454639e-06, + "loss": 0.8031, + "num_input_tokens_seen": 16233625, + "step": 760 + }, + { + "epoch": 0.09150483977634823, + "flos": 21002655823680.0, + "grad_norm": 3.2026807523487015, + "learning_rate": 3.960518737096054e-06, + "loss": 0.7638, + "num_input_tokens_seen": 16253255, + "step": 761 + }, + { + "epoch": 0.09162508266698731, + "flos": 22824335083200.0, + "grad_norm": 2.7914242615480336, + "learning_rate": 3.960364573328334e-06, + "loss": 0.7298, + "num_input_tokens_seen": 16272580, + "step": 762 + }, + { + "epoch": 0.0917453255576264, + "flos": 21693341733120.0, + "grad_norm": 3.17119421958358, + "learning_rate": 3.9602101121748675e-06, + "loss": 0.8877, + "num_input_tokens_seen": 16293075, + "step": 763 + }, + { + "epoch": 0.0918655684482655, + "flos": 14588487828960.0, + "grad_norm": 2.269820229390156, + "learning_rate": 3.960055353659085e-06, + "loss": 0.7243, + "num_input_tokens_seen": 16310185, + "step": 764 + }, + { + "epoch": 0.09198581133890459, + "flos": 23401301496000.0, + "grad_norm": 1.9062224839522788, + "learning_rate": 3.959900297804465e-06, + "loss": 0.8374, + "num_input_tokens_seen": 16330355, + "step": 765 + }, + { + "epoch": 0.09210605422954368, + "flos": 16771886189280.0, + "grad_norm": 2.1080428826821933, + "learning_rate": 3.9597449446345276e-06, + "loss": 0.7712, + "num_input_tokens_seen": 16347600, + "step": 766 + }, + { + "epoch": 0.09222629712018277, + "flos": 22641840063360.0, + "grad_norm": 2.7560607706797193, + "learning_rate": 3.95958929417284e-06, + "loss": 0.8321, + "num_input_tokens_seen": 16365755, + "step": 767 + }, + { + "epoch": 0.09234654001082186, + "flos": 69877422319680.0, + "grad_norm": 0.7286579310131155, + "learning_rate": 3.9594333464430145e-06, + "loss": 0.5875, + "num_input_tokens_seen": 16435245, + "step": 768 + }, + { + "epoch": 0.09246678290146094, + "flos": 19982928766080.0, + "grad_norm": 2.6228688989904745, + "learning_rate": 3.959277101468709e-06, + "loss": 0.8851, + "num_input_tokens_seen": 16454795, + "step": 769 + }, + { + "epoch": 0.09258702579210004, + "flos": 17722168668000.0, + "grad_norm": 2.5534426954900735, + "learning_rate": 3.959120559273624e-06, + "loss": 0.7852, + "num_input_tokens_seen": 16472580, + "step": 770 + }, + { + "epoch": 0.09270726868273914, + "flos": 20856741581280.0, + "grad_norm": 2.2353705475793832, + "learning_rate": 3.958963719881509e-06, + "loss": 0.8392, + "num_input_tokens_seen": 16490670, + "step": 771 + }, + { + "epoch": 0.09282751157337822, + "flos": 16991408023680.0, + "grad_norm": 2.501426480542859, + "learning_rate": 3.958806583316154e-06, + "loss": 0.9366, + "num_input_tokens_seen": 16508640, + "step": 772 + }, + { + "epoch": 0.09294775446401732, + "flos": 32476833992160.0, + "grad_norm": 4.213323196889861, + "learning_rate": 3.9586491496013985e-06, + "loss": 0.7886, + "num_input_tokens_seen": 16531035, + "step": 773 + }, + { + "epoch": 0.0930679973546564, + "flos": 18233779175520.0, + "grad_norm": 2.306230242057875, + "learning_rate": 3.958491418761124e-06, + "loss": 0.8309, + "num_input_tokens_seen": 16548605, + "step": 774 + }, + { + "epoch": 0.0931882402452955, + "flos": 21069424179840.0, + "grad_norm": 4.733803289660976, + "learning_rate": 3.958333390819258e-06, + "loss": 0.7277, + "num_input_tokens_seen": 16565535, + "step": 775 + }, + { + "epoch": 0.0933084831359346, + "flos": 24172657251840.0, + "grad_norm": 2.129506549718547, + "learning_rate": 3.9581750657997754e-06, + "loss": 0.8025, + "num_input_tokens_seen": 16584900, + "step": 776 + }, + { + "epoch": 0.09342872602657368, + "flos": 25443654852480.0, + "grad_norm": 2.1298908721477496, + "learning_rate": 3.95801644372669e-06, + "loss": 0.8978, + "num_input_tokens_seen": 16604805, + "step": 777 + }, + { + "epoch": 0.09354896891721277, + "flos": 23115977719200.0, + "grad_norm": 8.434250495503345, + "learning_rate": 3.957857524624068e-06, + "loss": 0.8454, + "num_input_tokens_seen": 16623685, + "step": 778 + }, + { + "epoch": 0.09366921180785186, + "flos": 24244815223200.0, + "grad_norm": 1.9481598265407576, + "learning_rate": 3.957698308516016e-06, + "loss": 0.8975, + "num_input_tokens_seen": 16644985, + "step": 779 + }, + { + "epoch": 0.09378945469849095, + "flos": 18703233441600.0, + "grad_norm": 3.7090003167443077, + "learning_rate": 3.957538795426688e-06, + "loss": 0.8276, + "num_input_tokens_seen": 16662410, + "step": 780 + }, + { + "epoch": 0.09390969758913004, + "flos": 23185533807360.0, + "grad_norm": 2.932309048444727, + "learning_rate": 3.9573789853802804e-06, + "loss": 0.7729, + "num_input_tokens_seen": 16680885, + "step": 781 + }, + { + "epoch": 0.09402994047976913, + "flos": 19618756461120.0, + "grad_norm": 3.698083721560387, + "learning_rate": 3.957218878401037e-06, + "loss": 0.7486, + "num_input_tokens_seen": 16699375, + "step": 782 + }, + { + "epoch": 0.09415018337040823, + "flos": 29378098461120.0, + "grad_norm": 2.2831230312686666, + "learning_rate": 3.957058474513246e-06, + "loss": 0.8984, + "num_input_tokens_seen": 16719230, + "step": 783 + }, + { + "epoch": 0.09427042626104731, + "flos": 24536941066080.0, + "grad_norm": 2.172095417940628, + "learning_rate": 3.956897773741241e-06, + "loss": 0.7875, + "num_input_tokens_seen": 16738220, + "step": 784 + }, + { + "epoch": 0.09439066915168641, + "flos": 26321816529600.0, + "grad_norm": 1.9168498394001736, + "learning_rate": 3.956736776109398e-06, + "loss": 0.7176, + "num_input_tokens_seen": 16759395, + "step": 785 + }, + { + "epoch": 0.09451091204232549, + "flos": 19399606324320.0, + "grad_norm": 2.196617417408114, + "learning_rate": 3.956575481642143e-06, + "loss": 0.842, + "num_input_tokens_seen": 16778285, + "step": 786 + }, + { + "epoch": 0.09463115493296459, + "flos": 25331496485760.0, + "grad_norm": 2.772193307243752, + "learning_rate": 3.956413890363943e-06, + "loss": 0.7499, + "num_input_tokens_seen": 16795265, + "step": 787 + }, + { + "epoch": 0.09475139782360369, + "flos": 10109123874240.0, + "grad_norm": 3.6242951332655977, + "learning_rate": 3.956252002299312e-06, + "loss": 0.8199, + "num_input_tokens_seen": 16811165, + "step": 788 + }, + { + "epoch": 0.09487164071424277, + "flos": 17206357977600.0, + "grad_norm": 2.6740948117038035, + "learning_rate": 3.956089817472807e-06, + "loss": 0.904, + "num_input_tokens_seen": 16828470, + "step": 789 + }, + { + "epoch": 0.09499188360488187, + "flos": 30805343497440.0, + "grad_norm": 2.940870907562297, + "learning_rate": 3.955927335909032e-06, + "loss": 0.8573, + "num_input_tokens_seen": 16851680, + "step": 790 + }, + { + "epoch": 0.09511212649552095, + "flos": 29309583126240.0, + "grad_norm": 2.3868573986682047, + "learning_rate": 3.955764557632634e-06, + "loss": 0.7594, + "num_input_tokens_seen": 16871490, + "step": 791 + }, + { + "epoch": 0.09523236938616005, + "flos": 10364185732800.0, + "grad_norm": 2.51769583628884, + "learning_rate": 3.955601482668309e-06, + "loss": 0.9474, + "num_input_tokens_seen": 16889350, + "step": 792 + }, + { + "epoch": 0.09535261227679913, + "flos": 19034653454400.0, + "grad_norm": 2.5881255421088656, + "learning_rate": 3.955438111040794e-06, + "loss": 0.8835, + "num_input_tokens_seen": 16907625, + "step": 793 + }, + { + "epoch": 0.09547285516743823, + "flos": 20893136509920.0, + "grad_norm": 2.4014135529213916, + "learning_rate": 3.955274442774873e-06, + "loss": 0.8006, + "num_input_tokens_seen": 16926205, + "step": 794 + }, + { + "epoch": 0.09559309805807732, + "flos": 30111126460800.0, + "grad_norm": 2.891133031792126, + "learning_rate": 3.9551104778953725e-06, + "loss": 0.7061, + "num_input_tokens_seen": 16946500, + "step": 795 + }, + { + "epoch": 0.0957133409487164, + "flos": 21036597548160.0, + "grad_norm": 1.8853952887852867, + "learning_rate": 3.954946216427167e-06, + "loss": 0.853, + "num_input_tokens_seen": 16966960, + "step": 796 + }, + { + "epoch": 0.0958335838393555, + "flos": 71195674152480.0, + "grad_norm": 0.8879419703811068, + "learning_rate": 3.954781658395176e-06, + "loss": 0.616, + "num_input_tokens_seen": 17023215, + "step": 797 + }, + { + "epoch": 0.09595382672999458, + "flos": 21841783519200.0, + "grad_norm": 2.189529458449958, + "learning_rate": 3.95461680382436e-06, + "loss": 0.9237, + "num_input_tokens_seen": 17042140, + "step": 798 + }, + { + "epoch": 0.09607406962063368, + "flos": 18668882849760.0, + "grad_norm": 2.992676378755141, + "learning_rate": 3.9544516527397295e-06, + "loss": 0.8642, + "num_input_tokens_seen": 17058490, + "step": 799 + }, + { + "epoch": 0.09619431251127276, + "flos": 22536335083680.0, + "grad_norm": 1.9430919060599032, + "learning_rate": 3.954286205166338e-06, + "loss": 0.8075, + "num_input_tokens_seen": 17079655, + "step": 800 + }, + { + "epoch": 0.09631455540191186, + "flos": 14226099672480.0, + "grad_norm": 2.9281781593797955, + "learning_rate": 3.954120461129282e-06, + "loss": 0.8398, + "num_input_tokens_seen": 17097065, + "step": 801 + }, + { + "epoch": 0.09643479829255096, + "flos": 20710790169120.0, + "grad_norm": 3.2043829703934055, + "learning_rate": 3.953954420653706e-06, + "loss": 0.8361, + "num_input_tokens_seen": 17114090, + "step": 802 + }, + { + "epoch": 0.09655504118319004, + "flos": 24385339850400.0, + "grad_norm": 2.035339112953421, + "learning_rate": 3.953788083764798e-06, + "loss": 0.8796, + "num_input_tokens_seen": 17133325, + "step": 803 + }, + { + "epoch": 0.09667528407382914, + "flos": 18415716648960.0, + "grad_norm": 2.350727123018955, + "learning_rate": 3.953621450487792e-06, + "loss": 0.9245, + "num_input_tokens_seen": 17151945, + "step": 804 + }, + { + "epoch": 0.09679552696446822, + "flos": 70715918129280.0, + "grad_norm": 0.8368227896667816, + "learning_rate": 3.953454520847964e-06, + "loss": 0.6117, + "num_input_tokens_seen": 17216790, + "step": 805 + }, + { + "epoch": 0.09691576985510732, + "flos": 21914201678880.0, + "grad_norm": 2.5517257647540266, + "learning_rate": 3.9532872948706395e-06, + "loss": 0.7387, + "num_input_tokens_seen": 17236285, + "step": 806 + }, + { + "epoch": 0.09703601274574641, + "flos": 17939571826080.0, + "grad_norm": 3.660006948579123, + "learning_rate": 3.9531197725811845e-06, + "loss": 0.8296, + "num_input_tokens_seen": 17251670, + "step": 807 + }, + { + "epoch": 0.0971562556363855, + "flos": 22130452574400.0, + "grad_norm": 2.4267670475805163, + "learning_rate": 3.952951954005013e-06, + "loss": 0.8799, + "num_input_tokens_seen": 17271355, + "step": 808 + }, + { + "epoch": 0.0972764985270246, + "flos": 25811692812480.0, + "grad_norm": 1.7988230221621093, + "learning_rate": 3.952783839167584e-06, + "loss": 0.8451, + "num_input_tokens_seen": 17291400, + "step": 809 + }, + { + "epoch": 0.09739674141766368, + "flos": 20310371614560.0, + "grad_norm": 3.2895304512345844, + "learning_rate": 3.952615428094398e-06, + "loss": 0.743, + "num_input_tokens_seen": 17310920, + "step": 810 + }, + { + "epoch": 0.09751698430830277, + "flos": 15720782120640.0, + "grad_norm": 1.9809604613716605, + "learning_rate": 3.952446720811004e-06, + "loss": 0.734, + "num_input_tokens_seen": 17329245, + "step": 811 + }, + { + "epoch": 0.09763722719894186, + "flos": 63625826353440.0, + "grad_norm": 0.8332757373615308, + "learning_rate": 3.952277717342995e-06, + "loss": 0.6358, + "num_input_tokens_seen": 17395680, + "step": 812 + }, + { + "epoch": 0.09775747008958095, + "flos": 22059372526080.0, + "grad_norm": 2.4119963955385724, + "learning_rate": 3.952108417716009e-06, + "loss": 0.8547, + "num_input_tokens_seen": 17415135, + "step": 813 + }, + { + "epoch": 0.09787771298022005, + "flos": 21255599005920.0, + "grad_norm": 2.4050115395505056, + "learning_rate": 3.951938821955727e-06, + "loss": 0.85, + "num_input_tokens_seen": 17434535, + "step": 814 + }, + { + "epoch": 0.09799795587085913, + "flos": 22023089106720.0, + "grad_norm": 3.0001255644905864, + "learning_rate": 3.9517689300878786e-06, + "loss": 0.7686, + "num_input_tokens_seen": 17454270, + "step": 815 + }, + { + "epoch": 0.09811819876149823, + "flos": 22132088043840.0, + "grad_norm": 1.7113720764353628, + "learning_rate": 3.951598742138236e-06, + "loss": 0.7821, + "num_input_tokens_seen": 17472995, + "step": 816 + }, + { + "epoch": 0.09823844165213731, + "flos": 22199153758080.0, + "grad_norm": 2.5652467846152742, + "learning_rate": 3.951428258132615e-06, + "loss": 0.7999, + "num_input_tokens_seen": 17491115, + "step": 817 + }, + { + "epoch": 0.09835868454277641, + "flos": 22455442218720.0, + "grad_norm": 2.473195865320522, + "learning_rate": 3.951257478096879e-06, + "loss": 0.8454, + "num_input_tokens_seen": 17509480, + "step": 818 + }, + { + "epoch": 0.0984789274334155, + "flos": 16339347228480.0, + "grad_norm": 3.5645225539043066, + "learning_rate": 3.951086402056936e-06, + "loss": 0.684, + "num_input_tokens_seen": 17524080, + "step": 819 + }, + { + "epoch": 0.09859917032405459, + "flos": 24208866331680.0, + "grad_norm": 2.471992777298434, + "learning_rate": 3.950915030038735e-06, + "loss": 0.8378, + "num_input_tokens_seen": 17543875, + "step": 820 + }, + { + "epoch": 0.09871941321469369, + "flos": 17395432044960.0, + "grad_norm": 2.7100723161233473, + "learning_rate": 3.9507433620682765e-06, + "loss": 0.8391, + "num_input_tokens_seen": 17560810, + "step": 821 + }, + { + "epoch": 0.09883965610533277, + "flos": 28436550876000.0, + "grad_norm": 1.827473842365808, + "learning_rate": 3.9505713981716e-06, + "loss": 0.8804, + "num_input_tokens_seen": 17583640, + "step": 822 + }, + { + "epoch": 0.09895989899597187, + "flos": 23659968821280.0, + "grad_norm": 10.401137187113545, + "learning_rate": 3.950399138374795e-06, + "loss": 0.8097, + "num_input_tokens_seen": 17602720, + "step": 823 + }, + { + "epoch": 0.09908014188661095, + "flos": 24644230194240.0, + "grad_norm": 1.9152554900849168, + "learning_rate": 3.95022658270399e-06, + "loss": 0.7471, + "num_input_tokens_seen": 17623085, + "step": 824 + }, + { + "epoch": 0.09920038477725004, + "flos": 14044459557120.0, + "grad_norm": 1.911107050936967, + "learning_rate": 3.9500537311853635e-06, + "loss": 0.7818, + "num_input_tokens_seen": 17641040, + "step": 825 + }, + { + "epoch": 0.09932062766788914, + "flos": 13388161409280.0, + "grad_norm": 3.897532019639635, + "learning_rate": 3.949880583845136e-06, + "loss": 0.833, + "num_input_tokens_seen": 17659115, + "step": 826 + }, + { + "epoch": 0.09944087055852822, + "flos": 19472990897760.0, + "grad_norm": 3.0749696931663575, + "learning_rate": 3.949707140709575e-06, + "loss": 0.8111, + "num_input_tokens_seen": 17678845, + "step": 827 + }, + { + "epoch": 0.09956111344916732, + "flos": 17723878476960.0, + "grad_norm": 2.7997868048020638, + "learning_rate": 3.949533401804991e-06, + "loss": 0.8354, + "num_input_tokens_seen": 17695750, + "step": 828 + }, + { + "epoch": 0.0996813563398064, + "flos": 17942025030240.0, + "grad_norm": 4.467860633541188, + "learning_rate": 3.949359367157739e-06, + "loss": 0.9091, + "num_input_tokens_seen": 17714445, + "step": 829 + }, + { + "epoch": 0.0998015992304455, + "flos": 17432235840960.0, + "grad_norm": 2.071216544213198, + "learning_rate": 3.949185036794222e-06, + "loss": 0.7567, + "num_input_tokens_seen": 17732455, + "step": 830 + }, + { + "epoch": 0.0999218421210846, + "flos": 25851804717120.0, + "grad_norm": 2.645496817587433, + "learning_rate": 3.949010410740884e-06, + "loss": 0.7898, + "num_input_tokens_seen": 17755280, + "step": 831 + }, + { + "epoch": 0.10004208501172368, + "flos": 21185671220160.0, + "grad_norm": 1.9769784843960474, + "learning_rate": 3.948835489024216e-06, + "loss": 0.8673, + "num_input_tokens_seen": 17775015, + "step": 832 + }, + { + "epoch": 0.10016232790236278, + "flos": 17323794450240.0, + "grad_norm": 3.366954821980003, + "learning_rate": 3.948660271670755e-06, + "loss": 0.9058, + "num_input_tokens_seen": 17792165, + "step": 833 + }, + { + "epoch": 0.10028257079300186, + "flos": 25629792508800.0, + "grad_norm": 2.450818088730094, + "learning_rate": 3.948484758707079e-06, + "loss": 0.8408, + "num_input_tokens_seen": 17811765, + "step": 834 + }, + { + "epoch": 0.10040281368364096, + "flos": 25120040489280.0, + "grad_norm": 3.0362259441630473, + "learning_rate": 3.948308950159815e-06, + "loss": 0.8356, + "num_input_tokens_seen": 17830445, + "step": 835 + }, + { + "epoch": 0.10052305657428004, + "flos": 17578298762400.0, + "grad_norm": 2.5593267191710103, + "learning_rate": 3.9481328460556326e-06, + "loss": 0.758, + "num_input_tokens_seen": 17847665, + "step": 836 + }, + { + "epoch": 0.10064329946491914, + "flos": 18633565844160.0, + "grad_norm": 2.194556489597649, + "learning_rate": 3.9479564464212455e-06, + "loss": 0.891, + "num_input_tokens_seen": 17866825, + "step": 837 + }, + { + "epoch": 0.10076354235555823, + "flos": 17174534929440.0, + "grad_norm": 3.3499189850989044, + "learning_rate": 3.947779751283414e-06, + "loss": 0.7618, + "num_input_tokens_seen": 17883995, + "step": 838 + }, + { + "epoch": 0.10088378524619732, + "flos": 22929282516480.0, + "grad_norm": 1.9610954920129042, + "learning_rate": 3.947602760668944e-06, + "loss": 0.7626, + "num_input_tokens_seen": 17903785, + "step": 839 + }, + { + "epoch": 0.10100402813683641, + "flos": 37831163024640.0, + "grad_norm": 2.264508412890264, + "learning_rate": 3.947425474604684e-06, + "loss": 0.7154, + "num_input_tokens_seen": 17927720, + "step": 840 + }, + { + "epoch": 0.1011242710274755, + "flos": 21512705201280.0, + "grad_norm": 2.3374910323843063, + "learning_rate": 3.947247893117528e-06, + "loss": 0.9212, + "num_input_tokens_seen": 17946225, + "step": 841 + }, + { + "epoch": 0.10124451391811459, + "flos": 13602070609920.0, + "grad_norm": 5.286566772350974, + "learning_rate": 3.947070016234413e-06, + "loss": 0.691, + "num_input_tokens_seen": 17962015, + "step": 842 + }, + { + "epoch": 0.10136475680875369, + "flos": 16625414400480.0, + "grad_norm": 2.8150091211093278, + "learning_rate": 3.946891843982326e-06, + "loss": 0.7491, + "num_input_tokens_seen": 17979680, + "step": 843 + }, + { + "epoch": 0.10148499969939277, + "flos": 19433213520960.0, + "grad_norm": 2.3407891454042136, + "learning_rate": 3.9467133763882935e-06, + "loss": 0.7468, + "num_input_tokens_seen": 17998145, + "step": 844 + }, + { + "epoch": 0.10160524259003187, + "flos": 21074850964800.0, + "grad_norm": 2.04709583285125, + "learning_rate": 3.9465346134793905e-06, + "loss": 0.8625, + "num_input_tokens_seen": 18017955, + "step": 845 + }, + { + "epoch": 0.10172548548067095, + "flos": 17687520718080.0, + "grad_norm": 2.5255718829963323, + "learning_rate": 3.9463555552827335e-06, + "loss": 0.7956, + "num_input_tokens_seen": 18035125, + "step": 846 + }, + { + "epoch": 0.10184572837131005, + "flos": 21074702285760.0, + "grad_norm": 2.760492940364766, + "learning_rate": 3.946176201825487e-06, + "loss": 0.8642, + "num_input_tokens_seen": 18053160, + "step": 847 + }, + { + "epoch": 0.10196597126194913, + "flos": 26030248233120.0, + "grad_norm": 2.0730474868465865, + "learning_rate": 3.9459965531348575e-06, + "loss": 0.8374, + "num_input_tokens_seen": 18072815, + "step": 848 + }, + { + "epoch": 0.10208621415258823, + "flos": 29272556311680.0, + "grad_norm": 2.7491976694619735, + "learning_rate": 3.945816609238098e-06, + "loss": 0.8482, + "num_input_tokens_seen": 18092225, + "step": 849 + }, + { + "epoch": 0.10220645704322733, + "flos": 23771755490400.0, + "grad_norm": 2.3483127177596783, + "learning_rate": 3.945636370162507e-06, + "loss": 0.8529, + "num_input_tokens_seen": 18112335, + "step": 850 + }, + { + "epoch": 0.10232669993386641, + "flos": 23185236449280.0, + "grad_norm": 2.0393011311589424, + "learning_rate": 3.945455835935425e-06, + "loss": 0.7916, + "num_input_tokens_seen": 18131520, + "step": 851 + }, + { + "epoch": 0.1024469428245055, + "flos": 22889914007040.0, + "grad_norm": 2.848693574996882, + "learning_rate": 3.94527500658424e-06, + "loss": 0.7525, + "num_input_tokens_seen": 18149185, + "step": 852 + }, + { + "epoch": 0.10256718571514459, + "flos": 31314500800800.0, + "grad_norm": 3.403706572189417, + "learning_rate": 3.945093882136382e-06, + "loss": 0.8137, + "num_input_tokens_seen": 18172960, + "step": 853 + }, + { + "epoch": 0.10268742860578368, + "flos": 23441227551840.0, + "grad_norm": 2.1797555345424984, + "learning_rate": 3.944912462619329e-06, + "loss": 0.8459, + "num_input_tokens_seen": 18191925, + "step": 854 + }, + { + "epoch": 0.10280767149642277, + "flos": 25483766757120.0, + "grad_norm": 3.2778980391112795, + "learning_rate": 3.9447307480606025e-06, + "loss": 0.8076, + "num_input_tokens_seen": 18212010, + "step": 855 + }, + { + "epoch": 0.10292791438706186, + "flos": 17323051055040.0, + "grad_norm": 2.7017758922238087, + "learning_rate": 3.944548738487767e-06, + "loss": 0.8992, + "num_input_tokens_seen": 18229525, + "step": 856 + }, + { + "epoch": 0.10304815727770096, + "flos": 27013914889920.0, + "grad_norm": 2.649322541930835, + "learning_rate": 3.944366433928434e-06, + "loss": 0.9071, + "num_input_tokens_seen": 18249505, + "step": 857 + }, + { + "epoch": 0.10316840016834004, + "flos": 22749984096000.0, + "grad_norm": 1.7109308808777857, + "learning_rate": 3.9441838344102594e-06, + "loss": 0.8386, + "num_input_tokens_seen": 18269230, + "step": 858 + }, + { + "epoch": 0.10328864305897914, + "flos": 20675064296160.0, + "grad_norm": 4.44470339391592, + "learning_rate": 3.944000939960943e-06, + "loss": 0.6783, + "num_input_tokens_seen": 18287955, + "step": 859 + }, + { + "epoch": 0.10340888594961822, + "flos": 28438074836160.0, + "grad_norm": 1.7163974266884618, + "learning_rate": 3.943817750608229e-06, + "loss": 0.7982, + "num_input_tokens_seen": 18310505, + "step": 860 + }, + { + "epoch": 0.10352912884025732, + "flos": 13351469122560.0, + "grad_norm": 2.866566603784588, + "learning_rate": 3.943634266379908e-06, + "loss": 0.8176, + "num_input_tokens_seen": 18327400, + "step": 861 + }, + { + "epoch": 0.10364937173089642, + "flos": 25522020173760.0, + "grad_norm": 2.360675276629175, + "learning_rate": 3.943450487303815e-06, + "loss": 0.8489, + "num_input_tokens_seen": 18347895, + "step": 862 + }, + { + "epoch": 0.1037696146215355, + "flos": 21185225183040.0, + "grad_norm": 16.44217792276163, + "learning_rate": 3.943266413407827e-06, + "loss": 0.8546, + "num_input_tokens_seen": 18367170, + "step": 863 + }, + { + "epoch": 0.1038898575121746, + "flos": 25770242796480.0, + "grad_norm": 2.016275108829061, + "learning_rate": 3.94308204471987e-06, + "loss": 0.8502, + "num_input_tokens_seen": 18386185, + "step": 864 + }, + { + "epoch": 0.10401010040281368, + "flos": 19035285340320.0, + "grad_norm": 3.0155854691992894, + "learning_rate": 3.942897381267912e-06, + "loss": 0.7488, + "num_input_tokens_seen": 18402550, + "step": 865 + }, + { + "epoch": 0.10413034329345278, + "flos": 16331653088160.0, + "grad_norm": 11.361299081587855, + "learning_rate": 3.942712423079965e-06, + "loss": 0.6607, + "num_input_tokens_seen": 18418940, + "step": 866 + }, + { + "epoch": 0.10425058618409186, + "flos": 17211970611360.0, + "grad_norm": 2.691916274867403, + "learning_rate": 3.942527170184088e-06, + "loss": 0.9008, + "num_input_tokens_seen": 18435560, + "step": 867 + }, + { + "epoch": 0.10437082907473096, + "flos": 17941802011680.0, + "grad_norm": 2.9953854356461465, + "learning_rate": 3.942341622608385e-06, + "loss": 0.7775, + "num_input_tokens_seen": 18452590, + "step": 868 + }, + { + "epoch": 0.10449107196537005, + "flos": 36231087106080.0, + "grad_norm": 1.6129226101202694, + "learning_rate": 3.942155780381001e-06, + "loss": 0.7762, + "num_input_tokens_seen": 18476325, + "step": 869 + }, + { + "epoch": 0.10461131485600914, + "flos": 23767927005120.0, + "grad_norm": 2.011450414446101, + "learning_rate": 3.94196964353013e-06, + "loss": 0.7609, + "num_input_tokens_seen": 18495775, + "step": 870 + }, + { + "epoch": 0.10473155774664823, + "flos": 18379284550560.0, + "grad_norm": 2.267164457113366, + "learning_rate": 3.941783212084008e-06, + "loss": 0.8085, + "num_input_tokens_seen": 18513530, + "step": 871 + }, + { + "epoch": 0.10485180063728732, + "flos": 25555961898240.0, + "grad_norm": 3.9789302948787513, + "learning_rate": 3.941596486070916e-06, + "loss": 0.7869, + "num_input_tokens_seen": 18531575, + "step": 872 + }, + { + "epoch": 0.10497204352792641, + "flos": 27049900951200.0, + "grad_norm": 17.043013067928857, + "learning_rate": 3.941409465519182e-06, + "loss": 0.5835, + "num_input_tokens_seen": 18552100, + "step": 873 + }, + { + "epoch": 0.10509228641856551, + "flos": 32811413434560.0, + "grad_norm": 1.7370519513194307, + "learning_rate": 3.941222150457176e-06, + "loss": 0.8522, + "num_input_tokens_seen": 18575355, + "step": 874 + }, + { + "epoch": 0.10521252930920459, + "flos": 14298926699520.0, + "grad_norm": 3.6667582471565834, + "learning_rate": 3.941034540913311e-06, + "loss": 0.7153, + "num_input_tokens_seen": 18592885, + "step": 875 + }, + { + "epoch": 0.10533277219984369, + "flos": 21655980390720.0, + "grad_norm": 1.7601414151165513, + "learning_rate": 3.940846636916051e-06, + "loss": 0.8248, + "num_input_tokens_seen": 18612920, + "step": 876 + }, + { + "epoch": 0.10545301509048277, + "flos": 22237853211840.0, + "grad_norm": 2.2904683766025475, + "learning_rate": 3.940658438493899e-06, + "loss": 0.8626, + "num_input_tokens_seen": 18630765, + "step": 877 + }, + { + "epoch": 0.10557325798112187, + "flos": 22167999765600.0, + "grad_norm": 2.559210922701566, + "learning_rate": 3.940469945675405e-06, + "loss": 0.7582, + "num_input_tokens_seen": 18649150, + "step": 878 + }, + { + "epoch": 0.10569350087176095, + "flos": 25738977294720.0, + "grad_norm": 2.119287779308964, + "learning_rate": 3.940281158489163e-06, + "loss": 0.9138, + "num_input_tokens_seen": 18668380, + "step": 879 + }, + { + "epoch": 0.10581374376240005, + "flos": 17286804805440.0, + "grad_norm": 2.1913786153699863, + "learning_rate": 3.940092076963812e-06, + "loss": 0.8295, + "num_input_tokens_seen": 18685875, + "step": 880 + }, + { + "epoch": 0.10593398665303914, + "flos": 34300037211840.0, + "grad_norm": 2.3496942037108037, + "learning_rate": 3.9399027011280355e-06, + "loss": 0.7911, + "num_input_tokens_seen": 18707970, + "step": 881 + }, + { + "epoch": 0.10605422954367823, + "flos": 23224716468000.0, + "grad_norm": 2.473492858999789, + "learning_rate": 3.939713031010561e-06, + "loss": 0.7768, + "num_input_tokens_seen": 18726335, + "step": 882 + }, + { + "epoch": 0.10617447243431732, + "flos": 22787494117440.0, + "grad_norm": 2.51840946744717, + "learning_rate": 3.939523066640163e-06, + "loss": 0.7765, + "num_input_tokens_seen": 18745230, + "step": 883 + }, + { + "epoch": 0.10629471532495641, + "flos": 24351583974720.0, + "grad_norm": 2.1083876948328104, + "learning_rate": 3.939332808045657e-06, + "loss": 0.8138, + "num_input_tokens_seen": 18764880, + "step": 884 + }, + { + "epoch": 0.1064149582155955, + "flos": 21075371341440.0, + "grad_norm": 1.8664642593227831, + "learning_rate": 3.939142255255906e-06, + "loss": 0.8514, + "num_input_tokens_seen": 18785765, + "step": 885 + }, + { + "epoch": 0.1065352011062346, + "flos": 20672425243200.0, + "grad_norm": 2.217406497433346, + "learning_rate": 3.938951408299817e-06, + "loss": 0.8666, + "num_input_tokens_seen": 18804605, + "step": 886 + }, + { + "epoch": 0.10665544399687368, + "flos": 62570224743840.0, + "grad_norm": 0.8005353968893699, + "learning_rate": 3.938760267206342e-06, + "loss": 0.5438, + "num_input_tokens_seen": 18866425, + "step": 887 + }, + { + "epoch": 0.10677568688751278, + "flos": 26102926581120.0, + "grad_norm": 2.787723850448143, + "learning_rate": 3.938568832004475e-06, + "loss": 0.7859, + "num_input_tokens_seen": 18885130, + "step": 888 + }, + { + "epoch": 0.10689592977815186, + "flos": 12768518378400.0, + "grad_norm": 2.657872501079506, + "learning_rate": 3.938377102723257e-06, + "loss": 0.7571, + "num_input_tokens_seen": 18902265, + "step": 889 + }, + { + "epoch": 0.10701617266879096, + "flos": 22095061229280.0, + "grad_norm": 4.649819868967267, + "learning_rate": 3.938185079391774e-06, + "loss": 0.834, + "num_input_tokens_seen": 18919310, + "step": 890 + }, + { + "epoch": 0.10713641555943004, + "flos": 19717124846880.0, + "grad_norm": 2.9610067475205635, + "learning_rate": 3.937992762039157e-06, + "loss": 1.0557, + "num_input_tokens_seen": 18932635, + "step": 891 + }, + { + "epoch": 0.10725665845006914, + "flos": 23918784825600.0, + "grad_norm": 1.934082647610985, + "learning_rate": 3.937800150694577e-06, + "loss": 0.8076, + "num_input_tokens_seen": 18953810, + "step": 892 + }, + { + "epoch": 0.10737690134070824, + "flos": 18525310302240.0, + "grad_norm": 2.3590466605597267, + "learning_rate": 3.937607245387255e-06, + "loss": 0.7614, + "num_input_tokens_seen": 18973175, + "step": 893 + }, + { + "epoch": 0.10749714423134732, + "flos": 22675038392640.0, + "grad_norm": 4.947222059248566, + "learning_rate": 3.937414046146455e-06, + "loss": 0.7203, + "num_input_tokens_seen": 18991130, + "step": 894 + }, + { + "epoch": 0.10761738712198642, + "flos": 21075966057600.0, + "grad_norm": 2.4803708995832685, + "learning_rate": 3.9372205530014845e-06, + "loss": 0.7536, + "num_input_tokens_seen": 19009610, + "step": 895 + }, + { + "epoch": 0.1077376300126255, + "flos": 23732647169280.0, + "grad_norm": 2.87414986024188, + "learning_rate": 3.937026765981696e-06, + "loss": 0.7105, + "num_input_tokens_seen": 19029680, + "step": 896 + }, + { + "epoch": 0.1078578729032646, + "flos": 20889605382720.0, + "grad_norm": 2.0279645906325325, + "learning_rate": 3.936832685116488e-06, + "loss": 0.7959, + "num_input_tokens_seen": 19047775, + "step": 897 + }, + { + "epoch": 0.10797811579390369, + "flos": 14808901737600.0, + "grad_norm": 2.5602854952464167, + "learning_rate": 3.936638310435301e-06, + "loss": 0.9037, + "num_input_tokens_seen": 19065640, + "step": 898 + }, + { + "epoch": 0.10809835868454278, + "flos": 19509162807840.0, + "grad_norm": 2.5699717750929034, + "learning_rate": 3.936443641967623e-06, + "loss": 0.8186, + "num_input_tokens_seen": 19084470, + "step": 899 + }, + { + "epoch": 0.10821860157518187, + "flos": 18416088346560.0, + "grad_norm": 2.5608959294704596, + "learning_rate": 3.936248679742983e-06, + "loss": 0.8323, + "num_input_tokens_seen": 19102965, + "step": 900 + }, + { + "epoch": 0.10833884446582095, + "flos": 49288720576800.0, + "grad_norm": 1.0530610718836009, + "learning_rate": 3.936053423790959e-06, + "loss": 0.7016, + "num_input_tokens_seen": 19151285, + "step": 901 + }, + { + "epoch": 0.10845908735646005, + "flos": 20382789774240.0, + "grad_norm": 1.8450780708315242, + "learning_rate": 3.935857874141168e-06, + "loss": 0.7736, + "num_input_tokens_seen": 19170120, + "step": 902 + }, + { + "epoch": 0.10857933024709913, + "flos": 14007953119200.0, + "grad_norm": 2.658624731931996, + "learning_rate": 3.935662030823279e-06, + "loss": 0.8365, + "num_input_tokens_seen": 19186305, + "step": 903 + }, + { + "epoch": 0.10869957313773823, + "flos": 13350093841440.0, + "grad_norm": 2.675592078879869, + "learning_rate": 3.935465893866998e-06, + "loss": 0.7236, + "num_input_tokens_seen": 19204530, + "step": 904 + }, + { + "epoch": 0.10881981602837733, + "flos": 25769945438400.0, + "grad_norm": 1.9039658165756224, + "learning_rate": 3.935269463302079e-06, + "loss": 0.8021, + "num_input_tokens_seen": 19222735, + "step": 905 + }, + { + "epoch": 0.10894005891901641, + "flos": 20747928492960.0, + "grad_norm": 1.9786780574056, + "learning_rate": 3.935072739158322e-06, + "loss": 0.7671, + "num_input_tokens_seen": 19242445, + "step": 906 + }, + { + "epoch": 0.10906030180965551, + "flos": 26611637847360.0, + "grad_norm": 1.844521700467987, + "learning_rate": 3.934875721465569e-06, + "loss": 0.7997, + "num_input_tokens_seen": 19262865, + "step": 907 + }, + { + "epoch": 0.10918054470029459, + "flos": 36481651423680.0, + "grad_norm": 2.982490783574844, + "learning_rate": 3.9346784102537076e-06, + "loss": 0.7189, + "num_input_tokens_seen": 19285000, + "step": 908 + }, + { + "epoch": 0.10930078759093369, + "flos": 21731483640480.0, + "grad_norm": 2.1808272905840576, + "learning_rate": 3.934480805552669e-06, + "loss": 0.7832, + "num_input_tokens_seen": 19306010, + "step": 909 + }, + { + "epoch": 0.10942103048157277, + "flos": 22569496243200.0, + "grad_norm": 5.112036149077737, + "learning_rate": 3.93428290739243e-06, + "loss": 0.8814, + "num_input_tokens_seen": 19325380, + "step": 910 + }, + { + "epoch": 0.10954127337221187, + "flos": 15024372068160.0, + "grad_norm": 10.660752405234028, + "learning_rate": 3.9340847158030125e-06, + "loss": 0.8002, + "num_input_tokens_seen": 19338960, + "step": 911 + }, + { + "epoch": 0.10966151626285096, + "flos": 21619994329440.0, + "grad_norm": 2.2561879368749946, + "learning_rate": 3.9338862308144814e-06, + "loss": 0.756, + "num_input_tokens_seen": 19357780, + "step": 912 + }, + { + "epoch": 0.10978175915349005, + "flos": 20092708268160.0, + "grad_norm": 1.7763090706116316, + "learning_rate": 3.933687452456946e-06, + "loss": 0.8464, + "num_input_tokens_seen": 19377680, + "step": 913 + }, + { + "epoch": 0.10990200204412914, + "flos": 20383570339200.0, + "grad_norm": 3.060987449343604, + "learning_rate": 3.933488380760562e-06, + "loss": 0.8603, + "num_input_tokens_seen": 19397040, + "step": 914 + }, + { + "epoch": 0.11002224493476823, + "flos": 17505211547040.0, + "grad_norm": 2.567930097204918, + "learning_rate": 3.9332890157555286e-06, + "loss": 0.8726, + "num_input_tokens_seen": 19413975, + "step": 915 + }, + { + "epoch": 0.11014248782540732, + "flos": 12185939331840.0, + "grad_norm": 2.813016547345943, + "learning_rate": 3.933089357472088e-06, + "loss": 0.7639, + "num_input_tokens_seen": 19432005, + "step": 916 + }, + { + "epoch": 0.11026273071604642, + "flos": 22351015162080.0, + "grad_norm": 2.0564789015607956, + "learning_rate": 3.932889405940529e-06, + "loss": 0.861, + "num_input_tokens_seen": 19450100, + "step": 917 + }, + { + "epoch": 0.1103829736066855, + "flos": 19800805443840.0, + "grad_norm": 2.5059915271381374, + "learning_rate": 3.932689161191184e-06, + "loss": 0.803, + "num_input_tokens_seen": 19467805, + "step": 918 + }, + { + "epoch": 0.1105032164973246, + "flos": 22637045164320.0, + "grad_norm": 2.783495221788469, + "learning_rate": 3.93248862325443e-06, + "loss": 0.8781, + "num_input_tokens_seen": 19486710, + "step": 919 + }, + { + "epoch": 0.11062345938796368, + "flos": 66388272633120.0, + "grad_norm": 0.9439270964936185, + "learning_rate": 3.932287792160688e-06, + "loss": 0.6441, + "num_input_tokens_seen": 19545170, + "step": 920 + }, + { + "epoch": 0.11074370227860278, + "flos": 21876468638880.0, + "grad_norm": 3.3093776742493914, + "learning_rate": 3.932086667940424e-06, + "loss": 0.8031, + "num_input_tokens_seen": 19561875, + "step": 921 + }, + { + "epoch": 0.11086394516924186, + "flos": 28617001559040.0, + "grad_norm": 2.3051243254920855, + "learning_rate": 3.93188525062415e-06, + "loss": 0.8213, + "num_input_tokens_seen": 19582340, + "step": 922 + }, + { + "epoch": 0.11098418805988096, + "flos": 24500137270080.0, + "grad_norm": 2.2495812980170196, + "learning_rate": 3.931683540242418e-06, + "loss": 0.8612, + "num_input_tokens_seen": 19603405, + "step": 923 + }, + { + "epoch": 0.11110443095052006, + "flos": 22927349688960.0, + "grad_norm": 2.79695427638389, + "learning_rate": 3.9314815368258295e-06, + "loss": 0.9104, + "num_input_tokens_seen": 19619885, + "step": 924 + }, + { + "epoch": 0.11122467384115914, + "flos": 18922941124800.0, + "grad_norm": 2.789910281116306, + "learning_rate": 3.9312792404050275e-06, + "loss": 0.7835, + "num_input_tokens_seen": 19637940, + "step": 925 + }, + { + "epoch": 0.11134491673179824, + "flos": 25047064783200.0, + "grad_norm": 2.0294770940036373, + "learning_rate": 3.9310766510107e-06, + "loss": 0.7765, + "num_input_tokens_seen": 19657115, + "step": 926 + }, + { + "epoch": 0.11146515962243732, + "flos": 24464076869280.0, + "grad_norm": 4.891809094687984, + "learning_rate": 3.9308737686735806e-06, + "loss": 0.9231, + "num_input_tokens_seen": 19677075, + "step": 927 + }, + { + "epoch": 0.11158540251307641, + "flos": 22311758161920.0, + "grad_norm": 2.265801258465753, + "learning_rate": 3.9306705934244455e-06, + "loss": 0.8271, + "num_input_tokens_seen": 19697155, + "step": 928 + }, + { + "epoch": 0.11170564540371551, + "flos": 19873966998720.0, + "grad_norm": 1.7641463037660194, + "learning_rate": 3.930467125294116e-06, + "loss": 0.8812, + "num_input_tokens_seen": 19716705, + "step": 929 + }, + { + "epoch": 0.1118258882943546, + "flos": 64493766346560.0, + "grad_norm": 0.9328854799076836, + "learning_rate": 3.930263364313458e-06, + "loss": 0.6044, + "num_input_tokens_seen": 19767275, + "step": 930 + }, + { + "epoch": 0.11194613118499369, + "flos": 17177545680000.0, + "grad_norm": 3.539685461865971, + "learning_rate": 3.930059310513384e-06, + "loss": 0.8306, + "num_input_tokens_seen": 19786315, + "step": 931 + }, + { + "epoch": 0.11206637407563277, + "flos": 31817487924000.0, + "grad_norm": 2.015884737534469, + "learning_rate": 3.929854963924846e-06, + "loss": 0.8378, + "num_input_tokens_seen": 19806580, + "step": 932 + }, + { + "epoch": 0.11218661696627187, + "flos": 21914313188160.0, + "grad_norm": 2.2585853389133423, + "learning_rate": 3.929650324578845e-06, + "loss": 0.7749, + "num_input_tokens_seen": 19826045, + "step": 933 + }, + { + "epoch": 0.11230685985691095, + "flos": 25840876807680.0, + "grad_norm": 3.026373962611406, + "learning_rate": 3.929445392506423e-06, + "loss": 0.8204, + "num_input_tokens_seen": 19843985, + "step": 934 + }, + { + "epoch": 0.11242710274755005, + "flos": 22199562625440.0, + "grad_norm": 2.1878999616917145, + "learning_rate": 3.92924016773867e-06, + "loss": 0.7603, + "num_input_tokens_seen": 19860480, + "step": 935 + }, + { + "epoch": 0.11254734563818915, + "flos": 17687000341440.0, + "grad_norm": 2.5996885223148216, + "learning_rate": 3.9290346503067175e-06, + "loss": 0.7398, + "num_input_tokens_seen": 19877145, + "step": 936 + }, + { + "epoch": 0.11266758852882823, + "flos": 54851123157600.0, + "grad_norm": 2.4787258963824486, + "learning_rate": 3.9288288402417415e-06, + "loss": 0.788, + "num_input_tokens_seen": 19903405, + "step": 937 + }, + { + "epoch": 0.11278783141946733, + "flos": 18851192020800.0, + "grad_norm": 3.1504478378445646, + "learning_rate": 3.928622737574964e-06, + "loss": 0.7094, + "num_input_tokens_seen": 19918740, + "step": 938 + }, + { + "epoch": 0.11290807431010641, + "flos": 26431335843360.0, + "grad_norm": 2.529744395134005, + "learning_rate": 3.928416342337652e-06, + "loss": 0.9132, + "num_input_tokens_seen": 19938475, + "step": 939 + }, + { + "epoch": 0.1130283172007455, + "flos": 22678123482720.0, + "grad_norm": 2.052497886442892, + "learning_rate": 3.928209654561113e-06, + "loss": 0.829, + "num_input_tokens_seen": 19959310, + "step": 940 + }, + { + "epoch": 0.1131485600913846, + "flos": 23186648900160.0, + "grad_norm": 3.06231252913922, + "learning_rate": 3.928002674276703e-06, + "loss": 0.8139, + "num_input_tokens_seen": 19978700, + "step": 941 + }, + { + "epoch": 0.11326880298202369, + "flos": 14044422387360.0, + "grad_norm": 2.2801165965868417, + "learning_rate": 3.92779540151582e-06, + "loss": 0.757, + "num_input_tokens_seen": 19995785, + "step": 942 + }, + { + "epoch": 0.11338904587266278, + "flos": 16302208904640.0, + "grad_norm": 3.9558366363706607, + "learning_rate": 3.927587836309907e-06, + "loss": 0.8568, + "num_input_tokens_seen": 20013615, + "step": 943 + }, + { + "epoch": 0.11350928876330187, + "flos": 24391324181760.0, + "grad_norm": 4.185517710176038, + "learning_rate": 3.927379978690452e-06, + "loss": 0.7851, + "num_input_tokens_seen": 20032880, + "step": 944 + }, + { + "epoch": 0.11362953165394096, + "flos": 24461958192960.0, + "grad_norm": 2.383751455945679, + "learning_rate": 3.927171828688987e-06, + "loss": 0.8708, + "num_input_tokens_seen": 20052405, + "step": 945 + }, + { + "epoch": 0.11374977454458005, + "flos": 24026408481600.0, + "grad_norm": 3.2747636899067154, + "learning_rate": 3.926963386337088e-06, + "loss": 0.822, + "num_input_tokens_seen": 20070635, + "step": 946 + }, + { + "epoch": 0.11387001743521914, + "flos": 39400791176160.0, + "grad_norm": 2.2883751784478927, + "learning_rate": 3.926754651666375e-06, + "loss": 0.7042, + "num_input_tokens_seen": 20091195, + "step": 947 + }, + { + "epoch": 0.11399026032585824, + "flos": 25046544406560.0, + "grad_norm": 3.754501438491633, + "learning_rate": 3.926545624708513e-06, + "loss": 0.7857, + "num_input_tokens_seen": 20110435, + "step": 948 + }, + { + "epoch": 0.11411050321649732, + "flos": 17936077868640.0, + "grad_norm": 2.043765436050031, + "learning_rate": 3.926336305495213e-06, + "loss": 0.8556, + "num_input_tokens_seen": 20128995, + "step": 949 + }, + { + "epoch": 0.11423074610713642, + "flos": 22423321812480.0, + "grad_norm": 3.096527702189148, + "learning_rate": 3.926126694058226e-06, + "loss": 0.8915, + "num_input_tokens_seen": 20148145, + "step": 950 + }, + { + "epoch": 0.1143509889977755, + "flos": 19689539151360.0, + "grad_norm": 2.2707667443081125, + "learning_rate": 3.92591679042935e-06, + "loss": 0.8212, + "num_input_tokens_seen": 20168035, + "step": 951 + }, + { + "epoch": 0.1144712318884146, + "flos": 19794412245120.0, + "grad_norm": 1.8781140217642693, + "learning_rate": 3.92570659464043e-06, + "loss": 0.8234, + "num_input_tokens_seen": 20186095, + "step": 952 + }, + { + "epoch": 0.1145914747790537, + "flos": 14918049353760.0, + "grad_norm": 1.9665043451380986, + "learning_rate": 3.925496106723349e-06, + "loss": 0.7981, + "num_input_tokens_seen": 20203535, + "step": 953 + }, + { + "epoch": 0.11471171766969278, + "flos": 19836717165600.0, + "grad_norm": 3.1006868216855303, + "learning_rate": 3.9252853267100405e-06, + "loss": 0.8392, + "num_input_tokens_seen": 20222780, + "step": 954 + }, + { + "epoch": 0.11483196056033187, + "flos": 22494216012000.0, + "grad_norm": 1.898638679299906, + "learning_rate": 3.9250742546324786e-06, + "loss": 0.8383, + "num_input_tokens_seen": 20243015, + "step": 955 + }, + { + "epoch": 0.11495220345097096, + "flos": 28179816378240.0, + "grad_norm": 2.4367124801413205, + "learning_rate": 3.924862890522683e-06, + "loss": 0.8692, + "num_input_tokens_seen": 20263025, + "step": 956 + }, + { + "epoch": 0.11507244634161005, + "flos": 17467366997760.0, + "grad_norm": 3.1794952409435395, + "learning_rate": 3.9246512344127174e-06, + "loss": 0.8651, + "num_input_tokens_seen": 20280685, + "step": 957 + }, + { + "epoch": 0.11519268923224914, + "flos": 22532692447200.0, + "grad_norm": 6.379673460254267, + "learning_rate": 3.9244392863346895e-06, + "loss": 0.8187, + "num_input_tokens_seen": 20300090, + "step": 958 + }, + { + "epoch": 0.11531293212288823, + "flos": 16958730071040.0, + "grad_norm": 2.1112269431120536, + "learning_rate": 3.9242270463207524e-06, + "loss": 0.9228, + "num_input_tokens_seen": 20318480, + "step": 959 + }, + { + "epoch": 0.11543317501352733, + "flos": 12404531922240.0, + "grad_norm": 3.380086668359277, + "learning_rate": 3.924014514403102e-06, + "loss": 0.8524, + "num_input_tokens_seen": 20334635, + "step": 960 + }, + { + "epoch": 0.11555341790416641, + "flos": 19793148473280.0, + "grad_norm": 2.4685538491658816, + "learning_rate": 3.92380169061398e-06, + "loss": 0.9148, + "num_input_tokens_seen": 20352415, + "step": 961 + }, + { + "epoch": 0.11567366079480551, + "flos": 25701987649920.0, + "grad_norm": 1.9909946834966472, + "learning_rate": 3.9235885749856705e-06, + "loss": 0.8412, + "num_input_tokens_seen": 20371095, + "step": 962 + }, + { + "epoch": 0.1157939036854446, + "flos": 18197384246880.0, + "grad_norm": 2.1145411770632063, + "learning_rate": 3.9233751675505035e-06, + "loss": 0.8294, + "num_input_tokens_seen": 20389805, + "step": 963 + }, + { + "epoch": 0.11591414657608369, + "flos": 23040697488000.0, + "grad_norm": 2.1092178851354113, + "learning_rate": 3.923161468340853e-06, + "loss": 0.8492, + "num_input_tokens_seen": 20409640, + "step": 964 + }, + { + "epoch": 0.11603438946672277, + "flos": 19433919746400.0, + "grad_norm": 3.454923825783685, + "learning_rate": 3.9229474773891374e-06, + "loss": 0.8164, + "num_input_tokens_seen": 20428180, + "step": 965 + }, + { + "epoch": 0.11615463235736187, + "flos": 26793463811520.0, + "grad_norm": 4.1660108177275355, + "learning_rate": 3.922733194727818e-06, + "loss": 0.8373, + "num_input_tokens_seen": 20447975, + "step": 966 + }, + { + "epoch": 0.11627487524800097, + "flos": 18553602223200.0, + "grad_norm": 2.1129348668441907, + "learning_rate": 3.922518620389402e-06, + "loss": 0.8739, + "num_input_tokens_seen": 20464840, + "step": 967 + }, + { + "epoch": 0.11639511813864005, + "flos": 18124668729120.0, + "grad_norm": 2.3459271300983424, + "learning_rate": 3.922303754406439e-06, + "loss": 0.8961, + "num_input_tokens_seen": 20482680, + "step": 968 + }, + { + "epoch": 0.11651536102927915, + "flos": 20892132926400.0, + "grad_norm": 1.8900834876497532, + "learning_rate": 3.922088596811526e-06, + "loss": 0.7916, + "num_input_tokens_seen": 20501810, + "step": 969 + }, + { + "epoch": 0.11663560391991823, + "flos": 16485187131360.0, + "grad_norm": 2.529264448392525, + "learning_rate": 3.9218731476373e-06, + "loss": 0.8727, + "num_input_tokens_seen": 20517995, + "step": 970 + }, + { + "epoch": 0.11675584681055733, + "flos": 19836717165600.0, + "grad_norm": 2.2524269578206932, + "learning_rate": 3.9216574069164455e-06, + "loss": 0.8474, + "num_input_tokens_seen": 20536970, + "step": 971 + }, + { + "epoch": 0.11687608970119642, + "flos": 21913012246560.0, + "grad_norm": 1.5906307887363584, + "learning_rate": 3.921441374681691e-06, + "loss": 0.8036, + "num_input_tokens_seen": 20557030, + "step": 972 + }, + { + "epoch": 0.1169963325918355, + "flos": 24026743009440.0, + "grad_norm": 2.106970766166147, + "learning_rate": 3.921225050965808e-06, + "loss": 0.6538, + "num_input_tokens_seen": 20576915, + "step": 973 + }, + { + "epoch": 0.1171165754824746, + "flos": 23334756158400.0, + "grad_norm": 2.6399990445915575, + "learning_rate": 3.921008435801612e-06, + "loss": 0.7475, + "num_input_tokens_seen": 20596000, + "step": 974 + }, + { + "epoch": 0.11723681837311369, + "flos": 18525867848640.0, + "grad_norm": 2.2183607592799213, + "learning_rate": 3.920791529221963e-06, + "loss": 0.7585, + "num_input_tokens_seen": 20614675, + "step": 975 + }, + { + "epoch": 0.11735706126375278, + "flos": 23516879480640.0, + "grad_norm": 2.4578991738258607, + "learning_rate": 3.920574331259768e-06, + "loss": 0.7633, + "num_input_tokens_seen": 20635875, + "step": 976 + }, + { + "epoch": 0.11747730415439187, + "flos": 22349528371680.0, + "grad_norm": 3.2106416663445905, + "learning_rate": 3.9203568419479716e-06, + "loss": 0.7921, + "num_input_tokens_seen": 20656430, + "step": 977 + }, + { + "epoch": 0.11759754704503096, + "flos": 22169040518880.0, + "grad_norm": 2.560605535118975, + "learning_rate": 3.92013906131957e-06, + "loss": 0.7535, + "num_input_tokens_seen": 20675360, + "step": 978 + }, + { + "epoch": 0.11771778993567006, + "flos": 22277890776960.0, + "grad_norm": 2.6020104191635256, + "learning_rate": 3.9199209894076e-06, + "loss": 0.8262, + "num_input_tokens_seen": 20695675, + "step": 979 + }, + { + "epoch": 0.11783803282630914, + "flos": 21257792021760.0, + "grad_norm": 2.8781952724099606, + "learning_rate": 3.919702626245142e-06, + "loss": 0.903, + "num_input_tokens_seen": 20715330, + "step": 980 + }, + { + "epoch": 0.11795827571694824, + "flos": 25335101952480.0, + "grad_norm": 2.1736379370559145, + "learning_rate": 3.919483971865322e-06, + "loss": 0.6619, + "num_input_tokens_seen": 20735325, + "step": 981 + }, + { + "epoch": 0.11807851860758732, + "flos": 23588405566080.0, + "grad_norm": 2.004190108809176, + "learning_rate": 3.91926502630131e-06, + "loss": 0.8793, + "num_input_tokens_seen": 20755940, + "step": 982 + }, + { + "epoch": 0.11819876149822642, + "flos": 24936727734720.0, + "grad_norm": 2.8878913118827807, + "learning_rate": 3.91904578958632e-06, + "loss": 0.7243, + "num_input_tokens_seen": 20775355, + "step": 983 + }, + { + "epoch": 0.11831900438886551, + "flos": 22970360834880.0, + "grad_norm": 2.4541574237262997, + "learning_rate": 3.918826261753608e-06, + "loss": 0.8406, + "num_input_tokens_seen": 20794935, + "step": 984 + }, + { + "epoch": 0.1184392472795046, + "flos": 27925051877760.0, + "grad_norm": 2.6557631102479577, + "learning_rate": 3.918606442836478e-06, + "loss": 0.7086, + "num_input_tokens_seen": 20817155, + "step": 985 + }, + { + "epoch": 0.1185594901701437, + "flos": 19870027004160.0, + "grad_norm": 2.022792168298051, + "learning_rate": 3.918386332868277e-06, + "loss": 0.7777, + "num_input_tokens_seen": 20835125, + "step": 986 + }, + { + "epoch": 0.11867973306078278, + "flos": 18885282424320.0, + "grad_norm": 1.9069237690868501, + "learning_rate": 3.918165931882394e-06, + "loss": 0.9446, + "num_input_tokens_seen": 20853525, + "step": 987 + }, + { + "epoch": 0.11879997595142187, + "flos": 16958692901280.0, + "grad_norm": 2.4378257924991225, + "learning_rate": 3.917945239912264e-06, + "loss": 0.7553, + "num_input_tokens_seen": 20871000, + "step": 988 + }, + { + "epoch": 0.11892021884206096, + "flos": 17505471735360.0, + "grad_norm": 2.3655886648554785, + "learning_rate": 3.917724256991367e-06, + "loss": 0.7569, + "num_input_tokens_seen": 20889745, + "step": 989 + }, + { + "epoch": 0.11904046173270005, + "flos": 30911406023520.0, + "grad_norm": 2.4058378783199488, + "learning_rate": 3.9175029831532245e-06, + "loss": 0.8172, + "num_input_tokens_seen": 20909260, + "step": 990 + }, + { + "epoch": 0.11916070462333915, + "flos": 20128508480640.0, + "grad_norm": 2.3572756288134302, + "learning_rate": 3.917281418431404e-06, + "loss": 0.8837, + "num_input_tokens_seen": 20928305, + "step": 991 + }, + { + "epoch": 0.11928094751397823, + "flos": 23517474196800.0, + "grad_norm": 2.323448245774684, + "learning_rate": 3.917059562859516e-06, + "loss": 0.7685, + "num_input_tokens_seen": 20947630, + "step": 992 + }, + { + "epoch": 0.11940119040461733, + "flos": 23873989531200.0, + "grad_norm": 2.3023403823978064, + "learning_rate": 3.916837416471218e-06, + "loss": 0.8926, + "num_input_tokens_seen": 20964250, + "step": 993 + }, + { + "epoch": 0.11952143329525641, + "flos": 13824974892480.0, + "grad_norm": 2.3242931193787197, + "learning_rate": 3.916614979300207e-06, + "loss": 0.724, + "num_input_tokens_seen": 20979095, + "step": 994 + }, + { + "epoch": 0.11964167618589551, + "flos": 26976776566080.0, + "grad_norm": 11.626398474140313, + "learning_rate": 3.9163922513802274e-06, + "loss": 0.7885, + "num_input_tokens_seen": 21001430, + "step": 995 + }, + { + "epoch": 0.1197619190765346, + "flos": 12549925788000.0, + "grad_norm": 2.8021933003418, + "learning_rate": 3.916169232745067e-06, + "loss": 0.8247, + "num_input_tokens_seen": 21019105, + "step": 996 + }, + { + "epoch": 0.11988216196717369, + "flos": 16885419837120.0, + "grad_norm": 2.98984780731205, + "learning_rate": 3.915945923428559e-06, + "loss": 0.9188, + "num_input_tokens_seen": 21035630, + "step": 997 + }, + { + "epoch": 0.12000240485781279, + "flos": 16193581665120.0, + "grad_norm": 2.3327271718633558, + "learning_rate": 3.915722323464577e-06, + "loss": 0.8302, + "num_input_tokens_seen": 21054235, + "step": 998 + }, + { + "epoch": 0.12012264774845187, + "flos": 49273629654240.0, + "grad_norm": 2.8554927965813754, + "learning_rate": 3.91549843288704e-06, + "loss": 0.7068, + "num_input_tokens_seen": 21077915, + "step": 999 + }, + { + "epoch": 0.12024289063909097, + "flos": 26940753335040.0, + "grad_norm": 2.582002994538572, + "learning_rate": 3.915274251729916e-06, + "loss": 0.792, + "num_input_tokens_seen": 21099205, + "step": 1000 + }, + { + "epoch": 0.12036313352973005, + "flos": 19509497335680.0, + "grad_norm": 2.0373907141835144, + "learning_rate": 3.91504978002721e-06, + "loss": 0.9027, + "num_input_tokens_seen": 21118260, + "step": 1001 + }, + { + "epoch": 0.12048337642036915, + "flos": 17243496301440.0, + "grad_norm": 2.135009416653238, + "learning_rate": 3.914825017812974e-06, + "loss": 0.7681, + "num_input_tokens_seen": 21134910, + "step": 1002 + }, + { + "epoch": 0.12060361931100824, + "flos": 22824297913440.0, + "grad_norm": 2.53599470994199, + "learning_rate": 3.9145999651213065e-06, + "loss": 0.7247, + "num_input_tokens_seen": 21154150, + "step": 1003 + }, + { + "epoch": 0.12072386220164733, + "flos": 16703036326560.0, + "grad_norm": 3.4170683802951314, + "learning_rate": 3.9143746219863465e-06, + "loss": 0.888, + "num_input_tokens_seen": 21171255, + "step": 1004 + }, + { + "epoch": 0.12084410509228642, + "flos": 55065732850080.0, + "grad_norm": 1.0122172773314633, + "learning_rate": 3.914148988442278e-06, + "loss": 0.6492, + "num_input_tokens_seen": 21227045, + "step": 1005 + }, + { + "epoch": 0.1209643479829255, + "flos": 26721863386560.0, + "grad_norm": 4.714702158447667, + "learning_rate": 3.91392306452333e-06, + "loss": 0.953, + "num_input_tokens_seen": 21244120, + "step": 1006 + }, + { + "epoch": 0.1210845908735646, + "flos": 11019554636640.0, + "grad_norm": 2.9910156505077055, + "learning_rate": 3.913696850263774e-06, + "loss": 0.6643, + "num_input_tokens_seen": 21258485, + "step": 1007 + }, + { + "epoch": 0.1212048337642037, + "flos": 20455393782720.0, + "grad_norm": 2.088789458253135, + "learning_rate": 3.913470345697929e-06, + "loss": 0.7955, + "num_input_tokens_seen": 21277115, + "step": 1008 + }, + { + "epoch": 0.12132507665484278, + "flos": 22314137026560.0, + "grad_norm": 2.217039711713119, + "learning_rate": 3.913243550860153e-06, + "loss": 0.8583, + "num_input_tokens_seen": 21294880, + "step": 1009 + }, + { + "epoch": 0.12144531954548188, + "flos": 29272890839520.0, + "grad_norm": 1.8934431903940592, + "learning_rate": 3.913016465784852e-06, + "loss": 0.7594, + "num_input_tokens_seen": 21315555, + "step": 1010 + }, + { + "epoch": 0.12156556243612096, + "flos": 20456025668640.0, + "grad_norm": 3.0269058403036677, + "learning_rate": 3.912789090506474e-06, + "loss": 0.7241, + "num_input_tokens_seen": 21334815, + "step": 1011 + }, + { + "epoch": 0.12168580532676006, + "flos": 16448643523680.0, + "grad_norm": 2.3443928154939915, + "learning_rate": 3.9125614250595114e-06, + "loss": 0.7207, + "num_input_tokens_seen": 21351025, + "step": 1012 + }, + { + "epoch": 0.12180604821739914, + "flos": 15319583001120.0, + "grad_norm": 2.9778215443856806, + "learning_rate": 3.912333469478502e-06, + "loss": 0.8926, + "num_input_tokens_seen": 21368990, + "step": 1013 + }, + { + "epoch": 0.12192629110803824, + "flos": 19290533047680.0, + "grad_norm": 2.1972932940966214, + "learning_rate": 3.912105223798025e-06, + "loss": 0.7781, + "num_input_tokens_seen": 21388410, + "step": 1014 + }, + { + "epoch": 0.12204653399867733, + "flos": 47657267647200.0, + "grad_norm": 0.9950376110061077, + "learning_rate": 3.9118766880527065e-06, + "loss": 0.6761, + "num_input_tokens_seen": 21442845, + "step": 1015 + }, + { + "epoch": 0.12216677688931642, + "flos": 18195971796000.0, + "grad_norm": 1.7533032712546999, + "learning_rate": 3.9116478622772145e-06, + "loss": 0.7405, + "num_input_tokens_seen": 21461420, + "step": 1016 + }, + { + "epoch": 0.12228701977995551, + "flos": 27486119718240.0, + "grad_norm": 1.7640512254893599, + "learning_rate": 3.911418746506261e-06, + "loss": 0.8849, + "num_input_tokens_seen": 21481550, + "step": 1017 + }, + { + "epoch": 0.1224072626705946, + "flos": 21766986494880.0, + "grad_norm": 1.9532938159464885, + "learning_rate": 3.911189340774604e-06, + "loss": 0.788, + "num_input_tokens_seen": 21501680, + "step": 1018 + }, + { + "epoch": 0.1225275055612337, + "flos": 20674209391680.0, + "grad_norm": 7.321619330390008, + "learning_rate": 3.910959645117043e-06, + "loss": 0.7946, + "num_input_tokens_seen": 21521015, + "step": 1019 + }, + { + "epoch": 0.12264774845187278, + "flos": 57663451255200.0, + "grad_norm": 0.8164375335686344, + "learning_rate": 3.910729659568423e-06, + "loss": 0.5668, + "num_input_tokens_seen": 21578200, + "step": 1020 + }, + { + "epoch": 0.12276799134251187, + "flos": 26358471646560.0, + "grad_norm": 1.9634652430656696, + "learning_rate": 3.9104993841636344e-06, + "loss": 0.826, + "num_input_tokens_seen": 21598770, + "step": 1021 + }, + { + "epoch": 0.12288823423315097, + "flos": 21033958495200.0, + "grad_norm": 6.683233022848696, + "learning_rate": 3.910268818937608e-06, + "loss": 0.8117, + "num_input_tokens_seen": 21616925, + "step": 1022 + }, + { + "epoch": 0.12300847712379005, + "flos": 12295124117760.0, + "grad_norm": 2.497210567684321, + "learning_rate": 3.9100379639253196e-06, + "loss": 0.8758, + "num_input_tokens_seen": 21633645, + "step": 1023 + }, + { + "epoch": 0.12312872001442915, + "flos": 16738390501920.0, + "grad_norm": 2.656928570408541, + "learning_rate": 3.909806819161791e-06, + "loss": 0.8652, + "num_input_tokens_seen": 21649120, + "step": 1024 + }, + { + "epoch": 0.12324896290506823, + "flos": 18378727004160.0, + "grad_norm": 1.982523677824034, + "learning_rate": 3.909575384682086e-06, + "loss": 0.8611, + "num_input_tokens_seen": 21668000, + "step": 1025 + }, + { + "epoch": 0.12336920579570733, + "flos": 18888441853920.0, + "grad_norm": 2.097481548059936, + "learning_rate": 3.9093436605213144e-06, + "loss": 0.6955, + "num_input_tokens_seen": 21688500, + "step": 1026 + }, + { + "epoch": 0.12348944868634643, + "flos": 23844210819840.0, + "grad_norm": 2.393173753984498, + "learning_rate": 3.909111646714627e-06, + "loss": 0.7967, + "num_input_tokens_seen": 21709345, + "step": 1027 + }, + { + "epoch": 0.12360969157698551, + "flos": 18999001920960.0, + "grad_norm": 2.1478956614323477, + "learning_rate": 3.9088793432972206e-06, + "loss": 0.7269, + "num_input_tokens_seen": 21728165, + "step": 1028 + }, + { + "epoch": 0.1237299344676246, + "flos": 13205963747520.0, + "grad_norm": 2.51444553186237, + "learning_rate": 3.908646750304336e-06, + "loss": 0.8189, + "num_input_tokens_seen": 21745850, + "step": 1029 + }, + { + "epoch": 0.12385017735826369, + "flos": 20456471705760.0, + "grad_norm": 1.7172424640831705, + "learning_rate": 3.908413867771257e-06, + "loss": 0.8757, + "num_input_tokens_seen": 21764760, + "step": 1030 + }, + { + "epoch": 0.12397042024890279, + "flos": 17323199734080.0, + "grad_norm": 1.7641528583695578, + "learning_rate": 3.908180695733311e-06, + "loss": 0.8082, + "num_input_tokens_seen": 21784250, + "step": 1031 + }, + { + "epoch": 0.12409066313954187, + "flos": 20383235811360.0, + "grad_norm": 2.0745289601115604, + "learning_rate": 3.907947234225871e-06, + "loss": 0.8263, + "num_input_tokens_seen": 21803260, + "step": 1032 + }, + { + "epoch": 0.12421090603018096, + "flos": 20706961683840.0, + "grad_norm": 2.0203981881435067, + "learning_rate": 3.907713483284352e-06, + "loss": 0.8744, + "num_input_tokens_seen": 21822735, + "step": 1033 + }, + { + "epoch": 0.12433114892082006, + "flos": 24463779511200.0, + "grad_norm": 2.7177588749946717, + "learning_rate": 3.907479442944216e-06, + "loss": 0.9779, + "num_input_tokens_seen": 21841140, + "step": 1034 + }, + { + "epoch": 0.12445139181145914, + "flos": 19654259315520.0, + "grad_norm": 2.167774748498905, + "learning_rate": 3.907245113240963e-06, + "loss": 0.923, + "num_input_tokens_seen": 21859260, + "step": 1035 + }, + { + "epoch": 0.12457163470209824, + "flos": 46355976692160.0, + "grad_norm": 2.125923853489339, + "learning_rate": 3.907010494210144e-06, + "loss": 0.7395, + "num_input_tokens_seen": 21881920, + "step": 1036 + }, + { + "epoch": 0.12469187759273732, + "flos": 20347249750080.0, + "grad_norm": 2.2062366776969187, + "learning_rate": 3.9067755858873495e-06, + "loss": 0.9245, + "num_input_tokens_seen": 21898720, + "step": 1037 + }, + { + "epoch": 0.12481212048337642, + "flos": 69125586421440.0, + "grad_norm": 0.8648636068736896, + "learning_rate": 3.906540388308214e-06, + "loss": 0.6285, + "num_input_tokens_seen": 21958305, + "step": 1038 + }, + { + "epoch": 0.12493236337401552, + "flos": 18197793114240.0, + "grad_norm": 2.169375071744003, + "learning_rate": 3.906304901508417e-06, + "loss": 0.8145, + "num_input_tokens_seen": 21977205, + "step": 1039 + }, + { + "epoch": 0.12505260626465461, + "flos": 30001235449440.0, + "grad_norm": 2.4973977850998197, + "learning_rate": 3.9060691255236835e-06, + "loss": 0.7623, + "num_input_tokens_seen": 21997770, + "step": 1040 + }, + { + "epoch": 0.1251728491552937, + "flos": 24399315680160.0, + "grad_norm": 2.6082834484786006, + "learning_rate": 3.905833060389778e-06, + "loss": 0.806, + "num_input_tokens_seen": 22021890, + "step": 1041 + }, + { + "epoch": 0.12529309204593278, + "flos": 27080720415840.0, + "grad_norm": 5.854120691012563, + "learning_rate": 3.905596706142513e-06, + "loss": 0.7828, + "num_input_tokens_seen": 22043300, + "step": 1042 + }, + { + "epoch": 0.12541333493657186, + "flos": 30729951756960.0, + "grad_norm": 4.299975355210762, + "learning_rate": 3.9053600628177435e-06, + "loss": 0.8643, + "num_input_tokens_seen": 22062910, + "step": 1043 + }, + { + "epoch": 0.12553357782721097, + "flos": 23625729738720.0, + "grad_norm": 3.0464596196223015, + "learning_rate": 3.905123130451367e-06, + "loss": 0.8517, + "num_input_tokens_seen": 22084010, + "step": 1044 + }, + { + "epoch": 0.12565382071785006, + "flos": 24828174834720.0, + "grad_norm": 2.647510827810535, + "learning_rate": 3.904885909079326e-06, + "loss": 0.7964, + "num_input_tokens_seen": 22102795, + "step": 1045 + }, + { + "epoch": 0.12577406360848914, + "flos": 21330284520960.0, + "grad_norm": 2.8443857677023723, + "learning_rate": 3.904648398737607e-06, + "loss": 0.7821, + "num_input_tokens_seen": 22121360, + "step": 1046 + }, + { + "epoch": 0.12589430649912825, + "flos": 36086064937920.0, + "grad_norm": 1.9638762926737583, + "learning_rate": 3.9044105994622406e-06, + "loss": 0.7822, + "num_input_tokens_seen": 22142505, + "step": 1047 + }, + { + "epoch": 0.12601454938976733, + "flos": 25301383246560.0, + "grad_norm": 2.295225420417303, + "learning_rate": 3.9041725112893005e-06, + "loss": 0.8194, + "num_input_tokens_seen": 22163730, + "step": 1048 + }, + { + "epoch": 0.12613479228040642, + "flos": 15538064082240.0, + "grad_norm": 4.583692686847684, + "learning_rate": 3.903934134254904e-06, + "loss": 0.751, + "num_input_tokens_seen": 22181520, + "step": 1049 + }, + { + "epoch": 0.1262550351710455, + "flos": 21439952513760.0, + "grad_norm": 2.596935476248467, + "learning_rate": 3.903695468395213e-06, + "loss": 0.8502, + "num_input_tokens_seen": 22199390, + "step": 1050 + }, + { + "epoch": 0.1263752780616846, + "flos": 31531346412480.0, + "grad_norm": 3.660443761845054, + "learning_rate": 3.903456513746434e-06, + "loss": 0.5556, + "num_input_tokens_seen": 22220365, + "step": 1051 + }, + { + "epoch": 0.1264955209523237, + "flos": 28726966909920.0, + "grad_norm": 2.121951440886541, + "learning_rate": 3.903217270344815e-06, + "loss": 0.8736, + "num_input_tokens_seen": 22240615, + "step": 1052 + }, + { + "epoch": 0.12661576384296278, + "flos": 29199134568480.0, + "grad_norm": 1.8630948083396242, + "learning_rate": 3.902977738226648e-06, + "loss": 0.8198, + "num_input_tokens_seen": 22261350, + "step": 1053 + }, + { + "epoch": 0.12673600673360189, + "flos": 20820606840960.0, + "grad_norm": 2.1112993293090097, + "learning_rate": 3.902737917428273e-06, + "loss": 0.915, + "num_input_tokens_seen": 22279515, + "step": 1054 + }, + { + "epoch": 0.12685624962424097, + "flos": 25227478296480.0, + "grad_norm": 2.112082545097878, + "learning_rate": 3.902497807986068e-06, + "loss": 0.8406, + "num_input_tokens_seen": 22299535, + "step": 1055 + }, + { + "epoch": 0.12697649251488005, + "flos": 27487643678400.0, + "grad_norm": 4.495174151516915, + "learning_rate": 3.902257409936458e-06, + "loss": 0.8402, + "num_input_tokens_seen": 22320300, + "step": 1056 + }, + { + "epoch": 0.12709673540551916, + "flos": 21221545772160.0, + "grad_norm": 1.939712491704911, + "learning_rate": 3.902016723315912e-06, + "loss": 0.8406, + "num_input_tokens_seen": 22338240, + "step": 1057 + }, + { + "epoch": 0.12721697829615825, + "flos": 25301420416320.0, + "grad_norm": 2.273580749822701, + "learning_rate": 3.901775748160941e-06, + "loss": 0.6897, + "num_input_tokens_seen": 22357420, + "step": 1058 + }, + { + "epoch": 0.12733722118679733, + "flos": 61855409926560.0, + "grad_norm": 0.7999242150650349, + "learning_rate": 3.901534484508101e-06, + "loss": 0.6089, + "num_input_tokens_seen": 22420095, + "step": 1059 + }, + { + "epoch": 0.1274574640774364, + "flos": 26937965603040.0, + "grad_norm": 2.095161531543981, + "learning_rate": 3.901292932393991e-06, + "loss": 0.7454, + "num_input_tokens_seen": 22438975, + "step": 1060 + }, + { + "epoch": 0.12757770696807552, + "flos": 22204543373280.0, + "grad_norm": 2.4820073199536923, + "learning_rate": 3.9010510918552555e-06, + "loss": 0.8493, + "num_input_tokens_seen": 22458050, + "step": 1061 + }, + { + "epoch": 0.1276979498587146, + "flos": 28506924698880.0, + "grad_norm": 8.031862703099781, + "learning_rate": 3.900808962928581e-06, + "loss": 0.7499, + "num_input_tokens_seen": 22475665, + "step": 1062 + }, + { + "epoch": 0.1278181927493537, + "flos": 17395580724000.0, + "grad_norm": 2.31633553871441, + "learning_rate": 3.900566545650698e-06, + "loss": 0.8961, + "num_input_tokens_seen": 22493335, + "step": 1063 + }, + { + "epoch": 0.1279384356399928, + "flos": 21108346652160.0, + "grad_norm": 2.5384466275308517, + "learning_rate": 3.900323840058381e-06, + "loss": 0.8203, + "num_input_tokens_seen": 22511045, + "step": 1064 + }, + { + "epoch": 0.12805867853063188, + "flos": 26538587801760.0, + "grad_norm": 3.842137192206357, + "learning_rate": 3.900080846188449e-06, + "loss": 0.8164, + "num_input_tokens_seen": 22529635, + "step": 1065 + }, + { + "epoch": 0.12817892142127096, + "flos": 16412917650720.0, + "grad_norm": 2.0369587419214663, + "learning_rate": 3.8998375640777625e-06, + "loss": 0.811, + "num_input_tokens_seen": 22547505, + "step": 1066 + }, + { + "epoch": 0.12829916431191005, + "flos": 60670217332800.0, + "grad_norm": 0.7060540754597862, + "learning_rate": 3.899593993763229e-06, + "loss": 0.5258, + "num_input_tokens_seen": 22608465, + "step": 1067 + }, + { + "epoch": 0.12841940720254916, + "flos": 29744054914560.0, + "grad_norm": 2.3945666959295333, + "learning_rate": 3.899350135281796e-06, + "loss": 0.8204, + "num_input_tokens_seen": 22629465, + "step": 1068 + }, + { + "epoch": 0.12853965009318824, + "flos": 25914447229920.0, + "grad_norm": 2.131213808171901, + "learning_rate": 3.8991059886704585e-06, + "loss": 0.7985, + "num_input_tokens_seen": 22648970, + "step": 1069 + }, + { + "epoch": 0.12865989298382732, + "flos": 30802555765440.0, + "grad_norm": 2.5551414902957736, + "learning_rate": 3.898861553966252e-06, + "loss": 0.8307, + "num_input_tokens_seen": 22668620, + "step": 1070 + }, + { + "epoch": 0.12878013587446643, + "flos": 25848719627040.0, + "grad_norm": 1.9099687043467133, + "learning_rate": 3.898616831206257e-06, + "loss": 0.8812, + "num_input_tokens_seen": 22689045, + "step": 1071 + }, + { + "epoch": 0.12890037876510552, + "flos": 23299848020160.0, + "grad_norm": 2.248664565613066, + "learning_rate": 3.8983718204276e-06, + "loss": 0.7711, + "num_input_tokens_seen": 22711265, + "step": 1072 + }, + { + "epoch": 0.1290206216557446, + "flos": 23553534597600.0, + "grad_norm": 2.1357363825938487, + "learning_rate": 3.898126521667446e-06, + "loss": 0.826, + "num_input_tokens_seen": 22731420, + "step": 1073 + }, + { + "epoch": 0.12914086454638368, + "flos": 24136039304640.0, + "grad_norm": 1.7396072644930878, + "learning_rate": 3.897880934963007e-06, + "loss": 0.8365, + "num_input_tokens_seen": 22750970, + "step": 1074 + }, + { + "epoch": 0.1292611074370228, + "flos": 20238287982720.0, + "grad_norm": 2.1864185384757935, + "learning_rate": 3.89763506035154e-06, + "loss": 0.7872, + "num_input_tokens_seen": 22768820, + "step": 1075 + }, + { + "epoch": 0.12938135032766188, + "flos": 27338309818080.0, + "grad_norm": 1.9172975566087813, + "learning_rate": 3.897388897870343e-06, + "loss": 0.8139, + "num_input_tokens_seen": 22789460, + "step": 1076 + }, + { + "epoch": 0.12950159321830096, + "flos": 29235046290240.0, + "grad_norm": 2.0084429393105268, + "learning_rate": 3.89714244755676e-06, + "loss": 0.752, + "num_input_tokens_seen": 22810820, + "step": 1077 + }, + { + "epoch": 0.12962183610894007, + "flos": 24499765572480.0, + "grad_norm": 2.9860905215319917, + "learning_rate": 3.896895709448175e-06, + "loss": 0.8643, + "num_input_tokens_seen": 22830570, + "step": 1078 + }, + { + "epoch": 0.12974207899957915, + "flos": 11200153998720.0, + "grad_norm": 2.682991504418298, + "learning_rate": 3.896648683582019e-06, + "loss": 0.773, + "num_input_tokens_seen": 22845905, + "step": 1079 + }, + { + "epoch": 0.12986232189021824, + "flos": 24682520780640.0, + "grad_norm": 2.8448779206322334, + "learning_rate": 3.896401369995766e-06, + "loss": 0.8091, + "num_input_tokens_seen": 22865710, + "step": 1080 + }, + { + "epoch": 0.12998256478085732, + "flos": 23881163294880.0, + "grad_norm": 2.4490906409627566, + "learning_rate": 3.896153768726932e-06, + "loss": 0.7955, + "num_input_tokens_seen": 22886020, + "step": 1081 + }, + { + "epoch": 0.13010280767149643, + "flos": 18597691292160.0, + "grad_norm": 3.881439273169017, + "learning_rate": 3.8959058798130806e-06, + "loss": 0.8775, + "num_input_tokens_seen": 22903995, + "step": 1082 + }, + { + "epoch": 0.1302230505621355, + "flos": 22751433716640.0, + "grad_norm": 2.4891965131967857, + "learning_rate": 3.895657703291814e-06, + "loss": 0.7517, + "num_input_tokens_seen": 22924100, + "step": 1083 + }, + { + "epoch": 0.1303432934527746, + "flos": 21293146197120.0, + "grad_norm": 3.9742706829357606, + "learning_rate": 3.895409239200781e-06, + "loss": 0.7968, + "num_input_tokens_seen": 22939985, + "step": 1084 + }, + { + "epoch": 0.1304635363434137, + "flos": 20892616133280.0, + "grad_norm": 2.356716856505219, + "learning_rate": 3.895160487577673e-06, + "loss": 0.9132, + "num_input_tokens_seen": 22957555, + "step": 1085 + }, + { + "epoch": 0.1305837792340528, + "flos": 63155480013120.0, + "grad_norm": 0.7888104631128499, + "learning_rate": 3.894911448460226e-06, + "loss": 0.6085, + "num_input_tokens_seen": 23016870, + "step": 1086 + }, + { + "epoch": 0.13070402212469187, + "flos": 26390852241120.0, + "grad_norm": 2.5639621935837806, + "learning_rate": 3.8946621218862195e-06, + "loss": 0.7296, + "num_input_tokens_seen": 23037870, + "step": 1087 + }, + { + "epoch": 0.13082426501533098, + "flos": 27633743769600.0, + "grad_norm": 2.137095527973613, + "learning_rate": 3.894412507893475e-06, + "loss": 0.8924, + "num_input_tokens_seen": 23058150, + "step": 1088 + }, + { + "epoch": 0.13094450790597006, + "flos": 24791222359680.0, + "grad_norm": 2.4658276608801395, + "learning_rate": 3.894162606519859e-06, + "loss": 0.7229, + "num_input_tokens_seen": 23077100, + "step": 1089 + }, + { + "epoch": 0.13106475079660915, + "flos": 19035471189120.0, + "grad_norm": 2.0502330228904797, + "learning_rate": 3.893912417803282e-06, + "loss": 0.7731, + "num_input_tokens_seen": 23096815, + "step": 1090 + }, + { + "epoch": 0.13118499368724823, + "flos": 28871691720000.0, + "grad_norm": 2.2276289677987826, + "learning_rate": 3.8936619417816975e-06, + "loss": 0.7647, + "num_input_tokens_seen": 23117145, + "step": 1091 + }, + { + "epoch": 0.13130523657788734, + "flos": 14263237996320.0, + "grad_norm": 4.747744978640085, + "learning_rate": 3.8934111784931015e-06, + "loss": 0.7196, + "num_input_tokens_seen": 23134595, + "step": 1092 + }, + { + "epoch": 0.13142547946852642, + "flos": 70073750223840.0, + "grad_norm": 0.9325385415035766, + "learning_rate": 3.893160127975535e-06, + "loss": 0.5909, + "num_input_tokens_seen": 23195285, + "step": 1093 + }, + { + "epoch": 0.1315457223591655, + "flos": 45739901958240.0, + "grad_norm": 2.4242360081551637, + "learning_rate": 3.8929087902670826e-06, + "loss": 0.8171, + "num_input_tokens_seen": 23216915, + "step": 1094 + }, + { + "epoch": 0.13166596524980462, + "flos": 62791530726720.0, + "grad_norm": 1.1110448493008274, + "learning_rate": 3.8926571654058715e-06, + "loss": 0.6073, + "num_input_tokens_seen": 23273560, + "step": 1095 + }, + { + "epoch": 0.1317862081404437, + "flos": 23552791202400.0, + "grad_norm": 2.51443813592705, + "learning_rate": 3.892405253430074e-06, + "loss": 0.7689, + "num_input_tokens_seen": 23291080, + "step": 1096 + }, + { + "epoch": 0.13190645103108278, + "flos": 20231671765440.0, + "grad_norm": 2.079134817650318, + "learning_rate": 3.892153054377904e-06, + "loss": 0.826, + "num_input_tokens_seen": 23308485, + "step": 1097 + }, + { + "epoch": 0.13202669392172187, + "flos": 53379077093280.0, + "grad_norm": 0.9348550360800287, + "learning_rate": 3.891900568287619e-06, + "loss": 0.5943, + "num_input_tokens_seen": 23360870, + "step": 1098 + }, + { + "epoch": 0.13214693681236098, + "flos": 15828814644000.0, + "grad_norm": 2.569363218747695, + "learning_rate": 3.891647795197523e-06, + "loss": 0.7205, + "num_input_tokens_seen": 23376910, + "step": 1099 + }, + { + "epoch": 0.13226717970300006, + "flos": 19326258920640.0, + "grad_norm": 5.9703383005159285, + "learning_rate": 3.8913947351459605e-06, + "loss": 0.688, + "num_input_tokens_seen": 23394450, + "step": 1100 + }, + { + "epoch": 0.13238742259363914, + "flos": 20668336569600.0, + "grad_norm": 3.525781498847072, + "learning_rate": 3.89114138817132e-06, + "loss": 0.6781, + "num_input_tokens_seen": 23411115, + "step": 1101 + }, + { + "epoch": 0.13250766548427825, + "flos": 21002432805120.0, + "grad_norm": 1.933510641758776, + "learning_rate": 3.890887754312035e-06, + "loss": 0.8436, + "num_input_tokens_seen": 23429360, + "step": 1102 + }, + { + "epoch": 0.13262790837491734, + "flos": 22605519474240.0, + "grad_norm": 1.967734059585835, + "learning_rate": 3.890633833606581e-06, + "loss": 0.8803, + "num_input_tokens_seen": 23449050, + "step": 1103 + }, + { + "epoch": 0.13274815126555642, + "flos": 19655114220000.0, + "grad_norm": 2.0808620927641464, + "learning_rate": 3.890379626093477e-06, + "loss": 0.6991, + "num_input_tokens_seen": 23468800, + "step": 1104 + }, + { + "epoch": 0.1328683941561955, + "flos": 21287013186720.0, + "grad_norm": 2.4452582892603165, + "learning_rate": 3.890125131811287e-06, + "loss": 0.9218, + "num_input_tokens_seen": 23486850, + "step": 1105 + }, + { + "epoch": 0.1329886370468346, + "flos": 13679804045280.0, + "grad_norm": 2.2140965622655515, + "learning_rate": 3.889870350798618e-06, + "loss": 0.7545, + "num_input_tokens_seen": 23504195, + "step": 1106 + }, + { + "epoch": 0.1331088799374737, + "flos": 21002395635360.0, + "grad_norm": 2.189916420478284, + "learning_rate": 3.889615283094119e-06, + "loss": 0.7865, + "num_input_tokens_seen": 23523385, + "step": 1107 + }, + { + "epoch": 0.13322912282811278, + "flos": 18234485400960.0, + "grad_norm": 2.6097107631270933, + "learning_rate": 3.889359928736485e-06, + "loss": 0.8476, + "num_input_tokens_seen": 23541330, + "step": 1108 + }, + { + "epoch": 0.1333493657187519, + "flos": 24425823452640.0, + "grad_norm": 2.464250641330366, + "learning_rate": 3.889104287764451e-06, + "loss": 0.9152, + "num_input_tokens_seen": 23561705, + "step": 1109 + }, + { + "epoch": 0.13346960860939097, + "flos": 22127107296000.0, + "grad_norm": 2.026799887913094, + "learning_rate": 3.888848360216798e-06, + "loss": 0.9063, + "num_input_tokens_seen": 23579350, + "step": 1110 + }, + { + "epoch": 0.13358985150003005, + "flos": 67834771605120.0, + "grad_norm": 0.838564088572976, + "learning_rate": 3.888592146132351e-06, + "loss": 0.5662, + "num_input_tokens_seen": 23640620, + "step": 1111 + }, + { + "epoch": 0.13371009439066917, + "flos": 26795842676160.0, + "grad_norm": 2.0771800022210876, + "learning_rate": 3.888335645549978e-06, + "loss": 0.7825, + "num_input_tokens_seen": 23661040, + "step": 1112 + }, + { + "epoch": 0.13383033728130825, + "flos": 26285124242880.0, + "grad_norm": 2.4530826249076942, + "learning_rate": 3.888078858508588e-06, + "loss": 0.8146, + "num_input_tokens_seen": 23680900, + "step": 1113 + }, + { + "epoch": 0.13395058017194733, + "flos": 22531651693920.0, + "grad_norm": 1.9921489038655542, + "learning_rate": 3.8878217850471365e-06, + "loss": 0.846, + "num_input_tokens_seen": 23700815, + "step": 1114 + }, + { + "epoch": 0.13407082306258641, + "flos": 25773922602720.0, + "grad_norm": 2.198991509685285, + "learning_rate": 3.887564425204621e-06, + "loss": 0.7415, + "num_input_tokens_seen": 23722500, + "step": 1115 + }, + { + "epoch": 0.13419106595322552, + "flos": 68240213810880.0, + "grad_norm": 0.8383078792214022, + "learning_rate": 3.887306779020083e-06, + "loss": 0.5449, + "num_input_tokens_seen": 23777155, + "step": 1116 + }, + { + "epoch": 0.1343113088438646, + "flos": 20419928098080.0, + "grad_norm": 2.7968946061965556, + "learning_rate": 3.887048846532608e-06, + "loss": 0.7088, + "num_input_tokens_seen": 23794130, + "step": 1117 + }, + { + "epoch": 0.1344315517345037, + "flos": 67293016422240.0, + "grad_norm": 0.7941530901244415, + "learning_rate": 3.8867906277813224e-06, + "loss": 0.5807, + "num_input_tokens_seen": 23852285, + "step": 1118 + }, + { + "epoch": 0.1345517946251428, + "flos": 40675728771360.0, + "grad_norm": 3.182794126477251, + "learning_rate": 3.886532122805399e-06, + "loss": 0.7387, + "num_input_tokens_seen": 23874765, + "step": 1119 + }, + { + "epoch": 0.13467203751578188, + "flos": 22784111669280.0, + "grad_norm": 2.2220209215966404, + "learning_rate": 3.886273331644053e-06, + "loss": 0.8992, + "num_input_tokens_seen": 23893635, + "step": 1120 + }, + { + "epoch": 0.13479228040642097, + "flos": 17067282971040.0, + "grad_norm": 2.2511287612930215, + "learning_rate": 3.886014254336542e-06, + "loss": 0.8225, + "num_input_tokens_seen": 23910110, + "step": 1121 + }, + { + "epoch": 0.13491252329706005, + "flos": 23696512428960.0, + "grad_norm": 2.1434318907025167, + "learning_rate": 3.885754890922168e-06, + "loss": 0.9281, + "num_input_tokens_seen": 23930440, + "step": 1122 + }, + { + "epoch": 0.13503276618769916, + "flos": 34077839154720.0, + "grad_norm": 1.9640911018273735, + "learning_rate": 3.885495241440277e-06, + "loss": 0.7837, + "num_input_tokens_seen": 23954535, + "step": 1123 + }, + { + "epoch": 0.13515300907833824, + "flos": 17687149020480.0, + "grad_norm": 1.9691188406198055, + "learning_rate": 3.885235305930257e-06, + "loss": 0.7401, + "num_input_tokens_seen": 23972735, + "step": 1124 + }, + { + "epoch": 0.13527325196897733, + "flos": 20231783274720.0, + "grad_norm": 1.9926683032702432, + "learning_rate": 3.884975084431539e-06, + "loss": 0.8554, + "num_input_tokens_seen": 23987685, + "step": 1125 + }, + { + "epoch": 0.13539349485961644, + "flos": 18160617620640.0, + "grad_norm": 2.5688452895559095, + "learning_rate": 3.8847145769836e-06, + "loss": 0.9224, + "num_input_tokens_seen": 24004105, + "step": 1126 + }, + { + "epoch": 0.13551373775025552, + "flos": 19289938331520.0, + "grad_norm": 10.112440820622187, + "learning_rate": 3.884453783625959e-06, + "loss": 0.669, + "num_input_tokens_seen": 24021715, + "step": 1127 + }, + { + "epoch": 0.1356339806408946, + "flos": 20821090047840.0, + "grad_norm": 2.6277720936755626, + "learning_rate": 3.884192704398176e-06, + "loss": 0.8482, + "num_input_tokens_seen": 24038915, + "step": 1128 + }, + { + "epoch": 0.13575422353153369, + "flos": 50403136213920.0, + "grad_norm": 1.8140465288009526, + "learning_rate": 3.883931339339858e-06, + "loss": 0.7458, + "num_input_tokens_seen": 24063300, + "step": 1129 + }, + { + "epoch": 0.1358744664221728, + "flos": 18125003256960.0, + "grad_norm": 2.1579412931529958, + "learning_rate": 3.883669688490654e-06, + "loss": 0.7899, + "num_input_tokens_seen": 24081070, + "step": 1130 + }, + { + "epoch": 0.13599470931281188, + "flos": 18416757402240.0, + "grad_norm": 2.259883788505247, + "learning_rate": 3.883407751890256e-06, + "loss": 0.8586, + "num_input_tokens_seen": 24099675, + "step": 1131 + }, + { + "epoch": 0.13611495220345096, + "flos": 26642271463200.0, + "grad_norm": 1.9449009594057953, + "learning_rate": 3.8831455295783994e-06, + "loss": 0.8578, + "num_input_tokens_seen": 24118925, + "step": 1132 + }, + { + "epoch": 0.13623519509409007, + "flos": 21655162656000.0, + "grad_norm": 2.1573199506225884, + "learning_rate": 3.882883021594864e-06, + "loss": 0.74, + "num_input_tokens_seen": 24137065, + "step": 1133 + }, + { + "epoch": 0.13635543798472916, + "flos": 14809422114240.0, + "grad_norm": 2.260102796664182, + "learning_rate": 3.8826202279794705e-06, + "loss": 0.8704, + "num_input_tokens_seen": 24154125, + "step": 1134 + }, + { + "epoch": 0.13647568087536824, + "flos": 22857496242720.0, + "grad_norm": 2.317195352655482, + "learning_rate": 3.882357148772085e-06, + "loss": 0.7063, + "num_input_tokens_seen": 24171550, + "step": 1135 + }, + { + "epoch": 0.13659592376600732, + "flos": 19909507022880.0, + "grad_norm": 3.6447862411215883, + "learning_rate": 3.882093784012617e-06, + "loss": 0.8431, + "num_input_tokens_seen": 24190190, + "step": 1136 + }, + { + "epoch": 0.13671616665664643, + "flos": 21397796272320.0, + "grad_norm": 1.9795404850808116, + "learning_rate": 3.881830133741019e-06, + "loss": 0.8442, + "num_input_tokens_seen": 24209055, + "step": 1137 + }, + { + "epoch": 0.13683640954728551, + "flos": 22745375045760.0, + "grad_norm": 3.0873244633369756, + "learning_rate": 3.881566197997285e-06, + "loss": 0.7628, + "num_input_tokens_seen": 24225490, + "step": 1138 + }, + { + "epoch": 0.1369566524379246, + "flos": 21694865693280.0, + "grad_norm": 2.0002167107475164, + "learning_rate": 3.881301976821456e-06, + "loss": 0.7514, + "num_input_tokens_seen": 24245520, + "step": 1139 + }, + { + "epoch": 0.1370768953285637, + "flos": 18597951480480.0, + "grad_norm": 2.3731899391031783, + "learning_rate": 3.881037470253612e-06, + "loss": 0.9061, + "num_input_tokens_seen": 24265035, + "step": 1140 + }, + { + "epoch": 0.1371971382192028, + "flos": 14918383881600.0, + "grad_norm": 2.7724631733766083, + "learning_rate": 3.88077267833388e-06, + "loss": 0.7924, + "num_input_tokens_seen": 24281550, + "step": 1141 + }, + { + "epoch": 0.13731738110984187, + "flos": 18996623056320.0, + "grad_norm": 2.1084622138918374, + "learning_rate": 3.880507601102427e-06, + "loss": 0.8386, + "num_input_tokens_seen": 24299485, + "step": 1142 + }, + { + "epoch": 0.13743762400048098, + "flos": 18161621204160.0, + "grad_norm": 2.900666153789882, + "learning_rate": 3.880242238599467e-06, + "loss": 0.8225, + "num_input_tokens_seen": 24316010, + "step": 1143 + }, + { + "epoch": 0.13755786689112007, + "flos": 21001949598240.0, + "grad_norm": 2.5843816496499437, + "learning_rate": 3.879976590865254e-06, + "loss": 0.8349, + "num_input_tokens_seen": 24335465, + "step": 1144 + }, + { + "epoch": 0.13767810978175915, + "flos": 21330135841920.0, + "grad_norm": 2.073042886491888, + "learning_rate": 3.879710657940087e-06, + "loss": 0.8767, + "num_input_tokens_seen": 24354415, + "step": 1145 + }, + { + "epoch": 0.13779835267239823, + "flos": 30548162962560.0, + "grad_norm": 3.985719222159974, + "learning_rate": 3.879444439864308e-06, + "loss": 0.7044, + "num_input_tokens_seen": 24373990, + "step": 1146 + }, + { + "epoch": 0.13791859556303734, + "flos": 22637305352640.0, + "grad_norm": 2.1354043603557997, + "learning_rate": 3.879177936678301e-06, + "loss": 0.8613, + "num_input_tokens_seen": 24392995, + "step": 1147 + }, + { + "epoch": 0.13803883845367643, + "flos": 35173441159680.0, + "grad_norm": 2.3668401985843417, + "learning_rate": 3.878911148422496e-06, + "loss": 0.7704, + "num_input_tokens_seen": 24414470, + "step": 1148 + }, + { + "epoch": 0.1381590813443155, + "flos": 31968531593280.0, + "grad_norm": 3.2471321363945225, + "learning_rate": 3.878644075137364e-06, + "loss": 0.7104, + "num_input_tokens_seen": 24435120, + "step": 1149 + }, + { + "epoch": 0.13827932423495462, + "flos": 17796519655200.0, + "grad_norm": 2.2810984521981084, + "learning_rate": 3.878376716863418e-06, + "loss": 0.7932, + "num_input_tokens_seen": 24451420, + "step": 1150 + }, + { + "epoch": 0.1383995671255937, + "flos": 19399420475520.0, + "grad_norm": 3.0540133851962743, + "learning_rate": 3.878109073641219e-06, + "loss": 0.7188, + "num_input_tokens_seen": 24470450, + "step": 1151 + }, + { + "epoch": 0.13851981001623279, + "flos": 28256137362720.0, + "grad_norm": 1.76321178700026, + "learning_rate": 3.877841145511366e-06, + "loss": 0.8121, + "num_input_tokens_seen": 24493190, + "step": 1152 + }, + { + "epoch": 0.13864005290687187, + "flos": 21182734809120.0, + "grad_norm": 1.8000774870194594, + "learning_rate": 3.8775729325145035e-06, + "loss": 0.8288, + "num_input_tokens_seen": 24512425, + "step": 1153 + }, + { + "epoch": 0.13876029579751098, + "flos": 71551551867360.0, + "grad_norm": 0.7905708098404953, + "learning_rate": 3.877304434691321e-06, + "loss": 0.6459, + "num_input_tokens_seen": 24579275, + "step": 1154 + }, + { + "epoch": 0.13888053868815006, + "flos": 21909889986720.0, + "grad_norm": 1.7991141921226115, + "learning_rate": 3.877035652082548e-06, + "loss": 0.7988, + "num_input_tokens_seen": 24598720, + "step": 1155 + }, + { + "epoch": 0.13900078157878915, + "flos": 19580651723520.0, + "grad_norm": 1.9103840261130178, + "learning_rate": 3.87676658472896e-06, + "loss": 0.8545, + "num_input_tokens_seen": 24617530, + "step": 1156 + }, + { + "epoch": 0.13912102446942826, + "flos": 22605705323040.0, + "grad_norm": 2.1305490857020155, + "learning_rate": 3.876497232671372e-06, + "loss": 0.8512, + "num_input_tokens_seen": 24637485, + "step": 1157 + }, + { + "epoch": 0.13924126736006734, + "flos": 29600779725120.0, + "grad_norm": 2.412569244655997, + "learning_rate": 3.876227595950647e-06, + "loss": 0.8338, + "num_input_tokens_seen": 24656915, + "step": 1158 + }, + { + "epoch": 0.13936151025070642, + "flos": 27378347383200.0, + "grad_norm": 2.1280650305455793, + "learning_rate": 3.875957674607686e-06, + "loss": 0.7881, + "num_input_tokens_seen": 24679190, + "step": 1159 + }, + { + "epoch": 0.1394817531413455, + "flos": 16376076684960.0, + "grad_norm": 1.867291371796177, + "learning_rate": 3.8756874686834386e-06, + "loss": 0.8803, + "num_input_tokens_seen": 24697405, + "step": 1160 + }, + { + "epoch": 0.13960199603198462, + "flos": 30877984675680.0, + "grad_norm": 2.0027417225383917, + "learning_rate": 3.875416978218893e-06, + "loss": 0.8042, + "num_input_tokens_seen": 24720600, + "step": 1161 + }, + { + "epoch": 0.1397222389226237, + "flos": 18088013612160.0, + "grad_norm": 2.6625968055856655, + "learning_rate": 3.8751462032550835e-06, + "loss": 0.8255, + "num_input_tokens_seen": 24737605, + "step": 1162 + }, + { + "epoch": 0.13984248181326278, + "flos": 16848801889920.0, + "grad_norm": 2.838387264053689, + "learning_rate": 3.874875143833085e-06, + "loss": 0.8354, + "num_input_tokens_seen": 24755845, + "step": 1163 + }, + { + "epoch": 0.1399627247039019, + "flos": 54043075115040.0, + "grad_norm": 2.2271405456826, + "learning_rate": 3.874603799994019e-06, + "loss": 0.6881, + "num_input_tokens_seen": 24779380, + "step": 1164 + }, + { + "epoch": 0.14008296759454097, + "flos": 11749014339360.0, + "grad_norm": 2.199993640717072, + "learning_rate": 3.874332171779046e-06, + "loss": 0.8728, + "num_input_tokens_seen": 24794260, + "step": 1165 + }, + { + "epoch": 0.14020321048518006, + "flos": 21986062292160.0, + "grad_norm": 1.7841220923671148, + "learning_rate": 3.874060259229373e-06, + "loss": 0.7573, + "num_input_tokens_seen": 24815835, + "step": 1166 + }, + { + "epoch": 0.14032345337581917, + "flos": 23370853728960.0, + "grad_norm": 2.636150835830159, + "learning_rate": 3.873788062386249e-06, + "loss": 0.935, + "num_input_tokens_seen": 24834095, + "step": 1167 + }, + { + "epoch": 0.14044369626645825, + "flos": 29603307268800.0, + "grad_norm": 2.6833378698970125, + "learning_rate": 3.873515581290965e-06, + "loss": 0.8222, + "num_input_tokens_seen": 24858900, + "step": 1168 + }, + { + "epoch": 0.14056393915709733, + "flos": 18306903560640.0, + "grad_norm": 2.6441841292102715, + "learning_rate": 3.8732428159848575e-06, + "loss": 0.7538, + "num_input_tokens_seen": 24877875, + "step": 1169 + }, + { + "epoch": 0.14068418204773642, + "flos": 26649259378080.0, + "grad_norm": 2.0435839104315265, + "learning_rate": 3.872969766509304e-06, + "loss": 0.7824, + "num_input_tokens_seen": 24898430, + "step": 1170 + }, + { + "epoch": 0.14080442493837553, + "flos": 65166456358560.0, + "grad_norm": 0.7626629186337526, + "learning_rate": 3.872696432905726e-06, + "loss": 0.5558, + "num_input_tokens_seen": 24959250, + "step": 1171 + }, + { + "epoch": 0.1409246678290146, + "flos": 25739832199200.0, + "grad_norm": 2.6635214516284473, + "learning_rate": 3.872422815215589e-06, + "loss": 0.7179, + "num_input_tokens_seen": 24978330, + "step": 1172 + }, + { + "epoch": 0.1410449107196537, + "flos": 21837174468960.0, + "grad_norm": 2.0678312583788507, + "learning_rate": 3.8721489134803994e-06, + "loss": 0.743, + "num_input_tokens_seen": 24997120, + "step": 1173 + }, + { + "epoch": 0.1411651536102928, + "flos": 16660211029440.0, + "grad_norm": 2.704984697732285, + "learning_rate": 3.871874727741707e-06, + "loss": 0.7264, + "num_input_tokens_seen": 25014350, + "step": 1174 + }, + { + "epoch": 0.1412853965009319, + "flos": 20962172221440.0, + "grad_norm": 2.0894093855328553, + "learning_rate": 3.871600258041108e-06, + "loss": 0.966, + "num_input_tokens_seen": 25033875, + "step": 1175 + }, + { + "epoch": 0.14140563939157097, + "flos": 20306134261920.0, + "grad_norm": 2.4333471476367414, + "learning_rate": 3.871325504420238e-06, + "loss": 0.8546, + "num_input_tokens_seen": 25052865, + "step": 1176 + }, + { + "epoch": 0.14152588228221005, + "flos": 21038753394240.0, + "grad_norm": 2.4416020429302683, + "learning_rate": 3.871050466920776e-06, + "loss": 0.8211, + "num_input_tokens_seen": 25072770, + "step": 1177 + }, + { + "epoch": 0.14164612517284916, + "flos": 18197235567840.0, + "grad_norm": 2.024965512017333, + "learning_rate": 3.870775145584447e-06, + "loss": 0.7966, + "num_input_tokens_seen": 25090550, + "step": 1178 + }, + { + "epoch": 0.14176636806348825, + "flos": 22711842188640.0, + "grad_norm": 4.79017653583945, + "learning_rate": 3.8704995404530145e-06, + "loss": 0.6463, + "num_input_tokens_seen": 25106175, + "step": 1179 + }, + { + "epoch": 0.14188661095412733, + "flos": 22059558374880.0, + "grad_norm": 1.8843809061289984, + "learning_rate": 3.87022365156829e-06, + "loss": 0.8536, + "num_input_tokens_seen": 25126890, + "step": 1180 + }, + { + "epoch": 0.14200685384476644, + "flos": 24317530740960.0, + "grad_norm": 2.20387861200019, + "learning_rate": 3.869947478972123e-06, + "loss": 0.8074, + "num_input_tokens_seen": 25147915, + "step": 1181 + }, + { + "epoch": 0.14212709673540552, + "flos": 23987597518560.0, + "grad_norm": 2.421058467916878, + "learning_rate": 3.869671022706412e-06, + "loss": 0.8255, + "num_input_tokens_seen": 25166645, + "step": 1182 + }, + { + "epoch": 0.1422473396260446, + "flos": 26393900161440.0, + "grad_norm": 1.9913515019813453, + "learning_rate": 3.869394282813092e-06, + "loss": 0.6523, + "num_input_tokens_seen": 25185845, + "step": 1183 + }, + { + "epoch": 0.1423675825166837, + "flos": 17031073891200.0, + "grad_norm": 4.58291094596518, + "learning_rate": 3.869117259334147e-06, + "loss": 0.8921, + "num_input_tokens_seen": 25203770, + "step": 1184 + }, + { + "epoch": 0.1424878254073228, + "flos": 17903920292640.0, + "grad_norm": 1.9022928658011184, + "learning_rate": 3.868839952311599e-06, + "loss": 0.8208, + "num_input_tokens_seen": 25221725, + "step": 1185 + }, + { + "epoch": 0.14260806829796188, + "flos": 20274682911360.0, + "grad_norm": 2.4985285243188073, + "learning_rate": 3.868562361787516e-06, + "loss": 0.8039, + "num_input_tokens_seen": 25240855, + "step": 1186 + }, + { + "epoch": 0.14272831118860096, + "flos": 23152595666400.0, + "grad_norm": 1.9388366518510618, + "learning_rate": 3.868284487804009e-06, + "loss": 0.6912, + "num_input_tokens_seen": 25260085, + "step": 1187 + }, + { + "epoch": 0.14284855407924008, + "flos": 27193064631360.0, + "grad_norm": 1.7616824803985434, + "learning_rate": 3.86800633040323e-06, + "loss": 0.7795, + "num_input_tokens_seen": 25280675, + "step": 1188 + }, + { + "epoch": 0.14296879696987916, + "flos": 28143793147200.0, + "grad_norm": 2.2733456307952085, + "learning_rate": 3.867727889627376e-06, + "loss": 0.7845, + "num_input_tokens_seen": 25300290, + "step": 1189 + }, + { + "epoch": 0.14308903986051824, + "flos": 19362653849280.0, + "grad_norm": 7.284554055204967, + "learning_rate": 3.867449165518687e-06, + "loss": 0.784, + "num_input_tokens_seen": 25316760, + "step": 1190 + }, + { + "epoch": 0.14320928275115732, + "flos": 17432719047840.0, + "grad_norm": 2.231602796343788, + "learning_rate": 3.867170158119444e-06, + "loss": 0.7116, + "num_input_tokens_seen": 25335760, + "step": 1191 + }, + { + "epoch": 0.14332952564179643, + "flos": 21435269124000.0, + "grad_norm": 2.4106489324036944, + "learning_rate": 3.866890867471972e-06, + "loss": 0.7582, + "num_input_tokens_seen": 25354230, + "step": 1192 + }, + { + "epoch": 0.14344976853243552, + "flos": 16373660650560.0, + "grad_norm": 2.544934803365569, + "learning_rate": 3.86661129361864e-06, + "loss": 0.8982, + "num_input_tokens_seen": 25368720, + "step": 1193 + }, + { + "epoch": 0.1435700114230746, + "flos": 18889817135040.0, + "grad_norm": 3.296888884745644, + "learning_rate": 3.866331436601859e-06, + "loss": 0.8656, + "num_input_tokens_seen": 25386715, + "step": 1194 + }, + { + "epoch": 0.1436902543137137, + "flos": 19727235021600.0, + "grad_norm": 2.5181968289770786, + "learning_rate": 3.866051296464083e-06, + "loss": 0.7385, + "num_input_tokens_seen": 25405550, + "step": 1195 + }, + { + "epoch": 0.1438104972043528, + "flos": 14664065418240.0, + "grad_norm": 2.3138036404105153, + "learning_rate": 3.86577087324781e-06, + "loss": 0.8564, + "num_input_tokens_seen": 25422040, + "step": 1196 + }, + { + "epoch": 0.14393074009499188, + "flos": 17068063536000.0, + "grad_norm": 3.0985977262235123, + "learning_rate": 3.865490166995578e-06, + "loss": 0.7779, + "num_input_tokens_seen": 25440270, + "step": 1197 + }, + { + "epoch": 0.144050982985631, + "flos": 30432213280320.0, + "grad_norm": 3.08734237948104, + "learning_rate": 3.86520917774997e-06, + "loss": 0.846, + "num_input_tokens_seen": 25459705, + "step": 1198 + }, + { + "epoch": 0.14417122587627007, + "flos": 17833100432640.0, + "grad_norm": 2.162669039607755, + "learning_rate": 3.864927905553614e-06, + "loss": 0.7479, + "num_input_tokens_seen": 25477615, + "step": 1199 + }, + { + "epoch": 0.14429146876690915, + "flos": 21582967514880.0, + "grad_norm": 3.710571270991098, + "learning_rate": 3.8646463504491765e-06, + "loss": 0.8917, + "num_input_tokens_seen": 25497750, + "step": 1200 + }, + { + "epoch": 0.14441171165754824, + "flos": 23224865147040.0, + "grad_norm": 1.9088284923150456, + "learning_rate": 3.8643645124793705e-06, + "loss": 0.8293, + "num_input_tokens_seen": 25516650, + "step": 1201 + }, + { + "epoch": 0.14453195454818735, + "flos": 42792544624320.0, + "grad_norm": 1.96194638747325, + "learning_rate": 3.8640823916869515e-06, + "loss": 0.7531, + "num_input_tokens_seen": 25540960, + "step": 1202 + }, + { + "epoch": 0.14465219743882643, + "flos": 27196855946880.0, + "grad_norm": 1.7356317451738852, + "learning_rate": 3.863799988114714e-06, + "loss": 0.7854, + "num_input_tokens_seen": 25562990, + "step": 1203 + }, + { + "epoch": 0.1447724403294655, + "flos": 16667013095520.0, + "grad_norm": 2.978093912256584, + "learning_rate": 3.863517301805502e-06, + "loss": 0.7071, + "num_input_tokens_seen": 25581380, + "step": 1204 + }, + { + "epoch": 0.14489268322010462, + "flos": 20044753544160.0, + "grad_norm": 2.6476903891254886, + "learning_rate": 3.863234332802196e-06, + "loss": 0.9676, + "num_input_tokens_seen": 25593585, + "step": 1205 + }, + { + "epoch": 0.1450129261107437, + "flos": 27086741916960.0, + "grad_norm": 5.051773953223055, + "learning_rate": 3.862951081147723e-06, + "loss": 0.7392, + "num_input_tokens_seen": 25613750, + "step": 1206 + }, + { + "epoch": 0.1451331690013828, + "flos": 25665518381760.0, + "grad_norm": 2.3528207853315526, + "learning_rate": 3.862667546885053e-06, + "loss": 0.7822, + "num_input_tokens_seen": 25632300, + "step": 1207 + }, + { + "epoch": 0.14525341189202187, + "flos": 25700463689760.0, + "grad_norm": 2.348832177089366, + "learning_rate": 3.8623837300571965e-06, + "loss": 0.7354, + "num_input_tokens_seen": 25651285, + "step": 1208 + }, + { + "epoch": 0.14537365478266098, + "flos": 23040846167040.0, + "grad_norm": 2.0288007788846176, + "learning_rate": 3.8620996307072085e-06, + "loss": 0.8402, + "num_input_tokens_seen": 25671990, + "step": 1209 + }, + { + "epoch": 0.14549389767330007, + "flos": 20565656491680.0, + "grad_norm": 2.0235178574324824, + "learning_rate": 3.861815248878188e-06, + "loss": 0.6426, + "num_input_tokens_seen": 25689475, + "step": 1210 + }, + { + "epoch": 0.14561414056393915, + "flos": 15100395694560.0, + "grad_norm": 2.567793890486496, + "learning_rate": 3.861530584613274e-06, + "loss": 0.8023, + "num_input_tokens_seen": 25707475, + "step": 1211 + }, + { + "epoch": 0.14573438345457826, + "flos": 19399829342880.0, + "grad_norm": 2.895497602216399, + "learning_rate": 3.86124563795565e-06, + "loss": 0.8258, + "num_input_tokens_seen": 25726290, + "step": 1212 + }, + { + "epoch": 0.14585462634521734, + "flos": 24792932168640.0, + "grad_norm": 2.038558605385568, + "learning_rate": 3.860960408948543e-06, + "loss": 0.7025, + "num_input_tokens_seen": 25748040, + "step": 1213 + }, + { + "epoch": 0.14597486923585642, + "flos": 15426686280480.0, + "grad_norm": 2.6486016056677255, + "learning_rate": 3.860674897635222e-06, + "loss": 0.8937, + "num_input_tokens_seen": 25764525, + "step": 1214 + }, + { + "epoch": 0.1460951121264955, + "flos": 16631547410880.0, + "grad_norm": 2.801186589447216, + "learning_rate": 3.860389104058998e-06, + "loss": 0.8355, + "num_input_tokens_seen": 25782755, + "step": 1215 + }, + { + "epoch": 0.14621535501713462, + "flos": 24827691627840.0, + "grad_norm": 2.4149958216787013, + "learning_rate": 3.860103028263227e-06, + "loss": 0.727, + "num_input_tokens_seen": 25805860, + "step": 1216 + }, + { + "epoch": 0.1463355979077737, + "flos": 25191863932800.0, + "grad_norm": 2.5197331783786487, + "learning_rate": 3.859816670291304e-06, + "loss": 0.6996, + "num_input_tokens_seen": 25824955, + "step": 1217 + }, + { + "epoch": 0.14645584079841278, + "flos": 22022643069600.0, + "grad_norm": 2.279143732805186, + "learning_rate": 3.859530030186672e-06, + "loss": 0.8987, + "num_input_tokens_seen": 25841965, + "step": 1218 + }, + { + "epoch": 0.1465760836890519, + "flos": 23590115375040.0, + "grad_norm": 2.9371808467123617, + "learning_rate": 3.859243107992813e-06, + "loss": 0.8255, + "num_input_tokens_seen": 25860770, + "step": 1219 + }, + { + "epoch": 0.14669632657969098, + "flos": 37353791599680.0, + "grad_norm": 2.6377671707573516, + "learning_rate": 3.858955903753252e-06, + "loss": 0.7865, + "num_input_tokens_seen": 25879410, + "step": 1220 + }, + { + "epoch": 0.14681656947033006, + "flos": 28325210244000.0, + "grad_norm": 1.618213051956264, + "learning_rate": 3.858668417511559e-06, + "loss": 0.8327, + "num_input_tokens_seen": 25902160, + "step": 1221 + }, + { + "epoch": 0.14693681236096917, + "flos": 18453226670400.0, + "grad_norm": 3.0743789728041677, + "learning_rate": 3.8583806493113445e-06, + "loss": 0.7664, + "num_input_tokens_seen": 25920505, + "step": 1222 + }, + { + "epoch": 0.14705705525160825, + "flos": 20748188681280.0, + "grad_norm": 2.512298184920013, + "learning_rate": 3.858092599196263e-06, + "loss": 0.8234, + "num_input_tokens_seen": 25938460, + "step": 1223 + }, + { + "epoch": 0.14717729814224734, + "flos": 29889894817440.0, + "grad_norm": 2.8384012076713976, + "learning_rate": 3.857804267210012e-06, + "loss": 0.8221, + "num_input_tokens_seen": 25957760, + "step": 1224 + }, + { + "epoch": 0.14729754103288642, + "flos": 20018914827360.0, + "grad_norm": 2.2528687308886246, + "learning_rate": 3.857515653396331e-06, + "loss": 0.8853, + "num_input_tokens_seen": 25974970, + "step": 1225 + }, + { + "epoch": 0.14741778392352553, + "flos": 19254063779520.0, + "grad_norm": 2.8200745204791855, + "learning_rate": 3.857226757799002e-06, + "loss": 0.8706, + "num_input_tokens_seen": 25992525, + "step": 1226 + }, + { + "epoch": 0.1475380268141646, + "flos": 25374730650240.0, + "grad_norm": 2.443127713400758, + "learning_rate": 3.85693758046185e-06, + "loss": 0.742, + "num_input_tokens_seen": 26010815, + "step": 1227 + }, + { + "epoch": 0.1476582697048037, + "flos": 20817819108960.0, + "grad_norm": 1.8434583142261918, + "learning_rate": 3.8566481214287435e-06, + "loss": 0.8285, + "num_input_tokens_seen": 26028935, + "step": 1228 + }, + { + "epoch": 0.1477785125954428, + "flos": 14008324816800.0, + "grad_norm": 2.1133800772582645, + "learning_rate": 3.8563583807435935e-06, + "loss": 0.9059, + "num_input_tokens_seen": 26045080, + "step": 1229 + }, + { + "epoch": 0.1478987554860819, + "flos": 20487254000640.0, + "grad_norm": 2.260864307244352, + "learning_rate": 3.856068358450353e-06, + "loss": 0.783, + "num_input_tokens_seen": 26065030, + "step": 1230 + }, + { + "epoch": 0.14801899837672097, + "flos": 17831576472480.0, + "grad_norm": 1.8814597434289682, + "learning_rate": 3.8557780545930186e-06, + "loss": 0.8602, + "num_input_tokens_seen": 26083445, + "step": 1231 + }, + { + "epoch": 0.14813924126736006, + "flos": 20851426305600.0, + "grad_norm": 1.997586701651579, + "learning_rate": 3.855487469215628e-06, + "loss": 0.7945, + "num_input_tokens_seen": 26102375, + "step": 1232 + }, + { + "epoch": 0.14825948415799917, + "flos": 37360184798400.0, + "grad_norm": 2.7745648738875075, + "learning_rate": 3.855196602362264e-06, + "loss": 0.722, + "num_input_tokens_seen": 26125295, + "step": 1233 + }, + { + "epoch": 0.14837972704863825, + "flos": 21983088711360.0, + "grad_norm": 2.4059680922208657, + "learning_rate": 3.854905454077051e-06, + "loss": 0.946, + "num_input_tokens_seen": 26142905, + "step": 1234 + }, + { + "epoch": 0.14849996993927733, + "flos": 20966372404320.0, + "grad_norm": 2.06089211414677, + "learning_rate": 3.854614024404155e-06, + "loss": 0.8822, + "num_input_tokens_seen": 26161215, + "step": 1235 + }, + { + "epoch": 0.14862021282991644, + "flos": 20019435204000.0, + "grad_norm": 2.12104429183821, + "learning_rate": 3.8543223133877865e-06, + "loss": 0.8919, + "num_input_tokens_seen": 26179810, + "step": 1236 + }, + { + "epoch": 0.14874045572055553, + "flos": 22679610273120.0, + "grad_norm": 1.93663036149357, + "learning_rate": 3.854030321072198e-06, + "loss": 0.8847, + "num_input_tokens_seen": 26199715, + "step": 1237 + }, + { + "epoch": 0.1488606986111946, + "flos": 25374916499040.0, + "grad_norm": 2.145551595123936, + "learning_rate": 3.853738047501682e-06, + "loss": 0.7362, + "num_input_tokens_seen": 26220275, + "step": 1238 + }, + { + "epoch": 0.1489809415018337, + "flos": 16994567453280.0, + "grad_norm": 2.0160074980005973, + "learning_rate": 3.85344549272058e-06, + "loss": 0.7773, + "num_input_tokens_seen": 26238255, + "step": 1239 + }, + { + "epoch": 0.1491011843924728, + "flos": 33611098281120.0, + "grad_norm": 1.851932408205485, + "learning_rate": 3.853152656773269e-06, + "loss": 0.8248, + "num_input_tokens_seen": 26259690, + "step": 1240 + }, + { + "epoch": 0.14922142728311188, + "flos": 21148793084640.0, + "grad_norm": 2.2981524385998595, + "learning_rate": 3.852859539704174e-06, + "loss": 0.8441, + "num_input_tokens_seen": 26278990, + "step": 1241 + }, + { + "epoch": 0.14934167017375097, + "flos": 29817625336800.0, + "grad_norm": 2.2521598833491456, + "learning_rate": 3.85256614155776e-06, + "loss": 0.7608, + "num_input_tokens_seen": 26299120, + "step": 1242 + }, + { + "epoch": 0.14946191306439008, + "flos": 16995050660160.0, + "grad_norm": 2.406786970194384, + "learning_rate": 3.852272462378535e-06, + "loss": 0.7427, + "num_input_tokens_seen": 26316995, + "step": 1243 + }, + { + "epoch": 0.14958215595502916, + "flos": 15646840000800.0, + "grad_norm": 2.0621088986505534, + "learning_rate": 3.85197850221105e-06, + "loss": 0.7812, + "num_input_tokens_seen": 26333295, + "step": 1244 + }, + { + "epoch": 0.14970239884566824, + "flos": 33060936998880.0, + "grad_norm": 1.8596597661812682, + "learning_rate": 3.851684261099899e-06, + "loss": 0.7584, + "num_input_tokens_seen": 26355035, + "step": 1245 + }, + { + "epoch": 0.14982264173630733, + "flos": 17795887769280.0, + "grad_norm": 2.3589113297754887, + "learning_rate": 3.851389739089718e-06, + "loss": 0.8682, + "num_input_tokens_seen": 26371655, + "step": 1246 + }, + { + "epoch": 0.14994288462694644, + "flos": 32362742797920.0, + "grad_norm": 2.084932578115092, + "learning_rate": 3.851094936225186e-06, + "loss": 0.8045, + "num_input_tokens_seen": 26392540, + "step": 1247 + }, + { + "epoch": 0.15006312751758552, + "flos": 31748935419360.0, + "grad_norm": 1.7014475970184273, + "learning_rate": 3.850799852551024e-06, + "loss": 0.7678, + "num_input_tokens_seen": 26414960, + "step": 1248 + }, + { + "epoch": 0.1501833704082246, + "flos": 16594446256800.0, + "grad_norm": 3.5777435765760863, + "learning_rate": 3.850504488111995e-06, + "loss": 0.8629, + "num_input_tokens_seen": 26431635, + "step": 1249 + }, + { + "epoch": 0.15030361329886371, + "flos": 23437399066560.0, + "grad_norm": 2.2037275856509653, + "learning_rate": 3.850208842952907e-06, + "loss": 0.8237, + "num_input_tokens_seen": 26450440, + "step": 1250 + }, + { + "epoch": 0.1504238561895028, + "flos": 25592691354720.0, + "grad_norm": 37.356711924114116, + "learning_rate": 3.849912917118608e-06, + "loss": 0.7966, + "num_input_tokens_seen": 26471480, + "step": 1251 + }, + { + "epoch": 0.15054409908014188, + "flos": 52021716939360.0, + "grad_norm": 0.8765590946450954, + "learning_rate": 3.849616710653992e-06, + "loss": 0.5927, + "num_input_tokens_seen": 26532950, + "step": 1252 + }, + { + "epoch": 0.150664341970781, + "flos": 18853385036640.0, + "grad_norm": 2.168945642271193, + "learning_rate": 3.84932022360399e-06, + "loss": 0.7524, + "num_input_tokens_seen": 26551615, + "step": 1253 + }, + { + "epoch": 0.15078458486142007, + "flos": 22131567667200.0, + "grad_norm": 2.9633967269736243, + "learning_rate": 3.849023456013581e-06, + "loss": 0.8468, + "num_input_tokens_seen": 26569175, + "step": 1254 + }, + { + "epoch": 0.15090482775205916, + "flos": 26613719353920.0, + "grad_norm": 2.4867829942808704, + "learning_rate": 3.848726407927784e-06, + "loss": 0.6254, + "num_input_tokens_seen": 26589560, + "step": 1255 + }, + { + "epoch": 0.15102507064269824, + "flos": 21767878569120.0, + "grad_norm": 2.970482250582452, + "learning_rate": 3.84842907939166e-06, + "loss": 0.8665, + "num_input_tokens_seen": 26608785, + "step": 1256 + }, + { + "epoch": 0.15114531353333735, + "flos": 22788311852160.0, + "grad_norm": 2.861800702228816, + "learning_rate": 3.8481314704503146e-06, + "loss": 0.7123, + "num_input_tokens_seen": 26628615, + "step": 1257 + }, + { + "epoch": 0.15126555642397643, + "flos": 19654631013120.0, + "grad_norm": 3.127981061034174, + "learning_rate": 3.847833581148895e-06, + "loss": 0.8829, + "num_input_tokens_seen": 26647525, + "step": 1258 + }, + { + "epoch": 0.15138579931461552, + "flos": 28684624819680.0, + "grad_norm": 3.399756866064403, + "learning_rate": 3.84753541153259e-06, + "loss": 0.8076, + "num_input_tokens_seen": 26667095, + "step": 1259 + }, + { + "epoch": 0.15150604220525463, + "flos": 22095395757120.0, + "grad_norm": 1.617085034353883, + "learning_rate": 3.847236961646633e-06, + "loss": 0.8307, + "num_input_tokens_seen": 26686075, + "step": 1260 + }, + { + "epoch": 0.1516262850958937, + "flos": 12950418682080.0, + "grad_norm": 3.4136384399466855, + "learning_rate": 3.846938231536296e-06, + "loss": 0.7826, + "num_input_tokens_seen": 26701695, + "step": 1261 + }, + { + "epoch": 0.1517465279865328, + "flos": 21765945741600.0, + "grad_norm": 2.6318022377248735, + "learning_rate": 3.8466392212468995e-06, + "loss": 0.8141, + "num_input_tokens_seen": 26721885, + "step": 1262 + }, + { + "epoch": 0.15186677087717187, + "flos": 58091790033120.0, + "grad_norm": 0.8210718908766717, + "learning_rate": 3.8463399308238e-06, + "loss": 0.6188, + "num_input_tokens_seen": 26780990, + "step": 1263 + }, + { + "epoch": 0.15198701376781099, + "flos": 32622673895040.0, + "grad_norm": 1.8302706654818406, + "learning_rate": 3.846040360312402e-06, + "loss": 0.6429, + "num_input_tokens_seen": 26804250, + "step": 1264 + }, + { + "epoch": 0.15210725665845007, + "flos": 28361828191200.0, + "grad_norm": 2.157863731798133, + "learning_rate": 3.8457405097581485e-06, + "loss": 0.8067, + "num_input_tokens_seen": 26823240, + "step": 1265 + }, + { + "epoch": 0.15222749954908915, + "flos": 19909767211200.0, + "grad_norm": 2.065034522814042, + "learning_rate": 3.8454403792065275e-06, + "loss": 0.7789, + "num_input_tokens_seen": 26842060, + "step": 1266 + }, + { + "epoch": 0.15234774243972826, + "flos": 21293778083040.0, + "grad_norm": 2.241787652174355, + "learning_rate": 3.845139968703068e-06, + "loss": 0.8547, + "num_input_tokens_seen": 26859580, + "step": 1267 + }, + { + "epoch": 0.15246798533036734, + "flos": 25920617410080.0, + "grad_norm": 3.025877921335264, + "learning_rate": 3.844839278293342e-06, + "loss": 0.8338, + "num_input_tokens_seen": 26880430, + "step": 1268 + }, + { + "epoch": 0.15258822822100643, + "flos": 25774926186240.0, + "grad_norm": 3.6304894023362846, + "learning_rate": 3.8445383080229654e-06, + "loss": 0.7725, + "num_input_tokens_seen": 26896445, + "step": 1269 + }, + { + "epoch": 0.1527084711116455, + "flos": 25229299614720.0, + "grad_norm": 2.46050168977155, + "learning_rate": 3.844237057937593e-06, + "loss": 0.7386, + "num_input_tokens_seen": 26915450, + "step": 1270 + }, + { + "epoch": 0.15282871400228462, + "flos": 29198205324480.0, + "grad_norm": 3.044455795390679, + "learning_rate": 3.843935528082926e-06, + "loss": 0.7826, + "num_input_tokens_seen": 26936475, + "step": 1271 + }, + { + "epoch": 0.1529489568929237, + "flos": 20853024605280.0, + "grad_norm": 2.128890683178465, + "learning_rate": 3.843633718504704e-06, + "loss": 0.8509, + "num_input_tokens_seen": 26954920, + "step": 1272 + }, + { + "epoch": 0.1530691997835628, + "flos": 20055198246720.0, + "grad_norm": 2.9145023126545553, + "learning_rate": 3.843331629248715e-06, + "loss": 0.905, + "num_input_tokens_seen": 26972520, + "step": 1273 + }, + { + "epoch": 0.1531894426742019, + "flos": 28718269186080.0, + "grad_norm": 2.3133580445976976, + "learning_rate": 3.843029260360782e-06, + "loss": 0.7676, + "num_input_tokens_seen": 26990890, + "step": 1274 + }, + { + "epoch": 0.15330968556484098, + "flos": 22204506203520.0, + "grad_norm": 8.101989087431035, + "learning_rate": 3.8427266118867755e-06, + "loss": 0.7874, + "num_input_tokens_seen": 27010640, + "step": 1275 + }, + { + "epoch": 0.15342992845548006, + "flos": 27488052545760.0, + "grad_norm": 2.4654882588924103, + "learning_rate": 3.842423683872608e-06, + "loss": 0.8246, + "num_input_tokens_seen": 27031215, + "step": 1276 + }, + { + "epoch": 0.15355017134611917, + "flos": 19581655307040.0, + "grad_norm": 2.367737457548281, + "learning_rate": 3.842120476364232e-06, + "loss": 0.774, + "num_input_tokens_seen": 27049665, + "step": 1277 + }, + { + "epoch": 0.15367041423675826, + "flos": 18452297426400.0, + "grad_norm": 4.303874581179436, + "learning_rate": 3.841816989407644e-06, + "loss": 0.8393, + "num_input_tokens_seen": 27065155, + "step": 1278 + }, + { + "epoch": 0.15379065712739734, + "flos": 41367938640960.0, + "grad_norm": 2.3287711047016324, + "learning_rate": 3.841513223048884e-06, + "loss": 0.7644, + "num_input_tokens_seen": 27085840, + "step": 1279 + }, + { + "epoch": 0.15391090001803642, + "flos": 22022717409120.0, + "grad_norm": 2.735639617581061, + "learning_rate": 3.841209177334031e-06, + "loss": 0.7804, + "num_input_tokens_seen": 27104200, + "step": 1280 + }, + { + "epoch": 0.15403114290867553, + "flos": 15428470428960.0, + "grad_norm": 2.0417484993656454, + "learning_rate": 3.84090485230921e-06, + "loss": 0.747, + "num_input_tokens_seen": 27122440, + "step": 1281 + }, + { + "epoch": 0.15415138579931462, + "flos": 17903585764800.0, + "grad_norm": 2.801685054415382, + "learning_rate": 3.840600248020588e-06, + "loss": 0.7643, + "num_input_tokens_seen": 27139380, + "step": 1282 + }, + { + "epoch": 0.1542716286899537, + "flos": 11413022446080.0, + "grad_norm": 2.755460952302149, + "learning_rate": 3.840295364514371e-06, + "loss": 0.7969, + "num_input_tokens_seen": 27156760, + "step": 1283 + }, + { + "epoch": 0.1543918715805928, + "flos": 17395580724000.0, + "grad_norm": 3.1891573916591844, + "learning_rate": 3.83999020183681e-06, + "loss": 0.7833, + "num_input_tokens_seen": 27175455, + "step": 1284 + }, + { + "epoch": 0.1545121144712319, + "flos": 17761128310080.0, + "grad_norm": 2.2838964918702844, + "learning_rate": 3.839684760034199e-06, + "loss": 0.7849, + "num_input_tokens_seen": 27193660, + "step": 1285 + }, + { + "epoch": 0.15463235736187098, + "flos": 28179704868960.0, + "grad_norm": 2.4301378623881407, + "learning_rate": 3.8393790391528716e-06, + "loss": 0.6539, + "num_input_tokens_seen": 27214355, + "step": 1286 + }, + { + "epoch": 0.15475260025251006, + "flos": 22824223573920.0, + "grad_norm": 1.9385653612375413, + "learning_rate": 3.8390730392392075e-06, + "loss": 0.8909, + "num_input_tokens_seen": 27234335, + "step": 1287 + }, + { + "epoch": 0.15487284314314917, + "flos": 17577852725280.0, + "grad_norm": 2.3550223311805105, + "learning_rate": 3.838766760339626e-06, + "loss": 0.7906, + "num_input_tokens_seen": 27252860, + "step": 1288 + }, + { + "epoch": 0.15499308603378825, + "flos": 20053339758720.0, + "grad_norm": 7.44042603452217, + "learning_rate": 3.838460202500587e-06, + "loss": 0.7948, + "num_input_tokens_seen": 27268650, + "step": 1289 + }, + { + "epoch": 0.15511332892442733, + "flos": 15895248472320.0, + "grad_norm": 2.3378192377624383, + "learning_rate": 3.838153365768599e-06, + "loss": 0.7421, + "num_input_tokens_seen": 27285160, + "step": 1290 + }, + { + "epoch": 0.15523357181506645, + "flos": 41222990812320.0, + "grad_norm": 2.4883178516612374, + "learning_rate": 3.837846250190206e-06, + "loss": 0.7566, + "num_input_tokens_seen": 27308025, + "step": 1291 + }, + { + "epoch": 0.15535381470570553, + "flos": 18452483275200.0, + "grad_norm": 2.6043365626546024, + "learning_rate": 3.837538855811998e-06, + "loss": 0.7694, + "num_input_tokens_seen": 27326200, + "step": 1292 + }, + { + "epoch": 0.1554740575963446, + "flos": 13898508144960.0, + "grad_norm": 4.701257886365236, + "learning_rate": 3.837231182680606e-06, + "loss": 0.7152, + "num_input_tokens_seen": 27344165, + "step": 1293 + }, + { + "epoch": 0.1555943004869837, + "flos": 20817224392800.0, + "grad_norm": 1.884301510923019, + "learning_rate": 3.836923230842706e-06, + "loss": 0.7622, + "num_input_tokens_seen": 27363280, + "step": 1294 + }, + { + "epoch": 0.1557145433776228, + "flos": 22054168759680.0, + "grad_norm": 2.313445130082744, + "learning_rate": 3.836615000345011e-06, + "loss": 0.8115, + "num_input_tokens_seen": 27381460, + "step": 1295 + }, + { + "epoch": 0.1558347862682619, + "flos": 19763778629280.0, + "grad_norm": 2.3024938456135198, + "learning_rate": 3.836306491234282e-06, + "loss": 0.7816, + "num_input_tokens_seen": 27399310, + "step": 1296 + }, + { + "epoch": 0.15595502915890097, + "flos": 17212565327520.0, + "grad_norm": 2.477321183635099, + "learning_rate": 3.835997703557317e-06, + "loss": 0.759, + "num_input_tokens_seen": 27416050, + "step": 1297 + }, + { + "epoch": 0.15607527204954008, + "flos": 19691434809120.0, + "grad_norm": 1.9167548726006107, + "learning_rate": 3.83568863736096e-06, + "loss": 0.7986, + "num_input_tokens_seen": 27434920, + "step": 1298 + }, + { + "epoch": 0.15619551494017916, + "flos": 18489324240960.0, + "grad_norm": 3.2077402499721677, + "learning_rate": 3.8353792926920975e-06, + "loss": 0.8968, + "num_input_tokens_seen": 27453570, + "step": 1299 + }, + { + "epoch": 0.15631575783081825, + "flos": 19873781149920.0, + "grad_norm": 2.700863283394022, + "learning_rate": 3.835069669597655e-06, + "loss": 0.8176, + "num_input_tokens_seen": 27471960, + "step": 1300 + }, + { + "epoch": 0.15643600072145733, + "flos": 20748040002240.0, + "grad_norm": 2.1144420658021796, + "learning_rate": 3.834759768124603e-06, + "loss": 0.7981, + "num_input_tokens_seen": 27490835, + "step": 1301 + }, + { + "epoch": 0.15655624361209644, + "flos": 18519809177760.0, + "grad_norm": 2.8172193139300963, + "learning_rate": 3.834449588319953e-06, + "loss": 0.7661, + "num_input_tokens_seen": 27508870, + "step": 1302 + }, + { + "epoch": 0.15667648650273552, + "flos": 25193462232480.0, + "grad_norm": 2.024286629566717, + "learning_rate": 3.834139130230758e-06, + "loss": 0.8498, + "num_input_tokens_seen": 27528175, + "step": 1303 + }, + { + "epoch": 0.1567967293933746, + "flos": 24791556887520.0, + "grad_norm": 2.214500441006177, + "learning_rate": 3.833828393904117e-06, + "loss": 0.8147, + "num_input_tokens_seen": 27548455, + "step": 1304 + }, + { + "epoch": 0.15691697228401372, + "flos": 19136999004480.0, + "grad_norm": 2.5713573987133187, + "learning_rate": 3.833517379387165e-06, + "loss": 0.7803, + "num_input_tokens_seen": 27564910, + "step": 1305 + }, + { + "epoch": 0.1570372151746528, + "flos": 24755459316960.0, + "grad_norm": 2.142150780650376, + "learning_rate": 3.833206086727085e-06, + "loss": 0.8926, + "num_input_tokens_seen": 27584260, + "step": 1306 + }, + { + "epoch": 0.15715745806529188, + "flos": 24828063325440.0, + "grad_norm": 2.7004442096803163, + "learning_rate": 3.8328945159710994e-06, + "loss": 0.7049, + "num_input_tokens_seen": 27604440, + "step": 1307 + }, + { + "epoch": 0.157277700955931, + "flos": 21840928614720.0, + "grad_norm": 2.0723708950273347, + "learning_rate": 3.832582667166473e-06, + "loss": 0.8849, + "num_input_tokens_seen": 27624250, + "step": 1308 + }, + { + "epoch": 0.15739794384657008, + "flos": 24498167272800.0, + "grad_norm": 1.9486248816452758, + "learning_rate": 3.8322705403605125e-06, + "loss": 0.814, + "num_input_tokens_seen": 27644595, + "step": 1309 + }, + { + "epoch": 0.15751818673720916, + "flos": 17719975652160.0, + "grad_norm": 2.165064720789829, + "learning_rate": 3.831958135600568e-06, + "loss": 0.8117, + "num_input_tokens_seen": 27662145, + "step": 1310 + }, + { + "epoch": 0.15763842962784824, + "flos": 17833137602400.0, + "grad_norm": 3.0530445724717192, + "learning_rate": 3.831645452934032e-06, + "loss": 0.7967, + "num_input_tokens_seen": 27680575, + "step": 1311 + }, + { + "epoch": 0.15775867251848735, + "flos": 26942277295200.0, + "grad_norm": 1.9797327603990005, + "learning_rate": 3.831332492408336e-06, + "loss": 0.802, + "num_input_tokens_seen": 27701985, + "step": 1312 + }, + { + "epoch": 0.15787891540912644, + "flos": 19212836782080.0, + "grad_norm": 2.1129513410750795, + "learning_rate": 3.831019254070957e-06, + "loss": 0.6936, + "num_input_tokens_seen": 27719770, + "step": 1313 + }, + { + "epoch": 0.15799915829976552, + "flos": 27231987103680.0, + "grad_norm": 2.754867234535892, + "learning_rate": 3.8307057379694135e-06, + "loss": 0.9509, + "num_input_tokens_seen": 27739105, + "step": 1314 + }, + { + "epoch": 0.15811940119040463, + "flos": 20376173556960.0, + "grad_norm": 2.7112389652490188, + "learning_rate": 3.830391944151264e-06, + "loss": 0.8248, + "num_input_tokens_seen": 27754985, + "step": 1315 + }, + { + "epoch": 0.1582396440810437, + "flos": 32552523090720.0, + "grad_norm": 2.119972227562424, + "learning_rate": 3.830077872664114e-06, + "loss": 0.6752, + "num_input_tokens_seen": 27776110, + "step": 1316 + }, + { + "epoch": 0.1583598869716828, + "flos": 33751883096640.0, + "grad_norm": 2.135110699823428, + "learning_rate": 3.829763523555604e-06, + "loss": 0.7324, + "num_input_tokens_seen": 27796510, + "step": 1317 + }, + { + "epoch": 0.15848012986232188, + "flos": 24646014342720.0, + "grad_norm": 3.4068822711369418, + "learning_rate": 3.829448896873423e-06, + "loss": 0.7804, + "num_input_tokens_seen": 27815610, + "step": 1318 + }, + { + "epoch": 0.158600372752961, + "flos": 22569793601280.0, + "grad_norm": 1.9717462536938388, + "learning_rate": 3.829133992665299e-06, + "loss": 0.7932, + "num_input_tokens_seen": 27834735, + "step": 1319 + }, + { + "epoch": 0.15872061564360007, + "flos": 27887876384160.0, + "grad_norm": 2.482744365547078, + "learning_rate": 3.828818810979002e-06, + "loss": 0.8881, + "num_input_tokens_seen": 27853065, + "step": 1320 + }, + { + "epoch": 0.15884085853423915, + "flos": 23662942402080.0, + "grad_norm": 1.96523934661109, + "learning_rate": 3.8285033518623454e-06, + "loss": 0.8047, + "num_input_tokens_seen": 27873315, + "step": 1321 + }, + { + "epoch": 0.15896110142487826, + "flos": 23079322602240.0, + "grad_norm": 3.1766947691062315, + "learning_rate": 3.8281876153631845e-06, + "loss": 0.8167, + "num_input_tokens_seen": 27890070, + "step": 1322 + }, + { + "epoch": 0.15908134431551735, + "flos": 14664511455360.0, + "grad_norm": 2.1474259648201017, + "learning_rate": 3.827871601529416e-06, + "loss": 0.6469, + "num_input_tokens_seen": 27908590, + "step": 1323 + }, + { + "epoch": 0.15920158720615643, + "flos": 20164829069760.0, + "grad_norm": 5.054910504523323, + "learning_rate": 3.827555310408979e-06, + "loss": 0.8084, + "num_input_tokens_seen": 27927985, + "step": 1324 + }, + { + "epoch": 0.1593218300967955, + "flos": 24791073680640.0, + "grad_norm": 2.0290550697339844, + "learning_rate": 3.827238742049854e-06, + "loss": 0.8272, + "num_input_tokens_seen": 27948280, + "step": 1325 + }, + { + "epoch": 0.15944207298743462, + "flos": 28288034750400.0, + "grad_norm": 2.326394788693464, + "learning_rate": 3.826921896500066e-06, + "loss": 0.5209, + "num_input_tokens_seen": 27969565, + "step": 1326 + }, + { + "epoch": 0.1595623158780737, + "flos": 22931958739200.0, + "grad_norm": 1.9662873139329964, + "learning_rate": 3.826604773807678e-06, + "loss": 0.7827, + "num_input_tokens_seen": 27987540, + "step": 1327 + }, + { + "epoch": 0.1596825587687128, + "flos": 19682142369120.0, + "grad_norm": 2.9234101041879272, + "learning_rate": 3.826287374020798e-06, + "loss": 0.7366, + "num_input_tokens_seen": 28002345, + "step": 1328 + }, + { + "epoch": 0.1598028016593519, + "flos": 22605370795200.0, + "grad_norm": 2.46563456525023, + "learning_rate": 3.825969697187575e-06, + "loss": 0.8248, + "num_input_tokens_seen": 28021675, + "step": 1329 + }, + { + "epoch": 0.15992304454999098, + "flos": 20453609634240.0, + "grad_norm": 2.3310325950723105, + "learning_rate": 3.8256517433562015e-06, + "loss": 0.6969, + "num_input_tokens_seen": 28039615, + "step": 1330 + }, + { + "epoch": 0.16004328744063007, + "flos": 17650977110400.0, + "grad_norm": 2.6636260897361046, + "learning_rate": 3.82533351257491e-06, + "loss": 0.9178, + "num_input_tokens_seen": 28057565, + "step": 1331 + }, + { + "epoch": 0.16016353033126918, + "flos": 24064290200640.0, + "grad_norm": 1.869979494767804, + "learning_rate": 3.825015004891975e-06, + "loss": 0.8857, + "num_input_tokens_seen": 28076305, + "step": 1332 + }, + { + "epoch": 0.16028377322190826, + "flos": 27595155825120.0, + "grad_norm": 2.3753854539184793, + "learning_rate": 3.824696220355716e-06, + "loss": 0.7582, + "num_input_tokens_seen": 28097895, + "step": 1333 + }, + { + "epoch": 0.16040401611254734, + "flos": 20931352756800.0, + "grad_norm": 1.7881143381226916, + "learning_rate": 3.824377159014491e-06, + "loss": 0.7904, + "num_input_tokens_seen": 28118790, + "step": 1334 + }, + { + "epoch": 0.16052425900318643, + "flos": 21216565024320.0, + "grad_norm": 3.4620673615123976, + "learning_rate": 3.824057820916702e-06, + "loss": 0.8505, + "num_input_tokens_seen": 28135995, + "step": 1335 + }, + { + "epoch": 0.16064450189382554, + "flos": 15502040851200.0, + "grad_norm": 2.6904208257567306, + "learning_rate": 3.8237382061107904e-06, + "loss": 0.7189, + "num_input_tokens_seen": 28152715, + "step": 1336 + }, + { + "epoch": 0.16076474478446462, + "flos": 21148272708000.0, + "grad_norm": 2.229617836712028, + "learning_rate": 3.823418314645243e-06, + "loss": 0.787, + "num_input_tokens_seen": 28171590, + "step": 1337 + }, + { + "epoch": 0.1608849876751037, + "flos": 18340250568960.0, + "grad_norm": 2.417319461744135, + "learning_rate": 3.823098146568588e-06, + "loss": 0.7543, + "num_input_tokens_seen": 28191295, + "step": 1338 + }, + { + "epoch": 0.1610052305657428, + "flos": 29454679633920.0, + "grad_norm": 1.8267769040770379, + "learning_rate": 3.822777701929394e-06, + "loss": 0.7158, + "num_input_tokens_seen": 28213120, + "step": 1339 + }, + { + "epoch": 0.1611254734563819, + "flos": 26759484917280.0, + "grad_norm": 2.033121391361199, + "learning_rate": 3.8224569807762714e-06, + "loss": 0.7339, + "num_input_tokens_seen": 28232440, + "step": 1340 + }, + { + "epoch": 0.16124571634702098, + "flos": 22387558769760.0, + "grad_norm": 2.2298259176209894, + "learning_rate": 3.822135983157873e-06, + "loss": 0.7632, + "num_input_tokens_seen": 28251715, + "step": 1341 + }, + { + "epoch": 0.16136595923766006, + "flos": 10983717254400.0, + "grad_norm": 2.270182478372041, + "learning_rate": 3.821814709122896e-06, + "loss": 0.8414, + "num_input_tokens_seen": 28269765, + "step": 1342 + }, + { + "epoch": 0.16148620212829917, + "flos": 21184704806400.0, + "grad_norm": 2.7716085931685206, + "learning_rate": 3.821493158720076e-06, + "loss": 0.8456, + "num_input_tokens_seen": 28288870, + "step": 1343 + }, + { + "epoch": 0.16160644501893826, + "flos": 16734487677120.0, + "grad_norm": 3.5229522979997805, + "learning_rate": 3.821171331998191e-06, + "loss": 0.7328, + "num_input_tokens_seen": 28305400, + "step": 1344 + }, + { + "epoch": 0.16172668790957734, + "flos": 64352870021760.0, + "grad_norm": 0.744793903018466, + "learning_rate": 3.820849229006064e-06, + "loss": 0.5446, + "num_input_tokens_seen": 28373150, + "step": 1345 + }, + { + "epoch": 0.16184693080021645, + "flos": 23224419109920.0, + "grad_norm": 2.722471386887858, + "learning_rate": 3.8205268497925564e-06, + "loss": 0.7122, + "num_input_tokens_seen": 28393620, + "step": 1346 + }, + { + "epoch": 0.16196717369085553, + "flos": 17426102830560.0, + "grad_norm": 2.5742674665226177, + "learning_rate": 3.8202041944065725e-06, + "loss": 0.7857, + "num_input_tokens_seen": 28409440, + "step": 1347 + }, + { + "epoch": 0.16208741658149461, + "flos": 23839044223200.0, + "grad_norm": 2.80058485803012, + "learning_rate": 3.819881262897061e-06, + "loss": 0.7421, + "num_input_tokens_seen": 28427575, + "step": 1348 + }, + { + "epoch": 0.1622076594721337, + "flos": 25847716043520.0, + "grad_norm": 2.4380649009872113, + "learning_rate": 3.819558055313008e-06, + "loss": 0.7347, + "num_input_tokens_seen": 28448260, + "step": 1349 + }, + { + "epoch": 0.1623279023627728, + "flos": 21508876716000.0, + "grad_norm": 2.2626938373295795, + "learning_rate": 3.819234571703444e-06, + "loss": 0.7781, + "num_input_tokens_seen": 28467085, + "step": 1350 + }, + { + "epoch": 0.1624481452534119, + "flos": 22053685552800.0, + "grad_norm": 1.8206319242189255, + "learning_rate": 3.8189108121174435e-06, + "loss": 0.8573, + "num_input_tokens_seen": 28486850, + "step": 1351 + }, + { + "epoch": 0.16256838814405097, + "flos": 27048600009600.0, + "grad_norm": 2.235327350956964, + "learning_rate": 3.818586776604118e-06, + "loss": 0.8378, + "num_input_tokens_seen": 28507490, + "step": 1352 + }, + { + "epoch": 0.16268863103469008, + "flos": 20091667514880.0, + "grad_norm": 2.7042172380684404, + "learning_rate": 3.818262465212625e-06, + "loss": 0.6148, + "num_input_tokens_seen": 28527615, + "step": 1353 + }, + { + "epoch": 0.16280887392532917, + "flos": 18306717711840.0, + "grad_norm": 2.404148521773581, + "learning_rate": 3.817937877992161e-06, + "loss": 0.7714, + "num_input_tokens_seen": 28544910, + "step": 1354 + }, + { + "epoch": 0.16292911681596825, + "flos": 11857641578880.0, + "grad_norm": 2.4930745522269135, + "learning_rate": 3.817613014991967e-06, + "loss": 0.8581, + "num_input_tokens_seen": 28561050, + "step": 1355 + }, + { + "epoch": 0.16304935970660733, + "flos": 26066048445600.0, + "grad_norm": 2.0224087136017523, + "learning_rate": 3.817287876261323e-06, + "loss": 0.7621, + "num_input_tokens_seen": 28581385, + "step": 1356 + }, + { + "epoch": 0.16316960259724644, + "flos": 29309843314560.0, + "grad_norm": 2.2239532708877965, + "learning_rate": 3.816962461849553e-06, + "loss": 0.8023, + "num_input_tokens_seen": 28603255, + "step": 1357 + }, + { + "epoch": 0.16328984548788553, + "flos": 20856964599840.0, + "grad_norm": 2.1885639328136373, + "learning_rate": 3.8166367718060235e-06, + "loss": 0.8453, + "num_input_tokens_seen": 28623145, + "step": 1358 + }, + { + "epoch": 0.1634100883785246, + "flos": 18015446773440.0, + "grad_norm": 4.77126751163698, + "learning_rate": 3.816310806180139e-06, + "loss": 0.7721, + "num_input_tokens_seen": 28641555, + "step": 1359 + }, + { + "epoch": 0.16353033126916372, + "flos": 24536978235840.0, + "grad_norm": 1.8653717150230755, + "learning_rate": 3.81598456502135e-06, + "loss": 0.8099, + "num_input_tokens_seen": 28661575, + "step": 1360 + }, + { + "epoch": 0.1636505741598028, + "flos": 19864228521600.0, + "grad_norm": 2.279115488243727, + "learning_rate": 3.8156580483791455e-06, + "loss": 0.8727, + "num_input_tokens_seen": 28676765, + "step": 1361 + }, + { + "epoch": 0.16377081705044189, + "flos": 28361865360960.0, + "grad_norm": 3.708677997359456, + "learning_rate": 3.815331256303059e-06, + "loss": 0.7695, + "num_input_tokens_seen": 28696795, + "step": 1362 + }, + { + "epoch": 0.163891059941081, + "flos": 21876765996960.0, + "grad_norm": 2.5122544353672582, + "learning_rate": 3.815004188842665e-06, + "loss": 0.7762, + "num_input_tokens_seen": 28714835, + "step": 1363 + }, + { + "epoch": 0.16401130283172008, + "flos": 26759224728960.0, + "grad_norm": 1.6461756949719646, + "learning_rate": 3.814676846047578e-06, + "loss": 0.7961, + "num_input_tokens_seen": 28735710, + "step": 1364 + }, + { + "epoch": 0.16413154572235916, + "flos": 32949893724960.0, + "grad_norm": 1.8885433622739418, + "learning_rate": 3.8143492279674565e-06, + "loss": 0.6982, + "num_input_tokens_seen": 28758405, + "step": 1365 + }, + { + "epoch": 0.16425178861299825, + "flos": 40057120760160.0, + "grad_norm": 0.8408118058149824, + "learning_rate": 3.8140213346519997e-06, + "loss": 0.5832, + "num_input_tokens_seen": 28809000, + "step": 1366 + }, + { + "epoch": 0.16437203150363736, + "flos": 25411051239360.0, + "grad_norm": 1.9268571738112357, + "learning_rate": 3.813693166150948e-06, + "loss": 0.7677, + "num_input_tokens_seen": 28829210, + "step": 1367 + }, + { + "epoch": 0.16449227439427644, + "flos": 23443309058400.0, + "grad_norm": 3.013956982321443, + "learning_rate": 3.813364722514086e-06, + "loss": 0.8586, + "num_input_tokens_seen": 28847545, + "step": 1368 + }, + { + "epoch": 0.16461251728491552, + "flos": 13525823964960.0, + "grad_norm": 2.3112587561681335, + "learning_rate": 3.8130360037912368e-06, + "loss": 0.8031, + "num_input_tokens_seen": 28863990, + "step": 1369 + }, + { + "epoch": 0.16473276017555463, + "flos": 22970435174400.0, + "grad_norm": 2.5115551932374447, + "learning_rate": 3.812707010032268e-06, + "loss": 0.8139, + "num_input_tokens_seen": 28883045, + "step": 1370 + }, + { + "epoch": 0.16485300306619372, + "flos": 24755199128640.0, + "grad_norm": 1.9041272210084599, + "learning_rate": 3.8123777412870863e-06, + "loss": 0.7943, + "num_input_tokens_seen": 28902545, + "step": 1371 + }, + { + "epoch": 0.1649732459568328, + "flos": 21076412094720.0, + "grad_norm": 2.135366063952806, + "learning_rate": 3.812048197605643e-06, + "loss": 0.7855, + "num_input_tokens_seen": 28921440, + "step": 1372 + }, + { + "epoch": 0.16509348884747188, + "flos": 20237841945600.0, + "grad_norm": 2.272516962204212, + "learning_rate": 3.8117183790379277e-06, + "loss": 0.8164, + "num_input_tokens_seen": 28939890, + "step": 1373 + }, + { + "epoch": 0.165213731738111, + "flos": 11020037843520.0, + "grad_norm": 3.2183325043065962, + "learning_rate": 3.811388285633976e-06, + "loss": 0.9395, + "num_input_tokens_seen": 28955875, + "step": 1374 + }, + { + "epoch": 0.16533397462875007, + "flos": 29929040308320.0, + "grad_norm": 2.7568151815770197, + "learning_rate": 3.811057917443861e-06, + "loss": 0.6205, + "num_input_tokens_seen": 28975140, + "step": 1375 + }, + { + "epoch": 0.16545421751938916, + "flos": 65463674458560.0, + "grad_norm": 0.8589181205730051, + "learning_rate": 3.8107272745177e-06, + "loss": 0.683, + "num_input_tokens_seen": 29047470, + "step": 1376 + }, + { + "epoch": 0.16557446041002827, + "flos": 22460237117760.0, + "grad_norm": 1.9854125666967959, + "learning_rate": 3.8103963569056513e-06, + "loss": 0.7863, + "num_input_tokens_seen": 29066605, + "step": 1377 + }, + { + "epoch": 0.16569470330066735, + "flos": 24567314493600.0, + "grad_norm": 1.7390117353380081, + "learning_rate": 3.8100651646579146e-06, + "loss": 0.8816, + "num_input_tokens_seen": 29085815, + "step": 1378 + }, + { + "epoch": 0.16581494619130643, + "flos": 14984743370400.0, + "grad_norm": 2.3373864824606763, + "learning_rate": 3.8097336978247317e-06, + "loss": 0.9288, + "num_input_tokens_seen": 29102520, + "step": 1379 + }, + { + "epoch": 0.16593518908194552, + "flos": 17395580724000.0, + "grad_norm": 2.6083404865918367, + "learning_rate": 3.8094019564563854e-06, + "loss": 0.8933, + "num_input_tokens_seen": 29119450, + "step": 1380 + }, + { + "epoch": 0.16605543197258463, + "flos": 20383533169440.0, + "grad_norm": 2.396023010264416, + "learning_rate": 3.809069940603201e-06, + "loss": 0.7494, + "num_input_tokens_seen": 29137405, + "step": 1381 + }, + { + "epoch": 0.1661756748632237, + "flos": 14189667574080.0, + "grad_norm": 2.5132960562466997, + "learning_rate": 3.8087376503155452e-06, + "loss": 0.7805, + "num_input_tokens_seen": 29154890, + "step": 1382 + }, + { + "epoch": 0.1662959177538628, + "flos": 65986181439360.0, + "grad_norm": 0.8906585285501201, + "learning_rate": 3.808405085643826e-06, + "loss": 0.5625, + "num_input_tokens_seen": 29219770, + "step": 1383 + }, + { + "epoch": 0.1664161606445019, + "flos": 20710976017920.0, + "grad_norm": 2.8927446171911004, + "learning_rate": 3.8080722466384925e-06, + "loss": 0.8901, + "num_input_tokens_seen": 29237620, + "step": 1384 + }, + { + "epoch": 0.166536403535141, + "flos": 25224616224960.0, + "grad_norm": 2.4278000740620054, + "learning_rate": 3.8077391333500376e-06, + "loss": 0.7145, + "num_input_tokens_seen": 29256760, + "step": 1385 + }, + { + "epoch": 0.16665664642578007, + "flos": 25411014069600.0, + "grad_norm": 2.3044115838875925, + "learning_rate": 3.8074057458289934e-06, + "loss": 0.7676, + "num_input_tokens_seen": 29277450, + "step": 1386 + }, + { + "epoch": 0.16677688931641918, + "flos": 22169077688640.0, + "grad_norm": 3.2330511212494466, + "learning_rate": 3.807072084125934e-06, + "loss": 0.8265, + "num_input_tokens_seen": 29299300, + "step": 1387 + }, + { + "epoch": 0.16689713220705826, + "flos": 16921889105280.0, + "grad_norm": 2.5334414191361, + "learning_rate": 3.806738148291477e-06, + "loss": 0.8088, + "num_input_tokens_seen": 29316125, + "step": 1388 + }, + { + "epoch": 0.16701737509769735, + "flos": 36192164633760.0, + "grad_norm": 2.7286844385635343, + "learning_rate": 3.8064039383762793e-06, + "loss": 0.7132, + "num_input_tokens_seen": 29338490, + "step": 1389 + }, + { + "epoch": 0.16713761798833643, + "flos": 23225422693440.0, + "grad_norm": 2.252262461475912, + "learning_rate": 3.8060694544310396e-06, + "loss": 0.7716, + "num_input_tokens_seen": 29357000, + "step": 1390 + }, + { + "epoch": 0.16725786087897554, + "flos": 25266326429280.0, + "grad_norm": 2.7534031185176544, + "learning_rate": 3.8057346965065006e-06, + "loss": 0.7873, + "num_input_tokens_seen": 29378230, + "step": 1391 + }, + { + "epoch": 0.16737810376961462, + "flos": 31785776385120.0, + "grad_norm": 1.8040260324735575, + "learning_rate": 3.805399664653443e-06, + "loss": 0.8444, + "num_input_tokens_seen": 29400370, + "step": 1392 + }, + { + "epoch": 0.1674983466602537, + "flos": 27921632259840.0, + "grad_norm": 4.353304228478464, + "learning_rate": 3.805064358922692e-06, + "loss": 0.7464, + "num_input_tokens_seen": 29418660, + "step": 1393 + }, + { + "epoch": 0.16761858955089282, + "flos": 21731595149760.0, + "grad_norm": 2.2859216174540626, + "learning_rate": 3.8047287793651136e-06, + "loss": 0.8132, + "num_input_tokens_seen": 29440105, + "step": 1394 + }, + { + "epoch": 0.1677388324415319, + "flos": 23771458132320.0, + "grad_norm": 1.9664079571406972, + "learning_rate": 3.8043929260316137e-06, + "loss": 0.8872, + "num_input_tokens_seen": 29458260, + "step": 1395 + }, + { + "epoch": 0.16785907533217098, + "flos": 20529558921120.0, + "grad_norm": 2.1542399892619786, + "learning_rate": 3.8040567989731417e-06, + "loss": 0.8376, + "num_input_tokens_seen": 29476205, + "step": 1396 + }, + { + "epoch": 0.16797931822281006, + "flos": 15647471886720.0, + "grad_norm": 5.4840456437118075, + "learning_rate": 3.8037203982406876e-06, + "loss": 0.8001, + "num_input_tokens_seen": 29494370, + "step": 1397 + }, + { + "epoch": 0.16809956111344918, + "flos": 16517493386400.0, + "grad_norm": 2.022699523439303, + "learning_rate": 3.8033837238852835e-06, + "loss": 0.7334, + "num_input_tokens_seen": 29511070, + "step": 1398 + }, + { + "epoch": 0.16821980400408826, + "flos": 23224827977280.0, + "grad_norm": 2.25850321094419, + "learning_rate": 3.8030467759580017e-06, + "loss": 0.6965, + "num_input_tokens_seen": 29531990, + "step": 1399 + }, + { + "epoch": 0.16834004689472734, + "flos": 20745029251680.0, + "grad_norm": 2.1306102967251497, + "learning_rate": 3.802709554509958e-06, + "loss": 0.8718, + "num_input_tokens_seen": 29549790, + "step": 1400 + }, + { + "epoch": 0.16846028978536645, + "flos": 26649036359520.0, + "grad_norm": 1.8524097805731226, + "learning_rate": 3.8023720595923083e-06, + "loss": 0.7912, + "num_input_tokens_seen": 29569045, + "step": 1401 + }, + { + "epoch": 0.16858053267600553, + "flos": 18816469731360.0, + "grad_norm": 8.62120702519601, + "learning_rate": 3.80203429125625e-06, + "loss": 0.8721, + "num_input_tokens_seen": 29587220, + "step": 1402 + }, + { + "epoch": 0.16870077556664462, + "flos": 27704712308640.0, + "grad_norm": 1.8789718521308012, + "learning_rate": 3.8016962495530225e-06, + "loss": 0.7024, + "num_input_tokens_seen": 29606410, + "step": 1403 + }, + { + "epoch": 0.1688210184572837, + "flos": 13711218226080.0, + "grad_norm": 2.559131920977276, + "learning_rate": 3.8013579345339063e-06, + "loss": 0.766, + "num_input_tokens_seen": 29621275, + "step": 1404 + }, + { + "epoch": 0.1689412613479228, + "flos": 26430852636480.0, + "grad_norm": 2.714711664952307, + "learning_rate": 3.801019346250224e-06, + "loss": 0.6954, + "num_input_tokens_seen": 29641420, + "step": 1405 + }, + { + "epoch": 0.1690615042385619, + "flos": 21108569670720.0, + "grad_norm": 5.203579169719349, + "learning_rate": 3.8006804847533395e-06, + "loss": 0.8407, + "num_input_tokens_seen": 29660935, + "step": 1406 + }, + { + "epoch": 0.16918174712920098, + "flos": 20819566087680.0, + "grad_norm": 2.574862548715055, + "learning_rate": 3.8003413500946556e-06, + "loss": 0.8556, + "num_input_tokens_seen": 29681045, + "step": 1407 + }, + { + "epoch": 0.1693019900198401, + "flos": 16959176108160.0, + "grad_norm": 2.8297027065093894, + "learning_rate": 3.8000019423256216e-06, + "loss": 0.8273, + "num_input_tokens_seen": 29698810, + "step": 1408 + }, + { + "epoch": 0.16942223291047917, + "flos": 26758481333760.0, + "grad_norm": 1.7721441316822428, + "learning_rate": 3.7996622614977234e-06, + "loss": 0.8811, + "num_input_tokens_seen": 29720000, + "step": 1409 + }, + { + "epoch": 0.16954247580111825, + "flos": 18557096180640.0, + "grad_norm": 3.2620676916175153, + "learning_rate": 3.799322307662492e-06, + "loss": 0.7925, + "num_input_tokens_seen": 29737820, + "step": 1410 + }, + { + "epoch": 0.16966271869175734, + "flos": 13963641031680.0, + "grad_norm": 3.503581530644052, + "learning_rate": 3.798982080871496e-06, + "loss": 0.8364, + "num_input_tokens_seen": 29752960, + "step": 1411 + }, + { + "epoch": 0.16978296158239645, + "flos": 37433346353280.0, + "grad_norm": 2.525732738572951, + "learning_rate": 3.798641581176349e-06, + "loss": 0.6793, + "num_input_tokens_seen": 29775240, + "step": 1412 + }, + { + "epoch": 0.16990320447303553, + "flos": 28288294938720.0, + "grad_norm": 2.2374970292226153, + "learning_rate": 3.7983008086287044e-06, + "loss": 0.7442, + "num_input_tokens_seen": 29796560, + "step": 1413 + }, + { + "epoch": 0.1700234473636746, + "flos": 20159216436000.0, + "grad_norm": 2.455046951715126, + "learning_rate": 3.797959763280257e-06, + "loss": 0.7956, + "num_input_tokens_seen": 29815245, + "step": 1414 + }, + { + "epoch": 0.17014369025431372, + "flos": 24822896728800.0, + "grad_norm": 2.1820913752471456, + "learning_rate": 3.797618445182743e-06, + "loss": 0.7879, + "num_input_tokens_seen": 29836640, + "step": 1415 + }, + { + "epoch": 0.1702639331449528, + "flos": 16443699945600.0, + "grad_norm": 2.4758487479150437, + "learning_rate": 3.79727685438794e-06, + "loss": 0.851, + "num_input_tokens_seen": 29850830, + "step": 1416 + }, + { + "epoch": 0.1703841760355919, + "flos": 52433961211200.0, + "grad_norm": 0.8473042754964155, + "learning_rate": 3.796934990947667e-06, + "loss": 0.6164, + "num_input_tokens_seen": 29912515, + "step": 1417 + }, + { + "epoch": 0.170504418926231, + "flos": 49300280372160.0, + "grad_norm": 0.8857151687690182, + "learning_rate": 3.7965928549137854e-06, + "loss": 0.6246, + "num_input_tokens_seen": 29969330, + "step": 1418 + }, + { + "epoch": 0.17062466181687008, + "flos": 25812213189120.0, + "grad_norm": 2.3756021439020594, + "learning_rate": 3.7962504463381953e-06, + "loss": 0.7759, + "num_input_tokens_seen": 29990820, + "step": 1419 + }, + { + "epoch": 0.17074490470750917, + "flos": 20930535022080.0, + "grad_norm": 2.0150937112073746, + "learning_rate": 3.7959077652728412e-06, + "loss": 0.7888, + "num_input_tokens_seen": 30009275, + "step": 1420 + }, + { + "epoch": 0.17086514759814825, + "flos": 20929865966400.0, + "grad_norm": 3.2494211414991714, + "learning_rate": 3.795564811769707e-06, + "loss": 0.7784, + "num_input_tokens_seen": 30027630, + "step": 1421 + }, + { + "epoch": 0.17098539048878736, + "flos": 28433168427840.0, + "grad_norm": 2.927036886876178, + "learning_rate": 3.795221585880818e-06, + "loss": 0.7805, + "num_input_tokens_seen": 30048310, + "step": 1422 + }, + { + "epoch": 0.17110563337942644, + "flos": 16266929068800.0, + "grad_norm": 2.6022902378706143, + "learning_rate": 3.794878087658242e-06, + "loss": 0.9118, + "num_input_tokens_seen": 30066080, + "step": 1423 + }, + { + "epoch": 0.17122587627006552, + "flos": 29631413340960.0, + "grad_norm": 2.0048369092559235, + "learning_rate": 3.7945343171540873e-06, + "loss": 0.7854, + "num_input_tokens_seen": 30086955, + "step": 1424 + }, + { + "epoch": 0.17134611916070464, + "flos": 25302126641760.0, + "grad_norm": 1.9561275748558242, + "learning_rate": 3.7941902744205033e-06, + "loss": 0.7877, + "num_input_tokens_seen": 30107990, + "step": 1425 + }, + { + "epoch": 0.17146636205134372, + "flos": 13934122508640.0, + "grad_norm": 2.3729817699402473, + "learning_rate": 3.7938459595096817e-06, + "loss": 0.8368, + "num_input_tokens_seen": 30126535, + "step": 1426 + }, + { + "epoch": 0.1715866049419828, + "flos": 23881014615840.0, + "grad_norm": 2.0971243697265143, + "learning_rate": 3.7935013724738545e-06, + "loss": 0.8631, + "num_input_tokens_seen": 30147035, + "step": 1427 + }, + { + "epoch": 0.17170684783262188, + "flos": 22676859710880.0, + "grad_norm": 2.6279603523271273, + "learning_rate": 3.7931565133652945e-06, + "loss": 0.7789, + "num_input_tokens_seen": 30167110, + "step": 1428 + }, + { + "epoch": 0.171827090723261, + "flos": 26575094239680.0, + "grad_norm": 2.5291564195436873, + "learning_rate": 3.792811382236317e-06, + "loss": 0.6775, + "num_input_tokens_seen": 30185620, + "step": 1429 + }, + { + "epoch": 0.17194733361390008, + "flos": 28108178783520.0, + "grad_norm": 2.330608298010913, + "learning_rate": 3.792465979139279e-06, + "loss": 0.7839, + "num_input_tokens_seen": 30208225, + "step": 1430 + }, + { + "epoch": 0.17206757650453916, + "flos": 65436689212800.0, + "grad_norm": 1.0866289201764692, + "learning_rate": 3.792120304126576e-06, + "loss": 0.6567, + "num_input_tokens_seen": 30271600, + "step": 1431 + }, + { + "epoch": 0.17218781939517827, + "flos": 22241681697120.0, + "grad_norm": 2.3000689247418724, + "learning_rate": 3.791774357250649e-06, + "loss": 0.8345, + "num_input_tokens_seen": 30290430, + "step": 1432 + }, + { + "epoch": 0.17230806228581735, + "flos": 14117360923680.0, + "grad_norm": 2.6601098234407754, + "learning_rate": 3.7914281385639757e-06, + "loss": 0.7875, + "num_input_tokens_seen": 30308305, + "step": 1433 + }, + { + "epoch": 0.17242830517645644, + "flos": 20675324484480.0, + "grad_norm": 2.0479464456663967, + "learning_rate": 3.7910816481190784e-06, + "loss": 0.7987, + "num_input_tokens_seen": 30328120, + "step": 1434 + }, + { + "epoch": 0.17254854806709552, + "flos": 30730397794080.0, + "grad_norm": 2.5078143063357494, + "learning_rate": 3.7907348859685193e-06, + "loss": 0.7473, + "num_input_tokens_seen": 30349825, + "step": 1435 + }, + { + "epoch": 0.17266879095773463, + "flos": 26576581030080.0, + "grad_norm": 3.354371927429464, + "learning_rate": 3.790387852164902e-06, + "loss": 0.8087, + "num_input_tokens_seen": 30370555, + "step": 1436 + }, + { + "epoch": 0.1727890338483737, + "flos": 20237395908480.0, + "grad_norm": 4.175297965192967, + "learning_rate": 3.7900405467608707e-06, + "loss": 0.7657, + "num_input_tokens_seen": 30390100, + "step": 1437 + }, + { + "epoch": 0.1729092767390128, + "flos": 18153704045280.0, + "grad_norm": 5.213351627002305, + "learning_rate": 3.7896929698091114e-06, + "loss": 0.7913, + "num_input_tokens_seen": 30402915, + "step": 1438 + }, + { + "epoch": 0.1730295196296519, + "flos": 26721119991360.0, + "grad_norm": 3.5063357908799553, + "learning_rate": 3.7893451213623518e-06, + "loss": 0.6802, + "num_input_tokens_seen": 30420145, + "step": 1439 + }, + { + "epoch": 0.173149762520291, + "flos": 23808447777120.0, + "grad_norm": 2.3794233389679507, + "learning_rate": 3.7889970014733606e-06, + "loss": 0.8248, + "num_input_tokens_seen": 30439250, + "step": 1440 + }, + { + "epoch": 0.17327000541093007, + "flos": 23334793328160.0, + "grad_norm": 1.990175075976566, + "learning_rate": 3.7886486101949463e-06, + "loss": 0.7812, + "num_input_tokens_seen": 30460430, + "step": 1441 + }, + { + "epoch": 0.17339024830156918, + "flos": 18195302740320.0, + "grad_norm": 2.7576065634976668, + "learning_rate": 3.7882999475799594e-06, + "loss": 0.8825, + "num_input_tokens_seen": 30478465, + "step": 1442 + }, + { + "epoch": 0.17351049119220827, + "flos": 23298695757600.0, + "grad_norm": 3.7549362754240074, + "learning_rate": 3.787951013681293e-06, + "loss": 0.812, + "num_input_tokens_seen": 30496510, + "step": 1443 + }, + { + "epoch": 0.17363073408284735, + "flos": 23769376625760.0, + "grad_norm": 4.422123662743799, + "learning_rate": 3.787601808551879e-06, + "loss": 0.7771, + "num_input_tokens_seen": 30516325, + "step": 1444 + }, + { + "epoch": 0.17375097697348643, + "flos": 18488952543360.0, + "grad_norm": 2.550847985882836, + "learning_rate": 3.7872523322446926e-06, + "loss": 0.8382, + "num_input_tokens_seen": 30535210, + "step": 1445 + }, + { + "epoch": 0.17387121986412554, + "flos": 38822040614880.0, + "grad_norm": 9.101054505673195, + "learning_rate": 3.7869025848127478e-06, + "loss": 0.6031, + "num_input_tokens_seen": 30559405, + "step": 1446 + }, + { + "epoch": 0.17399146275476463, + "flos": 20346655033920.0, + "grad_norm": 3.5896739898823258, + "learning_rate": 3.786552566309102e-06, + "loss": 0.8073, + "num_input_tokens_seen": 30577615, + "step": 1447 + }, + { + "epoch": 0.1741117056454037, + "flos": 19136775985920.0, + "grad_norm": 2.4257912750774127, + "learning_rate": 3.7862022767868517e-06, + "loss": 0.8633, + "num_input_tokens_seen": 30595205, + "step": 1448 + }, + { + "epoch": 0.17423194853604282, + "flos": 25337964024000.0, + "grad_norm": 2.1982870889600115, + "learning_rate": 3.7858517162991367e-06, + "loss": 0.8435, + "num_input_tokens_seen": 30615280, + "step": 1449 + }, + { + "epoch": 0.1743521914266819, + "flos": 25155803532000.0, + "grad_norm": 3.919195589714115, + "learning_rate": 3.7855008848991363e-06, + "loss": 0.6087, + "num_input_tokens_seen": 30633485, + "step": 1450 + }, + { + "epoch": 0.17447243431732098, + "flos": 25629829678560.0, + "grad_norm": 2.9024661463193304, + "learning_rate": 3.7851497826400714e-06, + "loss": 0.7751, + "num_input_tokens_seen": 30653625, + "step": 1451 + }, + { + "epoch": 0.17459267720796007, + "flos": 36229674655200.0, + "grad_norm": 2.631509701922038, + "learning_rate": 3.7847984095752034e-06, + "loss": 0.7608, + "num_input_tokens_seen": 30677080, + "step": 1452 + }, + { + "epoch": 0.17471292009859918, + "flos": 19982594238240.0, + "grad_norm": 2.485714149631691, + "learning_rate": 3.784446765757836e-06, + "loss": 0.8042, + "num_input_tokens_seen": 30695725, + "step": 1453 + }, + { + "epoch": 0.17483316298923826, + "flos": 27779360653920.0, + "grad_norm": 2.5413066629902987, + "learning_rate": 3.7840948512413133e-06, + "loss": 0.781, + "num_input_tokens_seen": 30713310, + "step": 1454 + }, + { + "epoch": 0.17495340587987734, + "flos": 43954431778560.0, + "grad_norm": 2.0620838092159275, + "learning_rate": 3.7837426660790196e-06, + "loss": 0.7898, + "num_input_tokens_seen": 30734725, + "step": 1455 + }, + { + "epoch": 0.17507364877051645, + "flos": 20855998186080.0, + "grad_norm": 2.077872813131562, + "learning_rate": 3.783390210324382e-06, + "loss": 0.8152, + "num_input_tokens_seen": 30754450, + "step": 1456 + }, + { + "epoch": 0.17519389166115554, + "flos": 24682223422560.0, + "grad_norm": 2.791153836949177, + "learning_rate": 3.7830374840308676e-06, + "loss": 0.7223, + "num_input_tokens_seen": 30774605, + "step": 1457 + }, + { + "epoch": 0.17531413455179462, + "flos": 23763838331520.0, + "grad_norm": 8.186521515361965, + "learning_rate": 3.7826844872519842e-06, + "loss": 0.8347, + "num_input_tokens_seen": 30792460, + "step": 1458 + }, + { + "epoch": 0.1754343774424337, + "flos": 24537238424160.0, + "grad_norm": 2.2338644658215405, + "learning_rate": 3.782331220041282e-06, + "loss": 0.7284, + "num_input_tokens_seen": 30812525, + "step": 1459 + }, + { + "epoch": 0.17555462033307281, + "flos": 18088868516640.0, + "grad_norm": 2.2111430257967903, + "learning_rate": 3.7819776824523504e-06, + "loss": 0.827, + "num_input_tokens_seen": 30830390, + "step": 1460 + }, + { + "epoch": 0.1756748632237119, + "flos": 28325321753280.0, + "grad_norm": 2.1362979342025255, + "learning_rate": 3.7816238745388213e-06, + "loss": 0.8402, + "num_input_tokens_seen": 30849935, + "step": 1461 + }, + { + "epoch": 0.17579510611435098, + "flos": 25695854639520.0, + "grad_norm": 2.113718491764397, + "learning_rate": 3.781269796354367e-06, + "loss": 0.8715, + "num_input_tokens_seen": 30869555, + "step": 1462 + }, + { + "epoch": 0.1759153490049901, + "flos": 18561630891360.0, + "grad_norm": 2.7124263841334497, + "learning_rate": 3.7809154479527006e-06, + "loss": 0.8593, + "num_input_tokens_seen": 30888760, + "step": 1463 + }, + { + "epoch": 0.17603559189562917, + "flos": 18592636204800.0, + "grad_norm": 2.333720095687542, + "learning_rate": 3.780560829387577e-06, + "loss": 0.8433, + "num_input_tokens_seen": 30907260, + "step": 1464 + }, + { + "epoch": 0.17615583478626826, + "flos": 60444004941600.0, + "grad_norm": 0.8532651625414057, + "learning_rate": 3.7802059407127915e-06, + "loss": 0.5787, + "num_input_tokens_seen": 30965810, + "step": 1465 + }, + { + "epoch": 0.17627607767690734, + "flos": 23582198216160.0, + "grad_norm": 2.418709833959993, + "learning_rate": 3.7798507819821797e-06, + "loss": 0.8641, + "num_input_tokens_seen": 30982935, + "step": 1466 + }, + { + "epoch": 0.17639632056754645, + "flos": 17613987465600.0, + "grad_norm": 2.9630988694740377, + "learning_rate": 3.7794953532496197e-06, + "loss": 0.787, + "num_input_tokens_seen": 30998840, + "step": 1467 + }, + { + "epoch": 0.17651656345818553, + "flos": 57850666834560.0, + "grad_norm": 0.8625224649572016, + "learning_rate": 3.7791396545690295e-06, + "loss": 0.5789, + "num_input_tokens_seen": 31060075, + "step": 1468 + }, + { + "epoch": 0.17663680634882462, + "flos": 22896716073120.0, + "grad_norm": 2.3261299545905234, + "learning_rate": 3.7787836859943685e-06, + "loss": 0.8084, + "num_input_tokens_seen": 31078800, + "step": 1469 + }, + { + "epoch": 0.17675704923946373, + "flos": 22605630983520.0, + "grad_norm": 3.1144591511919626, + "learning_rate": 3.7784274475796363e-06, + "loss": 0.7893, + "num_input_tokens_seen": 31097430, + "step": 1470 + }, + { + "epoch": 0.1768772921301028, + "flos": 27087671160960.0, + "grad_norm": 2.913204936232307, + "learning_rate": 3.7780709393788745e-06, + "loss": 0.7563, + "num_input_tokens_seen": 31117905, + "step": 1471 + }, + { + "epoch": 0.1769975350207419, + "flos": 19144841823840.0, + "grad_norm": 2.550526441199601, + "learning_rate": 3.777714161446165e-06, + "loss": 0.7538, + "num_input_tokens_seen": 31137610, + "step": 1472 + }, + { + "epoch": 0.177117777911381, + "flos": 36082831168800.0, + "grad_norm": 2.4471357574468064, + "learning_rate": 3.7773571138356304e-06, + "loss": 0.6951, + "num_input_tokens_seen": 31157780, + "step": 1473 + }, + { + "epoch": 0.17723802080202009, + "flos": 22059037998240.0, + "grad_norm": 2.3532515128323417, + "learning_rate": 3.776999796601435e-06, + "loss": 0.8892, + "num_input_tokens_seen": 31176540, + "step": 1474 + }, + { + "epoch": 0.17735826369265917, + "flos": 30179121419040.0, + "grad_norm": 2.1935939374660465, + "learning_rate": 3.776642209797783e-06, + "loss": 0.7276, + "num_input_tokens_seen": 31198370, + "step": 1475 + }, + { + "epoch": 0.17747850658329825, + "flos": 21367051147200.0, + "grad_norm": 2.3123907114958544, + "learning_rate": 3.7762843534789205e-06, + "loss": 0.7809, + "num_input_tokens_seen": 31217120, + "step": 1476 + }, + { + "epoch": 0.17759874947393736, + "flos": 16959250447680.0, + "grad_norm": 2.6577094164037227, + "learning_rate": 3.7759262276991343e-06, + "loss": 0.8866, + "num_input_tokens_seen": 31234730, + "step": 1477 + }, + { + "epoch": 0.17771899236457644, + "flos": 11529901372320.0, + "grad_norm": 3.5841621209329886, + "learning_rate": 3.7755678325127506e-06, + "loss": 0.8017, + "num_input_tokens_seen": 31252410, + "step": 1478 + }, + { + "epoch": 0.17783923525521553, + "flos": 18781004046720.0, + "grad_norm": 2.1066657858657605, + "learning_rate": 3.7752091679741393e-06, + "loss": 0.7528, + "num_input_tokens_seen": 31270200, + "step": 1479 + }, + { + "epoch": 0.17795947814585464, + "flos": 30364664359200.0, + "grad_norm": 3.3279284208742967, + "learning_rate": 3.774850234137708e-06, + "loss": 0.7864, + "num_input_tokens_seen": 31287495, + "step": 1480 + }, + { + "epoch": 0.17807972103649372, + "flos": 24354148688160.0, + "grad_norm": 2.4022447022899596, + "learning_rate": 3.7744910310579076e-06, + "loss": 0.8287, + "num_input_tokens_seen": 31306420, + "step": 1481 + }, + { + "epoch": 0.1781999639271328, + "flos": 20272564235040.0, + "grad_norm": 2.8378385367079653, + "learning_rate": 3.774131558789229e-06, + "loss": 0.8569, + "num_input_tokens_seen": 31325750, + "step": 1482 + }, + { + "epoch": 0.1783202068177719, + "flos": 15901901859360.0, + "grad_norm": 2.492814801231681, + "learning_rate": 3.773771817386203e-06, + "loss": 0.6958, + "num_input_tokens_seen": 31343080, + "step": 1483 + }, + { + "epoch": 0.178440449708411, + "flos": 20602348778400.0, + "grad_norm": 1.7085272826220443, + "learning_rate": 3.773411806903403e-06, + "loss": 0.7901, + "num_input_tokens_seen": 31362160, + "step": 1484 + }, + { + "epoch": 0.17856069259905008, + "flos": 21654976807200.0, + "grad_norm": 2.249053332071725, + "learning_rate": 3.7730515273954415e-06, + "loss": 0.9498, + "num_input_tokens_seen": 31380770, + "step": 1485 + }, + { + "epoch": 0.17868093548968916, + "flos": 26534684976960.0, + "grad_norm": 2.625546080691077, + "learning_rate": 3.772690978916973e-06, + "loss": 0.8495, + "num_input_tokens_seen": 31400525, + "step": 1486 + }, + { + "epoch": 0.17880117838032827, + "flos": 18554308448640.0, + "grad_norm": 2.6393763733267908, + "learning_rate": 3.772330161522693e-06, + "loss": 0.8668, + "num_input_tokens_seen": 31418435, + "step": 1487 + }, + { + "epoch": 0.17892142127096736, + "flos": 26503159286880.0, + "grad_norm": 3.8233838675057497, + "learning_rate": 3.7719690752673365e-06, + "loss": 0.7942, + "num_input_tokens_seen": 31439230, + "step": 1488 + }, + { + "epoch": 0.17904166416160644, + "flos": 23838263658240.0, + "grad_norm": 2.2556668021835904, + "learning_rate": 3.7716077202056796e-06, + "loss": 0.7811, + "num_input_tokens_seen": 31457705, + "step": 1489 + }, + { + "epoch": 0.17916190705224552, + "flos": 19107145953600.0, + "grad_norm": 5.001275281396498, + "learning_rate": 3.7712460963925404e-06, + "loss": 0.9386, + "num_input_tokens_seen": 31473645, + "step": 1490 + }, + { + "epoch": 0.17928214994288463, + "flos": 25115988985440.0, + "grad_norm": 2.3655192221773897, + "learning_rate": 3.7708842038827775e-06, + "loss": 0.7565, + "num_input_tokens_seen": 31492605, + "step": 1491 + }, + { + "epoch": 0.17940239283352372, + "flos": 22352724971040.0, + "grad_norm": 2.3250006891065005, + "learning_rate": 3.770522042731288e-06, + "loss": 0.8593, + "num_input_tokens_seen": 31514500, + "step": 1492 + }, + { + "epoch": 0.1795226357241628, + "flos": 23145310393440.0, + "grad_norm": 2.2896038365740456, + "learning_rate": 3.7701596129930122e-06, + "loss": 0.8801, + "num_input_tokens_seen": 31533225, + "step": 1493 + }, + { + "epoch": 0.1796428786148019, + "flos": 22058852149440.0, + "grad_norm": 2.647761705988055, + "learning_rate": 3.7697969147229315e-06, + "loss": 0.7381, + "num_input_tokens_seen": 31551385, + "step": 1494 + }, + { + "epoch": 0.179763121505441, + "flos": 21294186950400.0, + "grad_norm": 2.6234010640488776, + "learning_rate": 3.7694339479760647e-06, + "loss": 0.856, + "num_input_tokens_seen": 31570415, + "step": 1495 + }, + { + "epoch": 0.17988336439608008, + "flos": 68064186329280.0, + "grad_norm": 0.7759265728971154, + "learning_rate": 3.769070712807476e-06, + "loss": 0.573, + "num_input_tokens_seen": 31632445, + "step": 1496 + }, + { + "epoch": 0.18000360728671919, + "flos": 21913718472000.0, + "grad_norm": 2.0178058270513475, + "learning_rate": 3.768707209272266e-06, + "loss": 0.7907, + "num_input_tokens_seen": 31652575, + "step": 1497 + }, + { + "epoch": 0.18012385017735827, + "flos": 18959187374400.0, + "grad_norm": 2.9212577290202164, + "learning_rate": 3.768343437425579e-06, + "loss": 0.766, + "num_input_tokens_seen": 31671145, + "step": 1498 + }, + { + "epoch": 0.18024409306799735, + "flos": 19719689560320.0, + "grad_norm": 2.967180660454861, + "learning_rate": 3.7679793973225987e-06, + "loss": 0.8561, + "num_input_tokens_seen": 31686955, + "step": 1499 + }, + { + "epoch": 0.18036433595863643, + "flos": 67131373637760.0, + "grad_norm": 0.8561075690766449, + "learning_rate": 3.767615089018549e-06, + "loss": 0.6159, + "num_input_tokens_seen": 31749300, + "step": 1500 + }, + { + "epoch": 0.18048457884927555, + "flos": 18155413854240.0, + "grad_norm": 3.0788777781527368, + "learning_rate": 3.7672505125686966e-06, + "loss": 0.864, + "num_input_tokens_seen": 31765665, + "step": 1501 + }, + { + "epoch": 0.18060482173991463, + "flos": 15791304622560.0, + "grad_norm": 3.43887010492862, + "learning_rate": 3.7668856680283455e-06, + "loss": 0.8459, + "num_input_tokens_seen": 31782690, + "step": 1502 + }, + { + "epoch": 0.1807250646305537, + "flos": 18560924665920.0, + "grad_norm": 2.6899402983326137, + "learning_rate": 3.7665205554528437e-06, + "loss": 0.8234, + "num_input_tokens_seen": 31802045, + "step": 1503 + }, + { + "epoch": 0.18084530752119282, + "flos": 23116052058720.0, + "grad_norm": 2.088459105219344, + "learning_rate": 3.7661551748975782e-06, + "loss": 0.7458, + "num_input_tokens_seen": 31820220, + "step": 1504 + }, + { + "epoch": 0.1809655504118319, + "flos": 59717518819680.0, + "grad_norm": 0.816409097328879, + "learning_rate": 3.7657895264179772e-06, + "loss": 0.6048, + "num_input_tokens_seen": 31876195, + "step": 1505 + }, + { + "epoch": 0.181085793302471, + "flos": 44138227740000.0, + "grad_norm": 1.929960285628065, + "learning_rate": 3.765423610069509e-06, + "loss": 0.742, + "num_input_tokens_seen": 31901585, + "step": 1506 + }, + { + "epoch": 0.18120603619311007, + "flos": 34848228496800.0, + "grad_norm": 1.9880936989453322, + "learning_rate": 3.765057425907683e-06, + "loss": 0.7219, + "num_input_tokens_seen": 31923045, + "step": 1507 + }, + { + "epoch": 0.18132627908374918, + "flos": 21476050084320.0, + "grad_norm": 2.1162907013694117, + "learning_rate": 3.764690973988048e-06, + "loss": 0.7847, + "num_input_tokens_seen": 31943385, + "step": 1508 + }, + { + "epoch": 0.18144652197438826, + "flos": 29015933323200.0, + "grad_norm": 10.132151397573931, + "learning_rate": 3.7643242543661967e-06, + "loss": 0.7359, + "num_input_tokens_seen": 31966045, + "step": 1509 + }, + { + "epoch": 0.18156676486502735, + "flos": 68576726080800.0, + "grad_norm": 0.8158077254544688, + "learning_rate": 3.7639572670977573e-06, + "loss": 0.6053, + "num_input_tokens_seen": 32021540, + "step": 1510 + }, + { + "epoch": 0.18168700775566646, + "flos": 26433677538240.0, + "grad_norm": 1.7042570638892138, + "learning_rate": 3.7635900122384042e-06, + "loss": 0.7645, + "num_input_tokens_seen": 32044535, + "step": 1511 + }, + { + "epoch": 0.18180725064630554, + "flos": 14984185824000.0, + "grad_norm": 2.657098073772404, + "learning_rate": 3.7632224898438477e-06, + "loss": 0.8659, + "num_input_tokens_seen": 32061010, + "step": 1512 + }, + { + "epoch": 0.18192749353694462, + "flos": 19654333655040.0, + "grad_norm": 1.8070027086483573, + "learning_rate": 3.762854699969842e-06, + "loss": 0.7937, + "num_input_tokens_seen": 32081520, + "step": 1513 + }, + { + "epoch": 0.1820477364275837, + "flos": 20673131468640.0, + "grad_norm": 2.770388450377511, + "learning_rate": 3.762486642672179e-06, + "loss": 0.7321, + "num_input_tokens_seen": 32098460, + "step": 1514 + }, + { + "epoch": 0.18216797931822282, + "flos": 17104384125120.0, + "grad_norm": 2.38698425314623, + "learning_rate": 3.7621183180066946e-06, + "loss": 0.8701, + "num_input_tokens_seen": 32116420, + "step": 1515 + }, + { + "epoch": 0.1822882222088619, + "flos": 29200621358880.0, + "grad_norm": 1.8026836549313, + "learning_rate": 3.7617497260292625e-06, + "loss": 0.7389, + "num_input_tokens_seen": 32138475, + "step": 1516 + }, + { + "epoch": 0.18240846509950098, + "flos": 17677224694560.0, + "grad_norm": 3.2731671278556798, + "learning_rate": 3.7613808667957967e-06, + "loss": 0.7876, + "num_input_tokens_seen": 32151165, + "step": 1517 + }, + { + "epoch": 0.1825287079901401, + "flos": 14768975681760.0, + "grad_norm": 2.8955819183181326, + "learning_rate": 3.7610117403622547e-06, + "loss": 0.9117, + "num_input_tokens_seen": 32167725, + "step": 1518 + }, + { + "epoch": 0.18264895088077918, + "flos": 21914945074080.0, + "grad_norm": 1.8514616089708147, + "learning_rate": 3.7606423467846313e-06, + "loss": 0.9029, + "num_input_tokens_seen": 32187330, + "step": 1519 + }, + { + "epoch": 0.18276919377141826, + "flos": 20856964599840.0, + "grad_norm": 1.7534091304427764, + "learning_rate": 3.760272686118964e-06, + "loss": 0.7998, + "num_input_tokens_seen": 32205550, + "step": 1520 + }, + { + "epoch": 0.18288943666205737, + "flos": 21439209118560.0, + "grad_norm": 2.2398764366742654, + "learning_rate": 3.7599027584213297e-06, + "loss": 0.9278, + "num_input_tokens_seen": 32224430, + "step": 1521 + }, + { + "epoch": 0.18300967955269645, + "flos": 21508467848640.0, + "grad_norm": 3.865048211748229, + "learning_rate": 3.7595325637478465e-06, + "loss": 0.7837, + "num_input_tokens_seen": 32242455, + "step": 1522 + }, + { + "epoch": 0.18312992244333554, + "flos": 28834776414720.0, + "grad_norm": 1.8228202213023759, + "learning_rate": 3.7591621021546723e-06, + "loss": 0.8169, + "num_input_tokens_seen": 32264450, + "step": 1523 + }, + { + "epoch": 0.18325016533397462, + "flos": 20091407326560.0, + "grad_norm": 1.970044936330824, + "learning_rate": 3.7587913736980062e-06, + "loss": 0.8153, + "num_input_tokens_seen": 32284090, + "step": 1524 + }, + { + "epoch": 0.18337040822461373, + "flos": 23295759346560.0, + "grad_norm": 2.471669655624644, + "learning_rate": 3.7584203784340865e-06, + "loss": 0.8431, + "num_input_tokens_seen": 32303260, + "step": 1525 + }, + { + "epoch": 0.1834906511152528, + "flos": 24973754549280.0, + "grad_norm": 2.793853064903862, + "learning_rate": 3.7580491164191938e-06, + "loss": 0.8584, + "num_input_tokens_seen": 32321570, + "step": 1526 + }, + { + "epoch": 0.1836108940058919, + "flos": 67155205187520.0, + "grad_norm": 0.752340638024538, + "learning_rate": 3.757677587709648e-06, + "loss": 0.6127, + "num_input_tokens_seen": 32384835, + "step": 1527 + }, + { + "epoch": 0.183731136896531, + "flos": 25701838970880.0, + "grad_norm": 3.2649401078857565, + "learning_rate": 3.7573057923618095e-06, + "loss": 0.7557, + "num_input_tokens_seen": 32404930, + "step": 1528 + }, + { + "epoch": 0.1838513797871701, + "flos": 20420262625920.0, + "grad_norm": 2.187254973721253, + "learning_rate": 3.7569337304320793e-06, + "loss": 0.7435, + "num_input_tokens_seen": 32424515, + "step": 1529 + }, + { + "epoch": 0.18397162267780917, + "flos": 68466946578720.0, + "grad_norm": 0.9870686687594833, + "learning_rate": 3.756561401976899e-06, + "loss": 0.644, + "num_input_tokens_seen": 32484220, + "step": 1530 + }, + { + "epoch": 0.18409186556844825, + "flos": 31895184189600.0, + "grad_norm": 2.2300246777045736, + "learning_rate": 3.7561888070527514e-06, + "loss": 0.8282, + "num_input_tokens_seen": 32506260, + "step": 1531 + }, + { + "epoch": 0.18421210845908736, + "flos": 20091853363680.0, + "grad_norm": 2.1337392281795053, + "learning_rate": 3.7558159457161577e-06, + "loss": 0.8017, + "num_input_tokens_seen": 32524225, + "step": 1532 + }, + { + "epoch": 0.18433235134972645, + "flos": 23077389774720.0, + "grad_norm": 2.880632439419788, + "learning_rate": 3.755442818023681e-06, + "loss": 0.778, + "num_input_tokens_seen": 32543850, + "step": 1533 + }, + { + "epoch": 0.18445259424036553, + "flos": 18265156186560.0, + "grad_norm": 1.9864241465237422, + "learning_rate": 3.7550694240319246e-06, + "loss": 0.7624, + "num_input_tokens_seen": 32561725, + "step": 1534 + }, + { + "epoch": 0.18457283713100464, + "flos": 21293629404000.0, + "grad_norm": 2.5799517084377945, + "learning_rate": 3.7546957637975326e-06, + "loss": 0.7603, + "num_input_tokens_seen": 32579135, + "step": 1535 + }, + { + "epoch": 0.18469308002164372, + "flos": 20346617864160.0, + "grad_norm": 1.6311795771221618, + "learning_rate": 3.7543218373771873e-06, + "loss": 0.7418, + "num_input_tokens_seen": 32598460, + "step": 1536 + }, + { + "epoch": 0.1848133229122828, + "flos": 26398137514080.0, + "grad_norm": 1.5184502765049692, + "learning_rate": 3.753947644827615e-06, + "loss": 0.7825, + "num_input_tokens_seen": 32621920, + "step": 1537 + }, + { + "epoch": 0.1849335658029219, + "flos": 70446806101440.0, + "grad_norm": 0.9594067422293932, + "learning_rate": 3.753573186205579e-06, + "loss": 0.5721, + "num_input_tokens_seen": 32690040, + "step": 1538 + }, + { + "epoch": 0.185053808693561, + "flos": 17359445983680.0, + "grad_norm": 2.2226648313762287, + "learning_rate": 3.753198461567885e-06, + "loss": 0.7766, + "num_input_tokens_seen": 32707285, + "step": 1539 + }, + { + "epoch": 0.18517405158420008, + "flos": 28944555916800.0, + "grad_norm": 1.8717109442705029, + "learning_rate": 3.7528234709713783e-06, + "loss": 0.9182, + "num_input_tokens_seen": 32730030, + "step": 1540 + }, + { + "epoch": 0.18529429447483917, + "flos": 26756474166720.0, + "grad_norm": 2.9125447347586295, + "learning_rate": 3.7524482144729447e-06, + "loss": 0.8476, + "num_input_tokens_seen": 32749485, + "step": 1541 + }, + { + "epoch": 0.18541453736547828, + "flos": 13562144554080.0, + "grad_norm": 2.386070150848568, + "learning_rate": 3.7520726921295106e-06, + "loss": 0.8396, + "num_input_tokens_seen": 32766205, + "step": 1542 + }, + { + "epoch": 0.18553478025611736, + "flos": 23991091476000.0, + "grad_norm": 2.440962166642636, + "learning_rate": 3.751696903998042e-06, + "loss": 0.7239, + "num_input_tokens_seen": 32784840, + "step": 1543 + }, + { + "epoch": 0.18565502314675644, + "flos": 25848087741120.0, + "grad_norm": 2.3576973514739885, + "learning_rate": 3.7513208501355456e-06, + "loss": 0.6994, + "num_input_tokens_seen": 32805945, + "step": 1544 + }, + { + "epoch": 0.18577526603739553, + "flos": 19582138513920.0, + "grad_norm": 2.9738672424012553, + "learning_rate": 3.750944530599069e-06, + "loss": 0.8384, + "num_input_tokens_seen": 32825515, + "step": 1545 + }, + { + "epoch": 0.18589550892803464, + "flos": 18448468941120.0, + "grad_norm": 2.1766647106744097, + "learning_rate": 3.7505679454456992e-06, + "loss": 0.8052, + "num_input_tokens_seen": 32842125, + "step": 1546 + }, + { + "epoch": 0.18601575181867372, + "flos": 23516284764480.0, + "grad_norm": 2.1168843295103223, + "learning_rate": 3.750191094732564e-06, + "loss": 0.6978, + "num_input_tokens_seen": 32862990, + "step": 1547 + }, + { + "epoch": 0.1861359947093128, + "flos": 26322448415520.0, + "grad_norm": 3.0284497412683247, + "learning_rate": 3.7498139785168313e-06, + "loss": 0.7598, + "num_input_tokens_seen": 32883595, + "step": 1548 + }, + { + "epoch": 0.1862562375999519, + "flos": 23297840853120.0, + "grad_norm": 2.133327100665207, + "learning_rate": 3.749436596855709e-06, + "loss": 0.7722, + "num_input_tokens_seen": 32902895, + "step": 1549 + }, + { + "epoch": 0.186376480490591, + "flos": 16624373647200.0, + "grad_norm": 2.1313246683146043, + "learning_rate": 3.749058949806446e-06, + "loss": 0.9072, + "num_input_tokens_seen": 32919620, + "step": 1550 + }, + { + "epoch": 0.18649672338123008, + "flos": 21437945346720.0, + "grad_norm": 1.8850557884642456, + "learning_rate": 3.748681037426331e-06, + "loss": 0.8424, + "num_input_tokens_seen": 32938830, + "step": 1551 + }, + { + "epoch": 0.1866169662718692, + "flos": 12295421475840.0, + "grad_norm": 2.44661309995229, + "learning_rate": 3.7483028597726936e-06, + "loss": 0.9167, + "num_input_tokens_seen": 32955600, + "step": 1552 + }, + { + "epoch": 0.18673720916250827, + "flos": 23547364417440.0, + "grad_norm": 2.866472284986496, + "learning_rate": 3.7479244169029017e-06, + "loss": 0.6246, + "num_input_tokens_seen": 32975390, + "step": 1553 + }, + { + "epoch": 0.18685745205314735, + "flos": 19690245376800.0, + "grad_norm": 2.7123820828591843, + "learning_rate": 3.7475457088743658e-06, + "loss": 0.7304, + "num_input_tokens_seen": 32992640, + "step": 1554 + }, + { + "epoch": 0.18697769494378644, + "flos": 34154940704160.0, + "grad_norm": 2.295210046734366, + "learning_rate": 3.7471667357445348e-06, + "loss": 0.7493, + "num_input_tokens_seen": 33012470, + "step": 1555 + }, + { + "epoch": 0.18709793783442555, + "flos": 34192116197760.0, + "grad_norm": 2.681893466662746, + "learning_rate": 3.7467874975709e-06, + "loss": 0.724, + "num_input_tokens_seen": 33033275, + "step": 1556 + }, + { + "epoch": 0.18721818072506463, + "flos": 40675840280640.0, + "grad_norm": 2.7370435499657617, + "learning_rate": 3.7464079944109904e-06, + "loss": 0.7806, + "num_input_tokens_seen": 33055135, + "step": 1557 + }, + { + "epoch": 0.18733842361570371, + "flos": 22125508996320.0, + "grad_norm": 2.598858043338468, + "learning_rate": 3.746028226322376e-06, + "loss": 0.7763, + "num_input_tokens_seen": 33071015, + "step": 1558 + }, + { + "epoch": 0.18745866650634282, + "flos": 18887958647040.0, + "grad_norm": 2.063998400539695, + "learning_rate": 3.745648193362669e-06, + "loss": 0.7532, + "num_input_tokens_seen": 33090370, + "step": 1559 + }, + { + "epoch": 0.1875789093969819, + "flos": 19287001920480.0, + "grad_norm": 2.721912605385, + "learning_rate": 3.745267895589518e-06, + "loss": 0.7217, + "num_input_tokens_seen": 33108910, + "step": 1560 + }, + { + "epoch": 0.187699152287621, + "flos": 16994790471840.0, + "grad_norm": 2.217264779742173, + "learning_rate": 3.7448873330606154e-06, + "loss": 0.8221, + "num_input_tokens_seen": 33126680, + "step": 1561 + }, + { + "epoch": 0.18781939517826007, + "flos": 22314471554400.0, + "grad_norm": 3.3776734677599567, + "learning_rate": 3.7445065058336914e-06, + "loss": 0.8728, + "num_input_tokens_seen": 33145550, + "step": 1562 + }, + { + "epoch": 0.18793963806889918, + "flos": 14590829523840.0, + "grad_norm": 2.0467510622664085, + "learning_rate": 3.7441254139665176e-06, + "loss": 0.8634, + "num_input_tokens_seen": 33161735, + "step": 1563 + }, + { + "epoch": 0.18805988095953827, + "flos": 17432273010720.0, + "grad_norm": 2.155937725066574, + "learning_rate": 3.743744057516905e-06, + "loss": 0.8254, + "num_input_tokens_seen": 33179795, + "step": 1564 + }, + { + "epoch": 0.18818012385017735, + "flos": 15022402070880.0, + "grad_norm": 3.1783696473368708, + "learning_rate": 3.743362436542706e-06, + "loss": 0.8761, + "num_input_tokens_seen": 33194485, + "step": 1565 + }, + { + "epoch": 0.18830036674081646, + "flos": 47483141556960.0, + "grad_norm": 3.067556138896761, + "learning_rate": 3.7429805511018115e-06, + "loss": 0.7675, + "num_input_tokens_seen": 33216145, + "step": 1566 + }, + { + "epoch": 0.18842060963145554, + "flos": 30001495637760.0, + "grad_norm": 2.028886329456139, + "learning_rate": 3.7425984012521524e-06, + "loss": 0.7826, + "num_input_tokens_seen": 33236345, + "step": 1567 + }, + { + "epoch": 0.18854085252209463, + "flos": 70217694468960.0, + "grad_norm": 0.7454486474657079, + "learning_rate": 3.7422159870517025e-06, + "loss": 0.604, + "num_input_tokens_seen": 33301600, + "step": 1568 + }, + { + "epoch": 0.1886610954127337, + "flos": 21258981454080.0, + "grad_norm": 1.8258243658535336, + "learning_rate": 3.7418333085584717e-06, + "loss": 0.7873, + "num_input_tokens_seen": 33322785, + "step": 1569 + }, + { + "epoch": 0.18878133830337282, + "flos": 17242009511040.0, + "grad_norm": 2.2516289218738614, + "learning_rate": 3.7414503658305128e-06, + "loss": 0.9113, + "num_input_tokens_seen": 33340020, + "step": 1570 + }, + { + "epoch": 0.1889015811940119, + "flos": 25738271069280.0, + "grad_norm": 2.6784962489158994, + "learning_rate": 3.7410671589259185e-06, + "loss": 0.7796, + "num_input_tokens_seen": 33358620, + "step": 1571 + }, + { + "epoch": 0.18902182408465099, + "flos": 21002172616800.0, + "grad_norm": 2.2639399082647143, + "learning_rate": 3.7406836879028205e-06, + "loss": 0.7955, + "num_input_tokens_seen": 33378845, + "step": 1572 + }, + { + "epoch": 0.1891420669752901, + "flos": 22240269246240.0, + "grad_norm": 2.2094656962079604, + "learning_rate": 3.7402999528193907e-06, + "loss": 0.7706, + "num_input_tokens_seen": 33398345, + "step": 1573 + }, + { + "epoch": 0.18926230986592918, + "flos": 21986359650240.0, + "grad_norm": 2.972619997623066, + "learning_rate": 3.739915953733842e-06, + "loss": 0.8568, + "num_input_tokens_seen": 33416390, + "step": 1574 + }, + { + "epoch": 0.18938255275656826, + "flos": 24427681940640.0, + "grad_norm": 2.0327736641929484, + "learning_rate": 3.7395316907044264e-06, + "loss": 0.8177, + "num_input_tokens_seen": 33437175, + "step": 1575 + }, + { + "epoch": 0.18950279564720737, + "flos": 24392141916480.0, + "grad_norm": 1.953704992008973, + "learning_rate": 3.7391471637894364e-06, + "loss": 0.797, + "num_input_tokens_seen": 33459440, + "step": 1576 + }, + { + "epoch": 0.18962303853784646, + "flos": 19728164265600.0, + "grad_norm": 1.9394440280973717, + "learning_rate": 3.738762373047205e-06, + "loss": 0.8483, + "num_input_tokens_seen": 33479360, + "step": 1577 + }, + { + "epoch": 0.18974328142848554, + "flos": 21002023937760.0, + "grad_norm": 1.7198678283031739, + "learning_rate": 3.738377318536103e-06, + "loss": 0.8307, + "num_input_tokens_seen": 33499245, + "step": 1578 + }, + { + "epoch": 0.18986352431912462, + "flos": 12947742459360.0, + "grad_norm": 2.505141692884501, + "learning_rate": 3.7379920003145447e-06, + "loss": 0.7168, + "num_input_tokens_seen": 33513520, + "step": 1579 + }, + { + "epoch": 0.18998376720976373, + "flos": 23733873771360.0, + "grad_norm": 2.546821827713737, + "learning_rate": 3.7376064184409817e-06, + "loss": 0.8366, + "num_input_tokens_seen": 33533100, + "step": 1580 + }, + { + "epoch": 0.19010401010040281, + "flos": 22933185341280.0, + "grad_norm": 2.5401030770360036, + "learning_rate": 3.7372205729739063e-06, + "loss": 0.872, + "num_input_tokens_seen": 33554915, + "step": 1581 + }, + { + "epoch": 0.1902242529910419, + "flos": 19108409725440.0, + "grad_norm": 4.568509259645727, + "learning_rate": 3.7368344639718514e-06, + "loss": 0.7192, + "num_input_tokens_seen": 33572850, + "step": 1582 + }, + { + "epoch": 0.190344495881681, + "flos": 25447446168000.0, + "grad_norm": 1.9357802234030808, + "learning_rate": 3.7364480914933895e-06, + "loss": 0.8087, + "num_input_tokens_seen": 33593090, + "step": 1583 + }, + { + "epoch": 0.1904647387723201, + "flos": 26754281150880.0, + "grad_norm": 2.201159960385345, + "learning_rate": 3.7360614555971325e-06, + "loss": 0.8108, + "num_input_tokens_seen": 33612745, + "step": 1584 + }, + { + "epoch": 0.19058498166295917, + "flos": 23953432775520.0, + "grad_norm": 2.540402290746118, + "learning_rate": 3.735674556341733e-06, + "loss": 0.8541, + "num_input_tokens_seen": 33632360, + "step": 1585 + }, + { + "epoch": 0.19070522455359826, + "flos": 28252717744800.0, + "grad_norm": 2.4874892958214256, + "learning_rate": 3.7352873937858835e-06, + "loss": 0.8285, + "num_input_tokens_seen": 33653895, + "step": 1586 + }, + { + "epoch": 0.19082546744423737, + "flos": 25623250631040.0, + "grad_norm": 3.123684415460784, + "learning_rate": 3.734899967988316e-06, + "loss": 0.7205, + "num_input_tokens_seen": 33672075, + "step": 1587 + }, + { + "epoch": 0.19094571033487645, + "flos": 19691397639360.0, + "grad_norm": 2.1756438343752813, + "learning_rate": 3.7345122790078026e-06, + "loss": 0.8376, + "num_input_tokens_seen": 33689640, + "step": 1588 + }, + { + "epoch": 0.19106595322551553, + "flos": 21585457888800.0, + "grad_norm": 2.9856396936314984, + "learning_rate": 3.7341243269031556e-06, + "loss": 0.9299, + "num_input_tokens_seen": 33708630, + "step": 1589 + }, + { + "epoch": 0.19118619611615464, + "flos": 29853908756160.0, + "grad_norm": 1.7241039464330679, + "learning_rate": 3.7337361117332275e-06, + "loss": 0.7727, + "num_input_tokens_seen": 33730370, + "step": 1590 + }, + { + "epoch": 0.19130643900679373, + "flos": 17248848746880.0, + "grad_norm": 2.372394739032804, + "learning_rate": 3.7333476335569087e-06, + "loss": 0.7706, + "num_input_tokens_seen": 33748370, + "step": 1591 + }, + { + "epoch": 0.1914266818974328, + "flos": 24790776322560.0, + "grad_norm": 3.00703590573084, + "learning_rate": 3.7329588924331325e-06, + "loss": 0.6655, + "num_input_tokens_seen": 33769575, + "step": 1592 + }, + { + "epoch": 0.1915469247880719, + "flos": 18925468668480.0, + "grad_norm": 2.444242493157429, + "learning_rate": 3.732569888420871e-06, + "loss": 0.8252, + "num_input_tokens_seen": 33789110, + "step": 1593 + }, + { + "epoch": 0.191667167678711, + "flos": 21002953181760.0, + "grad_norm": 2.9721589443169427, + "learning_rate": 3.732180621579134e-06, + "loss": 0.8282, + "num_input_tokens_seen": 33808325, + "step": 1594 + }, + { + "epoch": 0.1917874105693501, + "flos": 34188324882240.0, + "grad_norm": 2.5980235363582045, + "learning_rate": 3.7317910919669745e-06, + "loss": 0.8118, + "num_input_tokens_seen": 33829520, + "step": 1595 + }, + { + "epoch": 0.19190765345998917, + "flos": 23516953820160.0, + "grad_norm": 2.589116565459843, + "learning_rate": 3.7314012996434826e-06, + "loss": 0.7658, + "num_input_tokens_seen": 33848250, + "step": 1596 + }, + { + "epoch": 0.19202789635062828, + "flos": 19832888680320.0, + "grad_norm": 2.348817458795475, + "learning_rate": 3.7310112446677907e-06, + "loss": 0.8141, + "num_input_tokens_seen": 33866160, + "step": 1597 + }, + { + "epoch": 0.19214813924126736, + "flos": 20893099340160.0, + "grad_norm": 2.784574086685122, + "learning_rate": 3.7306209270990695e-06, + "loss": 0.6874, + "num_input_tokens_seen": 33884165, + "step": 1598 + }, + { + "epoch": 0.19226838213190645, + "flos": 26321965208640.0, + "grad_norm": 2.0454032207561754, + "learning_rate": 3.7302303469965292e-06, + "loss": 0.8658, + "num_input_tokens_seen": 33903705, + "step": 1599 + }, + { + "epoch": 0.19238862502254553, + "flos": 20820792689760.0, + "grad_norm": 2.0053090916539693, + "learning_rate": 3.7298395044194206e-06, + "loss": 0.7053, + "num_input_tokens_seen": 33922515, + "step": 1600 + }, + { + "epoch": 0.19250886791318464, + "flos": 21695051542080.0, + "grad_norm": 2.7541744453293395, + "learning_rate": 3.7294483994270356e-06, + "loss": 0.9426, + "num_input_tokens_seen": 33940840, + "step": 1601 + }, + { + "epoch": 0.19262911080382372, + "flos": 23334421630560.0, + "grad_norm": 2.5350922986444666, + "learning_rate": 3.7290570320787033e-06, + "loss": 0.7832, + "num_input_tokens_seen": 33960420, + "step": 1602 + }, + { + "epoch": 0.1927493536944628, + "flos": 21912120172320.0, + "grad_norm": 3.5745853423681253, + "learning_rate": 3.728665402433793e-06, + "loss": 0.7151, + "num_input_tokens_seen": 33978990, + "step": 1603 + }, + { + "epoch": 0.19286959658510192, + "flos": 16522139606400.0, + "grad_norm": 3.3434503724089026, + "learning_rate": 3.7282735105517164e-06, + "loss": 0.8592, + "num_input_tokens_seen": 33997115, + "step": 1604 + }, + { + "epoch": 0.192989839475741, + "flos": 21586126944480.0, + "grad_norm": 2.3295887691420964, + "learning_rate": 3.727881356491922e-06, + "loss": 0.673, + "num_input_tokens_seen": 34015525, + "step": 1605 + }, + { + "epoch": 0.19311008236638008, + "flos": 19254063779520.0, + "grad_norm": 2.135663819935672, + "learning_rate": 3.7274889403139002e-06, + "loss": 0.7581, + "num_input_tokens_seen": 34033150, + "step": 1606 + }, + { + "epoch": 0.1932303252570192, + "flos": 28611054397440.0, + "grad_norm": 2.8342014179680852, + "learning_rate": 3.727096262077179e-06, + "loss": 0.7819, + "num_input_tokens_seen": 34052145, + "step": 1607 + }, + { + "epoch": 0.19335056814765827, + "flos": 18342852452160.0, + "grad_norm": 2.0339733940342386, + "learning_rate": 3.7267033218413285e-06, + "loss": 0.8502, + "num_input_tokens_seen": 34069940, + "step": 1608 + }, + { + "epoch": 0.19347081103829736, + "flos": 13242172827360.0, + "grad_norm": 2.904825168936155, + "learning_rate": 3.726310119665957e-06, + "loss": 0.8141, + "num_input_tokens_seen": 34086275, + "step": 1609 + }, + { + "epoch": 0.19359105392893644, + "flos": 20271077444640.0, + "grad_norm": 2.238738807126281, + "learning_rate": 3.725916655610713e-06, + "loss": 0.8535, + "num_input_tokens_seen": 34105605, + "step": 1610 + }, + { + "epoch": 0.19371129681957555, + "flos": 20455356612960.0, + "grad_norm": 5.783636806708566, + "learning_rate": 3.725522929735284e-06, + "loss": 0.7539, + "num_input_tokens_seen": 34122540, + "step": 1611 + }, + { + "epoch": 0.19383153971021463, + "flos": 30401765513280.0, + "grad_norm": 2.693850621160798, + "learning_rate": 3.725128942099399e-06, + "loss": 0.7433, + "num_input_tokens_seen": 34142940, + "step": 1612 + }, + { + "epoch": 0.19395178260085372, + "flos": 24534710880480.0, + "grad_norm": 1.8560066166340516, + "learning_rate": 3.7247346927628245e-06, + "loss": 0.7995, + "num_input_tokens_seen": 34162875, + "step": 1613 + }, + { + "epoch": 0.19407202549149283, + "flos": 28909090232160.0, + "grad_norm": 2.2593418736225592, + "learning_rate": 3.7243401817853694e-06, + "loss": 0.7878, + "num_input_tokens_seen": 34183565, + "step": 1614 + }, + { + "epoch": 0.1941922683821319, + "flos": 17978903165760.0, + "grad_norm": 3.052740622994246, + "learning_rate": 3.723945409226879e-06, + "loss": 0.7162, + "num_input_tokens_seen": 34202055, + "step": 1615 + }, + { + "epoch": 0.194312511272771, + "flos": 9706735322400.0, + "grad_norm": 3.330999574677833, + "learning_rate": 3.723550375147241e-06, + "loss": 0.803, + "num_input_tokens_seen": 34216830, + "step": 1616 + }, + { + "epoch": 0.19443275416341008, + "flos": 26977482791520.0, + "grad_norm": 2.4644415235525985, + "learning_rate": 3.7231550796063816e-06, + "loss": 0.7983, + "num_input_tokens_seen": 34236840, + "step": 1617 + }, + { + "epoch": 0.1945529970540492, + "flos": 15823982575200.0, + "grad_norm": 2.187086883772661, + "learning_rate": 3.722759522664266e-06, + "loss": 0.6528, + "num_input_tokens_seen": 34254140, + "step": 1618 + }, + { + "epoch": 0.19467323994468827, + "flos": 19287373618080.0, + "grad_norm": 2.0121844595198772, + "learning_rate": 3.7223637043809016e-06, + "loss": 0.8205, + "num_input_tokens_seen": 34273120, + "step": 1619 + }, + { + "epoch": 0.19479348283532735, + "flos": 24098677962240.0, + "grad_norm": 2.3605162026904334, + "learning_rate": 3.7219676248163322e-06, + "loss": 0.8643, + "num_input_tokens_seen": 34290285, + "step": 1620 + }, + { + "epoch": 0.19491372572596646, + "flos": 25738717106400.0, + "grad_norm": 2.177744009227568, + "learning_rate": 3.721571284030643e-06, + "loss": 0.9357, + "num_input_tokens_seen": 34310095, + "step": 1621 + }, + { + "epoch": 0.19503396861660555, + "flos": 19617195331200.0, + "grad_norm": 3.308344988157817, + "learning_rate": 3.7211746820839587e-06, + "loss": 0.7912, + "num_input_tokens_seen": 34327190, + "step": 1622 + }, + { + "epoch": 0.19515421150724463, + "flos": 21003027521280.0, + "grad_norm": 3.541486638299364, + "learning_rate": 3.7207778190364437e-06, + "loss": 0.8063, + "num_input_tokens_seen": 34345615, + "step": 1623 + }, + { + "epoch": 0.1952744543978837, + "flos": 32913796154400.0, + "grad_norm": 1.7140869695260859, + "learning_rate": 3.720380694948302e-06, + "loss": 0.7392, + "num_input_tokens_seen": 34368500, + "step": 1624 + }, + { + "epoch": 0.19539469728852282, + "flos": 63952823164800.0, + "grad_norm": 1.0308974848413974, + "learning_rate": 3.719983309879777e-06, + "loss": 0.7125, + "num_input_tokens_seen": 34428280, + "step": 1625 + }, + { + "epoch": 0.1955149401791619, + "flos": 13352063838720.0, + "grad_norm": 4.239480857016779, + "learning_rate": 3.719585663891151e-06, + "loss": 0.7757, + "num_input_tokens_seen": 34445815, + "step": 1626 + }, + { + "epoch": 0.195635183069801, + "flos": 18702044009280.0, + "grad_norm": 2.4637350015943382, + "learning_rate": 3.719187757042747e-06, + "loss": 0.7931, + "num_input_tokens_seen": 34463635, + "step": 1627 + }, + { + "epoch": 0.1957554259604401, + "flos": 69215362859040.0, + "grad_norm": 0.7311139444789396, + "learning_rate": 3.7187895893949275e-06, + "loss": 0.5492, + "num_input_tokens_seen": 34530105, + "step": 1628 + }, + { + "epoch": 0.19587566885107918, + "flos": 21039199431360.0, + "grad_norm": 3.115825154600779, + "learning_rate": 3.7183911610080937e-06, + "loss": 0.761, + "num_input_tokens_seen": 34546970, + "step": 1629 + }, + { + "epoch": 0.19599591174171827, + "flos": 22162758829440.0, + "grad_norm": 5.532309565380201, + "learning_rate": 3.7179924719426872e-06, + "loss": 0.7491, + "num_input_tokens_seen": 34564865, + "step": 1630 + }, + { + "epoch": 0.19611615463235738, + "flos": 23734356978240.0, + "grad_norm": 2.8913439707115813, + "learning_rate": 3.7175935222591885e-06, + "loss": 0.7587, + "num_input_tokens_seen": 34584165, + "step": 1631 + }, + { + "epoch": 0.19623639752299646, + "flos": 28577038333440.0, + "grad_norm": 2.3764391102758413, + "learning_rate": 3.717194312018118e-06, + "loss": 0.7475, + "num_input_tokens_seen": 34603150, + "step": 1632 + }, + { + "epoch": 0.19635664041363554, + "flos": 21002692993440.0, + "grad_norm": 1.998292074511975, + "learning_rate": 3.716794841280036e-06, + "loss": 0.7602, + "num_input_tokens_seen": 34621855, + "step": 1633 + }, + { + "epoch": 0.19647688330427462, + "flos": 18852864660000.0, + "grad_norm": 3.0056767795518544, + "learning_rate": 3.7163951101055407e-06, + "loss": 0.7754, + "num_input_tokens_seen": 34639915, + "step": 1634 + }, + { + "epoch": 0.19659712619491373, + "flos": 24207788408640.0, + "grad_norm": 2.215181516697109, + "learning_rate": 3.715995118555273e-06, + "loss": 0.7864, + "num_input_tokens_seen": 34659890, + "step": 1635 + }, + { + "epoch": 0.19671736908555282, + "flos": 24682037573760.0, + "grad_norm": 3.262652636936559, + "learning_rate": 3.71559486668991e-06, + "loss": 0.8591, + "num_input_tokens_seen": 34678750, + "step": 1636 + }, + { + "epoch": 0.1968376119761919, + "flos": 23808187588800.0, + "grad_norm": 1.90150147891031, + "learning_rate": 3.715194354570169e-06, + "loss": 0.772, + "num_input_tokens_seen": 34698755, + "step": 1637 + }, + { + "epoch": 0.196957854866831, + "flos": 18088162291200.0, + "grad_norm": 2.341558016318755, + "learning_rate": 3.714793582256809e-06, + "loss": 0.8303, + "num_input_tokens_seen": 34715820, + "step": 1638 + }, + { + "epoch": 0.1970780977574701, + "flos": 21622224515040.0, + "grad_norm": 2.6782165582912576, + "learning_rate": 3.7143925498106253e-06, + "loss": 0.8531, + "num_input_tokens_seen": 34734360, + "step": 1639 + }, + { + "epoch": 0.19719834064810918, + "flos": 20782056066240.0, + "grad_norm": 2.082868053614023, + "learning_rate": 3.7139912572924558e-06, + "loss": 0.7942, + "num_input_tokens_seen": 34753390, + "step": 1640 + }, + { + "epoch": 0.19731858353874826, + "flos": 23401078477440.0, + "grad_norm": 3.6840548181255532, + "learning_rate": 3.7135897047631744e-06, + "loss": 0.8099, + "num_input_tokens_seen": 34771275, + "step": 1641 + }, + { + "epoch": 0.19743882642938737, + "flos": 23953878812640.0, + "grad_norm": 2.9917353594729272, + "learning_rate": 3.713187892283698e-06, + "loss": 0.7605, + "num_input_tokens_seen": 34790885, + "step": 1642 + }, + { + "epoch": 0.19755906932002645, + "flos": 14982810542880.0, + "grad_norm": 4.959190914244559, + "learning_rate": 3.71278581991498e-06, + "loss": 0.8713, + "num_input_tokens_seen": 34806745, + "step": 1643 + }, + { + "epoch": 0.19767931221066554, + "flos": 19466486189760.0, + "grad_norm": 3.2062428256841335, + "learning_rate": 3.712383487718015e-06, + "loss": 0.7882, + "num_input_tokens_seen": 34824985, + "step": 1644 + }, + { + "epoch": 0.19779955510130465, + "flos": 25700984066400.0, + "grad_norm": 2.227289877420333, + "learning_rate": 3.7119808957538365e-06, + "loss": 0.8695, + "num_input_tokens_seen": 34843980, + "step": 1645 + }, + { + "epoch": 0.19791979799194373, + "flos": 20748077172000.0, + "grad_norm": 10.273597536182212, + "learning_rate": 3.711578044083517e-06, + "loss": 0.7999, + "num_input_tokens_seen": 34860780, + "step": 1646 + }, + { + "epoch": 0.1980400408825828, + "flos": 25552207752480.0, + "grad_norm": 2.0168320937390742, + "learning_rate": 3.7111749327681698e-06, + "loss": 0.7462, + "num_input_tokens_seen": 34880815, + "step": 1647 + }, + { + "epoch": 0.1981602837732219, + "flos": 23480224363680.0, + "grad_norm": 2.315268225733745, + "learning_rate": 3.7107715618689455e-06, + "loss": 0.8659, + "num_input_tokens_seen": 34899790, + "step": 1648 + }, + { + "epoch": 0.198280526663861, + "flos": 23152223968800.0, + "grad_norm": 1.6808929376179094, + "learning_rate": 3.710367931447035e-06, + "loss": 0.8334, + "num_input_tokens_seen": 34921570, + "step": 1649 + }, + { + "epoch": 0.1984007695545001, + "flos": 21658507934400.0, + "grad_norm": 2.6585195382494193, + "learning_rate": 3.70996404156367e-06, + "loss": 0.8627, + "num_input_tokens_seen": 34941205, + "step": 1650 + }, + { + "epoch": 0.19852101244513917, + "flos": 36012085648320.0, + "grad_norm": 2.1811394999586864, + "learning_rate": 3.7095598922801187e-06, + "loss": 0.7289, + "num_input_tokens_seen": 34963280, + "step": 1651 + }, + { + "epoch": 0.19864125533577828, + "flos": 23072632045440.0, + "grad_norm": 2.8459983573045635, + "learning_rate": 3.7091554836576914e-06, + "loss": 0.7593, + "num_input_tokens_seen": 34979955, + "step": 1652 + }, + { + "epoch": 0.19876149822641737, + "flos": 24573298824960.0, + "grad_norm": 2.1739704808649916, + "learning_rate": 3.708750815757736e-06, + "loss": 0.8262, + "num_input_tokens_seen": 35000725, + "step": 1653 + }, + { + "epoch": 0.19888174111705645, + "flos": 32150357557440.0, + "grad_norm": 2.906530387232643, + "learning_rate": 3.7083458886416407e-06, + "loss": 0.7302, + "num_input_tokens_seen": 35019920, + "step": 1654 + }, + { + "epoch": 0.19900198400769553, + "flos": 24573261655200.0, + "grad_norm": 4.125706929984138, + "learning_rate": 3.707940702370832e-06, + "loss": 0.8831, + "num_input_tokens_seen": 35040765, + "step": 1655 + }, + { + "epoch": 0.19912222689833464, + "flos": 67818825778080.0, + "grad_norm": 0.7644198527865617, + "learning_rate": 3.707535257006777e-06, + "loss": 0.5824, + "num_input_tokens_seen": 35106390, + "step": 1656 + }, + { + "epoch": 0.19924246978897373, + "flos": 15720038725440.0, + "grad_norm": 2.441800065893622, + "learning_rate": 3.707129552610981e-06, + "loss": 0.8819, + "num_input_tokens_seen": 35125040, + "step": 1657 + }, + { + "epoch": 0.1993627126796128, + "flos": 17541606475680.0, + "grad_norm": 3.4209236668977594, + "learning_rate": 3.70672358924499e-06, + "loss": 0.7411, + "num_input_tokens_seen": 35144280, + "step": 1658 + }, + { + "epoch": 0.19948295557025192, + "flos": 40784950727040.0, + "grad_norm": 2.2702657591070428, + "learning_rate": 3.706317366970386e-06, + "loss": 0.7887, + "num_input_tokens_seen": 35165280, + "step": 1659 + }, + { + "epoch": 0.199603198460891, + "flos": 25046507236800.0, + "grad_norm": 2.1283473409830522, + "learning_rate": 3.705910885848795e-06, + "loss": 0.8348, + "num_input_tokens_seen": 35183655, + "step": 1660 + }, + { + "epoch": 0.19972344135153008, + "flos": 20055495604800.0, + "grad_norm": 2.8963681465849187, + "learning_rate": 3.705504145941879e-06, + "loss": 0.8448, + "num_input_tokens_seen": 35201825, + "step": 1661 + }, + { + "epoch": 0.1998436842421692, + "flos": 23693352999360.0, + "grad_norm": 2.224319452512375, + "learning_rate": 3.7050971473113403e-06, + "loss": 0.7859, + "num_input_tokens_seen": 35221600, + "step": 1662 + }, + { + "epoch": 0.19996392713280828, + "flos": 36050450574240.0, + "grad_norm": 1.814702373093554, + "learning_rate": 3.7046898900189196e-06, + "loss": 0.7973, + "num_input_tokens_seen": 35245935, + "step": 1663 + }, + { + "epoch": 0.20008417002344736, + "flos": 23623759741440.0, + "grad_norm": 1.9891004392447267, + "learning_rate": 3.704282374126398e-06, + "loss": 0.8292, + "num_input_tokens_seen": 35265615, + "step": 1664 + }, + { + "epoch": 0.20020441291408644, + "flos": 21840854275200.0, + "grad_norm": 2.199860555391019, + "learning_rate": 3.7038745996955954e-06, + "loss": 0.8715, + "num_input_tokens_seen": 35284960, + "step": 1665 + }, + { + "epoch": 0.20032465580472555, + "flos": 23146351146720.0, + "grad_norm": 3.523923988336543, + "learning_rate": 3.703466566788371e-06, + "loss": 0.7127, + "num_input_tokens_seen": 35304090, + "step": 1666 + }, + { + "epoch": 0.20044489869536464, + "flos": 23839192902240.0, + "grad_norm": 3.0173937102326023, + "learning_rate": 3.703058275466622e-06, + "loss": 0.7432, + "num_input_tokens_seen": 35323325, + "step": 1667 + }, + { + "epoch": 0.20056514158600372, + "flos": 21913904320800.0, + "grad_norm": 2.3053805069047084, + "learning_rate": 3.7026497257922877e-06, + "loss": 0.772, + "num_input_tokens_seen": 35343635, + "step": 1668 + }, + { + "epoch": 0.20068538447664283, + "flos": 23845028554560.0, + "grad_norm": 8.226332499158952, + "learning_rate": 3.7022409178273436e-06, + "loss": 0.8545, + "num_input_tokens_seen": 35364615, + "step": 1669 + }, + { + "epoch": 0.2008056273672819, + "flos": 18416608723200.0, + "grad_norm": 3.8334761606781, + "learning_rate": 3.7018318516338054e-06, + "loss": 0.7851, + "num_input_tokens_seen": 35383775, + "step": 1670 + }, + { + "epoch": 0.200925870257921, + "flos": 23625692568960.0, + "grad_norm": 3.1684136507495135, + "learning_rate": 3.7014225272737284e-06, + "loss": 0.8201, + "num_input_tokens_seen": 35403000, + "step": 1671 + }, + { + "epoch": 0.20104611314856008, + "flos": 16194139211520.0, + "grad_norm": 2.8222382540844038, + "learning_rate": 3.701012944809207e-06, + "loss": 0.736, + "num_input_tokens_seen": 35420115, + "step": 1672 + }, + { + "epoch": 0.2011663560391992, + "flos": 21366828128640.0, + "grad_norm": 2.8310405392669926, + "learning_rate": 3.700603104302374e-06, + "loss": 0.7894, + "num_input_tokens_seen": 35439925, + "step": 1673 + }, + { + "epoch": 0.20128659892983827, + "flos": 62140956721920.0, + "grad_norm": 0.919739528976938, + "learning_rate": 3.7001930058154027e-06, + "loss": 0.5597, + "num_input_tokens_seen": 35504885, + "step": 1674 + }, + { + "epoch": 0.20140684182047736, + "flos": 28397925761760.0, + "grad_norm": 7.95853070984653, + "learning_rate": 3.6997826494105037e-06, + "loss": 0.7978, + "num_input_tokens_seen": 35525330, + "step": 1675 + }, + { + "epoch": 0.20152708471111647, + "flos": 28033790626560.0, + "grad_norm": 2.168873311255662, + "learning_rate": 3.6993720351499286e-06, + "loss": 0.6931, + "num_input_tokens_seen": 35546455, + "step": 1676 + }, + { + "epoch": 0.20164732760175555, + "flos": 23516061745920.0, + "grad_norm": 2.1487614058035955, + "learning_rate": 3.6989611630959666e-06, + "loss": 0.7698, + "num_input_tokens_seen": 35565010, + "step": 1677 + }, + { + "epoch": 0.20176757049239463, + "flos": 70999383418080.0, + "grad_norm": 0.845288816109663, + "learning_rate": 3.6985500333109474e-06, + "loss": 0.5821, + "num_input_tokens_seen": 35633340, + "step": 1678 + }, + { + "epoch": 0.20188781338303372, + "flos": 21399431741760.0, + "grad_norm": 3.959362311218692, + "learning_rate": 3.6981386458572385e-06, + "loss": 0.7689, + "num_input_tokens_seen": 35651315, + "step": 1679 + }, + { + "epoch": 0.20200805627367283, + "flos": 11530012881600.0, + "grad_norm": 7.367713561546757, + "learning_rate": 3.6977270007972468e-06, + "loss": 0.7552, + "num_input_tokens_seen": 35668850, + "step": 1680 + }, + { + "epoch": 0.2021282991643119, + "flos": 28504322815680.0, + "grad_norm": 2.6452945751432346, + "learning_rate": 3.6973150981934196e-06, + "loss": 0.7249, + "num_input_tokens_seen": 35690080, + "step": 1681 + }, + { + "epoch": 0.202248542054951, + "flos": 17897675772960.0, + "grad_norm": 3.062149339179827, + "learning_rate": 3.6969029381082415e-06, + "loss": 0.84, + "num_input_tokens_seen": 35706075, + "step": 1682 + }, + { + "epoch": 0.2023687849455901, + "flos": 19836308298240.0, + "grad_norm": 2.4790955113903834, + "learning_rate": 3.696490520604237e-06, + "loss": 0.7951, + "num_input_tokens_seen": 35724765, + "step": 1683 + }, + { + "epoch": 0.20248902783622919, + "flos": 22532469428640.0, + "grad_norm": 1.8132205057221298, + "learning_rate": 3.696077845743968e-06, + "loss": 0.8079, + "num_input_tokens_seen": 35744355, + "step": 1684 + }, + { + "epoch": 0.20260927072686827, + "flos": 22677231408480.0, + "grad_norm": 5.076925000943973, + "learning_rate": 3.69566491359004e-06, + "loss": 0.7315, + "num_input_tokens_seen": 35761000, + "step": 1685 + }, + { + "epoch": 0.20272951361750738, + "flos": 50950323915360.0, + "grad_norm": 3.1451560969345342, + "learning_rate": 3.695251724205092e-06, + "loss": 0.6912, + "num_input_tokens_seen": 35785360, + "step": 1686 + }, + { + "epoch": 0.20284975650814646, + "flos": 26540037422400.0, + "grad_norm": 2.3143761730680446, + "learning_rate": 3.6948382776518054e-06, + "loss": 0.8628, + "num_input_tokens_seen": 35806065, + "step": 1687 + }, + { + "epoch": 0.20296999939878554, + "flos": 16011495512640.0, + "grad_norm": 2.6129714029457545, + "learning_rate": 3.6944245739929e-06, + "loss": 0.7952, + "num_input_tokens_seen": 35822225, + "step": 1688 + }, + { + "epoch": 0.20309024228942463, + "flos": 19175847137280.0, + "grad_norm": 2.484250815304744, + "learning_rate": 3.6940106132911332e-06, + "loss": 0.7207, + "num_input_tokens_seen": 35839410, + "step": 1689 + }, + { + "epoch": 0.20321048518006374, + "flos": 22788497700960.0, + "grad_norm": 2.3339139919761696, + "learning_rate": 3.6935963956093037e-06, + "loss": 0.8919, + "num_input_tokens_seen": 35859495, + "step": 1690 + }, + { + "epoch": 0.20333072807070282, + "flos": 19069115555520.0, + "grad_norm": 12.841363615552513, + "learning_rate": 3.6931819210102474e-06, + "loss": 0.6882, + "num_input_tokens_seen": 35878890, + "step": 1691 + }, + { + "epoch": 0.2034509709613419, + "flos": 18154893477600.0, + "grad_norm": 2.1510210628923634, + "learning_rate": 3.6927671895568402e-06, + "loss": 0.8432, + "num_input_tokens_seen": 35897950, + "step": 1692 + }, + { + "epoch": 0.20357121385198101, + "flos": 22890917590560.0, + "grad_norm": 4.475790812767975, + "learning_rate": 3.692352201311996e-06, + "loss": 0.866, + "num_input_tokens_seen": 35916800, + "step": 1693 + }, + { + "epoch": 0.2036914567426201, + "flos": 20892244435680.0, + "grad_norm": 5.06934769536659, + "learning_rate": 3.6919369563386687e-06, + "loss": 0.7675, + "num_input_tokens_seen": 35936600, + "step": 1694 + }, + { + "epoch": 0.20381169963325918, + "flos": 15497022933600.0, + "grad_norm": 2.7404727628175976, + "learning_rate": 3.69152145469985e-06, + "loss": 0.7879, + "num_input_tokens_seen": 35953645, + "step": 1695 + }, + { + "epoch": 0.20393194252389826, + "flos": 28791505080480.0, + "grad_norm": 4.414315059492017, + "learning_rate": 3.691105696458572e-06, + "loss": 0.8247, + "num_input_tokens_seen": 35970940, + "step": 1696 + }, + { + "epoch": 0.20405218541453737, + "flos": 22456297123200.0, + "grad_norm": 3.220462676743036, + "learning_rate": 3.690689681677904e-06, + "loss": 0.6809, + "num_input_tokens_seen": 35989410, + "step": 1697 + }, + { + "epoch": 0.20417242830517646, + "flos": 25338075533280.0, + "grad_norm": 1.9643343498352068, + "learning_rate": 3.690273410420956e-06, + "loss": 0.8897, + "num_input_tokens_seen": 36009175, + "step": 1698 + }, + { + "epoch": 0.20429267119581554, + "flos": 14772135111360.0, + "grad_norm": 3.0753734702977984, + "learning_rate": 3.689856882750875e-06, + "loss": 0.7696, + "num_input_tokens_seen": 36024375, + "step": 1699 + }, + { + "epoch": 0.20441291408645465, + "flos": 17756444920320.0, + "grad_norm": 1.911071745179971, + "learning_rate": 3.6894400987308486e-06, + "loss": 0.784, + "num_input_tokens_seen": 36041895, + "step": 1700 + }, + { + "epoch": 0.20453315697709373, + "flos": 16412360104320.0, + "grad_norm": 2.754313034423549, + "learning_rate": 3.6890230584241024e-06, + "loss": 0.8497, + "num_input_tokens_seen": 36058825, + "step": 1701 + }, + { + "epoch": 0.20465339986773282, + "flos": 66617904642240.0, + "grad_norm": 1.08491072379723, + "learning_rate": 3.6886057618939016e-06, + "loss": 0.6627, + "num_input_tokens_seen": 36121645, + "step": 1702 + }, + { + "epoch": 0.2047736427583719, + "flos": 41913565212480.0, + "grad_norm": 2.385524119881852, + "learning_rate": 3.6881882092035492e-06, + "loss": 0.6908, + "num_input_tokens_seen": 36143030, + "step": 1703 + }, + { + "epoch": 0.204893885649011, + "flos": 69840289729440.0, + "grad_norm": 0.9401285832299651, + "learning_rate": 3.6877704004163873e-06, + "loss": 0.6128, + "num_input_tokens_seen": 36203435, + "step": 1704 + }, + { + "epoch": 0.2050141285396501, + "flos": 22168297123680.0, + "grad_norm": 2.2684671399042657, + "learning_rate": 3.6873523355957984e-06, + "loss": 0.7773, + "num_input_tokens_seen": 36222435, + "step": 1705 + }, + { + "epoch": 0.20513437143028918, + "flos": 46218382742400.0, + "grad_norm": 1.0062427567878804, + "learning_rate": 3.686934014805201e-06, + "loss": 0.6412, + "num_input_tokens_seen": 36273065, + "step": 1706 + }, + { + "epoch": 0.20525461432092829, + "flos": 21872900341920.0, + "grad_norm": 2.023231407480754, + "learning_rate": 3.6865154381080552e-06, + "loss": 0.8084, + "num_input_tokens_seen": 36293750, + "step": 1707 + }, + { + "epoch": 0.20537485721156737, + "flos": 21183961411200.0, + "grad_norm": 2.5155742000402617, + "learning_rate": 3.6860966055678585e-06, + "loss": 0.8247, + "num_input_tokens_seen": 36310105, + "step": 1708 + }, + { + "epoch": 0.20549510010220645, + "flos": 20162338695840.0, + "grad_norm": 2.1330787576334562, + "learning_rate": 3.685677517248147e-06, + "loss": 0.8647, + "num_input_tokens_seen": 36328995, + "step": 1709 + }, + { + "epoch": 0.20561534299284553, + "flos": 16992448776960.0, + "grad_norm": 2.4673747807570336, + "learning_rate": 3.6852581732124967e-06, + "loss": 0.8032, + "num_input_tokens_seen": 36346340, + "step": 1710 + }, + { + "epoch": 0.20573558588348465, + "flos": 22857644921760.0, + "grad_norm": 3.9612979160635686, + "learning_rate": 3.6848385735245213e-06, + "loss": 0.7648, + "num_input_tokens_seen": 36365430, + "step": 1711 + }, + { + "epoch": 0.20585582877412373, + "flos": 24607798095840.0, + "grad_norm": 2.011476372496838, + "learning_rate": 3.6844187182478734e-06, + "loss": 0.8666, + "num_input_tokens_seen": 36386285, + "step": 1712 + }, + { + "epoch": 0.2059760716647628, + "flos": 24172322724000.0, + "grad_norm": 2.2290666385442686, + "learning_rate": 3.683998607446246e-06, + "loss": 0.751, + "num_input_tokens_seen": 36404935, + "step": 1713 + }, + { + "epoch": 0.20609631455540192, + "flos": 20200666452000.0, + "grad_norm": 2.8143069079059604, + "learning_rate": 3.6835782411833686e-06, + "loss": 0.7519, + "num_input_tokens_seen": 36424455, + "step": 1714 + }, + { + "epoch": 0.206216557446041, + "flos": 19836642826080.0, + "grad_norm": 2.257893913939542, + "learning_rate": 3.68315761952301e-06, + "loss": 0.742, + "num_input_tokens_seen": 36441685, + "step": 1715 + }, + { + "epoch": 0.2063368003366801, + "flos": 24061799826720.0, + "grad_norm": 2.1588295144091245, + "learning_rate": 3.6827367425289797e-06, + "loss": 0.8307, + "num_input_tokens_seen": 36461460, + "step": 1716 + }, + { + "epoch": 0.2064570432273192, + "flos": 20311189349280.0, + "grad_norm": 2.6113577065686107, + "learning_rate": 3.6823156102651225e-06, + "loss": 0.7247, + "num_input_tokens_seen": 36479615, + "step": 1717 + }, + { + "epoch": 0.20657728611795828, + "flos": 20491045316160.0, + "grad_norm": 7.8231819063844386, + "learning_rate": 3.6818942227953257e-06, + "loss": 0.7064, + "num_input_tokens_seen": 36500160, + "step": 1718 + }, + { + "epoch": 0.20669752900859736, + "flos": 21768175927200.0, + "grad_norm": 2.1972696397532765, + "learning_rate": 3.681472580183512e-06, + "loss": 0.6905, + "num_input_tokens_seen": 36518810, + "step": 1719 + }, + { + "epoch": 0.20681777189923645, + "flos": 15100618713120.0, + "grad_norm": 2.135502387086342, + "learning_rate": 3.6810506824936455e-06, + "loss": 0.8625, + "num_input_tokens_seen": 36536890, + "step": 1720 + }, + { + "epoch": 0.20693801478987556, + "flos": 56400153528480.0, + "grad_norm": 1.0987485290621652, + "learning_rate": 3.680628529789726e-06, + "loss": 0.6259, + "num_input_tokens_seen": 36589300, + "step": 1721 + }, + { + "epoch": 0.20705825768051464, + "flos": 21583264872960.0, + "grad_norm": 2.2136864708304325, + "learning_rate": 3.680206122135796e-06, + "loss": 0.8648, + "num_input_tokens_seen": 36609905, + "step": 1722 + }, + { + "epoch": 0.20717850057115372, + "flos": 25811878661280.0, + "grad_norm": 1.938146790865097, + "learning_rate": 3.6797834595959323e-06, + "loss": 0.7849, + "num_input_tokens_seen": 36629805, + "step": 1723 + }, + { + "epoch": 0.20729874346179283, + "flos": 29088946199040.0, + "grad_norm": 3.691566165104287, + "learning_rate": 3.679360542234254e-06, + "loss": 0.7775, + "num_input_tokens_seen": 36649150, + "step": 1724 + }, + { + "epoch": 0.20741898635243192, + "flos": 28981731410400.0, + "grad_norm": 1.9181908752723935, + "learning_rate": 3.678937370114916e-06, + "loss": 0.7281, + "num_input_tokens_seen": 36668955, + "step": 1725 + }, + { + "epoch": 0.207539229243071, + "flos": 15537395026560.0, + "grad_norm": 2.24898082114693, + "learning_rate": 3.678513943302114e-06, + "loss": 0.7904, + "num_input_tokens_seen": 36685730, + "step": 1726 + }, + { + "epoch": 0.20765947213371008, + "flos": 20491788711360.0, + "grad_norm": 2.0777934103846305, + "learning_rate": 3.678090261860082e-06, + "loss": 0.8501, + "num_input_tokens_seen": 36705460, + "step": 1727 + }, + { + "epoch": 0.2077797150243492, + "flos": 19326853636800.0, + "grad_norm": 2.3711134581398823, + "learning_rate": 3.6776663258530906e-06, + "loss": 0.7782, + "num_input_tokens_seen": 36724110, + "step": 1728 + }, + { + "epoch": 0.20789995791498828, + "flos": 21798586524480.0, + "grad_norm": 2.190432517483125, + "learning_rate": 3.6772421353454516e-06, + "loss": 0.7142, + "num_input_tokens_seen": 36742705, + "step": 1729 + }, + { + "epoch": 0.20802020080562736, + "flos": 23115754700640.0, + "grad_norm": 3.9921872611148412, + "learning_rate": 3.6768176904015153e-06, + "loss": 0.8835, + "num_input_tokens_seen": 36762510, + "step": 1730 + }, + { + "epoch": 0.20814044369626647, + "flos": 23039582395200.0, + "grad_norm": 2.3599062694104678, + "learning_rate": 3.6763929910856674e-06, + "loss": 0.6003, + "num_input_tokens_seen": 36780960, + "step": 1731 + }, + { + "epoch": 0.20826068658690555, + "flos": 19580019837600.0, + "grad_norm": 2.5661187090690922, + "learning_rate": 3.6759680374623365e-06, + "loss": 0.7795, + "num_input_tokens_seen": 36798875, + "step": 1732 + }, + { + "epoch": 0.20838092947754464, + "flos": 25337703835680.0, + "grad_norm": 3.145389112893473, + "learning_rate": 3.675542829595986e-06, + "loss": 0.7551, + "num_input_tokens_seen": 36818540, + "step": 1733 + }, + { + "epoch": 0.20850117236818372, + "flos": 24027337725600.0, + "grad_norm": 1.6135623862420208, + "learning_rate": 3.6751173675511213e-06, + "loss": 0.7912, + "num_input_tokens_seen": 36840585, + "step": 1734 + }, + { + "epoch": 0.20862141525882283, + "flos": 20048953727040.0, + "grad_norm": 2.469387406743077, + "learning_rate": 3.674691651392283e-06, + "loss": 0.873, + "num_input_tokens_seen": 36858255, + "step": 1735 + }, + { + "epoch": 0.2087416581494619, + "flos": 38958811096320.0, + "grad_norm": 5.819990356542195, + "learning_rate": 3.674265681184053e-06, + "loss": 0.7561, + "num_input_tokens_seen": 36881515, + "step": 1736 + }, + { + "epoch": 0.208861901040101, + "flos": 26064524485440.0, + "grad_norm": 1.88231433910345, + "learning_rate": 3.6738394569910504e-06, + "loss": 0.8628, + "num_input_tokens_seen": 36902055, + "step": 1737 + }, + { + "epoch": 0.2089821439307401, + "flos": 28357851026880.0, + "grad_norm": 2.5171532579457327, + "learning_rate": 3.6734129788779333e-06, + "loss": 0.8329, + "num_input_tokens_seen": 36921230, + "step": 1738 + }, + { + "epoch": 0.2091023868213792, + "flos": 21039571128960.0, + "grad_norm": 1.9222447109499416, + "learning_rate": 3.6729862469093976e-06, + "loss": 0.9072, + "num_input_tokens_seen": 36940325, + "step": 1739 + }, + { + "epoch": 0.20922262971201827, + "flos": 22423396152000.0, + "grad_norm": 3.167006491112087, + "learning_rate": 3.6725592611501782e-06, + "loss": 0.8283, + "num_input_tokens_seen": 36959000, + "step": 1740 + }, + { + "epoch": 0.20934287260265738, + "flos": 27812187285600.0, + "grad_norm": 2.4477791821985675, + "learning_rate": 3.6721320216650496e-06, + "loss": 0.7651, + "num_input_tokens_seen": 36979615, + "step": 1741 + }, + { + "epoch": 0.20946311549329646, + "flos": 16411988406720.0, + "grad_norm": 2.1867207910460733, + "learning_rate": 3.6717045285188215e-06, + "loss": 0.8388, + "num_input_tokens_seen": 36997310, + "step": 1742 + }, + { + "epoch": 0.20958335838393555, + "flos": 22460608815360.0, + "grad_norm": 3.832688581451686, + "learning_rate": 3.671276781776346e-06, + "loss": 0.8673, + "num_input_tokens_seen": 37015965, + "step": 1743 + }, + { + "epoch": 0.20970360127457463, + "flos": 25188518654400.0, + "grad_norm": 2.3499375111814187, + "learning_rate": 3.6708487815025128e-06, + "loss": 0.6725, + "num_input_tokens_seen": 37036545, + "step": 1744 + }, + { + "epoch": 0.20982384416521374, + "flos": 18452892142560.0, + "grad_norm": 2.6386847054531697, + "learning_rate": 3.6704205277622463e-06, + "loss": 0.7408, + "num_input_tokens_seen": 37054985, + "step": 1745 + }, + { + "epoch": 0.20994408705585282, + "flos": 25337183459040.0, + "grad_norm": 1.8481000242329153, + "learning_rate": 3.6699920206205146e-06, + "loss": 0.8089, + "num_input_tokens_seen": 37075725, + "step": 1746 + }, + { + "epoch": 0.2100643299464919, + "flos": 21290395634880.0, + "grad_norm": 1.9446058713420022, + "learning_rate": 3.669563260142321e-06, + "loss": 0.8222, + "num_input_tokens_seen": 37094615, + "step": 1747 + }, + { + "epoch": 0.21018457283713102, + "flos": 19326816467040.0, + "grad_norm": 3.2960604705541003, + "learning_rate": 3.6691342463927083e-06, + "loss": 0.8452, + "num_input_tokens_seen": 37113610, + "step": 1748 + }, + { + "epoch": 0.2103048157277701, + "flos": 28289410031520.0, + "grad_norm": 2.0073563469477937, + "learning_rate": 3.668704979436758e-06, + "loss": 0.8193, + "num_input_tokens_seen": 37133705, + "step": 1749 + }, + { + "epoch": 0.21042505861840918, + "flos": 17432496029280.0, + "grad_norm": 2.544814136066047, + "learning_rate": 3.668275459339588e-06, + "loss": 0.7884, + "num_input_tokens_seen": 37152185, + "step": 1750 + }, + { + "epoch": 0.21054530150904827, + "flos": 14189741913600.0, + "grad_norm": 2.3522808207161994, + "learning_rate": 3.667845686166358e-06, + "loss": 0.8024, + "num_input_tokens_seen": 37169110, + "step": 1751 + }, + { + "epoch": 0.21066554439968738, + "flos": 18591483942240.0, + "grad_norm": 2.3060026909778575, + "learning_rate": 3.6674156599822634e-06, + "loss": 0.8627, + "num_input_tokens_seen": 37184905, + "step": 1752 + }, + { + "epoch": 0.21078578729032646, + "flos": 23625023513280.0, + "grad_norm": 2.588089246505224, + "learning_rate": 3.666985380852539e-06, + "loss": 0.816, + "num_input_tokens_seen": 37203910, + "step": 1753 + }, + { + "epoch": 0.21090603018096554, + "flos": 29304082001760.0, + "grad_norm": 2.5273034683625766, + "learning_rate": 3.6665548488424576e-06, + "loss": 0.7463, + "num_input_tokens_seen": 37222670, + "step": 1754 + }, + { + "epoch": 0.21102627307160465, + "flos": 23228098916160.0, + "grad_norm": 1.8170006220495811, + "learning_rate": 3.6661240640173307e-06, + "loss": 0.8786, + "num_input_tokens_seen": 37244140, + "step": 1755 + }, + { + "epoch": 0.21114651596224374, + "flos": 54555757812000.0, + "grad_norm": 0.8612748386914955, + "learning_rate": 3.6656930264425085e-06, + "loss": 0.5788, + "num_input_tokens_seen": 37308185, + "step": 1756 + }, + { + "epoch": 0.21126675885288282, + "flos": 21512556522240.0, + "grad_norm": 1.9423275149046413, + "learning_rate": 3.665261736183378e-06, + "loss": 0.7525, + "num_input_tokens_seen": 37328220, + "step": 1757 + }, + { + "epoch": 0.2113870017435219, + "flos": 10946504591040.0, + "grad_norm": 4.19438698219085, + "learning_rate": 3.664830193305366e-06, + "loss": 0.8906, + "num_input_tokens_seen": 37345995, + "step": 1758 + }, + { + "epoch": 0.211507244634161, + "flos": 16630172129760.0, + "grad_norm": 2.890217190405393, + "learning_rate": 3.6643983978739373e-06, + "loss": 0.7694, + "num_input_tokens_seen": 37362610, + "step": 1759 + }, + { + "epoch": 0.2116274875248001, + "flos": 20923993144320.0, + "grad_norm": 2.33183249388566, + "learning_rate": 3.663966349954596e-06, + "loss": 0.8246, + "num_input_tokens_seen": 37382790, + "step": 1760 + }, + { + "epoch": 0.21174773041543918, + "flos": 68099168807040.0, + "grad_norm": 0.7936269724862381, + "learning_rate": 3.6635340496128816e-06, + "loss": 0.5968, + "num_input_tokens_seen": 37439720, + "step": 1761 + }, + { + "epoch": 0.2118679733060783, + "flos": 20638855216320.0, + "grad_norm": 2.0709402608097545, + "learning_rate": 3.6631014969143747e-06, + "loss": 0.926, + "num_input_tokens_seen": 37459410, + "step": 1762 + }, + { + "epoch": 0.21198821619671737, + "flos": 23190291536640.0, + "grad_norm": 2.571047228586736, + "learning_rate": 3.662668691924693e-06, + "loss": 0.8875, + "num_input_tokens_seen": 37480460, + "step": 1763 + }, + { + "epoch": 0.21210845908735645, + "flos": 24463482153120.0, + "grad_norm": 2.3141743552446883, + "learning_rate": 3.6622356347094927e-06, + "loss": 0.7136, + "num_input_tokens_seen": 37502025, + "step": 1764 + }, + { + "epoch": 0.21222870197799554, + "flos": 27050978874240.0, + "grad_norm": 6.548230451961961, + "learning_rate": 3.6618023253344684e-06, + "loss": 0.7892, + "num_input_tokens_seen": 37520885, + "step": 1765 + }, + { + "epoch": 0.21234894486863465, + "flos": 16849916982720.0, + "grad_norm": 1.6984223702887744, + "learning_rate": 3.6613687638653527e-06, + "loss": 0.8344, + "num_input_tokens_seen": 37539055, + "step": 1766 + }, + { + "epoch": 0.21246918775927373, + "flos": 23441301891360.0, + "grad_norm": 1.9808609854319832, + "learning_rate": 3.660934950367916e-06, + "loss": 0.7778, + "num_input_tokens_seen": 37558300, + "step": 1767 + }, + { + "epoch": 0.21258943064991281, + "flos": 22350308936640.0, + "grad_norm": 2.768662158959027, + "learning_rate": 3.660500884907968e-06, + "loss": 0.8329, + "num_input_tokens_seen": 37579040, + "step": 1768 + }, + { + "epoch": 0.21270967354055192, + "flos": 59355130663200.0, + "grad_norm": 0.8295907777777629, + "learning_rate": 3.660066567551356e-06, + "loss": 0.6, + "num_input_tokens_seen": 37639185, + "step": 1769 + }, + { + "epoch": 0.212829916431191, + "flos": 21513299917440.0, + "grad_norm": 3.8155079216768004, + "learning_rate": 3.6596319983639657e-06, + "loss": 0.8439, + "num_input_tokens_seen": 37657765, + "step": 1770 + }, + { + "epoch": 0.2129501593218301, + "flos": 28946191386240.0, + "grad_norm": 1.8023013238497838, + "learning_rate": 3.6591971774117214e-06, + "loss": 0.8639, + "num_input_tokens_seen": 37680740, + "step": 1771 + }, + { + "epoch": 0.2130704022124692, + "flos": 18780223481760.0, + "grad_norm": 2.242218761691366, + "learning_rate": 3.6587621047605833e-06, + "loss": 0.8054, + "num_input_tokens_seen": 37697910, + "step": 1772 + }, + { + "epoch": 0.21319064510310828, + "flos": 13898508144960.0, + "grad_norm": 2.768397347959679, + "learning_rate": 3.6583267804765542e-06, + "loss": 0.8653, + "num_input_tokens_seen": 37712245, + "step": 1773 + }, + { + "epoch": 0.21331088799374737, + "flos": 20929791626880.0, + "grad_norm": 2.178953532584361, + "learning_rate": 3.6578912046256702e-06, + "loss": 0.8577, + "num_input_tokens_seen": 37730765, + "step": 1774 + }, + { + "epoch": 0.21343113088438645, + "flos": 18598137329280.0, + "grad_norm": 2.1743417905863005, + "learning_rate": 3.6574553772740083e-06, + "loss": 0.7668, + "num_input_tokens_seen": 37748695, + "step": 1775 + }, + { + "epoch": 0.21355137377502556, + "flos": 67317176766240.0, + "grad_norm": 0.8590521812998763, + "learning_rate": 3.657019298487684e-06, + "loss": 0.6181, + "num_input_tokens_seen": 37813705, + "step": 1776 + }, + { + "epoch": 0.21367161666566464, + "flos": 34482420722400.0, + "grad_norm": 4.455630163285272, + "learning_rate": 3.6565829683328495e-06, + "loss": 0.8369, + "num_input_tokens_seen": 37836330, + "step": 1777 + }, + { + "epoch": 0.21379185955630373, + "flos": 18962309634240.0, + "grad_norm": 2.34214605761733, + "learning_rate": 3.6561463868756965e-06, + "loss": 0.8571, + "num_input_tokens_seen": 37855030, + "step": 1778 + }, + { + "epoch": 0.21391210244694284, + "flos": 28177957890240.0, + "grad_norm": 4.294971835273882, + "learning_rate": 3.655709554182452e-06, + "loss": 0.7808, + "num_input_tokens_seen": 37875370, + "step": 1779 + }, + { + "epoch": 0.21403234533758192, + "flos": 17430191504160.0, + "grad_norm": 2.306011065586402, + "learning_rate": 3.6552724703193855e-06, + "loss": 0.8468, + "num_input_tokens_seen": 37892560, + "step": 1780 + }, + { + "epoch": 0.214152588228221, + "flos": 51564125560320.0, + "grad_norm": 0.7946156279734466, + "learning_rate": 3.654835135352801e-06, + "loss": 0.5595, + "num_input_tokens_seen": 37947370, + "step": 1781 + }, + { + "epoch": 0.21427283111886009, + "flos": 19468233168480.0, + "grad_norm": 1.9513907095047258, + "learning_rate": 3.654397549349043e-06, + "loss": 0.8773, + "num_input_tokens_seen": 37964745, + "step": 1782 + }, + { + "epoch": 0.2143930740094992, + "flos": 20056201830240.0, + "grad_norm": 2.3332018701667163, + "learning_rate": 3.653959712374491e-06, + "loss": 0.7523, + "num_input_tokens_seen": 37982850, + "step": 1783 + }, + { + "epoch": 0.21451331690013828, + "flos": 21767469701760.0, + "grad_norm": 1.7855441884362993, + "learning_rate": 3.6535216244955663e-06, + "loss": 0.8284, + "num_input_tokens_seen": 38002745, + "step": 1784 + }, + { + "epoch": 0.21463355979077736, + "flos": 32806655705280.0, + "grad_norm": 6.9288248344452885, + "learning_rate": 3.653083285778726e-06, + "loss": 0.7091, + "num_input_tokens_seen": 38024115, + "step": 1785 + }, + { + "epoch": 0.21475380268141647, + "flos": 21513076898880.0, + "grad_norm": 2.4321882619394644, + "learning_rate": 3.6526446962904653e-06, + "loss": 0.8155, + "num_input_tokens_seen": 38042830, + "step": 1786 + }, + { + "epoch": 0.21487404557205556, + "flos": 32112810366240.0, + "grad_norm": 1.7954779459895664, + "learning_rate": 3.652205856097318e-06, + "loss": 0.7441, + "num_input_tokens_seen": 38066015, + "step": 1787 + }, + { + "epoch": 0.21499428846269464, + "flos": 12655096239840.0, + "grad_norm": 2.232016539132973, + "learning_rate": 3.651766765265856e-06, + "loss": 0.791, + "num_input_tokens_seen": 38083025, + "step": 1788 + }, + { + "epoch": 0.21511453135333372, + "flos": 23437807933920.0, + "grad_norm": 2.6555961252870603, + "learning_rate": 3.65132742386269e-06, + "loss": 0.8135, + "num_input_tokens_seen": 38098245, + "step": 1789 + }, + { + "epoch": 0.21523477424397283, + "flos": 26904953122560.0, + "grad_norm": 1.8717919562105707, + "learning_rate": 3.6508878319544656e-06, + "loss": 0.8455, + "num_input_tokens_seen": 38118460, + "step": 1790 + }, + { + "epoch": 0.21535501713461191, + "flos": 18889222418880.0, + "grad_norm": 2.952257301375161, + "learning_rate": 3.65044798960787e-06, + "loss": 0.817, + "num_input_tokens_seen": 38136800, + "step": 1791 + }, + { + "epoch": 0.215475260025251, + "flos": 17869681210080.0, + "grad_norm": 2.1970640098473915, + "learning_rate": 3.650007896889627e-06, + "loss": 0.7803, + "num_input_tokens_seen": 38155380, + "step": 1792 + }, + { + "epoch": 0.2155955029158901, + "flos": 16630655336640.0, + "grad_norm": 1.970346862832135, + "learning_rate": 3.6495675538664974e-06, + "loss": 0.8043, + "num_input_tokens_seen": 38174395, + "step": 1793 + }, + { + "epoch": 0.2157157458065292, + "flos": 23587364812800.0, + "grad_norm": 1.929274047543506, + "learning_rate": 3.649126960605282e-06, + "loss": 0.8256, + "num_input_tokens_seen": 38195380, + "step": 1794 + }, + { + "epoch": 0.21583598869716827, + "flos": 22095470096640.0, + "grad_norm": 2.7286237996966523, + "learning_rate": 3.6486861171728174e-06, + "loss": 0.8317, + "num_input_tokens_seen": 38213775, + "step": 1795 + }, + { + "epoch": 0.21595623158780738, + "flos": 23407731864480.0, + "grad_norm": 2.2858912663402786, + "learning_rate": 3.6482450236359803e-06, + "loss": 0.7845, + "num_input_tokens_seen": 38235630, + "step": 1796 + }, + { + "epoch": 0.21607647447844647, + "flos": 26867926308000.0, + "grad_norm": 2.992181763008534, + "learning_rate": 3.647803680061683e-06, + "loss": 0.7728, + "num_input_tokens_seen": 38256045, + "step": 1797 + }, + { + "epoch": 0.21619671736908555, + "flos": 14474656823040.0, + "grad_norm": 4.703349871661557, + "learning_rate": 3.6473620865168776e-06, + "loss": 0.7513, + "num_input_tokens_seen": 38272915, + "step": 1798 + }, + { + "epoch": 0.21631696025972463, + "flos": 17905592931840.0, + "grad_norm": 3.3801944460042312, + "learning_rate": 3.646920243068554e-06, + "loss": 0.8162, + "num_input_tokens_seen": 38290810, + "step": 1799 + }, + { + "epoch": 0.21643720315036374, + "flos": 24427384582560.0, + "grad_norm": 2.452882596813888, + "learning_rate": 3.6464781497837384e-06, + "loss": 0.7479, + "num_input_tokens_seen": 38312785, + "step": 1800 + }, + { + "epoch": 0.21655744604100283, + "flos": 28433540125440.0, + "grad_norm": 1.9620041093151366, + "learning_rate": 3.6460358067294965e-06, + "loss": 0.7274, + "num_input_tokens_seen": 38334015, + "step": 1801 + }, + { + "epoch": 0.2166776889316419, + "flos": 20124085279200.0, + "grad_norm": 2.106482436941052, + "learning_rate": 3.645593213972932e-06, + "loss": 0.7757, + "num_input_tokens_seen": 38352920, + "step": 1802 + }, + { + "epoch": 0.21679793182228102, + "flos": 15171921780000.0, + "grad_norm": 3.121650579288522, + "learning_rate": 3.6451503715811852e-06, + "loss": 0.7996, + "num_input_tokens_seen": 38371230, + "step": 1803 + }, + { + "epoch": 0.2169181747129201, + "flos": 17359780511520.0, + "grad_norm": 3.4696072730514365, + "learning_rate": 3.6447072796214345e-06, + "loss": 0.8013, + "num_input_tokens_seen": 38389795, + "step": 1804 + }, + { + "epoch": 0.21703841760355919, + "flos": 58677417000000.0, + "grad_norm": 1.0902998322013104, + "learning_rate": 3.644263938160898e-06, + "loss": 0.6318, + "num_input_tokens_seen": 38445760, + "step": 1805 + }, + { + "epoch": 0.21715866049419827, + "flos": 22386406507200.0, + "grad_norm": 3.6895850800664993, + "learning_rate": 3.6438203472668293e-06, + "loss": 0.7189, + "num_input_tokens_seen": 38465405, + "step": 1806 + }, + { + "epoch": 0.21727890338483738, + "flos": 17213234383200.0, + "grad_norm": 2.1579107992776625, + "learning_rate": 3.6433765070065206e-06, + "loss": 0.8189, + "num_input_tokens_seen": 38483195, + "step": 1807 + }, + { + "epoch": 0.21739914627547646, + "flos": 13415524086240.0, + "grad_norm": 3.2299137030727114, + "learning_rate": 3.6429324174473025e-06, + "loss": 0.8779, + "num_input_tokens_seen": 38496735, + "step": 1808 + }, + { + "epoch": 0.21751938916611555, + "flos": 20929605778080.0, + "grad_norm": 2.645924143596984, + "learning_rate": 3.6424880786565425e-06, + "loss": 0.8463, + "num_input_tokens_seen": 38512880, + "step": 1809 + }, + { + "epoch": 0.21763963205675466, + "flos": 27560359196160.0, + "grad_norm": 2.9653869854969828, + "learning_rate": 3.6420434907016482e-06, + "loss": 0.7976, + "num_input_tokens_seen": 38532570, + "step": 1810 + }, + { + "epoch": 0.21775987494739374, + "flos": 21399691930080.0, + "grad_norm": 2.315190271092967, + "learning_rate": 3.6415986536500606e-06, + "loss": 0.8075, + "num_input_tokens_seen": 38550900, + "step": 1811 + }, + { + "epoch": 0.21788011783803282, + "flos": 18306903560640.0, + "grad_norm": 1.9188086385806427, + "learning_rate": 3.641153567569263e-06, + "loss": 0.807, + "num_input_tokens_seen": 38569215, + "step": 1812 + }, + { + "epoch": 0.2180003607286719, + "flos": 30219270493440.0, + "grad_norm": 2.5357383212633238, + "learning_rate": 3.640708232526774e-06, + "loss": 0.9523, + "num_input_tokens_seen": 38587870, + "step": 1813 + }, + { + "epoch": 0.21812060361931102, + "flos": 25443654852480.0, + "grad_norm": 3.0913009968834353, + "learning_rate": 3.6402626485901504e-06, + "loss": 0.7852, + "num_input_tokens_seen": 38606965, + "step": 1814 + }, + { + "epoch": 0.2182408465099501, + "flos": 21876877506240.0, + "grad_norm": 2.3521978084074284, + "learning_rate": 3.639816815826988e-06, + "loss": 0.7813, + "num_input_tokens_seen": 38626290, + "step": 1815 + }, + { + "epoch": 0.21836108940058918, + "flos": 23623425213600.0, + "grad_norm": 3.1415443403340904, + "learning_rate": 3.6393707343049176e-06, + "loss": 0.7767, + "num_input_tokens_seen": 38646620, + "step": 1816 + }, + { + "epoch": 0.2184813322912283, + "flos": 24645679814880.0, + "grad_norm": 3.01336791013049, + "learning_rate": 3.6389244040916104e-06, + "loss": 0.7356, + "num_input_tokens_seen": 38665695, + "step": 1817 + }, + { + "epoch": 0.21860157518186737, + "flos": 26536431955680.0, + "grad_norm": 2.6233972470238163, + "learning_rate": 3.6384778252547747e-06, + "loss": 0.7944, + "num_input_tokens_seen": 38681575, + "step": 1818 + }, + { + "epoch": 0.21872181807250646, + "flos": 20856481392960.0, + "grad_norm": 2.7002848439609832, + "learning_rate": 3.638030997862155e-06, + "loss": 0.7778, + "num_input_tokens_seen": 38700510, + "step": 1819 + }, + { + "epoch": 0.21884206096314554, + "flos": 61122493436160.0, + "grad_norm": 0.7659189866504377, + "learning_rate": 3.6375839219815356e-06, + "loss": 0.5946, + "num_input_tokens_seen": 38758310, + "step": 1820 + }, + { + "epoch": 0.21896230385378465, + "flos": 23440149628800.0, + "grad_norm": 2.467727739207617, + "learning_rate": 3.6371365976807375e-06, + "loss": 0.8289, + "num_input_tokens_seen": 38778705, + "step": 1821 + }, + { + "epoch": 0.21908254674442373, + "flos": 25045986860160.0, + "grad_norm": 1.8822977462470525, + "learning_rate": 3.6366890250276185e-06, + "loss": 0.8341, + "num_input_tokens_seen": 38799660, + "step": 1822 + }, + { + "epoch": 0.21920278963506282, + "flos": 23480150024160.0, + "grad_norm": 2.3296925432750575, + "learning_rate": 3.6362412040900764e-06, + "loss": 0.9014, + "num_input_tokens_seen": 38819450, + "step": 1823 + }, + { + "epoch": 0.21932303252570193, + "flos": 29200286831040.0, + "grad_norm": 2.6204392913732897, + "learning_rate": 3.635793134936044e-06, + "loss": 0.8101, + "num_input_tokens_seen": 38840460, + "step": 1824 + }, + { + "epoch": 0.219443275416341, + "flos": 20776926639360.0, + "grad_norm": 1.760119606852337, + "learning_rate": 3.635344817633494e-06, + "loss": 0.7279, + "num_input_tokens_seen": 38859775, + "step": 1825 + }, + { + "epoch": 0.2195635183069801, + "flos": 14481050021760.0, + "grad_norm": 2.700845098692509, + "learning_rate": 3.634896252250436e-06, + "loss": 0.7513, + "num_input_tokens_seen": 38876540, + "step": 1826 + }, + { + "epoch": 0.2196837611976192, + "flos": 24208791992160.0, + "grad_norm": 1.9866481586955362, + "learning_rate": 3.6344474388549157e-06, + "loss": 0.8225, + "num_input_tokens_seen": 38896195, + "step": 1827 + }, + { + "epoch": 0.2198040040882583, + "flos": 18048719442240.0, + "grad_norm": 3.7012561315442554, + "learning_rate": 3.6339983775150183e-06, + "loss": 0.8025, + "num_input_tokens_seen": 38912755, + "step": 1828 + }, + { + "epoch": 0.21992424697889737, + "flos": 17758860954720.0, + "grad_norm": 3.009386636586731, + "learning_rate": 3.6335490682988664e-06, + "loss": 0.8471, + "num_input_tokens_seen": 38928365, + "step": 1829 + }, + { + "epoch": 0.22004448986953645, + "flos": 17613392749440.0, + "grad_norm": 2.258967437495944, + "learning_rate": 3.63309951127462e-06, + "loss": 0.83, + "num_input_tokens_seen": 38945275, + "step": 1830 + }, + { + "epoch": 0.22016473276017556, + "flos": 22243800373440.0, + "grad_norm": 6.879939021793981, + "learning_rate": 3.6326497065104757e-06, + "loss": 0.7542, + "num_input_tokens_seen": 38965740, + "step": 1831 + }, + { + "epoch": 0.22028497565081465, + "flos": 25519641309120.0, + "grad_norm": 2.243306482200659, + "learning_rate": 3.6321996540746697e-06, + "loss": 0.7795, + "num_input_tokens_seen": 38984815, + "step": 1832 + }, + { + "epoch": 0.22040521854145373, + "flos": 36195175384320.0, + "grad_norm": 1.9701269874695848, + "learning_rate": 3.6317493540354733e-06, + "loss": 0.8051, + "num_input_tokens_seen": 39008230, + "step": 1833 + }, + { + "epoch": 0.22052546143209284, + "flos": 11821804196640.0, + "grad_norm": 3.091225104318488, + "learning_rate": 3.6312988064611976e-06, + "loss": 0.7689, + "num_input_tokens_seen": 39026540, + "step": 1834 + }, + { + "epoch": 0.22064570432273192, + "flos": 24174701588640.0, + "grad_norm": 2.098418880331752, + "learning_rate": 3.6308480114201896e-06, + "loss": 0.8105, + "num_input_tokens_seen": 39048460, + "step": 1835 + }, + { + "epoch": 0.220765947213371, + "flos": 17905853120160.0, + "grad_norm": 2.004535657999906, + "learning_rate": 3.630396968980835e-06, + "loss": 0.7647, + "num_input_tokens_seen": 39066335, + "step": 1836 + }, + { + "epoch": 0.2208861901040101, + "flos": 26718555277920.0, + "grad_norm": 2.748818230296068, + "learning_rate": 3.6299456792115575e-06, + "loss": 0.8328, + "num_input_tokens_seen": 39087590, + "step": 1837 + }, + { + "epoch": 0.2210064329946492, + "flos": 17792505321120.0, + "grad_norm": 2.033149418043669, + "learning_rate": 3.629494142180815e-06, + "loss": 0.8109, + "num_input_tokens_seen": 39106335, + "step": 1838 + }, + { + "epoch": 0.22112667588528828, + "flos": 17941839181440.0, + "grad_norm": 2.7984093234945613, + "learning_rate": 3.6290423579571075e-06, + "loss": 0.8518, + "num_input_tokens_seen": 39122875, + "step": 1839 + }, + { + "epoch": 0.22124691877592736, + "flos": 18343075470720.0, + "grad_norm": 1.9186008755765707, + "learning_rate": 3.6285903266089694e-06, + "loss": 0.8032, + "num_input_tokens_seen": 39142150, + "step": 1840 + }, + { + "epoch": 0.22136716166656648, + "flos": 20783580026400.0, + "grad_norm": 2.6591160290649736, + "learning_rate": 3.628138048204974e-06, + "loss": 0.774, + "num_input_tokens_seen": 39162835, + "step": 1841 + }, + { + "epoch": 0.22148740455720556, + "flos": 17650568243040.0, + "grad_norm": 2.181596036641487, + "learning_rate": 3.6276855228137304e-06, + "loss": 0.7648, + "num_input_tokens_seen": 39181040, + "step": 1842 + }, + { + "epoch": 0.22160764744784464, + "flos": 21695720597760.0, + "grad_norm": 2.843374374230599, + "learning_rate": 3.6272327505038874e-06, + "loss": 0.8176, + "num_input_tokens_seen": 39197505, + "step": 1843 + }, + { + "epoch": 0.22172789033848372, + "flos": 23730565662720.0, + "grad_norm": 2.688722288165365, + "learning_rate": 3.626779731344131e-06, + "loss": 0.783, + "num_input_tokens_seen": 39217295, + "step": 1844 + }, + { + "epoch": 0.22184813322912283, + "flos": 16958172524640.0, + "grad_norm": 7.7754843312615165, + "learning_rate": 3.6263264654031814e-06, + "loss": 0.8553, + "num_input_tokens_seen": 39234195, + "step": 1845 + }, + { + "epoch": 0.22196837611976192, + "flos": 61735668928800.0, + "grad_norm": 0.7101201399717599, + "learning_rate": 3.6258729527498008e-06, + "loss": 0.5901, + "num_input_tokens_seen": 39295040, + "step": 1846 + }, + { + "epoch": 0.222088619010401, + "flos": 25521462627360.0, + "grad_norm": 2.75395698230088, + "learning_rate": 3.6254191934527854e-06, + "loss": 0.6489, + "num_input_tokens_seen": 39314395, + "step": 1847 + }, + { + "epoch": 0.2222088619010401, + "flos": 19291127763840.0, + "grad_norm": 2.0518212578690096, + "learning_rate": 3.6249651875809715e-06, + "loss": 0.653, + "num_input_tokens_seen": 39334275, + "step": 1848 + }, + { + "epoch": 0.2223291047916792, + "flos": 19072163475840.0, + "grad_norm": 2.2229232225497197, + "learning_rate": 3.62451093520323e-06, + "loss": 0.8939, + "num_input_tokens_seen": 39352180, + "step": 1849 + }, + { + "epoch": 0.22244934768231828, + "flos": 20820681180480.0, + "grad_norm": 2.3592746536439764, + "learning_rate": 3.6240564363884714e-06, + "loss": 0.905, + "num_input_tokens_seen": 39372125, + "step": 1850 + }, + { + "epoch": 0.2225695905729574, + "flos": 15610779600000.0, + "grad_norm": 2.3292989139397626, + "learning_rate": 3.623601691205643e-06, + "loss": 0.704, + "num_input_tokens_seen": 39391200, + "step": 1851 + }, + { + "epoch": 0.22268983346359647, + "flos": 25337220628800.0, + "grad_norm": 2.29631648117455, + "learning_rate": 3.623146699723729e-06, + "loss": 0.8149, + "num_input_tokens_seen": 39410660, + "step": 1852 + }, + { + "epoch": 0.22281007635423555, + "flos": 13242767543520.0, + "grad_norm": 2.377687740354293, + "learning_rate": 3.6226914620117507e-06, + "loss": 0.7773, + "num_input_tokens_seen": 39429280, + "step": 1853 + }, + { + "epoch": 0.22293031924487464, + "flos": 15318802436160.0, + "grad_norm": 3.149018416282125, + "learning_rate": 3.622235978138768e-06, + "loss": 0.8089, + "num_input_tokens_seen": 39446905, + "step": 1854 + }, + { + "epoch": 0.22305056213551375, + "flos": 22532395089120.0, + "grad_norm": 2.483873232667463, + "learning_rate": 3.621780248173877e-06, + "loss": 0.8144, + "num_input_tokens_seen": 39465105, + "step": 1855 + }, + { + "epoch": 0.22317080502615283, + "flos": 64787601998400.0, + "grad_norm": 0.8343155357516505, + "learning_rate": 3.6213242721862125e-06, + "loss": 0.6107, + "num_input_tokens_seen": 39523560, + "step": 1856 + }, + { + "epoch": 0.2232910479167919, + "flos": 25739088804000.0, + "grad_norm": 1.8487167702062017, + "learning_rate": 3.620868050244945e-06, + "loss": 0.751, + "num_input_tokens_seen": 39544040, + "step": 1857 + }, + { + "epoch": 0.22341129080743102, + "flos": 23218323269280.0, + "grad_norm": 2.2219896976359155, + "learning_rate": 3.6204115824192817e-06, + "loss": 0.7756, + "num_input_tokens_seen": 39561515, + "step": 1858 + }, + { + "epoch": 0.2235315336980701, + "flos": 21184258769280.0, + "grad_norm": 17.013399116923072, + "learning_rate": 3.619954868778471e-06, + "loss": 0.7622, + "num_input_tokens_seen": 39578210, + "step": 1859 + }, + { + "epoch": 0.2236517765887092, + "flos": 19873595301120.0, + "grad_norm": 2.004002381039726, + "learning_rate": 3.6194979093917944e-06, + "loss": 0.8304, + "num_input_tokens_seen": 39597625, + "step": 1860 + }, + { + "epoch": 0.22377201947934827, + "flos": 23181407964000.0, + "grad_norm": 2.1197084852639527, + "learning_rate": 3.6190407043285724e-06, + "loss": 0.8713, + "num_input_tokens_seen": 39615360, + "step": 1861 + }, + { + "epoch": 0.22389226236998738, + "flos": 26755730771520.0, + "grad_norm": 2.209661971775434, + "learning_rate": 3.618583253658163e-06, + "loss": 0.7559, + "num_input_tokens_seen": 39635460, + "step": 1862 + }, + { + "epoch": 0.22401250526062647, + "flos": 24135779116320.0, + "grad_norm": 2.417866003660607, + "learning_rate": 3.618125557449961e-06, + "loss": 0.8605, + "num_input_tokens_seen": 39653515, + "step": 1863 + }, + { + "epoch": 0.22413274815126555, + "flos": 16735825788480.0, + "grad_norm": 2.9766110393741605, + "learning_rate": 3.6176676157733983e-06, + "loss": 0.8308, + "num_input_tokens_seen": 39668525, + "step": 1864 + }, + { + "epoch": 0.22425299104190466, + "flos": 21329689804800.0, + "grad_norm": 2.5041141989019944, + "learning_rate": 3.6172094286979443e-06, + "loss": 0.7572, + "num_input_tokens_seen": 39685895, + "step": 1865 + }, + { + "epoch": 0.22437323393254374, + "flos": 32119500923040.0, + "grad_norm": 1.7574864742143845, + "learning_rate": 3.6167509962931064e-06, + "loss": 0.8159, + "num_input_tokens_seen": 39711595, + "step": 1866 + }, + { + "epoch": 0.22449347682318282, + "flos": 17977156187040.0, + "grad_norm": 3.709506560051806, + "learning_rate": 3.6162923186284276e-06, + "loss": 0.7675, + "num_input_tokens_seen": 39727555, + "step": 1867 + }, + { + "epoch": 0.2246137197138219, + "flos": 18670666998240.0, + "grad_norm": 3.5949523241386094, + "learning_rate": 3.6158333957734888e-06, + "loss": 0.8585, + "num_input_tokens_seen": 39746145, + "step": 1868 + }, + { + "epoch": 0.22473396260446102, + "flos": 15568363170240.0, + "grad_norm": 3.550518030329012, + "learning_rate": 3.6153742277979088e-06, + "loss": 0.8304, + "num_input_tokens_seen": 39763575, + "step": 1869 + }, + { + "epoch": 0.2248542054951001, + "flos": 14445063960480.0, + "grad_norm": 2.8245553941175374, + "learning_rate": 3.6149148147713434e-06, + "loss": 0.7835, + "num_input_tokens_seen": 39781810, + "step": 1870 + }, + { + "epoch": 0.22497444838573918, + "flos": 19216442248800.0, + "grad_norm": 2.2798339420212472, + "learning_rate": 3.614455156763484e-06, + "loss": 0.8639, + "num_input_tokens_seen": 39800115, + "step": 1871 + }, + { + "epoch": 0.2250946912763783, + "flos": 16886311911360.0, + "grad_norm": 2.8054893431283494, + "learning_rate": 3.613995253844061e-06, + "loss": 0.7127, + "num_input_tokens_seen": 39817795, + "step": 1872 + }, + { + "epoch": 0.22521493416701738, + "flos": 24646051512480.0, + "grad_norm": 2.0964432659919976, + "learning_rate": 3.6135351060828414e-06, + "loss": 0.8098, + "num_input_tokens_seen": 39838270, + "step": 1873 + }, + { + "epoch": 0.22533517705765646, + "flos": 17796631164480.0, + "grad_norm": 2.4812673649792822, + "learning_rate": 3.6130747135496285e-06, + "loss": 0.688, + "num_input_tokens_seen": 39857270, + "step": 1874 + }, + { + "epoch": 0.22545541994829554, + "flos": 33642884159520.0, + "grad_norm": 2.1740298113353806, + "learning_rate": 3.6126140763142646e-06, + "loss": 0.6643, + "num_input_tokens_seen": 39882300, + "step": 1875 + }, + { + "epoch": 0.22557566283893465, + "flos": 19144284277440.0, + "grad_norm": 3.8762751296335036, + "learning_rate": 3.6121531944466275e-06, + "loss": 0.8584, + "num_input_tokens_seen": 39899625, + "step": 1876 + }, + { + "epoch": 0.22569590572957374, + "flos": 20743505291520.0, + "grad_norm": 2.2840652861208253, + "learning_rate": 3.611692068016633e-06, + "loss": 0.7887, + "num_input_tokens_seen": 39915955, + "step": 1877 + }, + { + "epoch": 0.22581614862021282, + "flos": 18416497213920.0, + "grad_norm": 2.8023902602067596, + "learning_rate": 3.611230697094233e-06, + "loss": 0.7504, + "num_input_tokens_seen": 39932655, + "step": 1878 + }, + { + "epoch": 0.22593639151085193, + "flos": 20019844071360.0, + "grad_norm": 1.9894218845337621, + "learning_rate": 3.6107690817494173e-06, + "loss": 0.869, + "num_input_tokens_seen": 39952875, + "step": 1879 + }, + { + "epoch": 0.226056634401491, + "flos": 13096630282560.0, + "grad_norm": 3.506184427392775, + "learning_rate": 3.6103072220522117e-06, + "loss": 0.7064, + "num_input_tokens_seen": 39968405, + "step": 1880 + }, + { + "epoch": 0.2261768772921301, + "flos": 18961306050720.0, + "grad_norm": 2.8170817883758152, + "learning_rate": 3.609845118072682e-06, + "loss": 0.9215, + "num_input_tokens_seen": 39987075, + "step": 1881 + }, + { + "epoch": 0.2262971201827692, + "flos": 19945567423680.0, + "grad_norm": 3.196388119550444, + "learning_rate": 3.6093827698809276e-06, + "loss": 0.7959, + "num_input_tokens_seen": 40006175, + "step": 1882 + }, + { + "epoch": 0.2264173630734083, + "flos": 16630915524960.0, + "grad_norm": 2.525160351939859, + "learning_rate": 3.6089201775470864e-06, + "loss": 0.8484, + "num_input_tokens_seen": 40021630, + "step": 1883 + }, + { + "epoch": 0.22653760596404737, + "flos": 24354929253120.0, + "grad_norm": 1.6032562975442306, + "learning_rate": 3.6084573411413334e-06, + "loss": 0.7762, + "num_input_tokens_seen": 40041025, + "step": 1884 + }, + { + "epoch": 0.22665784885468646, + "flos": 18306457523520.0, + "grad_norm": 2.9776756628124015, + "learning_rate": 3.607994260733881e-06, + "loss": 0.8098, + "num_input_tokens_seen": 40060265, + "step": 1885 + }, + { + "epoch": 0.22677809174532557, + "flos": 24023583579840.0, + "grad_norm": 2.5555754352109648, + "learning_rate": 3.6075309363949776e-06, + "loss": 0.7458, + "num_input_tokens_seen": 40079435, + "step": 1886 + }, + { + "epoch": 0.22689833463596465, + "flos": 20345131073760.0, + "grad_norm": 2.4373052470337027, + "learning_rate": 3.6070673681949094e-06, + "loss": 0.8099, + "num_input_tokens_seen": 40097800, + "step": 1887 + }, + { + "epoch": 0.22701857752660373, + "flos": 30077259075840.0, + "grad_norm": 2.2614987386581475, + "learning_rate": 3.606603556203999e-06, + "loss": 0.8094, + "num_input_tokens_seen": 40122745, + "step": 1888 + }, + { + "epoch": 0.22713882041724284, + "flos": 22460051268960.0, + "grad_norm": 2.1709589501448687, + "learning_rate": 3.6061395004926066e-06, + "loss": 0.8375, + "num_input_tokens_seen": 40141680, + "step": 1889 + }, + { + "epoch": 0.22725906330788193, + "flos": 20492271918240.0, + "grad_norm": 2.575714371483767, + "learning_rate": 3.605675201131129e-06, + "loss": 0.8453, + "num_input_tokens_seen": 40160140, + "step": 1890 + }, + { + "epoch": 0.227379306198521, + "flos": 18962904350400.0, + "grad_norm": 2.907475530992113, + "learning_rate": 3.60521065819e-06, + "loss": 0.7984, + "num_input_tokens_seen": 40177450, + "step": 1891 + }, + { + "epoch": 0.2274995490891601, + "flos": 21767209513440.0, + "grad_norm": 2.05747060520686, + "learning_rate": 3.60474587173969e-06, + "loss": 0.8778, + "num_input_tokens_seen": 40195935, + "step": 1892 + }, + { + "epoch": 0.2276197919797992, + "flos": 19031382515520.0, + "grad_norm": 2.347713933099585, + "learning_rate": 3.6042808418507084e-06, + "loss": 0.8412, + "num_input_tokens_seen": 40213620, + "step": 1893 + }, + { + "epoch": 0.22774003487043828, + "flos": 18780000463200.0, + "grad_norm": 2.225264428531132, + "learning_rate": 3.6038155685935976e-06, + "loss": 0.7688, + "num_input_tokens_seen": 40230190, + "step": 1894 + }, + { + "epoch": 0.22786027776107737, + "flos": 22969654609440.0, + "grad_norm": 2.6493942648080804, + "learning_rate": 3.6033500520389404e-06, + "loss": 0.7065, + "num_input_tokens_seen": 40246860, + "step": 1895 + }, + { + "epoch": 0.22798052065171648, + "flos": 66611474273760.0, + "grad_norm": 0.7968693901313337, + "learning_rate": 3.6028842922573553e-06, + "loss": 0.6477, + "num_input_tokens_seen": 40311005, + "step": 1896 + }, + { + "epoch": 0.22810076354235556, + "flos": 61992514935840.0, + "grad_norm": 0.8649347272952939, + "learning_rate": 3.602418289319497e-06, + "loss": 0.6287, + "num_input_tokens_seen": 40369560, + "step": 1897 + }, + { + "epoch": 0.22822100643299464, + "flos": 23842129313280.0, + "grad_norm": 2.4116575242585223, + "learning_rate": 3.601952043296059e-06, + "loss": 0.7355, + "num_input_tokens_seen": 40389555, + "step": 1898 + }, + { + "epoch": 0.22834124932363373, + "flos": 20961205807680.0, + "grad_norm": 2.5261611648032876, + "learning_rate": 3.6014855542577696e-06, + "loss": 0.8025, + "num_input_tokens_seen": 40406765, + "step": 1899 + }, + { + "epoch": 0.22846149221427284, + "flos": 24865982214240.0, + "grad_norm": 2.4694073924472337, + "learning_rate": 3.6010188222753943e-06, + "loss": 0.845, + "num_input_tokens_seen": 40427535, + "step": 1900 + }, + { + "epoch": 0.22858173510491192, + "flos": 56052521766720.0, + "grad_norm": 0.9050351583385183, + "learning_rate": 3.6005518474197372e-06, + "loss": 0.6408, + "num_input_tokens_seen": 40479580, + "step": 1901 + }, + { + "epoch": 0.228701977995551, + "flos": 24136262323200.0, + "grad_norm": 2.3271148333442424, + "learning_rate": 3.6000846297616373e-06, + "loss": 0.784, + "num_input_tokens_seen": 40497880, + "step": 1902 + }, + { + "epoch": 0.22882222088619011, + "flos": 21357015312000.0, + "grad_norm": 2.6848765143443436, + "learning_rate": 3.5996171693719717e-06, + "loss": 0.7237, + "num_input_tokens_seen": 40513975, + "step": 1903 + }, + { + "epoch": 0.2289424637768292, + "flos": 64497260304000.0, + "grad_norm": 0.8483656977554805, + "learning_rate": 3.5991494663216528e-06, + "loss": 0.6474, + "num_input_tokens_seen": 40576960, + "step": 1904 + }, + { + "epoch": 0.22906270666746828, + "flos": 22131753516000.0, + "grad_norm": 2.0984993204766047, + "learning_rate": 3.5986815206816314e-06, + "loss": 0.8755, + "num_input_tokens_seen": 40595780, + "step": 1905 + }, + { + "epoch": 0.2291829495581074, + "flos": 25735297488480.0, + "grad_norm": 3.0537810480738785, + "learning_rate": 3.598213332522895e-06, + "loss": 0.7488, + "num_input_tokens_seen": 40615810, + "step": 1906 + }, + { + "epoch": 0.22930319244874647, + "flos": 31127879937600.0, + "grad_norm": 2.2749536446679928, + "learning_rate": 3.597744901916466e-06, + "loss": 0.775, + "num_input_tokens_seen": 40637095, + "step": 1907 + }, + { + "epoch": 0.22942343533938556, + "flos": 23220702133920.0, + "grad_norm": 2.0443251940713814, + "learning_rate": 3.5972762289334058e-06, + "loss": 0.7674, + "num_input_tokens_seen": 40656725, + "step": 1908 + }, + { + "epoch": 0.22954367823002464, + "flos": 14590234807680.0, + "grad_norm": 2.3316838246882847, + "learning_rate": 3.5968073136448116e-06, + "loss": 0.8489, + "num_input_tokens_seen": 40674745, + "step": 1909 + }, + { + "epoch": 0.22966392112066375, + "flos": 16739802952800.0, + "grad_norm": 2.172731299340372, + "learning_rate": 3.596338156121818e-06, + "loss": 0.9118, + "num_input_tokens_seen": 40693630, + "step": 1910 + }, + { + "epoch": 0.22978416401130283, + "flos": 67377774942240.0, + "grad_norm": 0.7436191604033574, + "learning_rate": 3.595868756435595e-06, + "loss": 0.593, + "num_input_tokens_seen": 40761310, + "step": 1911 + }, + { + "epoch": 0.22990440690194192, + "flos": 19836828674880.0, + "grad_norm": 2.531109558730542, + "learning_rate": 3.5953991146573504e-06, + "loss": 0.804, + "num_input_tokens_seen": 40779955, + "step": 1912 + }, + { + "epoch": 0.23002464979258103, + "flos": 13270204560000.0, + "grad_norm": 4.58422549263336, + "learning_rate": 3.5949292308583294e-06, + "loss": 0.8356, + "num_input_tokens_seen": 40793200, + "step": 1913 + }, + { + "epoch": 0.2301448926832201, + "flos": 22132125213600.0, + "grad_norm": 5.8965409823591175, + "learning_rate": 3.594459105109811e-06, + "loss": 0.8129, + "num_input_tokens_seen": 40812380, + "step": 1914 + }, + { + "epoch": 0.2302651355738592, + "flos": 20674952786880.0, + "grad_norm": 2.1670060503302637, + "learning_rate": 3.593988737483115e-06, + "loss": 0.8176, + "num_input_tokens_seen": 40832320, + "step": 1915 + }, + { + "epoch": 0.23038537846449827, + "flos": 18562522965600.0, + "grad_norm": 4.981804163798536, + "learning_rate": 3.5935181280495947e-06, + "loss": 0.7833, + "num_input_tokens_seen": 40850900, + "step": 1916 + }, + { + "epoch": 0.23050562135513739, + "flos": 64132418943360.0, + "grad_norm": 0.8017594852035117, + "learning_rate": 3.5930472768806412e-06, + "loss": 0.5429, + "num_input_tokens_seen": 40909570, + "step": 1917 + }, + { + "epoch": 0.23062586424577647, + "flos": 17288886312000.0, + "grad_norm": 2.1742909119638596, + "learning_rate": 3.5925761840476826e-06, + "loss": 0.7693, + "num_input_tokens_seen": 40928180, + "step": 1918 + }, + { + "epoch": 0.23074610713641555, + "flos": 27816052940640.0, + "grad_norm": 3.0095414382051504, + "learning_rate": 3.592104849622183e-06, + "loss": 0.8175, + "num_input_tokens_seen": 40950435, + "step": 1919 + }, + { + "epoch": 0.23086635002705466, + "flos": 28800462992640.0, + "grad_norm": 1.7190854525494403, + "learning_rate": 3.591633273675644e-06, + "loss": 0.7327, + "num_input_tokens_seen": 40972670, + "step": 1920 + }, + { + "epoch": 0.23098659291769374, + "flos": 62832874967040.0, + "grad_norm": 0.9116774522700115, + "learning_rate": 3.591161456279602e-06, + "loss": 0.582, + "num_input_tokens_seen": 41018215, + "step": 1921 + }, + { + "epoch": 0.23110683580833283, + "flos": 23443234718880.0, + "grad_norm": 1.8848689075692542, + "learning_rate": 3.590689397505633e-06, + "loss": 0.8022, + "num_input_tokens_seen": 41039125, + "step": 1922 + }, + { + "epoch": 0.2312270786989719, + "flos": 27231949933920.0, + "grad_norm": 2.081330773788009, + "learning_rate": 3.590217097425347e-06, + "loss": 0.8691, + "num_input_tokens_seen": 41059025, + "step": 1923 + }, + { + "epoch": 0.23134732158961102, + "flos": 13242656034240.0, + "grad_norm": 2.6103778670812687, + "learning_rate": 3.589744556110391e-06, + "loss": 0.7133, + "num_input_tokens_seen": 41077015, + "step": 1924 + }, + { + "epoch": 0.2314675644802501, + "flos": 36924449238240.0, + "grad_norm": 1.8647170702605185, + "learning_rate": 3.58927177363245e-06, + "loss": 0.8442, + "num_input_tokens_seen": 41099840, + "step": 1925 + }, + { + "epoch": 0.2315878073708892, + "flos": 23808113249280.0, + "grad_norm": 4.2103692345430375, + "learning_rate": 3.5887987500632447e-06, + "loss": 0.7246, + "num_input_tokens_seen": 41117845, + "step": 1926 + }, + { + "epoch": 0.2317080502615283, + "flos": 23006495575200.0, + "grad_norm": 2.4798835124068677, + "learning_rate": 3.5883254854745325e-06, + "loss": 0.8391, + "num_input_tokens_seen": 41136235, + "step": 1927 + }, + { + "epoch": 0.23182829315216738, + "flos": 11238890622240.0, + "grad_norm": 2.305396846941369, + "learning_rate": 3.587851979938107e-06, + "loss": 0.7493, + "num_input_tokens_seen": 41153125, + "step": 1928 + }, + { + "epoch": 0.23194853604280646, + "flos": 19800693934560.0, + "grad_norm": 2.3218953805596323, + "learning_rate": 3.5873782335257985e-06, + "loss": 0.7755, + "num_input_tokens_seen": 41171290, + "step": 1929 + }, + { + "epoch": 0.23206877893344555, + "flos": 15283373921280.0, + "grad_norm": 2.4585923419435614, + "learning_rate": 3.5869042463094744e-06, + "loss": 0.7845, + "num_input_tokens_seen": 41189005, + "step": 1930 + }, + { + "epoch": 0.23218902182408466, + "flos": 22679052726720.0, + "grad_norm": 2.076646321759849, + "learning_rate": 3.586430018361038e-06, + "loss": 0.772, + "num_input_tokens_seen": 41208775, + "step": 1931 + }, + { + "epoch": 0.23230926471472374, + "flos": 22678272161760.0, + "grad_norm": 3.115467486135191, + "learning_rate": 3.5859555497524283e-06, + "loss": 0.7647, + "num_input_tokens_seen": 41226010, + "step": 1932 + }, + { + "epoch": 0.23242950760536282, + "flos": 20346432015360.0, + "grad_norm": 2.000688196034421, + "learning_rate": 3.5854808405556237e-06, + "loss": 0.9169, + "num_input_tokens_seen": 41245245, + "step": 1933 + }, + { + "epoch": 0.23254975049600193, + "flos": 16884267574560.0, + "grad_norm": 2.8125126308581394, + "learning_rate": 3.5850058908426355e-06, + "loss": 0.7561, + "num_input_tokens_seen": 41263795, + "step": 1934 + }, + { + "epoch": 0.23266999338664102, + "flos": 23261520264000.0, + "grad_norm": 2.6229902475796028, + "learning_rate": 3.584530700685514e-06, + "loss": 0.853, + "num_input_tokens_seen": 41284055, + "step": 1935 + }, + { + "epoch": 0.2327902362772801, + "flos": 19541989439520.0, + "grad_norm": 28.751627204291086, + "learning_rate": 3.5840552701563448e-06, + "loss": 0.8864, + "num_input_tokens_seen": 41300255, + "step": 1936 + }, + { + "epoch": 0.2329104791679192, + "flos": 16703779721760.0, + "grad_norm": 2.3770111001685934, + "learning_rate": 3.5835795993272513e-06, + "loss": 0.8199, + "num_input_tokens_seen": 41318540, + "step": 1937 + }, + { + "epoch": 0.2330307220585583, + "flos": 22132088043840.0, + "grad_norm": 2.064923221186035, + "learning_rate": 3.583103688270391e-06, + "loss": 0.7165, + "num_input_tokens_seen": 41338320, + "step": 1938 + }, + { + "epoch": 0.23315096494919738, + "flos": 19289789652480.0, + "grad_norm": 3.360424073115159, + "learning_rate": 3.58262753705796e-06, + "loss": 0.8946, + "num_input_tokens_seen": 41353210, + "step": 1939 + }, + { + "epoch": 0.23327120783983646, + "flos": 52955793402720.0, + "grad_norm": 0.7627547823375614, + "learning_rate": 3.5821511457621902e-06, + "loss": 0.5552, + "num_input_tokens_seen": 41410510, + "step": 1940 + }, + { + "epoch": 0.23339145073047557, + "flos": 17102599976640.0, + "grad_norm": 7.074509394050156, + "learning_rate": 3.5816745144553497e-06, + "loss": 0.8177, + "num_input_tokens_seen": 41425830, + "step": 1941 + }, + { + "epoch": 0.23351169362111465, + "flos": 13060495542240.0, + "grad_norm": 2.3308922623554476, + "learning_rate": 3.5811976432097424e-06, + "loss": 0.7577, + "num_input_tokens_seen": 41443500, + "step": 1942 + }, + { + "epoch": 0.23363193651175373, + "flos": 15828777474240.0, + "grad_norm": 5.029084302243042, + "learning_rate": 3.58072053209771e-06, + "loss": 0.8452, + "num_input_tokens_seen": 41460055, + "step": 1943 + }, + { + "epoch": 0.23375217940239285, + "flos": 20995519229760.0, + "grad_norm": 2.6403348864936986, + "learning_rate": 3.5802431811916296e-06, + "loss": 0.789, + "num_input_tokens_seen": 41476665, + "step": 1944 + }, + { + "epoch": 0.23387242229303193, + "flos": 20565321963840.0, + "grad_norm": 2.077139369727798, + "learning_rate": 3.579765590563916e-06, + "loss": 0.8074, + "num_input_tokens_seen": 41496065, + "step": 1945 + }, + { + "epoch": 0.233992665183671, + "flos": 24244740883680.0, + "grad_norm": 2.659188743012721, + "learning_rate": 3.579287760287017e-06, + "loss": 0.8176, + "num_input_tokens_seen": 41516815, + "step": 1946 + }, + { + "epoch": 0.2341129080743101, + "flos": 30111349479360.0, + "grad_norm": 1.9908176275022056, + "learning_rate": 3.578809690433421e-06, + "loss": 0.7297, + "num_input_tokens_seen": 41538525, + "step": 1947 + }, + { + "epoch": 0.2342331509649492, + "flos": 22751582395680.0, + "grad_norm": 2.5148545501088377, + "learning_rate": 3.578331381075651e-06, + "loss": 0.8128, + "num_input_tokens_seen": 41559025, + "step": 1948 + }, + { + "epoch": 0.2343533938555883, + "flos": 23589371979840.0, + "grad_norm": 2.5165430135433153, + "learning_rate": 3.5778528322862646e-06, + "loss": 0.6911, + "num_input_tokens_seen": 41577845, + "step": 1949 + }, + { + "epoch": 0.23447363674622737, + "flos": 24535119747840.0, + "grad_norm": 2.0835332265959536, + "learning_rate": 3.5773740441378585e-06, + "loss": 0.8664, + "num_input_tokens_seen": 41598600, + "step": 1950 + }, + { + "epoch": 0.23459387963686648, + "flos": 53063571471360.0, + "grad_norm": 10.692423796115742, + "learning_rate": 3.5768950167030633e-06, + "loss": 0.7412, + "num_input_tokens_seen": 41623300, + "step": 1951 + }, + { + "epoch": 0.23471412252750556, + "flos": 23917149356160.0, + "grad_norm": 2.1940536419301586, + "learning_rate": 3.576415750054548e-06, + "loss": 0.7903, + "num_input_tokens_seen": 41643420, + "step": 1952 + }, + { + "epoch": 0.23483436541814465, + "flos": 15684350022240.0, + "grad_norm": 2.1216580534243494, + "learning_rate": 3.5759362442650172e-06, + "loss": 0.8579, + "num_input_tokens_seen": 41660170, + "step": 1953 + }, + { + "epoch": 0.23495460830878373, + "flos": 24900555824640.0, + "grad_norm": 2.805374273941809, + "learning_rate": 3.5754564994072113e-06, + "loss": 0.8548, + "num_input_tokens_seen": 41679010, + "step": 1954 + }, + { + "epoch": 0.23507485119942284, + "flos": 30438903837120.0, + "grad_norm": 2.615468112337689, + "learning_rate": 3.5749765155539067e-06, + "loss": 0.5985, + "num_input_tokens_seen": 41699495, + "step": 1955 + }, + { + "epoch": 0.23519509409006192, + "flos": 18302926396320.0, + "grad_norm": 3.1807520628465737, + "learning_rate": 3.574496292777917e-06, + "loss": 0.9254, + "num_input_tokens_seen": 41717705, + "step": 1956 + }, + { + "epoch": 0.235315336980701, + "flos": 29600668215840.0, + "grad_norm": 2.8053738490099214, + "learning_rate": 3.574015831152092e-06, + "loss": 0.7195, + "num_input_tokens_seen": 41738120, + "step": 1957 + }, + { + "epoch": 0.23543557987134012, + "flos": 18525310302240.0, + "grad_norm": 2.4469014328348573, + "learning_rate": 3.573535130749316e-06, + "loss": 0.8306, + "num_input_tokens_seen": 41756070, + "step": 1958 + }, + { + "epoch": 0.2355558227619792, + "flos": 24643635478080.0, + "grad_norm": 1.7601408055724646, + "learning_rate": 3.5730541916425127e-06, + "loss": 0.737, + "num_input_tokens_seen": 41777550, + "step": 1959 + }, + { + "epoch": 0.23567606565261828, + "flos": 21913458283680.0, + "grad_norm": 2.1372778124524006, + "learning_rate": 3.572573013904639e-06, + "loss": 0.8603, + "num_input_tokens_seen": 41797460, + "step": 1960 + }, + { + "epoch": 0.2357963085432574, + "flos": 13570173222240.0, + "grad_norm": 2.4260287197013803, + "learning_rate": 3.572091597608689e-06, + "loss": 0.9198, + "num_input_tokens_seen": 41815585, + "step": 1961 + }, + { + "epoch": 0.23591655143389648, + "flos": 22057068000960.0, + "grad_norm": 5.37007394437831, + "learning_rate": 3.571609942827694e-06, + "loss": 0.7396, + "num_input_tokens_seen": 41834700, + "step": 1962 + }, + { + "epoch": 0.23603679432453556, + "flos": 16993117832640.0, + "grad_norm": 3.8197323099182743, + "learning_rate": 3.57112804963472e-06, + "loss": 0.8808, + "num_input_tokens_seen": 41852915, + "step": 1963 + }, + { + "epoch": 0.23615703721517464, + "flos": 19143986919360.0, + "grad_norm": 2.02368115464107, + "learning_rate": 3.57064591810287e-06, + "loss": 0.7681, + "num_input_tokens_seen": 41870415, + "step": 1964 + }, + { + "epoch": 0.23627728010581375, + "flos": 19071605929440.0, + "grad_norm": 3.6091099761501155, + "learning_rate": 3.570163548305284e-06, + "loss": 0.8089, + "num_input_tokens_seen": 41889345, + "step": 1965 + }, + { + "epoch": 0.23639752299645284, + "flos": 14262011394240.0, + "grad_norm": 2.753532481700025, + "learning_rate": 3.569680940315135e-06, + "loss": 0.7024, + "num_input_tokens_seen": 41905745, + "step": 1966 + }, + { + "epoch": 0.23651776588709192, + "flos": 22860692842080.0, + "grad_norm": 2.1781711568005555, + "learning_rate": 3.5691980942056356e-06, + "loss": 0.8201, + "num_input_tokens_seen": 41925115, + "step": 1967 + }, + { + "epoch": 0.23663800877773103, + "flos": 18598025820000.0, + "grad_norm": 2.104264008439035, + "learning_rate": 3.5687150100500332e-06, + "loss": 0.7949, + "num_input_tokens_seen": 41944815, + "step": 1968 + }, + { + "epoch": 0.2367582516683701, + "flos": 25519195272000.0, + "grad_norm": 1.9044976857262188, + "learning_rate": 3.568231687921611e-06, + "loss": 0.746, + "num_input_tokens_seen": 41964990, + "step": 1969 + }, + { + "epoch": 0.2368784945590092, + "flos": 23261780452320.0, + "grad_norm": 1.6322206583586154, + "learning_rate": 3.5677481278936883e-06, + "loss": 0.8023, + "num_input_tokens_seen": 41984570, + "step": 1970 + }, + { + "epoch": 0.23699873744964828, + "flos": 69759837168000.0, + "grad_norm": 0.8356859416894361, + "learning_rate": 3.5672643300396214e-06, + "loss": 0.5781, + "num_input_tokens_seen": 42053715, + "step": 1971 + }, + { + "epoch": 0.2371189803402874, + "flos": 21803864630400.0, + "grad_norm": 2.2483403124558374, + "learning_rate": 3.566780294432802e-06, + "loss": 0.6718, + "num_input_tokens_seen": 42070890, + "step": 1972 + }, + { + "epoch": 0.23723922323092647, + "flos": 21877212034080.0, + "grad_norm": 2.730772424491428, + "learning_rate": 3.566296021146657e-06, + "loss": 0.752, + "num_input_tokens_seen": 42090270, + "step": 1973 + }, + { + "epoch": 0.23735946612156555, + "flos": 32661819385920.0, + "grad_norm": 2.103371561229985, + "learning_rate": 3.565811510254652e-06, + "loss": 0.7316, + "num_input_tokens_seen": 42111150, + "step": 1974 + }, + { + "epoch": 0.23747970901220466, + "flos": 70446285724800.0, + "grad_norm": 0.8420922276780445, + "learning_rate": 3.5653267618302845e-06, + "loss": 0.5819, + "num_input_tokens_seen": 42178730, + "step": 1975 + }, + { + "epoch": 0.23759995190284375, + "flos": 20820049294560.0, + "grad_norm": 3.2459647730280836, + "learning_rate": 3.564841775947093e-06, + "loss": 0.8597, + "num_input_tokens_seen": 42199620, + "step": 1976 + }, + { + "epoch": 0.23772019479348283, + "flos": 32874130286880.0, + "grad_norm": 2.6384644105617157, + "learning_rate": 3.5643565526786475e-06, + "loss": 0.7692, + "num_input_tokens_seen": 42219560, + "step": 1977 + }, + { + "epoch": 0.2378404376841219, + "flos": 32296271799840.0, + "grad_norm": 1.8672807742009012, + "learning_rate": 3.5638710920985574e-06, + "loss": 0.7731, + "num_input_tokens_seen": 42241020, + "step": 1978 + }, + { + "epoch": 0.23796068057476102, + "flos": 22964562352320.0, + "grad_norm": 2.121432057864511, + "learning_rate": 3.5633853942804655e-06, + "loss": 0.8248, + "num_input_tokens_seen": 42259345, + "step": 1979 + }, + { + "epoch": 0.2380809234654001, + "flos": 13461954850080.0, + "grad_norm": 4.123106088217628, + "learning_rate": 3.5628994592980527e-06, + "loss": 0.7653, + "num_input_tokens_seen": 42277250, + "step": 1980 + }, + { + "epoch": 0.2382011663560392, + "flos": 16847798306400.0, + "grad_norm": 2.068997408083974, + "learning_rate": 3.562413287225034e-06, + "loss": 0.7019, + "num_input_tokens_seen": 42295360, + "step": 1981 + }, + { + "epoch": 0.2383214092466783, + "flos": 18414861744480.0, + "grad_norm": 2.6583457308712326, + "learning_rate": 3.5619268781351623e-06, + "loss": 0.889, + "num_input_tokens_seen": 42313430, + "step": 1982 + }, + { + "epoch": 0.23844165213731738, + "flos": 19727569549440.0, + "grad_norm": 2.5798213117657722, + "learning_rate": 3.5614402321022256e-06, + "loss": 0.7731, + "num_input_tokens_seen": 42331260, + "step": 1983 + }, + { + "epoch": 0.23856189502795647, + "flos": 23333938423680.0, + "grad_norm": 2.193147725519959, + "learning_rate": 3.5609533492000463e-06, + "loss": 0.8687, + "num_input_tokens_seen": 42350150, + "step": 1984 + }, + { + "epoch": 0.23868213791859555, + "flos": 23441413400640.0, + "grad_norm": 2.9576065531840947, + "learning_rate": 3.560466229502485e-06, + "loss": 0.7859, + "num_input_tokens_seen": 42369495, + "step": 1985 + }, + { + "epoch": 0.23880238080923466, + "flos": 16593740031360.0, + "grad_norm": 2.354778739108252, + "learning_rate": 3.5599788730834384e-06, + "loss": 0.8937, + "num_input_tokens_seen": 42387455, + "step": 1986 + }, + { + "epoch": 0.23892262369987374, + "flos": 17323794450240.0, + "grad_norm": 3.363650134477849, + "learning_rate": 3.559491280016836e-06, + "loss": 0.7897, + "num_input_tokens_seen": 42405400, + "step": 1987 + }, + { + "epoch": 0.23904286659051283, + "flos": 22278039456000.0, + "grad_norm": 4.833399388530587, + "learning_rate": 3.5590034503766465e-06, + "loss": 0.7075, + "num_input_tokens_seen": 42425065, + "step": 1988 + }, + { + "epoch": 0.23916310948115194, + "flos": 21148012519680.0, + "grad_norm": 3.1490362842706845, + "learning_rate": 3.558515384236874e-06, + "loss": 0.8178, + "num_input_tokens_seen": 42442495, + "step": 1989 + }, + { + "epoch": 0.23928335237179102, + "flos": 14117398093440.0, + "grad_norm": 2.0389650986599714, + "learning_rate": 3.558027081671556e-06, + "loss": 0.8361, + "num_input_tokens_seen": 42459480, + "step": 1990 + }, + { + "epoch": 0.2394035952624301, + "flos": 23735620750080.0, + "grad_norm": 2.527335259778533, + "learning_rate": 3.557538542754769e-06, + "loss": 0.6863, + "num_input_tokens_seen": 42479175, + "step": 1991 + }, + { + "epoch": 0.2395238381530692, + "flos": 24171542159040.0, + "grad_norm": 2.0598511303191667, + "learning_rate": 3.557049767560623e-06, + "loss": 0.671, + "num_input_tokens_seen": 42498090, + "step": 1992 + }, + { + "epoch": 0.2396440810437083, + "flos": 25259338514400.0, + "grad_norm": 2.122711857083845, + "learning_rate": 3.5565607561632655e-06, + "loss": 0.8598, + "num_input_tokens_seen": 42516930, + "step": 1993 + }, + { + "epoch": 0.23976432393434738, + "flos": 28501943951040.0, + "grad_norm": 2.760942277110813, + "learning_rate": 3.5560715086368787e-06, + "loss": 0.7936, + "num_input_tokens_seen": 42534800, + "step": 1994 + }, + { + "epoch": 0.23988456682498646, + "flos": 19466374680480.0, + "grad_norm": 2.212298079075225, + "learning_rate": 3.5555820250556816e-06, + "loss": 0.8196, + "num_input_tokens_seen": 42552000, + "step": 1995 + }, + { + "epoch": 0.24000480971562557, + "flos": 20237433078240.0, + "grad_norm": 3.1136581594796557, + "learning_rate": 3.5550923054939278e-06, + "loss": 0.6943, + "num_input_tokens_seen": 42571455, + "step": 1996 + }, + { + "epoch": 0.24012505260626466, + "flos": 25406590868160.0, + "grad_norm": 2.030499363632121, + "learning_rate": 3.5546023500259083e-06, + "loss": 0.7441, + "num_input_tokens_seen": 42592550, + "step": 1997 + }, + { + "epoch": 0.24024529549690374, + "flos": 15531076167360.0, + "grad_norm": 2.7914134997303877, + "learning_rate": 3.5541121587259477e-06, + "loss": 0.8069, + "num_input_tokens_seen": 42610385, + "step": 1998 + }, + { + "epoch": 0.24036553838754285, + "flos": 57041317850400.0, + "grad_norm": 0.8639715550385032, + "learning_rate": 3.553621731668408e-06, + "loss": 0.5788, + "num_input_tokens_seen": 42673595, + "step": 1999 + }, + { + "epoch": 0.24048578127818193, + "flos": 24934088681760.0, + "grad_norm": 4.784195335597751, + "learning_rate": 3.553131068927688e-06, + "loss": 0.8356, + "num_input_tokens_seen": 42692000, + "step": 2000 + }, + { + "epoch": 0.24060602416882101, + "flos": 23297580664800.0, + "grad_norm": 1.7336081106683603, + "learning_rate": 3.552640170578219e-06, + "loss": 0.8069, + "num_input_tokens_seen": 42712250, + "step": 2001 + }, + { + "epoch": 0.2407262670594601, + "flos": 14153235475680.0, + "grad_norm": 2.654779184070554, + "learning_rate": 3.5521490366944703e-06, + "loss": 0.7817, + "num_input_tokens_seen": 42729900, + "step": 2002 + }, + { + "epoch": 0.2408465099500992, + "flos": 13643855153760.0, + "grad_norm": 2.4631966046115705, + "learning_rate": 3.5516576673509474e-06, + "loss": 0.803, + "num_input_tokens_seen": 42747900, + "step": 2003 + }, + { + "epoch": 0.2409667528407383, + "flos": 31203122999040.0, + "grad_norm": 1.9723175060137246, + "learning_rate": 3.5511660626221896e-06, + "loss": 0.862, + "num_input_tokens_seen": 42768540, + "step": 2004 + }, + { + "epoch": 0.24108699573137737, + "flos": 22168222784160.0, + "grad_norm": 2.3766895107261083, + "learning_rate": 3.5506742225827744e-06, + "loss": 0.8902, + "num_input_tokens_seen": 42785395, + "step": 2005 + }, + { + "epoch": 0.24120723862201648, + "flos": 26065602408480.0, + "grad_norm": 3.2108172094826513, + "learning_rate": 3.5501821473073116e-06, + "loss": 0.8987, + "num_input_tokens_seen": 42801980, + "step": 2006 + }, + { + "epoch": 0.24132748151265557, + "flos": 18598434687360.0, + "grad_norm": 2.361253306526385, + "learning_rate": 3.54968983687045e-06, + "loss": 0.8695, + "num_input_tokens_seen": 42818850, + "step": 2007 + }, + { + "epoch": 0.24144772440329465, + "flos": 15246049748640.0, + "grad_norm": 2.993862231227468, + "learning_rate": 3.549197291346872e-06, + "loss": 0.8969, + "num_input_tokens_seen": 42835375, + "step": 2008 + }, + { + "epoch": 0.24156796729393373, + "flos": 23989902043680.0, + "grad_norm": 2.2760129413203294, + "learning_rate": 3.548704510811297e-06, + "loss": 0.7958, + "num_input_tokens_seen": 42854570, + "step": 2009 + }, + { + "epoch": 0.24168821018457284, + "flos": 26249138181600.0, + "grad_norm": 3.5572861112418113, + "learning_rate": 3.5482114953384787e-06, + "loss": 0.7458, + "num_input_tokens_seen": 42873800, + "step": 2010 + }, + { + "epoch": 0.24180845307521193, + "flos": 18197272737600.0, + "grad_norm": 3.0697828617407623, + "learning_rate": 3.5477182450032077e-06, + "loss": 0.8461, + "num_input_tokens_seen": 42892400, + "step": 2011 + }, + { + "epoch": 0.241928695965851, + "flos": 20420113946880.0, + "grad_norm": 2.691728769142998, + "learning_rate": 3.5472247598803097e-06, + "loss": 0.8318, + "num_input_tokens_seen": 42911745, + "step": 2012 + }, + { + "epoch": 0.24204893885649012, + "flos": 25519864327680.0, + "grad_norm": 2.992600534482405, + "learning_rate": 3.546731040044645e-06, + "loss": 0.8572, + "num_input_tokens_seen": 42926915, + "step": 2013 + }, + { + "epoch": 0.2421691817471292, + "flos": 30616009241760.0, + "grad_norm": 2.2719656777945967, + "learning_rate": 3.546237085571112e-06, + "loss": 0.753, + "num_input_tokens_seen": 42945350, + "step": 2014 + }, + { + "epoch": 0.24228942463776829, + "flos": 21913681302240.0, + "grad_norm": 2.3201980305702383, + "learning_rate": 3.5457428965346425e-06, + "loss": 0.7269, + "num_input_tokens_seen": 42964290, + "step": 2015 + }, + { + "epoch": 0.2424096675284074, + "flos": 33935604718560.0, + "grad_norm": 1.8558765485254398, + "learning_rate": 3.545248473010205e-06, + "loss": 0.7486, + "num_input_tokens_seen": 42987095, + "step": 2016 + }, + { + "epoch": 0.24252991041904648, + "flos": 21622521873120.0, + "grad_norm": 2.004994117556954, + "learning_rate": 3.544753815072802e-06, + "loss": 0.879, + "num_input_tokens_seen": 43005750, + "step": 2017 + }, + { + "epoch": 0.24265015330968556, + "flos": 21838772768640.0, + "grad_norm": 2.1328768877643585, + "learning_rate": 3.544258922797474e-06, + "loss": 0.8873, + "num_input_tokens_seen": 43023870, + "step": 2018 + }, + { + "epoch": 0.24277039620032465, + "flos": 25591501922400.0, + "grad_norm": 2.682361303123042, + "learning_rate": 3.543763796259295e-06, + "loss": 0.7826, + "num_input_tokens_seen": 43043825, + "step": 2019 + }, + { + "epoch": 0.24289063909096376, + "flos": 26248915163040.0, + "grad_norm": 2.5891658304207192, + "learning_rate": 3.5432684355333754e-06, + "loss": 0.9111, + "num_input_tokens_seen": 43062480, + "step": 2020 + }, + { + "epoch": 0.24301088198160284, + "flos": 25038887436000.0, + "grad_norm": 2.0910406090155798, + "learning_rate": 3.5427728406948613e-06, + "loss": 0.7695, + "num_input_tokens_seen": 43081595, + "step": 2021 + }, + { + "epoch": 0.24313112487224192, + "flos": 69801621711840.0, + "grad_norm": 0.7477137767699537, + "learning_rate": 3.542277011818934e-06, + "loss": 0.5786, + "num_input_tokens_seen": 43145430, + "step": 2022 + }, + { + "epoch": 0.24325136776288103, + "flos": 40603756648800.0, + "grad_norm": 2.472761733470319, + "learning_rate": 3.5417809489808104e-06, + "loss": 0.7415, + "num_input_tokens_seen": 43167040, + "step": 2023 + }, + { + "epoch": 0.24337161065352012, + "flos": 25010558345280.0, + "grad_norm": 1.9532938159464885, + "learning_rate": 3.5412846522557422e-06, + "loss": 0.7267, + "num_input_tokens_seen": 43187930, + "step": 2024 + }, + { + "epoch": 0.2434918535441592, + "flos": 18634606597440.0, + "grad_norm": 2.083296800928878, + "learning_rate": 3.540788121719018e-06, + "loss": 0.7373, + "num_input_tokens_seen": 43206350, + "step": 2025 + }, + { + "epoch": 0.24361209643479828, + "flos": 23880940276320.0, + "grad_norm": 2.1252046374353135, + "learning_rate": 3.5402913574459604e-06, + "loss": 0.8232, + "num_input_tokens_seen": 43226255, + "step": 2026 + }, + { + "epoch": 0.2437323393254374, + "flos": 28616369673120.0, + "grad_norm": 1.9345851629959099, + "learning_rate": 3.5397943595119297e-06, + "loss": 0.8563, + "num_input_tokens_seen": 43247675, + "step": 2027 + }, + { + "epoch": 0.24385258221607647, + "flos": 23516656462080.0, + "grad_norm": 3.5410117198251356, + "learning_rate": 3.5392971279923177e-06, + "loss": 0.7766, + "num_input_tokens_seen": 43265895, + "step": 2028 + }, + { + "epoch": 0.24397282510671556, + "flos": 25299747777120.0, + "grad_norm": 2.5911407215061595, + "learning_rate": 3.5387996629625557e-06, + "loss": 0.8299, + "num_input_tokens_seen": 43283485, + "step": 2029 + }, + { + "epoch": 0.24409306799735467, + "flos": 65094707254560.0, + "grad_norm": 0.8636405162734259, + "learning_rate": 3.5383019644981083e-06, + "loss": 0.5503, + "num_input_tokens_seen": 43347180, + "step": 2030 + }, + { + "epoch": 0.24421331088799375, + "flos": 19509274317120.0, + "grad_norm": 2.8924900945627376, + "learning_rate": 3.5378040326744763e-06, + "loss": 0.7314, + "num_input_tokens_seen": 43366665, + "step": 2031 + }, + { + "epoch": 0.24433355377863283, + "flos": 21038493205920.0, + "grad_norm": 2.247989604097812, + "learning_rate": 3.5373058675671946e-06, + "loss": 0.8571, + "num_input_tokens_seen": 43384670, + "step": 2032 + }, + { + "epoch": 0.24445379666927192, + "flos": 22604813248800.0, + "grad_norm": 2.7918176372946832, + "learning_rate": 3.536807469251836e-06, + "loss": 0.72, + "num_input_tokens_seen": 43403825, + "step": 2033 + }, + { + "epoch": 0.24457403955991103, + "flos": 21221285583840.0, + "grad_norm": 2.2777731541325648, + "learning_rate": 3.5363088378040055e-06, + "loss": 0.828, + "num_input_tokens_seen": 43422715, + "step": 2034 + }, + { + "epoch": 0.2446942824505501, + "flos": 66901964647200.0, + "grad_norm": 0.7645926229829013, + "learning_rate": 3.5358099732993463e-06, + "loss": 0.6433, + "num_input_tokens_seen": 43481825, + "step": 2035 + }, + { + "epoch": 0.2448145253411892, + "flos": 20381972039520.0, + "grad_norm": 2.5947863794796397, + "learning_rate": 3.535310875813535e-06, + "loss": 0.8977, + "num_input_tokens_seen": 43500220, + "step": 2036 + }, + { + "epoch": 0.2449347682318283, + "flos": 28763436178080.0, + "grad_norm": 3.7517950212328572, + "learning_rate": 3.5348115454222843e-06, + "loss": 0.8176, + "num_input_tokens_seen": 43522805, + "step": 2037 + }, + { + "epoch": 0.2450550111224674, + "flos": 22496817895200.0, + "grad_norm": 2.4859257302763886, + "learning_rate": 3.5343119822013425e-06, + "loss": 0.8617, + "num_input_tokens_seen": 43541915, + "step": 2038 + }, + { + "epoch": 0.24517525401310647, + "flos": 21727952513280.0, + "grad_norm": 4.891440229140983, + "learning_rate": 3.533812186226493e-06, + "loss": 0.7776, + "num_input_tokens_seen": 43562625, + "step": 2039 + }, + { + "epoch": 0.24529549690374555, + "flos": 25007361745920.0, + "grad_norm": 2.0905963433871895, + "learning_rate": 3.5333121575735545e-06, + "loss": 0.7561, + "num_input_tokens_seen": 43582265, + "step": 2040 + }, + { + "epoch": 0.24541573979438466, + "flos": 32077456190880.0, + "grad_norm": 2.1661624566299693, + "learning_rate": 3.532811896318381e-06, + "loss": 0.7537, + "num_input_tokens_seen": 43604335, + "step": 2041 + }, + { + "epoch": 0.24553598268502375, + "flos": 31312493633760.0, + "grad_norm": 2.930668455563602, + "learning_rate": 3.5323114025368615e-06, + "loss": 0.8217, + "num_input_tokens_seen": 43619400, + "step": 2042 + }, + { + "epoch": 0.24565622557566283, + "flos": 14007655761120.0, + "grad_norm": 2.5909825463658587, + "learning_rate": 3.53181067630492e-06, + "loss": 0.8213, + "num_input_tokens_seen": 43636870, + "step": 2043 + }, + { + "epoch": 0.24577646846630194, + "flos": 16557939818880.0, + "grad_norm": 2.6498439958825215, + "learning_rate": 3.5313097176985175e-06, + "loss": 0.7604, + "num_input_tokens_seen": 43655180, + "step": 2044 + }, + { + "epoch": 0.24589671135694102, + "flos": 18780669518880.0, + "grad_norm": 2.0778679939761875, + "learning_rate": 3.5308085267936482e-06, + "loss": 0.8062, + "num_input_tokens_seen": 43672895, + "step": 2045 + }, + { + "epoch": 0.2460169542475801, + "flos": 19910138908800.0, + "grad_norm": 1.910624066503295, + "learning_rate": 3.530307103666342e-06, + "loss": 0.8983, + "num_input_tokens_seen": 43691975, + "step": 2046 + }, + { + "epoch": 0.24613719713821922, + "flos": 24136671190560.0, + "grad_norm": 1.951541716661188, + "learning_rate": 3.5298054483926658e-06, + "loss": 0.8043, + "num_input_tokens_seen": 43712670, + "step": 2047 + }, + { + "epoch": 0.2462574400288583, + "flos": 30177448779840.0, + "grad_norm": 3.6637949245982258, + "learning_rate": 3.5293035610487187e-06, + "loss": 0.8264, + "num_input_tokens_seen": 43732035, + "step": 2048 + }, + { + "epoch": 0.24637768291949738, + "flos": 68844760185600.0, + "grad_norm": 0.7264858738680399, + "learning_rate": 3.5288014417106374e-06, + "loss": 0.6194, + "num_input_tokens_seen": 43798055, + "step": 2049 + }, + { + "epoch": 0.24649792581013646, + "flos": 34334462143200.0, + "grad_norm": 1.853392843153236, + "learning_rate": 3.528299090454593e-06, + "loss": 0.7532, + "num_input_tokens_seen": 43818590, + "step": 2050 + }, + { + "epoch": 0.24661816870077558, + "flos": 19652326488000.0, + "grad_norm": 2.5284706209545473, + "learning_rate": 3.527796507356792e-06, + "loss": 0.8252, + "num_input_tokens_seen": 43832480, + "step": 2051 + }, + { + "epoch": 0.24673841159141466, + "flos": 19973450477280.0, + "grad_norm": 3.588566234192634, + "learning_rate": 3.527293692493475e-06, + "loss": 0.9049, + "num_input_tokens_seen": 43848345, + "step": 2052 + }, + { + "epoch": 0.24685865448205374, + "flos": 21615459618720.0, + "grad_norm": 2.493798384041706, + "learning_rate": 3.52679064594092e-06, + "loss": 0.7339, + "num_input_tokens_seen": 43865685, + "step": 2053 + }, + { + "epoch": 0.24697889737269285, + "flos": 17934665417760.0, + "grad_norm": 3.7748013829408107, + "learning_rate": 3.5262873677754375e-06, + "loss": 0.7525, + "num_input_tokens_seen": 43883690, + "step": 2054 + }, + { + "epoch": 0.24709914026333193, + "flos": 27305408846880.0, + "grad_norm": 2.628262082401887, + "learning_rate": 3.5257838580733745e-06, + "loss": 0.8067, + "num_input_tokens_seen": 43903295, + "step": 2055 + }, + { + "epoch": 0.24721938315397102, + "flos": 19247856429600.0, + "grad_norm": 8.75431804377823, + "learning_rate": 3.5252801169111138e-06, + "loss": 0.873, + "num_input_tokens_seen": 43920960, + "step": 2056 + }, + { + "epoch": 0.2473396260446101, + "flos": 23152670005920.0, + "grad_norm": 1.8803144796601003, + "learning_rate": 3.524776144365072e-06, + "loss": 0.7972, + "num_input_tokens_seen": 43939415, + "step": 2057 + }, + { + "epoch": 0.2474598689352492, + "flos": 21112286646720.0, + "grad_norm": 2.5997579461911795, + "learning_rate": 3.5242719405117016e-06, + "loss": 0.7941, + "num_input_tokens_seen": 43959220, + "step": 2058 + }, + { + "epoch": 0.2475801118258883, + "flos": 21617280936960.0, + "grad_norm": 46.796668926925236, + "learning_rate": 3.5237675054274893e-06, + "loss": 0.7486, + "num_input_tokens_seen": 43978420, + "step": 2059 + }, + { + "epoch": 0.24770035471652738, + "flos": 22642100251680.0, + "grad_norm": 2.209672653672747, + "learning_rate": 3.5232628391889584e-06, + "loss": 0.807, + "num_input_tokens_seen": 43996710, + "step": 2060 + }, + { + "epoch": 0.2478205976071665, + "flos": 22132162383360.0, + "grad_norm": 2.9563071247412536, + "learning_rate": 3.522757941872666e-06, + "loss": 0.6457, + "num_input_tokens_seen": 44014865, + "step": 2061 + }, + { + "epoch": 0.24794084049780557, + "flos": 24937359620640.0, + "grad_norm": 1.8663597027677588, + "learning_rate": 3.5222528135552042e-06, + "loss": 0.8248, + "num_input_tokens_seen": 44036325, + "step": 2062 + }, + { + "epoch": 0.24806108338844465, + "flos": 18270359952960.0, + "grad_norm": 2.13964888321538, + "learning_rate": 3.521747454313201e-06, + "loss": 0.8057, + "num_input_tokens_seen": 44055005, + "step": 2063 + }, + { + "epoch": 0.24818132627908374, + "flos": 19254732835200.0, + "grad_norm": 2.1221916490563966, + "learning_rate": 3.521241864223319e-06, + "loss": 0.6666, + "num_input_tokens_seen": 44073965, + "step": 2064 + }, + { + "epoch": 0.24830156916972285, + "flos": 70185351044160.0, + "grad_norm": 0.7963127695970443, + "learning_rate": 3.5207360433622552e-06, + "loss": 0.6187, + "num_input_tokens_seen": 44135965, + "step": 2065 + }, + { + "epoch": 0.24842181206036193, + "flos": 40350999315360.0, + "grad_norm": 2.1168057143228665, + "learning_rate": 3.5202299918067437e-06, + "loss": 0.7445, + "num_input_tokens_seen": 44159560, + "step": 2066 + }, + { + "epoch": 0.248542054951001, + "flos": 20053748626080.0, + "grad_norm": 2.941878930521394, + "learning_rate": 3.519723709633551e-06, + "loss": 0.6994, + "num_input_tokens_seen": 44178320, + "step": 2067 + }, + { + "epoch": 0.24866229784164012, + "flos": 23480335872960.0, + "grad_norm": 1.9997100619918005, + "learning_rate": 3.519217196919479e-06, + "loss": 0.8366, + "num_input_tokens_seen": 44197305, + "step": 2068 + }, + { + "epoch": 0.2487825407322792, + "flos": 19836791505120.0, + "grad_norm": 1.9663729168109312, + "learning_rate": 3.518710453741367e-06, + "loss": 0.7368, + "num_input_tokens_seen": 44216505, + "step": 2069 + }, + { + "epoch": 0.2489027836229183, + "flos": 21986173801440.0, + "grad_norm": 2.2316408301242685, + "learning_rate": 3.518203480176086e-06, + "loss": 0.6779, + "num_input_tokens_seen": 44235435, + "step": 2070 + }, + { + "epoch": 0.2490230265135574, + "flos": 23261037057120.0, + "grad_norm": 2.0476446670268658, + "learning_rate": 3.517696276300545e-06, + "loss": 0.8074, + "num_input_tokens_seen": 44256975, + "step": 2071 + }, + { + "epoch": 0.24914326940419648, + "flos": 19798463748960.0, + "grad_norm": 3.094069030958256, + "learning_rate": 3.517188842191685e-06, + "loss": 0.6972, + "num_input_tokens_seen": 44275125, + "step": 2072 + }, + { + "epoch": 0.24926351229483557, + "flos": 20200406263680.0, + "grad_norm": 1.7687475224669664, + "learning_rate": 3.5166811779264837e-06, + "loss": 0.7373, + "num_input_tokens_seen": 44295445, + "step": 2073 + }, + { + "epoch": 0.24938375518547465, + "flos": 23261260075680.0, + "grad_norm": 2.321651085470165, + "learning_rate": 3.5161732835819545e-06, + "loss": 0.7818, + "num_input_tokens_seen": 44314755, + "step": 2074 + }, + { + "epoch": 0.24950399807611376, + "flos": 17287102163520.0, + "grad_norm": 2.654991929065341, + "learning_rate": 3.515665159235143e-06, + "loss": 0.8303, + "num_input_tokens_seen": 44333640, + "step": 2075 + }, + { + "epoch": 0.24962424096675284, + "flos": 18997738149120.0, + "grad_norm": 2.3723579568418054, + "learning_rate": 3.5151568049631318e-06, + "loss": 0.7472, + "num_input_tokens_seen": 44352355, + "step": 2076 + }, + { + "epoch": 0.24974448385739192, + "flos": 33350683977120.0, + "grad_norm": 2.014666425483196, + "learning_rate": 3.5146482208430385e-06, + "loss": 0.8029, + "num_input_tokens_seen": 44374185, + "step": 2077 + }, + { + "epoch": 0.24986472674803104, + "flos": 29964171465120.0, + "grad_norm": 2.045956704122559, + "learning_rate": 3.514139406952014e-06, + "loss": 0.6803, + "num_input_tokens_seen": 44396210, + "step": 2078 + }, + { + "epoch": 0.24998496963867012, + "flos": 26574871221120.0, + "grad_norm": 2.000485122972397, + "learning_rate": 3.5136303633672454e-06, + "loss": 0.835, + "num_input_tokens_seen": 44416340, + "step": 2079 + }, + { + "epoch": 0.25010521252930923, + "flos": 23520447777600.0, + "grad_norm": 1.8827306326008169, + "learning_rate": 3.5131210901659544e-06, + "loss": 0.75, + "num_input_tokens_seen": 44438095, + "step": 2080 + }, + { + "epoch": 0.2502254554199483, + "flos": 23407694694720.0, + "grad_norm": 2.6867921473667917, + "learning_rate": 3.5126115874253967e-06, + "loss": 0.8225, + "num_input_tokens_seen": 44457650, + "step": 2081 + }, + { + "epoch": 0.2503456983105874, + "flos": 28720202013600.0, + "grad_norm": 11.479641350118206, + "learning_rate": 3.5121018552228644e-06, + "loss": 0.8082, + "num_input_tokens_seen": 44476155, + "step": 2082 + }, + { + "epoch": 0.2504659412012265, + "flos": 18743531195040.0, + "grad_norm": 2.121713228196751, + "learning_rate": 3.5115918936356827e-06, + "loss": 0.7616, + "num_input_tokens_seen": 44492910, + "step": 2083 + }, + { + "epoch": 0.25058618409186556, + "flos": 16849433775840.0, + "grad_norm": 2.332564658753979, + "learning_rate": 3.5110817027412123e-06, + "loss": 0.788, + "num_input_tokens_seen": 44512000, + "step": 2084 + }, + { + "epoch": 0.25070642698250467, + "flos": 24390580786560.0, + "grad_norm": 4.342171361563096, + "learning_rate": 3.5105712826168493e-06, + "loss": 0.6862, + "num_input_tokens_seen": 44531650, + "step": 2085 + }, + { + "epoch": 0.2508266698731437, + "flos": 20231857614240.0, + "grad_norm": 2.162443030163711, + "learning_rate": 3.5100606333400235e-06, + "loss": 0.7117, + "num_input_tokens_seen": 44548785, + "step": 2086 + }, + { + "epoch": 0.25094691276378284, + "flos": 19217929039200.0, + "grad_norm": 2.3892455266048667, + "learning_rate": 3.5095497549882006e-06, + "loss": 0.7736, + "num_input_tokens_seen": 44567230, + "step": 2087 + }, + { + "epoch": 0.25106715565442195, + "flos": 26904618594720.0, + "grad_norm": 2.980452111274002, + "learning_rate": 3.50903864763888e-06, + "loss": 0.7235, + "num_input_tokens_seen": 44588020, + "step": 2088 + }, + { + "epoch": 0.251187398545061, + "flos": 48289739978880.0, + "grad_norm": 7.708421971266842, + "learning_rate": 3.5085273113695965e-06, + "loss": 0.7638, + "num_input_tokens_seen": 44612590, + "step": 2089 + }, + { + "epoch": 0.2513076414357001, + "flos": 26977371282240.0, + "grad_norm": 3.3260989043944504, + "learning_rate": 3.508015746257919e-06, + "loss": 0.7855, + "num_input_tokens_seen": 44631630, + "step": 2090 + }, + { + "epoch": 0.2514278843263392, + "flos": 19435889743680.0, + "grad_norm": 2.515901728034494, + "learning_rate": 3.5075039523814518e-06, + "loss": 0.8344, + "num_input_tokens_seen": 44650820, + "step": 2091 + }, + { + "epoch": 0.2515481272169783, + "flos": 16841888314560.0, + "grad_norm": 2.702295429628163, + "learning_rate": 3.506991929817834e-06, + "loss": 0.8173, + "num_input_tokens_seen": 44667540, + "step": 2092 + }, + { + "epoch": 0.2516683701076174, + "flos": 23698519596000.0, + "grad_norm": 1.9753472131624932, + "learning_rate": 3.506479678644738e-06, + "loss": 0.8239, + "num_input_tokens_seen": 44687935, + "step": 2093 + }, + { + "epoch": 0.2517886129982565, + "flos": 27596308087680.0, + "grad_norm": 2.5873354910186497, + "learning_rate": 3.505967198939873e-06, + "loss": 0.7377, + "num_input_tokens_seen": 44704975, + "step": 2094 + }, + { + "epoch": 0.25190885588889556, + "flos": 38049606936000.0, + "grad_norm": 2.0999461484997943, + "learning_rate": 3.5054544907809813e-06, + "loss": 0.7799, + "num_input_tokens_seen": 44725475, + "step": 2095 + }, + { + "epoch": 0.25202909877953467, + "flos": 22237964721120.0, + "grad_norm": 2.242956687498858, + "learning_rate": 3.50494155424584e-06, + "loss": 0.8043, + "num_input_tokens_seen": 44744380, + "step": 2096 + }, + { + "epoch": 0.2521493416701738, + "flos": 21730665905760.0, + "grad_norm": 1.860244739899788, + "learning_rate": 3.504428389412262e-06, + "loss": 0.8336, + "num_input_tokens_seen": 44765190, + "step": 2097 + }, + { + "epoch": 0.25226958456081283, + "flos": 27706979664000.0, + "grad_norm": 2.415602109733983, + "learning_rate": 3.5039149963580927e-06, + "loss": 0.7353, + "num_input_tokens_seen": 44785210, + "step": 2098 + }, + { + "epoch": 0.25238982745145194, + "flos": 30687832685280.0, + "grad_norm": 3.0699655273688986, + "learning_rate": 3.503401375161215e-06, + "loss": 0.7032, + "num_input_tokens_seen": 44804955, + "step": 2099 + }, + { + "epoch": 0.252510070342091, + "flos": 20237433078240.0, + "grad_norm": 2.060883090535221, + "learning_rate": 3.502887525899544e-06, + "loss": 0.8343, + "num_input_tokens_seen": 44823935, + "step": 2100 + }, + { + "epoch": 0.2526303132327301, + "flos": 22714890108960.0, + "grad_norm": 2.4996200272764897, + "learning_rate": 3.50237344865103e-06, + "loss": 0.8285, + "num_input_tokens_seen": 44844935, + "step": 2101 + }, + { + "epoch": 0.2527505561233692, + "flos": 30220013888640.0, + "grad_norm": 2.431135917281697, + "learning_rate": 3.501859143493658e-06, + "loss": 0.7602, + "num_input_tokens_seen": 44865565, + "step": 2102 + }, + { + "epoch": 0.2528707990140083, + "flos": 58409308555680.0, + "grad_norm": 0.9105648711597373, + "learning_rate": 3.5013446105054488e-06, + "loss": 0.6045, + "num_input_tokens_seen": 44918485, + "step": 2103 + }, + { + "epoch": 0.2529910419046474, + "flos": 24609879602400.0, + "grad_norm": 1.784072914540825, + "learning_rate": 3.5008298497644555e-06, + "loss": 0.7489, + "num_input_tokens_seen": 44937530, + "step": 2104 + }, + { + "epoch": 0.2531112847952865, + "flos": 23808038909760.0, + "grad_norm": 3.506699825270304, + "learning_rate": 3.500314861348767e-06, + "loss": 0.8793, + "num_input_tokens_seen": 44958165, + "step": 2105 + }, + { + "epoch": 0.25323152768592555, + "flos": 16120940486880.0, + "grad_norm": 3.2307150953249724, + "learning_rate": 3.499799645336507e-06, + "loss": 0.7748, + "num_input_tokens_seen": 44975060, + "step": 2106 + }, + { + "epoch": 0.25335177057656466, + "flos": 28364504413920.0, + "grad_norm": 1.723128550307553, + "learning_rate": 3.4992842018058336e-06, + "loss": 0.8683, + "num_input_tokens_seen": 44997960, + "step": 2107 + }, + { + "epoch": 0.25347201346720377, + "flos": 18772640850720.0, + "grad_norm": 2.2584433548744394, + "learning_rate": 3.4987685308349384e-06, + "loss": 0.8863, + "num_input_tokens_seen": 45015690, + "step": 2108 + }, + { + "epoch": 0.2535922563578428, + "flos": 15793088771040.0, + "grad_norm": 2.5093917865120448, + "learning_rate": 3.4982526325020497e-06, + "loss": 0.6151, + "num_input_tokens_seen": 45032660, + "step": 2109 + }, + { + "epoch": 0.25371249924848194, + "flos": 16295518347840.0, + "grad_norm": 2.657425743637398, + "learning_rate": 3.4977365068854273e-06, + "loss": 0.8231, + "num_input_tokens_seen": 45048280, + "step": 2110 + }, + { + "epoch": 0.25383274213912105, + "flos": 21730517226720.0, + "grad_norm": 2.0840932858724797, + "learning_rate": 3.4972201540633676e-06, + "loss": 0.7343, + "num_input_tokens_seen": 45067635, + "step": 2111 + }, + { + "epoch": 0.2539529850297601, + "flos": 21366307752000.0, + "grad_norm": 2.1190489296477923, + "learning_rate": 3.4967035741142008e-06, + "loss": 0.8511, + "num_input_tokens_seen": 45086455, + "step": 2112 + }, + { + "epoch": 0.2540732279203992, + "flos": 25192347139680.0, + "grad_norm": 2.8728593237347573, + "learning_rate": 3.4961867671162917e-06, + "loss": 0.8188, + "num_input_tokens_seen": 45106385, + "step": 2113 + }, + { + "epoch": 0.2541934708110383, + "flos": 19399717833600.0, + "grad_norm": 3.1598696961879944, + "learning_rate": 3.4956697331480402e-06, + "loss": 0.7744, + "num_input_tokens_seen": 45124955, + "step": 2114 + }, + { + "epoch": 0.2543137137016774, + "flos": 23915105019360.0, + "grad_norm": 1.8992658150293014, + "learning_rate": 3.495152472287879e-06, + "loss": 0.7982, + "num_input_tokens_seen": 45145665, + "step": 2115 + }, + { + "epoch": 0.2544339565923165, + "flos": 25556630953920.0, + "grad_norm": 1.9305193262642024, + "learning_rate": 3.4946349846142766e-06, + "loss": 0.7364, + "num_input_tokens_seen": 45164805, + "step": 2116 + }, + { + "epoch": 0.25455419948295555, + "flos": 21658507934400.0, + "grad_norm": 2.520704744280241, + "learning_rate": 3.4941172702057353e-06, + "loss": 0.7593, + "num_input_tokens_seen": 45184865, + "step": 2117 + }, + { + "epoch": 0.25467444237359466, + "flos": 26212743252960.0, + "grad_norm": 2.3571268448554137, + "learning_rate": 3.4935993291407924e-06, + "loss": 0.8041, + "num_input_tokens_seen": 45203690, + "step": 2118 + }, + { + "epoch": 0.25479468526423377, + "flos": 26941013523360.0, + "grad_norm": 3.271315731427503, + "learning_rate": 3.4930811614980183e-06, + "loss": 0.712, + "num_input_tokens_seen": 45225385, + "step": 2119 + }, + { + "epoch": 0.2549149281548728, + "flos": 23442193965600.0, + "grad_norm": 1.840505298808957, + "learning_rate": 3.4925627673560198e-06, + "loss": 0.791, + "num_input_tokens_seen": 45246045, + "step": 2120 + }, + { + "epoch": 0.25503517104551193, + "flos": 25775335053600.0, + "grad_norm": 2.2196625458212735, + "learning_rate": 3.4920441467934357e-06, + "loss": 0.8815, + "num_input_tokens_seen": 45266560, + "step": 2121 + }, + { + "epoch": 0.25515541393615104, + "flos": 26606694269280.0, + "grad_norm": 3.1090152091423917, + "learning_rate": 3.491525299888941e-06, + "loss": 0.8301, + "num_input_tokens_seen": 45285405, + "step": 2122 + }, + { + "epoch": 0.2552756568267901, + "flos": 65862197355360.0, + "grad_norm": 0.8892132462012972, + "learning_rate": 3.491006226721244e-06, + "loss": 0.6259, + "num_input_tokens_seen": 45349615, + "step": 2123 + }, + { + "epoch": 0.2553958997174292, + "flos": 17905592931840.0, + "grad_norm": 2.1458502710161578, + "learning_rate": 3.4904869273690882e-06, + "loss": 0.7721, + "num_input_tokens_seen": 45367505, + "step": 2124 + }, + { + "epoch": 0.2555161426080683, + "flos": 23334384460800.0, + "grad_norm": 1.982545264402009, + "learning_rate": 3.489967401911251e-06, + "loss": 0.8899, + "num_input_tokens_seen": 45386805, + "step": 2125 + }, + { + "epoch": 0.2556363854987074, + "flos": 40566581155200.0, + "grad_norm": 2.043383583562806, + "learning_rate": 3.4894476504265428e-06, + "loss": 0.6984, + "num_input_tokens_seen": 45411045, + "step": 2126 + }, + { + "epoch": 0.2557566283893465, + "flos": 67922100572160.0, + "grad_norm": 0.7371513560944181, + "learning_rate": 3.4889276729938104e-06, + "loss": 0.5435, + "num_input_tokens_seen": 45469575, + "step": 2127 + }, + { + "epoch": 0.2558768712799856, + "flos": 22603214949120.0, + "grad_norm": 7.987942908100694, + "learning_rate": 3.488407469691934e-06, + "loss": 0.8068, + "num_input_tokens_seen": 45490270, + "step": 2128 + }, + { + "epoch": 0.25599711417062465, + "flos": 26358545986080.0, + "grad_norm": 3.06207256837611, + "learning_rate": 3.487887040599828e-06, + "loss": 0.809, + "num_input_tokens_seen": 45510070, + "step": 2129 + }, + { + "epoch": 0.25611735706126376, + "flos": 22819205656320.0, + "grad_norm": 2.927459927878347, + "learning_rate": 3.4873663857964407e-06, + "loss": 0.7605, + "num_input_tokens_seen": 45527285, + "step": 2130 + }, + { + "epoch": 0.2562375999519028, + "flos": 23334458800320.0, + "grad_norm": 2.0814433744023657, + "learning_rate": 3.4868455053607556e-06, + "loss": 0.6642, + "num_input_tokens_seen": 45546900, + "step": 2131 + }, + { + "epoch": 0.2563578428425419, + "flos": 22824297913440.0, + "grad_norm": 8.243151047055386, + "learning_rate": 3.486324399371789e-06, + "loss": 0.7124, + "num_input_tokens_seen": 45566200, + "step": 2132 + }, + { + "epoch": 0.25647808573318104, + "flos": 21623005080000.0, + "grad_norm": 1.9752753971129438, + "learning_rate": 3.485803067908593e-06, + "loss": 0.7847, + "num_input_tokens_seen": 45585710, + "step": 2133 + }, + { + "epoch": 0.2565983286238201, + "flos": 33682103989920.0, + "grad_norm": 3.0829485747694285, + "learning_rate": 3.485281511050253e-06, + "loss": 0.7993, + "num_input_tokens_seen": 45607845, + "step": 2134 + }, + { + "epoch": 0.2567185715144592, + "flos": 16192986948960.0, + "grad_norm": 3.200575395351299, + "learning_rate": 3.484759728875889e-06, + "loss": 0.8995, + "num_input_tokens_seen": 45623210, + "step": 2135 + }, + { + "epoch": 0.2568388144050983, + "flos": 17433722631360.0, + "grad_norm": 2.0749132391014773, + "learning_rate": 3.4842377214646543e-06, + "loss": 0.8124, + "num_input_tokens_seen": 45641505, + "step": 2136 + }, + { + "epoch": 0.25695905729573737, + "flos": 20857670825280.0, + "grad_norm": 1.9780280066841953, + "learning_rate": 3.483715488895737e-06, + "loss": 0.665, + "num_input_tokens_seen": 45661835, + "step": 2137 + }, + { + "epoch": 0.2570793001863765, + "flos": 24681888894720.0, + "grad_norm": 1.9717730367327346, + "learning_rate": 3.48319303124836e-06, + "loss": 0.7867, + "num_input_tokens_seen": 45682215, + "step": 2138 + }, + { + "epoch": 0.2571995430770156, + "flos": 26868335175360.0, + "grad_norm": 2.433577193055385, + "learning_rate": 3.4826703486017798e-06, + "loss": 0.6689, + "num_input_tokens_seen": 45701840, + "step": 2139 + }, + { + "epoch": 0.25731978596765465, + "flos": 19764373345440.0, + "grad_norm": 1.9803396931569515, + "learning_rate": 3.4821474410352867e-06, + "loss": 0.77, + "num_input_tokens_seen": 45720500, + "step": 2140 + }, + { + "epoch": 0.25744002885829376, + "flos": 70463612566560.0, + "grad_norm": 0.9306940992537791, + "learning_rate": 3.481624308628205e-06, + "loss": 0.6262, + "num_input_tokens_seen": 45781870, + "step": 2141 + }, + { + "epoch": 0.25756027174893287, + "flos": 18012436022880.0, + "grad_norm": 3.496661637711401, + "learning_rate": 3.481100951459893e-06, + "loss": 1.0082, + "num_input_tokens_seen": 45794890, + "step": 2142 + }, + { + "epoch": 0.2576805146395719, + "flos": 22641654214560.0, + "grad_norm": 1.8914738790718686, + "learning_rate": 3.4805773696097453e-06, + "loss": 0.7855, + "num_input_tokens_seen": 45815780, + "step": 2143 + }, + { + "epoch": 0.25780075753021103, + "flos": 16448866542240.0, + "grad_norm": 2.882995103464722, + "learning_rate": 3.4800535631571874e-06, + "loss": 0.8798, + "num_input_tokens_seen": 45831310, + "step": 2144 + }, + { + "epoch": 0.25792100042085014, + "flos": 22787679966240.0, + "grad_norm": 3.7999571546849586, + "learning_rate": 3.4795295321816804e-06, + "loss": 0.7646, + "num_input_tokens_seen": 45850135, + "step": 2145 + }, + { + "epoch": 0.2580412433114892, + "flos": 18670666998240.0, + "grad_norm": 3.694463329856489, + "learning_rate": 3.47900527676272e-06, + "loss": 0.912, + "num_input_tokens_seen": 45868470, + "step": 2146 + }, + { + "epoch": 0.2581614862021283, + "flos": 14262977808000.0, + "grad_norm": 2.334140354195676, + "learning_rate": 3.478480796979835e-06, + "loss": 0.8831, + "num_input_tokens_seen": 45886355, + "step": 2147 + }, + { + "epoch": 0.25828172909276736, + "flos": 29455980575520.0, + "grad_norm": 2.359694200358971, + "learning_rate": 3.4779560929125894e-06, + "loss": 0.7768, + "num_input_tokens_seen": 45907900, + "step": 2148 + }, + { + "epoch": 0.2584019719834065, + "flos": 67018620554880.0, + "grad_norm": 0.6919228514997875, + "learning_rate": 3.4774311646405783e-06, + "loss": 0.569, + "num_input_tokens_seen": 45977805, + "step": 2149 + }, + { + "epoch": 0.2585222148740456, + "flos": 22860841521120.0, + "grad_norm": 2.191021292154264, + "learning_rate": 3.476906012243435e-06, + "loss": 0.8403, + "num_input_tokens_seen": 45996715, + "step": 2150 + }, + { + "epoch": 0.25864245776468464, + "flos": 28867565876640.0, + "grad_norm": 2.5614136044409417, + "learning_rate": 3.476380635800824e-06, + "loss": 0.8135, + "num_input_tokens_seen": 46017915, + "step": 2151 + }, + { + "epoch": 0.25876270065532375, + "flos": 14772246620640.0, + "grad_norm": 2.2680959972672725, + "learning_rate": 3.475855035392444e-06, + "loss": 0.8618, + "num_input_tokens_seen": 46033235, + "step": 2152 + }, + { + "epoch": 0.25888294354596286, + "flos": 60380178730080.0, + "grad_norm": 2.041497891774655, + "learning_rate": 3.475329211098029e-06, + "loss": 0.7161, + "num_input_tokens_seen": 46056390, + "step": 2153 + }, + { + "epoch": 0.2590031864366019, + "flos": 27811072192800.0, + "grad_norm": 5.002742968620615, + "learning_rate": 3.4748031629973453e-06, + "loss": 0.8234, + "num_input_tokens_seen": 46077120, + "step": 2154 + }, + { + "epoch": 0.25912342932724103, + "flos": 62334045123360.0, + "grad_norm": 0.9278512981419076, + "learning_rate": 3.4742768911701944e-06, + "loss": 0.5656, + "num_input_tokens_seen": 46139815, + "step": 2155 + }, + { + "epoch": 0.25924367221788014, + "flos": 12361186248480.0, + "grad_norm": 5.133821106727794, + "learning_rate": 3.4737503956964113e-06, + "loss": 0.7061, + "num_input_tokens_seen": 46152120, + "step": 2156 + }, + { + "epoch": 0.2593639151085192, + "flos": 14554620444000.0, + "grad_norm": 2.292595230171322, + "learning_rate": 3.473223676655865e-06, + "loss": 0.6695, + "num_input_tokens_seen": 46170120, + "step": 2157 + }, + { + "epoch": 0.2594841579991583, + "flos": 15209357461920.0, + "grad_norm": 2.508488644156586, + "learning_rate": 3.472696734128459e-06, + "loss": 0.7964, + "num_input_tokens_seen": 46187985, + "step": 2158 + }, + { + "epoch": 0.2596044008897974, + "flos": 23589818016960.0, + "grad_norm": 2.246630264557072, + "learning_rate": 3.4721695681941286e-06, + "loss": 0.7578, + "num_input_tokens_seen": 46207025, + "step": 2159 + }, + { + "epoch": 0.25972464378043647, + "flos": 13752854090880.0, + "grad_norm": 2.6740506778271596, + "learning_rate": 3.471642178932845e-06, + "loss": 0.8258, + "num_input_tokens_seen": 46225670, + "step": 2160 + }, + { + "epoch": 0.2598448866710756, + "flos": 19545706415520.0, + "grad_norm": 4.02797122066941, + "learning_rate": 3.471114566424613e-06, + "loss": 0.9001, + "num_input_tokens_seen": 46244130, + "step": 2161 + }, + { + "epoch": 0.25996512956171464, + "flos": 21622336024320.0, + "grad_norm": 2.0286122728607046, + "learning_rate": 3.4705867307494715e-06, + "loss": 0.7536, + "num_input_tokens_seen": 46263200, + "step": 2162 + }, + { + "epoch": 0.26008537245235375, + "flos": 18197570095680.0, + "grad_norm": 4.98033662508569, + "learning_rate": 3.470058671987492e-06, + "loss": 0.8472, + "num_input_tokens_seen": 46281465, + "step": 2163 + }, + { + "epoch": 0.26020561534299286, + "flos": 24610139790720.0, + "grad_norm": 3.2765368808514257, + "learning_rate": 3.4695303902187805e-06, + "loss": 0.8467, + "num_input_tokens_seen": 46301020, + "step": 2164 + }, + { + "epoch": 0.2603258582336319, + "flos": 25736672769600.0, + "grad_norm": 1.9892270100324465, + "learning_rate": 3.469001885523478e-06, + "loss": 0.7844, + "num_input_tokens_seen": 46321740, + "step": 2165 + }, + { + "epoch": 0.260446101124271, + "flos": 28724996912640.0, + "grad_norm": 1.75187296774369, + "learning_rate": 3.4684731579817568e-06, + "loss": 0.8119, + "num_input_tokens_seen": 46342730, + "step": 2166 + }, + { + "epoch": 0.26056634401491013, + "flos": 25630015527360.0, + "grad_norm": 1.8136732316284232, + "learning_rate": 3.4679442076738247e-06, + "loss": 0.768, + "num_input_tokens_seen": 46362445, + "step": 2167 + }, + { + "epoch": 0.2606865869055492, + "flos": 27013989229440.0, + "grad_norm": 2.270496891774812, + "learning_rate": 3.4674150346799245e-06, + "loss": 0.8327, + "num_input_tokens_seen": 46382145, + "step": 2168 + }, + { + "epoch": 0.2608068297961883, + "flos": 17687446378560.0, + "grad_norm": 2.402342559069827, + "learning_rate": 3.4668856390803295e-06, + "loss": 0.8011, + "num_input_tokens_seen": 46400025, + "step": 2169 + }, + { + "epoch": 0.2609270726868274, + "flos": 18525235962720.0, + "grad_norm": 4.439296519936345, + "learning_rate": 3.4663560209553495e-06, + "loss": 0.8998, + "num_input_tokens_seen": 46418090, + "step": 2170 + }, + { + "epoch": 0.26104731557746647, + "flos": 21804533686080.0, + "grad_norm": 2.236833087265213, + "learning_rate": 3.4658261803853267e-06, + "loss": 0.7907, + "num_input_tokens_seen": 46436015, + "step": 2171 + }, + { + "epoch": 0.2611675584681056, + "flos": 21658693783200.0, + "grad_norm": 2.259845597048072, + "learning_rate": 3.4652961174506383e-06, + "loss": 0.809, + "num_input_tokens_seen": 46455885, + "step": 2172 + }, + { + "epoch": 0.2612878013587447, + "flos": 71759588245920.0, + "grad_norm": 0.9599584849142913, + "learning_rate": 3.464765832231694e-06, + "loss": 0.5798, + "num_input_tokens_seen": 46510050, + "step": 2173 + }, + { + "epoch": 0.26140804424938374, + "flos": 20198584945440.0, + "grad_norm": 1.921954951910544, + "learning_rate": 3.4642353248089373e-06, + "loss": 0.7014, + "num_input_tokens_seen": 46528810, + "step": 2174 + }, + { + "epoch": 0.26152828714002285, + "flos": 25520496213600.0, + "grad_norm": 2.1129438938359284, + "learning_rate": 3.463704595262846e-06, + "loss": 0.8019, + "num_input_tokens_seen": 46549690, + "step": 2175 + }, + { + "epoch": 0.26164853003066196, + "flos": 25410047655840.0, + "grad_norm": 3.11372492942074, + "learning_rate": 3.463173643673931e-06, + "loss": 0.7055, + "num_input_tokens_seen": 46572215, + "step": 2176 + }, + { + "epoch": 0.261768772921301, + "flos": 53868045483360.0, + "grad_norm": 0.9063810056509267, + "learning_rate": 3.4626424701227387e-06, + "loss": 0.635, + "num_input_tokens_seen": 46627275, + "step": 2177 + }, + { + "epoch": 0.26188901581194013, + "flos": 70586550163680.0, + "grad_norm": 0.8308876667666958, + "learning_rate": 3.4621110746898452e-06, + "loss": 0.5575, + "num_input_tokens_seen": 46695135, + "step": 2178 + }, + { + "epoch": 0.2620092587025792, + "flos": 21039533959200.0, + "grad_norm": 1.801970711718793, + "learning_rate": 3.4615794574558654e-06, + "loss": 0.7474, + "num_input_tokens_seen": 46714145, + "step": 2179 + }, + { + "epoch": 0.2621295015932183, + "flos": 18343484338080.0, + "grad_norm": 4.386821579173796, + "learning_rate": 3.4610476185014436e-06, + "loss": 0.8377, + "num_input_tokens_seen": 46731405, + "step": 2180 + }, + { + "epoch": 0.2622497444838574, + "flos": 23626175775840.0, + "grad_norm": 2.4512818817135136, + "learning_rate": 3.4605155579072597e-06, + "loss": 0.7958, + "num_input_tokens_seen": 46751260, + "step": 2181 + }, + { + "epoch": 0.26236998737449646, + "flos": 22092087648480.0, + "grad_norm": 2.5396932434715533, + "learning_rate": 3.459983275754027e-06, + "loss": 0.7146, + "num_input_tokens_seen": 46770155, + "step": 2182 + }, + { + "epoch": 0.26249023026513557, + "flos": 17869755549600.0, + "grad_norm": 3.343138451669516, + "learning_rate": 3.4594507721224918e-06, + "loss": 0.7921, + "num_input_tokens_seen": 46788805, + "step": 2183 + }, + { + "epoch": 0.2626104731557747, + "flos": 18306866390880.0, + "grad_norm": 1.832246096768869, + "learning_rate": 3.4589180470934353e-06, + "loss": 0.8188, + "num_input_tokens_seen": 46808670, + "step": 2184 + }, + { + "epoch": 0.26273071604641374, + "flos": 19290198519840.0, + "grad_norm": 2.045083346545472, + "learning_rate": 3.4583851007476713e-06, + "loss": 0.7707, + "num_input_tokens_seen": 46827140, + "step": 2185 + }, + { + "epoch": 0.26285095893705285, + "flos": 18301216587360.0, + "grad_norm": 2.308773543687534, + "learning_rate": 3.4578519331660464e-06, + "loss": 0.6845, + "num_input_tokens_seen": 46844135, + "step": 2186 + }, + { + "epoch": 0.26297120182769196, + "flos": 20165052088320.0, + "grad_norm": 2.2526427748941815, + "learning_rate": 3.4573185444294426e-06, + "loss": 0.8224, + "num_input_tokens_seen": 46862500, + "step": 2187 + }, + { + "epoch": 0.263091444718331, + "flos": 22386369337440.0, + "grad_norm": 2.058353764771077, + "learning_rate": 3.456784934618774e-06, + "loss": 0.7865, + "num_input_tokens_seen": 46883665, + "step": 2188 + }, + { + "epoch": 0.2632116876089701, + "flos": 18997700979360.0, + "grad_norm": 2.2459530993666017, + "learning_rate": 3.4562511038149897e-06, + "loss": 0.8037, + "num_input_tokens_seen": 46899160, + "step": 2189 + }, + { + "epoch": 0.26333193049960923, + "flos": 67212415181760.0, + "grad_norm": 0.8705233087459211, + "learning_rate": 3.4557170520990705e-06, + "loss": 0.5776, + "num_input_tokens_seen": 46959835, + "step": 2190 + }, + { + "epoch": 0.2634521733902483, + "flos": 25013048719200.0, + "grad_norm": 2.36803257528014, + "learning_rate": 3.4551827795520324e-06, + "loss": 0.8645, + "num_input_tokens_seen": 46982720, + "step": 2191 + }, + { + "epoch": 0.2635724162808874, + "flos": 20565619321920.0, + "grad_norm": 1.9868563059515068, + "learning_rate": 3.4546482862549226e-06, + "loss": 0.8497, + "num_input_tokens_seen": 47003035, + "step": 2192 + }, + { + "epoch": 0.2636926591715265, + "flos": 19217222813760.0, + "grad_norm": 2.3932830394925717, + "learning_rate": 3.4541135722888253e-06, + "loss": 0.7863, + "num_input_tokens_seen": 47019625, + "step": 2193 + }, + { + "epoch": 0.26381290206216557, + "flos": 28763584857120.0, + "grad_norm": 3.1325787007521853, + "learning_rate": 3.453578637734854e-06, + "loss": 0.7991, + "num_input_tokens_seen": 47040815, + "step": 2194 + }, + { + "epoch": 0.2639331449528047, + "flos": 24973085493600.0, + "grad_norm": 1.9178604748030466, + "learning_rate": 3.4530434826741605e-06, + "loss": 0.7852, + "num_input_tokens_seen": 47061755, + "step": 2195 + }, + { + "epoch": 0.26405338784344373, + "flos": 46468618265760.0, + "grad_norm": 3.115824236377442, + "learning_rate": 3.452508107187926e-06, + "loss": 0.6888, + "num_input_tokens_seen": 47085130, + "step": 2196 + }, + { + "epoch": 0.26417363073408284, + "flos": 21149276291520.0, + "grad_norm": 3.049536848079224, + "learning_rate": 3.451972511357366e-06, + "loss": 0.7745, + "num_input_tokens_seen": 47104675, + "step": 2197 + }, + { + "epoch": 0.26429387362472195, + "flos": 22642360440000.0, + "grad_norm": 2.2989457326410334, + "learning_rate": 3.45143669526373e-06, + "loss": 0.8552, + "num_input_tokens_seen": 47124435, + "step": 2198 + }, + { + "epoch": 0.264414116515361, + "flos": 67084422497280.0, + "grad_norm": 0.7946394810061429, + "learning_rate": 3.450900658988302e-06, + "loss": 0.6317, + "num_input_tokens_seen": 47185985, + "step": 2199 + }, + { + "epoch": 0.2645343594060001, + "flos": 25627896851040.0, + "grad_norm": 2.2329278274127184, + "learning_rate": 3.450364402612397e-06, + "loss": 0.7789, + "num_input_tokens_seen": 47203140, + "step": 2200 + }, + { + "epoch": 0.26465460229663923, + "flos": 22459679571360.0, + "grad_norm": 2.1270654401733355, + "learning_rate": 3.449827926217366e-06, + "loss": 0.838, + "num_input_tokens_seen": 47222575, + "step": 2201 + }, + { + "epoch": 0.2647748451872783, + "flos": 29345866545600.0, + "grad_norm": 2.1799237523174906, + "learning_rate": 3.449291229884591e-06, + "loss": 0.8029, + "num_input_tokens_seen": 47243815, + "step": 2202 + }, + { + "epoch": 0.2648950880779174, + "flos": 26759410577760.0, + "grad_norm": 1.9554875847293844, + "learning_rate": 3.4487543136954887e-06, + "loss": 0.8669, + "num_input_tokens_seen": 47263435, + "step": 2203 + }, + { + "epoch": 0.2650153309685565, + "flos": 28799645257920.0, + "grad_norm": 1.7903331259073523, + "learning_rate": 3.448217177731509e-06, + "loss": 0.9154, + "num_input_tokens_seen": 47283920, + "step": 2204 + }, + { + "epoch": 0.26513557385919556, + "flos": 20274311213760.0, + "grad_norm": 2.6105599825020707, + "learning_rate": 3.4476798220741348e-06, + "loss": 0.7813, + "num_input_tokens_seen": 47302800, + "step": 2205 + }, + { + "epoch": 0.26525581674983467, + "flos": 17651162959200.0, + "grad_norm": 1.801749939113048, + "learning_rate": 3.4471422468048826e-06, + "loss": 0.779, + "num_input_tokens_seen": 47321525, + "step": 2206 + }, + { + "epoch": 0.2653760596404738, + "flos": 26795285129760.0, + "grad_norm": 3.8006295686508342, + "learning_rate": 3.4466044520053022e-06, + "loss": 0.7265, + "num_input_tokens_seen": 47340530, + "step": 2207 + }, + { + "epoch": 0.26549630253111284, + "flos": 22749538058880.0, + "grad_norm": 2.4507129254726294, + "learning_rate": 3.446066437756977e-06, + "loss": 0.599, + "num_input_tokens_seen": 47359735, + "step": 2208 + }, + { + "epoch": 0.26561654542175195, + "flos": 23516321934240.0, + "grad_norm": 6.809920916441218, + "learning_rate": 3.4455282041415224e-06, + "loss": 0.7489, + "num_input_tokens_seen": 47378945, + "step": 2209 + }, + { + "epoch": 0.265736788312391, + "flos": 26868112156800.0, + "grad_norm": 2.3063135650062496, + "learning_rate": 3.4449897512405894e-06, + "loss": 0.8703, + "num_input_tokens_seen": 47398095, + "step": 2210 + }, + { + "epoch": 0.2658570312030301, + "flos": 23443569246720.0, + "grad_norm": 2.306073719100238, + "learning_rate": 3.444451079135859e-06, + "loss": 0.7562, + "num_input_tokens_seen": 47417605, + "step": 2211 + }, + { + "epoch": 0.2659772740936692, + "flos": 21834981453120.0, + "grad_norm": 4.266457724422243, + "learning_rate": 3.4439121879090493e-06, + "loss": 0.7437, + "num_input_tokens_seen": 47435385, + "step": 2212 + }, + { + "epoch": 0.2660975169843083, + "flos": 19764930891840.0, + "grad_norm": 2.4378848627893666, + "learning_rate": 3.4433730776419082e-06, + "loss": 0.8348, + "num_input_tokens_seen": 47454310, + "step": 2213 + }, + { + "epoch": 0.2662177598749474, + "flos": 28977196699680.0, + "grad_norm": 3.9804785731141057, + "learning_rate": 3.4428337484162183e-06, + "loss": 0.803, + "num_input_tokens_seen": 47475120, + "step": 2214 + }, + { + "epoch": 0.2663380027655865, + "flos": 21731483640480.0, + "grad_norm": 2.1073262755989717, + "learning_rate": 3.442294200313797e-06, + "loss": 0.8449, + "num_input_tokens_seen": 47493150, + "step": 2215 + }, + { + "epoch": 0.26645824565622556, + "flos": 66885195351840.0, + "grad_norm": 0.7724601658794344, + "learning_rate": 3.4417544334164916e-06, + "loss": 0.5269, + "num_input_tokens_seen": 47557295, + "step": 2216 + }, + { + "epoch": 0.26657848854686467, + "flos": 25227998673120.0, + "grad_norm": 1.9619209285633028, + "learning_rate": 3.4412144478061854e-06, + "loss": 0.7732, + "num_input_tokens_seen": 47579705, + "step": 2217 + }, + { + "epoch": 0.2666987314375038, + "flos": 23662905232320.0, + "grad_norm": 1.9966065705457314, + "learning_rate": 3.4406742435647925e-06, + "loss": 0.7546, + "num_input_tokens_seen": 47599770, + "step": 2218 + }, + { + "epoch": 0.26681897432814283, + "flos": 27010049234880.0, + "grad_norm": 2.251404535813281, + "learning_rate": 3.440133820774263e-06, + "loss": 0.7892, + "num_input_tokens_seen": 47619580, + "step": 2219 + }, + { + "epoch": 0.26693921721878194, + "flos": 28944332898240.0, + "grad_norm": 2.5736536265296786, + "learning_rate": 3.439593179516578e-06, + "loss": 0.8236, + "num_input_tokens_seen": 47639490, + "step": 2220 + }, + { + "epoch": 0.26705946010942105, + "flos": 21477165177120.0, + "grad_norm": 2.0687903224553503, + "learning_rate": 3.4390523198737524e-06, + "loss": 0.8109, + "num_input_tokens_seen": 47658650, + "step": 2221 + }, + { + "epoch": 0.2671797030000601, + "flos": 21440658739200.0, + "grad_norm": 1.996598510226382, + "learning_rate": 3.4385112419278333e-06, + "loss": 0.7374, + "num_input_tokens_seen": 47679875, + "step": 2222 + }, + { + "epoch": 0.2672999458906992, + "flos": 64098217030560.0, + "grad_norm": 0.7951985470538806, + "learning_rate": 3.4379699457609033e-06, + "loss": 0.6472, + "num_input_tokens_seen": 47737700, + "step": 2223 + }, + { + "epoch": 0.26742018878133833, + "flos": 16885010969760.0, + "grad_norm": 2.4942540894989382, + "learning_rate": 3.4374284314550755e-06, + "loss": 0.9031, + "num_input_tokens_seen": 47756740, + "step": 2224 + }, + { + "epoch": 0.2675404316719774, + "flos": 20638074651360.0, + "grad_norm": 2.691872698887182, + "learning_rate": 3.436886699092498e-06, + "loss": 0.8089, + "num_input_tokens_seen": 47775255, + "step": 2225 + }, + { + "epoch": 0.2676606745626165, + "flos": 17459858706240.0, + "grad_norm": 2.8060880686506326, + "learning_rate": 3.4363447487553502e-06, + "loss": 0.7122, + "num_input_tokens_seen": 47788570, + "step": 2226 + }, + { + "epoch": 0.26778091745325555, + "flos": 27812633322720.0, + "grad_norm": 2.6846672920369676, + "learning_rate": 3.4358025805258455e-06, + "loss": 0.7807, + "num_input_tokens_seen": 47808275, + "step": 2227 + }, + { + "epoch": 0.26790116034389466, + "flos": 20925963141600.0, + "grad_norm": 2.6971096957779768, + "learning_rate": 3.435260194486232e-06, + "loss": 0.8339, + "num_input_tokens_seen": 47826245, + "step": 2228 + }, + { + "epoch": 0.2680214032345338, + "flos": 18015000736320.0, + "grad_norm": 2.7662923406436652, + "learning_rate": 3.4347175907187875e-06, + "loss": 0.8218, + "num_input_tokens_seen": 47843235, + "step": 2229 + }, + { + "epoch": 0.26814164612517283, + "flos": 22387410090720.0, + "grad_norm": 2.2881795071963476, + "learning_rate": 3.4341747693058254e-06, + "loss": 0.88, + "num_input_tokens_seen": 47861710, + "step": 2230 + }, + { + "epoch": 0.26826188901581194, + "flos": 35576127069600.0, + "grad_norm": 3.261588636532811, + "learning_rate": 3.4336317303296916e-06, + "loss": 0.7788, + "num_input_tokens_seen": 47882685, + "step": 2231 + }, + { + "epoch": 0.26838213190645105, + "flos": 17614730860800.0, + "grad_norm": 2.1631339935406753, + "learning_rate": 3.4330884738727635e-06, + "loss": 0.7549, + "num_input_tokens_seen": 47900860, + "step": 2232 + }, + { + "epoch": 0.2685023747970901, + "flos": 22642434779520.0, + "grad_norm": 2.025690540264478, + "learning_rate": 3.4325450000174535e-06, + "loss": 0.7063, + "num_input_tokens_seen": 47917500, + "step": 2233 + }, + { + "epoch": 0.2686226176877292, + "flos": 20091853363680.0, + "grad_norm": 1.9735939139072451, + "learning_rate": 3.4320013088462067e-06, + "loss": 0.7426, + "num_input_tokens_seen": 47934860, + "step": 2234 + }, + { + "epoch": 0.2687428605783683, + "flos": 21840817105440.0, + "grad_norm": 2.102138511253625, + "learning_rate": 3.431457400441499e-06, + "loss": 0.8169, + "num_input_tokens_seen": 47955255, + "step": 2235 + }, + { + "epoch": 0.2688631034690074, + "flos": 69842817273120.0, + "grad_norm": 0.914719989070592, + "learning_rate": 3.4309132748858424e-06, + "loss": 0.6074, + "num_input_tokens_seen": 48016245, + "step": 2236 + }, + { + "epoch": 0.2689833463596465, + "flos": 22823963385600.0, + "grad_norm": 1.8279690798266715, + "learning_rate": 3.430368932261779e-06, + "loss": 0.837, + "num_input_tokens_seen": 48036600, + "step": 2237 + }, + { + "epoch": 0.2691035892502856, + "flos": 17175761531520.0, + "grad_norm": 2.205719090198138, + "learning_rate": 3.429824372651886e-06, + "loss": 0.7504, + "num_input_tokens_seen": 48054750, + "step": 2238 + }, + { + "epoch": 0.26922383214092466, + "flos": 17722205837760.0, + "grad_norm": 11.675333636660827, + "learning_rate": 3.4292795961387732e-06, + "loss": 0.8298, + "num_input_tokens_seen": 48072650, + "step": 2239 + }, + { + "epoch": 0.26934407503156377, + "flos": 16150198821600.0, + "grad_norm": 3.4897004216985623, + "learning_rate": 3.4287346028050818e-06, + "loss": 0.8775, + "num_input_tokens_seen": 48088720, + "step": 2240 + }, + { + "epoch": 0.2694643179222028, + "flos": 23698965633120.0, + "grad_norm": 1.6999381110201712, + "learning_rate": 3.4281893927334866e-06, + "loss": 0.7988, + "num_input_tokens_seen": 48108630, + "step": 2241 + }, + { + "epoch": 0.26958456081284193, + "flos": 24682780968960.0, + "grad_norm": 2.0947440904715378, + "learning_rate": 3.4276439660066963e-06, + "loss": 0.753, + "num_input_tokens_seen": 48128330, + "step": 2242 + }, + { + "epoch": 0.26970480370348104, + "flos": 18086787010080.0, + "grad_norm": 2.6549951618674927, + "learning_rate": 3.427098322707452e-06, + "loss": 0.8515, + "num_input_tokens_seen": 48144295, + "step": 2243 + }, + { + "epoch": 0.2698250465941201, + "flos": 10801408083360.0, + "grad_norm": 2.8230632354486063, + "learning_rate": 3.426552462918526e-06, + "loss": 0.8952, + "num_input_tokens_seen": 48162230, + "step": 2244 + }, + { + "epoch": 0.2699452894847592, + "flos": 17283422357280.0, + "grad_norm": 2.89048164630811, + "learning_rate": 3.426006386722726e-06, + "loss": 0.7318, + "num_input_tokens_seen": 48180290, + "step": 2245 + }, + { + "epoch": 0.2700655323753983, + "flos": 18052473588000.0, + "grad_norm": 5.174706919751327, + "learning_rate": 3.4254600942028914e-06, + "loss": 0.9233, + "num_input_tokens_seen": 48198165, + "step": 2246 + }, + { + "epoch": 0.2701857752660374, + "flos": 18160766299680.0, + "grad_norm": 2.7713081029600413, + "learning_rate": 3.424913585441893e-06, + "loss": 0.8265, + "num_input_tokens_seen": 48216000, + "step": 2247 + }, + { + "epoch": 0.2703060181566765, + "flos": 16296113064000.0, + "grad_norm": 2.2980831368383523, + "learning_rate": 3.4243668605226374e-06, + "loss": 0.8717, + "num_input_tokens_seen": 48233025, + "step": 2248 + }, + { + "epoch": 0.2704262610473156, + "flos": 19544442643680.0, + "grad_norm": 3.4306286191860407, + "learning_rate": 3.423819919528061e-06, + "loss": 0.8321, + "num_input_tokens_seen": 48251110, + "step": 2249 + }, + { + "epoch": 0.27054650393795465, + "flos": 20711161866720.0, + "grad_norm": 1.8480591284346266, + "learning_rate": 3.4232727625411355e-06, + "loss": 0.7812, + "num_input_tokens_seen": 48270215, + "step": 2250 + }, + { + "epoch": 0.27066674682859376, + "flos": 18631670186400.0, + "grad_norm": 1.9657807135348144, + "learning_rate": 3.4227253896448626e-06, + "loss": 0.8657, + "num_input_tokens_seen": 48288075, + "step": 2251 + }, + { + "epoch": 0.2707869897192329, + "flos": 22969877628000.0, + "grad_norm": 2.631403786702428, + "learning_rate": 3.42217780092228e-06, + "loss": 0.8206, + "num_input_tokens_seen": 48306415, + "step": 2252 + }, + { + "epoch": 0.27090723260987193, + "flos": 58240343328480.0, + "grad_norm": 0.8083886148944005, + "learning_rate": 3.421629996456456e-06, + "loss": 0.6032, + "num_input_tokens_seen": 48365035, + "step": 2253 + }, + { + "epoch": 0.27102747550051104, + "flos": 11967718439040.0, + "grad_norm": 2.063487654519407, + "learning_rate": 3.421081976330491e-06, + "loss": 0.8224, + "num_input_tokens_seen": 48383550, + "step": 2254 + }, + { + "epoch": 0.27114771839115015, + "flos": 19871997001440.0, + "grad_norm": 1.9500278788798926, + "learning_rate": 3.4205337406275207e-06, + "loss": 0.8787, + "num_input_tokens_seen": 48403670, + "step": 2255 + }, + { + "epoch": 0.2712679612817892, + "flos": 18305862807360.0, + "grad_norm": 3.4209677133561485, + "learning_rate": 3.4199852894307114e-06, + "loss": 0.7491, + "num_input_tokens_seen": 48420420, + "step": 2256 + }, + { + "epoch": 0.2713882041724283, + "flos": 24425228736480.0, + "grad_norm": 2.5889900376948543, + "learning_rate": 3.419436622823262e-06, + "loss": 0.7913, + "num_input_tokens_seen": 48441140, + "step": 2257 + }, + { + "epoch": 0.27150844706306737, + "flos": 23006384065920.0, + "grad_norm": 1.974633161060063, + "learning_rate": 3.4188877408884063e-06, + "loss": 0.7431, + "num_input_tokens_seen": 48461845, + "step": 2258 + }, + { + "epoch": 0.2716286899537065, + "flos": 22531354335840.0, + "grad_norm": 3.1116824382374797, + "learning_rate": 3.4183386437094088e-06, + "loss": 0.6475, + "num_input_tokens_seen": 48478510, + "step": 2259 + }, + { + "epoch": 0.2717489328443456, + "flos": 13096555943040.0, + "grad_norm": 2.8311109428698487, + "learning_rate": 3.417789331369565e-06, + "loss": 0.8225, + "num_input_tokens_seen": 48494300, + "step": 2260 + }, + { + "epoch": 0.27186917573498465, + "flos": 29236681759680.0, + "grad_norm": 2.2691343291975623, + "learning_rate": 3.4172398039522088e-06, + "loss": 0.9143, + "num_input_tokens_seen": 48515585, + "step": 2261 + }, + { + "epoch": 0.27198941862562376, + "flos": 25994968397280.0, + "grad_norm": 1.896843951755173, + "learning_rate": 3.4166900615407e-06, + "loss": 0.7962, + "num_input_tokens_seen": 48533500, + "step": 2262 + }, + { + "epoch": 0.27210966151626287, + "flos": 32733865848000.0, + "grad_norm": 2.0291256873701204, + "learning_rate": 3.416140104218436e-06, + "loss": 0.7469, + "num_input_tokens_seen": 48552865, + "step": 2263 + }, + { + "epoch": 0.2722299044069019, + "flos": 65377614996960.0, + "grad_norm": 0.8441141896696018, + "learning_rate": 3.4155899320688437e-06, + "loss": 0.6962, + "num_input_tokens_seen": 48618940, + "step": 2264 + }, + { + "epoch": 0.27235014729754103, + "flos": 15312186218880.0, + "grad_norm": 5.02098486369605, + "learning_rate": 3.415039545175384e-06, + "loss": 0.7456, + "num_input_tokens_seen": 48634665, + "step": 2265 + }, + { + "epoch": 0.27247039018818014, + "flos": 21841151633280.0, + "grad_norm": 3.4337873350043293, + "learning_rate": 3.414488943621551e-06, + "loss": 0.6489, + "num_input_tokens_seen": 48653330, + "step": 2266 + }, + { + "epoch": 0.2725906330788192, + "flos": 18668808510240.0, + "grad_norm": 2.8499003543499764, + "learning_rate": 3.41393812749087e-06, + "loss": 0.7396, + "num_input_tokens_seen": 48671375, + "step": 2267 + }, + { + "epoch": 0.2727108759694583, + "flos": 17860351600320.0, + "grad_norm": 2.8919292085248456, + "learning_rate": 3.4133870968668984e-06, + "loss": 0.7226, + "num_input_tokens_seen": 48686175, + "step": 2268 + }, + { + "epoch": 0.2728311188600974, + "flos": 24426641187360.0, + "grad_norm": 3.084669562500083, + "learning_rate": 3.412835851833229e-06, + "loss": 0.7838, + "num_input_tokens_seen": 48708370, + "step": 2269 + }, + { + "epoch": 0.2729513617507365, + "flos": 30948804535680.0, + "grad_norm": 2.0458577664140605, + "learning_rate": 3.4122843924734834e-06, + "loss": 0.7782, + "num_input_tokens_seen": 48730670, + "step": 2270 + }, + { + "epoch": 0.2730716046413756, + "flos": 19067294237280.0, + "grad_norm": 2.67300426867851, + "learning_rate": 3.411732718871319e-06, + "loss": 0.8808, + "num_input_tokens_seen": 48746510, + "step": 2271 + }, + { + "epoch": 0.27319184753201464, + "flos": 26905993875840.0, + "grad_norm": 1.7552024622365257, + "learning_rate": 3.4111808311104227e-06, + "loss": 0.7864, + "num_input_tokens_seen": 48768825, + "step": 2272 + }, + { + "epoch": 0.27331209042265375, + "flos": 31714213129920.0, + "grad_norm": 2.56022422032116, + "learning_rate": 3.410628729274517e-06, + "loss": 0.6939, + "num_input_tokens_seen": 48790345, + "step": 2273 + }, + { + "epoch": 0.27343233331329286, + "flos": 25702210668480.0, + "grad_norm": 2.7825357272724034, + "learning_rate": 3.4100764134473546e-06, + "loss": 0.8278, + "num_input_tokens_seen": 48809630, + "step": 2274 + }, + { + "epoch": 0.2735525762039319, + "flos": 24354520385760.0, + "grad_norm": 2.770522441966574, + "learning_rate": 3.4095238837127215e-06, + "loss": 0.8477, + "num_input_tokens_seen": 48828770, + "step": 2275 + }, + { + "epoch": 0.27367281909457103, + "flos": 14444915281440.0, + "grad_norm": 2.1367598116484174, + "learning_rate": 3.4089711401544355e-06, + "loss": 0.7945, + "num_input_tokens_seen": 48846365, + "step": 2276 + }, + { + "epoch": 0.27379306198521014, + "flos": 23443420567680.0, + "grad_norm": 2.567923783760745, + "learning_rate": 3.4084181828563486e-06, + "loss": 0.6752, + "num_input_tokens_seen": 48863085, + "step": 2277 + }, + { + "epoch": 0.2739133048758492, + "flos": 17433388103520.0, + "grad_norm": 1.7748569404010484, + "learning_rate": 3.4078650119023428e-06, + "loss": 0.7055, + "num_input_tokens_seen": 48881400, + "step": 2278 + }, + { + "epoch": 0.2740335477664883, + "flos": 19245366055680.0, + "grad_norm": 2.4895729054082296, + "learning_rate": 3.4073116273763337e-06, + "loss": 0.7458, + "num_input_tokens_seen": 48895845, + "step": 2279 + }, + { + "epoch": 0.2741537906571274, + "flos": 26067349387200.0, + "grad_norm": 1.9835958678394625, + "learning_rate": 3.40675802936227e-06, + "loss": 0.8105, + "num_input_tokens_seen": 48916630, + "step": 2280 + }, + { + "epoch": 0.27427403354776647, + "flos": 34114940308800.0, + "grad_norm": 2.1397312275388085, + "learning_rate": 3.4062042179441318e-06, + "loss": 0.7201, + "num_input_tokens_seen": 48937100, + "step": 2281 + }, + { + "epoch": 0.2743942764384056, + "flos": 18740148746880.0, + "grad_norm": 1.8999961451441159, + "learning_rate": 3.4056501932059314e-06, + "loss": 0.8081, + "num_input_tokens_seen": 48955175, + "step": 2282 + }, + { + "epoch": 0.2745145193290447, + "flos": 64810727322720.0, + "grad_norm": 0.7759981633906899, + "learning_rate": 3.405095955231715e-06, + "loss": 0.581, + "num_input_tokens_seen": 49006830, + "step": 2283 + }, + { + "epoch": 0.27463476221968375, + "flos": 16113803892960.0, + "grad_norm": 3.457678782708436, + "learning_rate": 3.4045415041055585e-06, + "loss": 0.9437, + "num_input_tokens_seen": 49022470, + "step": 2284 + }, + { + "epoch": 0.27475500511032286, + "flos": 10362141396000.0, + "grad_norm": 2.9260873061509587, + "learning_rate": 3.4039868399115728e-06, + "loss": 0.7851, + "num_input_tokens_seen": 49037110, + "step": 2285 + }, + { + "epoch": 0.27487524800096197, + "flos": 17286470277600.0, + "grad_norm": 2.869200122234534, + "learning_rate": 3.4034319627339003e-06, + "loss": 0.8017, + "num_input_tokens_seen": 49053975, + "step": 2286 + }, + { + "epoch": 0.274995490891601, + "flos": 27081129283200.0, + "grad_norm": 7.276059375724095, + "learning_rate": 3.402876872656715e-06, + "loss": 0.6974, + "num_input_tokens_seen": 49072295, + "step": 2287 + }, + { + "epoch": 0.27511573378224013, + "flos": 23402416588800.0, + "grad_norm": 2.220265475371792, + "learning_rate": 3.402321569764223e-06, + "loss": 0.9, + "num_input_tokens_seen": 49091600, + "step": 2288 + }, + { + "epoch": 0.2752359766728792, + "flos": 16698092748480.0, + "grad_norm": 1.9379236311928532, + "learning_rate": 3.4017660541406635e-06, + "loss": 0.8341, + "num_input_tokens_seen": 49107665, + "step": 2289 + }, + { + "epoch": 0.2753562195635183, + "flos": 25261382851200.0, + "grad_norm": 2.3546294936114807, + "learning_rate": 3.4012103258703092e-06, + "loss": 0.7396, + "num_input_tokens_seen": 49126420, + "step": 2290 + }, + { + "epoch": 0.2754764624541574, + "flos": 27299573194560.0, + "grad_norm": 1.9997683629364704, + "learning_rate": 3.4006543850374616e-06, + "loss": 0.8288, + "num_input_tokens_seen": 49141470, + "step": 2291 + }, + { + "epoch": 0.27559670534479647, + "flos": 17213903438880.0, + "grad_norm": 2.3451011323573843, + "learning_rate": 3.400098231726458e-06, + "loss": 0.7521, + "num_input_tokens_seen": 49158570, + "step": 2292 + }, + { + "epoch": 0.2757169482354356, + "flos": 21907548291840.0, + "grad_norm": 2.440862830146508, + "learning_rate": 3.3995418660216657e-06, + "loss": 0.8704, + "num_input_tokens_seen": 49176025, + "step": 2293 + }, + { + "epoch": 0.2758371911260747, + "flos": 20820941368800.0, + "grad_norm": 2.700052797719636, + "learning_rate": 3.3989852880074848e-06, + "loss": 0.8031, + "num_input_tokens_seen": 49197455, + "step": 2294 + }, + { + "epoch": 0.27595743401671374, + "flos": 69170233036800.0, + "grad_norm": 0.7477368541545302, + "learning_rate": 3.398428497768348e-06, + "loss": 0.6053, + "num_input_tokens_seen": 49262625, + "step": 2295 + }, + { + "epoch": 0.27607767690735285, + "flos": 21184741976160.0, + "grad_norm": 1.9537335477735445, + "learning_rate": 3.3978714953887205e-06, + "loss": 0.7241, + "num_input_tokens_seen": 49282380, + "step": 2296 + }, + { + "epoch": 0.27619791979799196, + "flos": 24789735569280.0, + "grad_norm": 1.9200061923165601, + "learning_rate": 3.397314280953098e-06, + "loss": 0.8607, + "num_input_tokens_seen": 49303205, + "step": 2297 + }, + { + "epoch": 0.276318162688631, + "flos": 24718097974560.0, + "grad_norm": 2.399425787580733, + "learning_rate": 3.3967568545460108e-06, + "loss": 0.8011, + "num_input_tokens_seen": 49322745, + "step": 2298 + }, + { + "epoch": 0.27643840557927013, + "flos": 18124185522240.0, + "grad_norm": 2.593769670894441, + "learning_rate": 3.3961992162520185e-06, + "loss": 0.8097, + "num_input_tokens_seen": 49341370, + "step": 2299 + }, + { + "epoch": 0.27655864846990924, + "flos": 24788285948640.0, + "grad_norm": 2.072385966036279, + "learning_rate": 3.3956413661557156e-06, + "loss": 0.7154, + "num_input_tokens_seen": 49361545, + "step": 2300 + }, + { + "epoch": 0.2766788913605483, + "flos": 20237693266560.0, + "grad_norm": 2.6220008428879233, + "learning_rate": 3.3950833043417273e-06, + "loss": 0.6671, + "num_input_tokens_seen": 49379690, + "step": 2301 + }, + { + "epoch": 0.2767991342511874, + "flos": 21439617985920.0, + "grad_norm": 2.61405652781441, + "learning_rate": 3.3945250308947105e-06, + "loss": 0.7336, + "num_input_tokens_seen": 49395995, + "step": 2302 + }, + { + "epoch": 0.2769193771418265, + "flos": 65907853287840.0, + "grad_norm": 1.3018220970037782, + "learning_rate": 3.3939665458993556e-06, + "loss": 0.6828, + "num_input_tokens_seen": 49450415, + "step": 2303 + }, + { + "epoch": 0.27703962003246557, + "flos": 20674915617120.0, + "grad_norm": 2.5130035768749863, + "learning_rate": 3.3934078494403843e-06, + "loss": 0.7693, + "num_input_tokens_seen": 49469990, + "step": 2304 + }, + { + "epoch": 0.2771598629231047, + "flos": 22897124940480.0, + "grad_norm": 2.3542509710920148, + "learning_rate": 3.3928489416025495e-06, + "loss": 0.8098, + "num_input_tokens_seen": 49490835, + "step": 2305 + }, + { + "epoch": 0.27728010581374374, + "flos": 18343335659040.0, + "grad_norm": 3.1663261698985736, + "learning_rate": 3.392289822470638e-06, + "loss": 0.7893, + "num_input_tokens_seen": 49508815, + "step": 2306 + }, + { + "epoch": 0.27740034870438285, + "flos": 19399978021920.0, + "grad_norm": 2.3497696946292104, + "learning_rate": 3.3917304921294674e-06, + "loss": 0.7555, + "num_input_tokens_seen": 49526980, + "step": 2307 + }, + { + "epoch": 0.27752059159502196, + "flos": 21583859589120.0, + "grad_norm": 3.1747684904984363, + "learning_rate": 3.3911709506638876e-06, + "loss": 0.8049, + "num_input_tokens_seen": 49546290, + "step": 2308 + }, + { + "epoch": 0.277640834485661, + "flos": 26569630284960.0, + "grad_norm": 2.2854164431863664, + "learning_rate": 3.390611198158781e-06, + "loss": 0.8118, + "num_input_tokens_seen": 49564165, + "step": 2309 + }, + { + "epoch": 0.2777610773763001, + "flos": 19464776380800.0, + "grad_norm": 2.355536768296533, + "learning_rate": 3.3900512346990612e-06, + "loss": 0.9023, + "num_input_tokens_seen": 49580155, + "step": 2310 + }, + { + "epoch": 0.27788132026693924, + "flos": 38235521573760.0, + "grad_norm": 1.9200652991555367, + "learning_rate": 3.389491060369674e-06, + "loss": 0.6585, + "num_input_tokens_seen": 49605830, + "step": 2311 + }, + { + "epoch": 0.2780015631575783, + "flos": 22350569124960.0, + "grad_norm": 2.069124161458714, + "learning_rate": 3.388930675255598e-06, + "loss": 0.8916, + "num_input_tokens_seen": 49625680, + "step": 2312 + }, + { + "epoch": 0.2781218060482174, + "flos": 12186050841120.0, + "grad_norm": 3.7631526444349004, + "learning_rate": 3.388370079441843e-06, + "loss": 0.7929, + "num_input_tokens_seen": 49641195, + "step": 2313 + }, + { + "epoch": 0.2782420489388565, + "flos": 18081657583200.0, + "grad_norm": 2.115775224995287, + "learning_rate": 3.3878092730134505e-06, + "loss": 0.9272, + "num_input_tokens_seen": 49659180, + "step": 2314 + }, + { + "epoch": 0.27836229182949557, + "flos": 18488394996960.0, + "grad_norm": 1.894369846036187, + "learning_rate": 3.3872482560554947e-06, + "loss": 0.8091, + "num_input_tokens_seen": 49677955, + "step": 2315 + }, + { + "epoch": 0.2784825347201347, + "flos": 66983563737600.0, + "grad_norm": 0.8044948532456707, + "learning_rate": 3.386687028653082e-06, + "loss": 0.5689, + "num_input_tokens_seen": 49740800, + "step": 2316 + }, + { + "epoch": 0.2786027776107738, + "flos": 22598531559360.0, + "grad_norm": 1.972296774564002, + "learning_rate": 3.386125590891349e-06, + "loss": 0.8541, + "num_input_tokens_seen": 49757675, + "step": 2317 + }, + { + "epoch": 0.27872302050141284, + "flos": 15756470823840.0, + "grad_norm": 2.4252960653575206, + "learning_rate": 3.3855639428554657e-06, + "loss": 0.8369, + "num_input_tokens_seen": 49775205, + "step": 2318 + }, + { + "epoch": 0.27884326339205195, + "flos": 22095358587360.0, + "grad_norm": 2.2740978841792656, + "learning_rate": 3.385002084630635e-06, + "loss": 0.8012, + "num_input_tokens_seen": 49794175, + "step": 2319 + }, + { + "epoch": 0.278963506282691, + "flos": 20529410242080.0, + "grad_norm": 5.583254040207219, + "learning_rate": 3.384440016302088e-06, + "loss": 0.848, + "num_input_tokens_seen": 49812850, + "step": 2320 + }, + { + "epoch": 0.2790837491733301, + "flos": 21911525456160.0, + "grad_norm": 5.358221969910635, + "learning_rate": 3.3838777379550923e-06, + "loss": 0.6272, + "num_input_tokens_seen": 49832295, + "step": 2321 + }, + { + "epoch": 0.27920399206396923, + "flos": 26249138181600.0, + "grad_norm": 2.190238546330409, + "learning_rate": 3.383315249674944e-06, + "loss": 0.7853, + "num_input_tokens_seen": 49850860, + "step": 2322 + }, + { + "epoch": 0.2793242349546083, + "flos": 25364248777920.0, + "grad_norm": 2.2630805098051336, + "learning_rate": 3.3827525515469715e-06, + "loss": 0.8588, + "num_input_tokens_seen": 49865765, + "step": 2323 + }, + { + "epoch": 0.2794444778452474, + "flos": 20820606840960.0, + "grad_norm": 3.125441710726021, + "learning_rate": 3.3821896436565367e-06, + "loss": 0.7084, + "num_input_tokens_seen": 49885425, + "step": 2324 + }, + { + "epoch": 0.2795647207358865, + "flos": 21545308814400.0, + "grad_norm": 4.3233051661679465, + "learning_rate": 3.381626526089032e-06, + "loss": 0.7038, + "num_input_tokens_seen": 49904990, + "step": 2325 + }, + { + "epoch": 0.27968496362652556, + "flos": 21441141946080.0, + "grad_norm": 2.1900349778817922, + "learning_rate": 3.3810631989298815e-06, + "loss": 0.7919, + "num_input_tokens_seen": 49924600, + "step": 2326 + }, + { + "epoch": 0.2798052065171647, + "flos": 23225459863200.0, + "grad_norm": 2.329997289107551, + "learning_rate": 3.3804996622645423e-06, + "loss": 0.8419, + "num_input_tokens_seen": 49942600, + "step": 2327 + }, + { + "epoch": 0.2799254494078038, + "flos": 21508430678880.0, + "grad_norm": 2.675440500657526, + "learning_rate": 3.3799359161785015e-06, + "loss": 0.8974, + "num_input_tokens_seen": 49962250, + "step": 2328 + }, + { + "epoch": 0.28004569229844284, + "flos": 26356018442400.0, + "grad_norm": 3.6154816726855223, + "learning_rate": 3.3793719607572798e-06, + "loss": 0.8596, + "num_input_tokens_seen": 49983215, + "step": 2329 + }, + { + "epoch": 0.28016593518908195, + "flos": 33499460291040.0, + "grad_norm": 2.635268608025393, + "learning_rate": 3.378807796086428e-06, + "loss": 0.7685, + "num_input_tokens_seen": 50006075, + "step": 2330 + }, + { + "epoch": 0.28028617807972106, + "flos": 15319099794240.0, + "grad_norm": 2.9644257221873045, + "learning_rate": 3.37824342225153e-06, + "loss": 0.769, + "num_input_tokens_seen": 50022495, + "step": 2331 + }, + { + "epoch": 0.2804064209703601, + "flos": 25483729587360.0, + "grad_norm": 2.37041120430535, + "learning_rate": 3.3776788393382006e-06, + "loss": 0.7803, + "num_input_tokens_seen": 50041975, + "step": 2332 + }, + { + "epoch": 0.2805266638609992, + "flos": 29309657465760.0, + "grad_norm": 2.5208160673028326, + "learning_rate": 3.3771140474320872e-06, + "loss": 0.7646, + "num_input_tokens_seen": 50061925, + "step": 2333 + }, + { + "epoch": 0.28064690675163834, + "flos": 21432853089600.0, + "grad_norm": 2.0321788131478926, + "learning_rate": 3.3765490466188664e-06, + "loss": 0.793, + "num_input_tokens_seen": 50079325, + "step": 2334 + }, + { + "epoch": 0.2807671496422774, + "flos": 20965666178880.0, + "grad_norm": 4.581557513606948, + "learning_rate": 3.3759838369842508e-06, + "loss": 0.7417, + "num_input_tokens_seen": 50097400, + "step": 2335 + }, + { + "epoch": 0.2808873925329165, + "flos": 21475864235520.0, + "grad_norm": 2.3480413187735794, + "learning_rate": 3.375418418613981e-06, + "loss": 0.7312, + "num_input_tokens_seen": 50116425, + "step": 2336 + }, + { + "epoch": 0.28100763542355556, + "flos": 16048039120320.0, + "grad_norm": 2.563788857465463, + "learning_rate": 3.374852791593831e-06, + "loss": 0.8348, + "num_input_tokens_seen": 50132625, + "step": 2337 + }, + { + "epoch": 0.28112787831419467, + "flos": 19035471189120.0, + "grad_norm": 3.2524709844797712, + "learning_rate": 3.374286956009605e-06, + "loss": 0.5423, + "num_input_tokens_seen": 50151190, + "step": 2338 + }, + { + "epoch": 0.2812481212048338, + "flos": 12805582362720.0, + "grad_norm": 2.151921225612749, + "learning_rate": 3.3737209119471405e-06, + "loss": 0.7503, + "num_input_tokens_seen": 50167700, + "step": 2339 + }, + { + "epoch": 0.28136836409547283, + "flos": 15610928279040.0, + "grad_norm": 4.605657183777622, + "learning_rate": 3.373154659492306e-06, + "loss": 0.6401, + "num_input_tokens_seen": 50185840, + "step": 2340 + }, + { + "epoch": 0.28148860698611194, + "flos": 19905158160960.0, + "grad_norm": 1.9679124049487124, + "learning_rate": 3.3725881987310016e-06, + "loss": 0.8532, + "num_input_tokens_seen": 50203375, + "step": 2341 + }, + { + "epoch": 0.28160884987675106, + "flos": 17432421689760.0, + "grad_norm": 2.4541050599358285, + "learning_rate": 3.372021529749159e-06, + "loss": 0.8795, + "num_input_tokens_seen": 50222435, + "step": 2342 + }, + { + "epoch": 0.2817290927673901, + "flos": 16810771491840.0, + "grad_norm": 2.5459217557728153, + "learning_rate": 3.3714546526327405e-06, + "loss": 0.9269, + "num_input_tokens_seen": 50240435, + "step": 2343 + }, + { + "epoch": 0.2818493356580292, + "flos": 15392261349120.0, + "grad_norm": 2.3505424238740904, + "learning_rate": 3.3708875674677423e-06, + "loss": 0.8803, + "num_input_tokens_seen": 50258410, + "step": 2344 + }, + { + "epoch": 0.28196957854866833, + "flos": 20383347320640.0, + "grad_norm": 2.8471411303212206, + "learning_rate": 3.37032027434019e-06, + "loss": 0.8396, + "num_input_tokens_seen": 50277930, + "step": 2345 + }, + { + "epoch": 0.2820898214393074, + "flos": 19944601009920.0, + "grad_norm": 2.7325009272539775, + "learning_rate": 3.369752773336141e-06, + "loss": 0.831, + "num_input_tokens_seen": 50297410, + "step": 2346 + }, + { + "epoch": 0.2822100643299465, + "flos": 22496111669760.0, + "grad_norm": 1.8146098780597448, + "learning_rate": 3.3691850645416864e-06, + "loss": 0.7837, + "num_input_tokens_seen": 50317120, + "step": 2347 + }, + { + "epoch": 0.2823303072205856, + "flos": 11530496088480.0, + "grad_norm": 3.1632862787260905, + "learning_rate": 3.368617148042945e-06, + "loss": 0.8317, + "num_input_tokens_seen": 50334790, + "step": 2348 + }, + { + "epoch": 0.28245055011122466, + "flos": 18233927854560.0, + "grad_norm": 4.395628552239452, + "learning_rate": 3.368049023926071e-06, + "loss": 0.8462, + "num_input_tokens_seen": 50353785, + "step": 2349 + }, + { + "epoch": 0.2825707930018638, + "flos": 24572889957600.0, + "grad_norm": 3.4911852599580198, + "learning_rate": 3.3674806922772476e-06, + "loss": 0.8348, + "num_input_tokens_seen": 50374670, + "step": 2350 + }, + { + "epoch": 0.28269103589250283, + "flos": 25190823179520.0, + "grad_norm": 2.3923238066367145, + "learning_rate": 3.3669121531826904e-06, + "loss": 0.7515, + "num_input_tokens_seen": 50395650, + "step": 2351 + }, + { + "epoch": 0.28281127878314194, + "flos": 19254100949280.0, + "grad_norm": 5.308310618369441, + "learning_rate": 3.366343406728647e-06, + "loss": 0.8302, + "num_input_tokens_seen": 50415100, + "step": 2352 + }, + { + "epoch": 0.28293152167378105, + "flos": 23844247989600.0, + "grad_norm": 2.028412700637236, + "learning_rate": 3.3657744530013946e-06, + "loss": 0.6865, + "num_input_tokens_seen": 50434335, + "step": 2353 + }, + { + "epoch": 0.2830517645644201, + "flos": 43803499618560.0, + "grad_norm": 2.118619540088723, + "learning_rate": 3.3652052920872437e-06, + "loss": 0.7181, + "num_input_tokens_seen": 50457200, + "step": 2354 + }, + { + "epoch": 0.2831720074550592, + "flos": 26613570674880.0, + "grad_norm": 2.293768826025433, + "learning_rate": 3.3646359240725355e-06, + "loss": 0.8508, + "num_input_tokens_seen": 50476390, + "step": 2355 + }, + { + "epoch": 0.2832922503456983, + "flos": 30985422482880.0, + "grad_norm": 2.134455903559955, + "learning_rate": 3.364066349043643e-06, + "loss": 0.6746, + "num_input_tokens_seen": 50497595, + "step": 2356 + }, + { + "epoch": 0.2834124932363374, + "flos": 20376359405760.0, + "grad_norm": 2.2695498441108457, + "learning_rate": 3.363496567086969e-06, + "loss": 0.8204, + "num_input_tokens_seen": 50514690, + "step": 2357 + }, + { + "epoch": 0.2835327361269765, + "flos": 39328112828160.0, + "grad_norm": 3.732060054087347, + "learning_rate": 3.3629265782889506e-06, + "loss": 0.7528, + "num_input_tokens_seen": 50536515, + "step": 2358 + }, + { + "epoch": 0.2836529790176156, + "flos": 30218081061120.0, + "grad_norm": 2.2064522582974124, + "learning_rate": 3.362356382736054e-06, + "loss": 0.7214, + "num_input_tokens_seen": 50557120, + "step": 2359 + }, + { + "epoch": 0.28377322190825466, + "flos": 12659965478400.0, + "grad_norm": 3.091153374876717, + "learning_rate": 3.361785980514777e-06, + "loss": 0.9102, + "num_input_tokens_seen": 50573320, + "step": 2360 + }, + { + "epoch": 0.28389346479889377, + "flos": 18270657311040.0, + "grad_norm": 18.433166201467873, + "learning_rate": 3.361215371711649e-06, + "loss": 0.7646, + "num_input_tokens_seen": 50592415, + "step": 2361 + }, + { + "epoch": 0.2840137076895329, + "flos": 20377697517120.0, + "grad_norm": 3.2142646425555585, + "learning_rate": 3.3606445564132326e-06, + "loss": 0.8337, + "num_input_tokens_seen": 50609710, + "step": 2362 + }, + { + "epoch": 0.28413395058017193, + "flos": 20019472373760.0, + "grad_norm": 2.1414572775438248, + "learning_rate": 3.360073534706118e-06, + "loss": 0.8239, + "num_input_tokens_seen": 50626865, + "step": 2363 + }, + { + "epoch": 0.28425419347081105, + "flos": 37609150816320.0, + "grad_norm": 2.4601121275244093, + "learning_rate": 3.35950230667693e-06, + "loss": 0.7644, + "num_input_tokens_seen": 50648560, + "step": 2364 + }, + { + "epoch": 0.28437443636145016, + "flos": 13825532438880.0, + "grad_norm": 2.58868225657389, + "learning_rate": 3.358930872412323e-06, + "loss": 0.8659, + "num_input_tokens_seen": 50665725, + "step": 2365 + }, + { + "epoch": 0.2844946792520892, + "flos": 22715298976320.0, + "grad_norm": 3.489373833739646, + "learning_rate": 3.3583592319989825e-06, + "loss": 0.8122, + "num_input_tokens_seen": 50685095, + "step": 2366 + }, + { + "epoch": 0.2846149221427283, + "flos": 32369396184960.0, + "grad_norm": 2.264159155889228, + "learning_rate": 3.357787385523627e-06, + "loss": 0.6908, + "num_input_tokens_seen": 50706500, + "step": 2367 + }, + { + "epoch": 0.2847351650333674, + "flos": 28435138425120.0, + "grad_norm": 2.066935768860565, + "learning_rate": 3.3572153330730048e-06, + "loss": 0.8256, + "num_input_tokens_seen": 50727595, + "step": 2368 + }, + { + "epoch": 0.2848554079240065, + "flos": 55673667406560.0, + "grad_norm": 0.8381438684010856, + "learning_rate": 3.3566430747338956e-06, + "loss": 0.6474, + "num_input_tokens_seen": 50782800, + "step": 2369 + }, + { + "epoch": 0.2849756508146456, + "flos": 11820205896960.0, + "grad_norm": 2.3415895804571867, + "learning_rate": 3.35607061059311e-06, + "loss": 0.8633, + "num_input_tokens_seen": 50797730, + "step": 2370 + }, + { + "epoch": 0.28509589370528465, + "flos": 25119185584800.0, + "grad_norm": 2.052594297150279, + "learning_rate": 3.3554979407374917e-06, + "loss": 0.7488, + "num_input_tokens_seen": 50819840, + "step": 2371 + }, + { + "epoch": 0.28521613659592376, + "flos": 19946347988640.0, + "grad_norm": 1.7648403483472432, + "learning_rate": 3.3549250652539134e-06, + "loss": 0.7351, + "num_input_tokens_seen": 50838935, + "step": 2372 + }, + { + "epoch": 0.2853363794865629, + "flos": 23334570309600.0, + "grad_norm": 1.939436652199052, + "learning_rate": 3.3543519842292794e-06, + "loss": 0.8122, + "num_input_tokens_seen": 50858590, + "step": 2373 + }, + { + "epoch": 0.28545662237720193, + "flos": 19833520566240.0, + "grad_norm": 3.9779225479955693, + "learning_rate": 3.353778697750527e-06, + "loss": 0.8329, + "num_input_tokens_seen": 50876995, + "step": 2374 + }, + { + "epoch": 0.28557686526784104, + "flos": 23844879875520.0, + "grad_norm": 2.030122297201688, + "learning_rate": 3.353205205904622e-06, + "loss": 0.8955, + "num_input_tokens_seen": 50896105, + "step": 2375 + }, + { + "epoch": 0.28569710815848015, + "flos": 44825494031520.0, + "grad_norm": 2.1478290597961887, + "learning_rate": 3.3526315087785637e-06, + "loss": 0.7165, + "num_input_tokens_seen": 50917220, + "step": 2376 + }, + { + "epoch": 0.2858173510491192, + "flos": 26791159286400.0, + "grad_norm": 1.804396568727032, + "learning_rate": 3.3520576064593805e-06, + "loss": 0.8107, + "num_input_tokens_seen": 50937175, + "step": 2377 + }, + { + "epoch": 0.2859375939397583, + "flos": 23115829040160.0, + "grad_norm": 2.1122214997127733, + "learning_rate": 3.3514834990341337e-06, + "loss": 0.8176, + "num_input_tokens_seen": 50957500, + "step": 2378 + }, + { + "epoch": 0.2860578368303974, + "flos": 12112740607200.0, + "grad_norm": 5.368049208141202, + "learning_rate": 3.3509091865899144e-06, + "loss": 0.9291, + "num_input_tokens_seen": 50972690, + "step": 2379 + }, + { + "epoch": 0.2861780797210365, + "flos": 19909618532160.0, + "grad_norm": 2.5702666768450504, + "learning_rate": 3.350334669213846e-06, + "loss": 0.705, + "num_input_tokens_seen": 50990095, + "step": 2380 + }, + { + "epoch": 0.2862983226116756, + "flos": 27523592569920.0, + "grad_norm": 2.224155342851695, + "learning_rate": 3.3497599469930816e-06, + "loss": 0.758, + "num_input_tokens_seen": 51008305, + "step": 2381 + }, + { + "epoch": 0.28641856550231465, + "flos": 22022457220800.0, + "grad_norm": 2.42840630707265, + "learning_rate": 3.349185020014807e-06, + "loss": 0.8314, + "num_input_tokens_seen": 51025570, + "step": 2382 + }, + { + "epoch": 0.28653880839295376, + "flos": 22346666300160.0, + "grad_norm": 2.2550711175230527, + "learning_rate": 3.348609888366237e-06, + "loss": 0.7448, + "num_input_tokens_seen": 51044585, + "step": 2383 + }, + { + "epoch": 0.28665905128359287, + "flos": 23334570309600.0, + "grad_norm": 3.9965304586337274, + "learning_rate": 3.348034552134619e-06, + "loss": 0.6337, + "num_input_tokens_seen": 51063990, + "step": 2384 + }, + { + "epoch": 0.2867792941742319, + "flos": 20851723663680.0, + "grad_norm": 2.2408577302016286, + "learning_rate": 3.3474590114072316e-06, + "loss": 0.8421, + "num_input_tokens_seen": 51081990, + "step": 2385 + }, + { + "epoch": 0.28689953706487104, + "flos": 20634394845120.0, + "grad_norm": 2.132182157382033, + "learning_rate": 3.3468832662713836e-06, + "loss": 0.8361, + "num_input_tokens_seen": 51100235, + "step": 2386 + }, + { + "epoch": 0.28701977995551015, + "flos": 12657623783520.0, + "grad_norm": 2.7449599811020127, + "learning_rate": 3.346307316814415e-06, + "loss": 0.842, + "num_input_tokens_seen": 51115405, + "step": 2387 + }, + { + "epoch": 0.2871400228461492, + "flos": 21222066148800.0, + "grad_norm": 2.036123796255517, + "learning_rate": 3.3457311631236965e-06, + "loss": 0.7585, + "num_input_tokens_seen": 51136390, + "step": 2388 + }, + { + "epoch": 0.2872602657367883, + "flos": 25083311032800.0, + "grad_norm": 2.1601835114619634, + "learning_rate": 3.345154805286631e-06, + "loss": 0.8489, + "num_input_tokens_seen": 51156650, + "step": 2389 + }, + { + "epoch": 0.2873805086274274, + "flos": 16622701008000.0, + "grad_norm": 2.7500138715914346, + "learning_rate": 3.344578243390651e-06, + "loss": 0.7644, + "num_input_tokens_seen": 51172210, + "step": 2390 + }, + { + "epoch": 0.2875007515180665, + "flos": 17395357705440.0, + "grad_norm": 2.814830746754329, + "learning_rate": 3.3440014775232206e-06, + "loss": 0.7898, + "num_input_tokens_seen": 51190265, + "step": 2391 + }, + { + "epoch": 0.2876209944087056, + "flos": 23400446591520.0, + "grad_norm": 2.272426419719312, + "learning_rate": 3.343424507771834e-06, + "loss": 0.7127, + "num_input_tokens_seen": 51208475, + "step": 2392 + }, + { + "epoch": 0.2877412372993447, + "flos": 13716124634400.0, + "grad_norm": 3.013435794584968, + "learning_rate": 3.342847334224018e-06, + "loss": 0.8661, + "num_input_tokens_seen": 51225875, + "step": 2393 + }, + { + "epoch": 0.28786148018998375, + "flos": 57997584660480.0, + "grad_norm": 0.9564242559494553, + "learning_rate": 3.342269956967329e-06, + "loss": 0.6233, + "num_input_tokens_seen": 51287780, + "step": 2394 + }, + { + "epoch": 0.28798172308062286, + "flos": 23401338665760.0, + "grad_norm": 3.4554351397921015, + "learning_rate": 3.341692376089355e-06, + "loss": 0.7124, + "num_input_tokens_seen": 51303335, + "step": 2395 + }, + { + "epoch": 0.288101965971262, + "flos": 25073758404480.0, + "grad_norm": 3.4219636252435874, + "learning_rate": 3.3411145916777146e-06, + "loss": 0.8361, + "num_input_tokens_seen": 51317495, + "step": 2396 + }, + { + "epoch": 0.28822220886190103, + "flos": 16229493386880.0, + "grad_norm": 3.7412761940439028, + "learning_rate": 3.3405366038200566e-06, + "loss": 0.9084, + "num_input_tokens_seen": 51336430, + "step": 2397 + }, + { + "epoch": 0.28834245175254014, + "flos": 24500546137440.0, + "grad_norm": 2.3506031804389376, + "learning_rate": 3.3399584126040617e-06, + "loss": 0.8502, + "num_input_tokens_seen": 51355930, + "step": 2398 + }, + { + "epoch": 0.2884626946431792, + "flos": 24536272010400.0, + "grad_norm": 2.228244721565513, + "learning_rate": 3.339380018117441e-06, + "loss": 0.909, + "num_input_tokens_seen": 51376765, + "step": 2399 + }, + { + "epoch": 0.2885829375338183, + "flos": 16521173192640.0, + "grad_norm": 2.9352500902557166, + "learning_rate": 3.3388014204479366e-06, + "loss": 0.7851, + "num_input_tokens_seen": 51394570, + "step": 2400 + }, + { + "epoch": 0.2887031804244574, + "flos": 24026817348960.0, + "grad_norm": 4.60866590315032, + "learning_rate": 3.338222619683321e-06, + "loss": 0.917, + "num_input_tokens_seen": 51414535, + "step": 2401 + }, + { + "epoch": 0.2888234233150965, + "flos": 23297320476480.0, + "grad_norm": 2.4188413262683888, + "learning_rate": 3.337643615911398e-06, + "loss": 0.7327, + "num_input_tokens_seen": 51434160, + "step": 2402 + }, + { + "epoch": 0.2889436662057356, + "flos": 22240938301920.0, + "grad_norm": 2.4522442056480673, + "learning_rate": 3.3370644092200026e-06, + "loss": 0.7919, + "num_input_tokens_seen": 51451435, + "step": 2403 + }, + { + "epoch": 0.2890639090963747, + "flos": 21586089774720.0, + "grad_norm": 2.595767603777315, + "learning_rate": 3.3364849996969985e-06, + "loss": 0.7865, + "num_input_tokens_seen": 51471455, + "step": 2404 + }, + { + "epoch": 0.28918415198701375, + "flos": 28544546229600.0, + "grad_norm": 20.850468457857474, + "learning_rate": 3.335905387430283e-06, + "loss": 0.8559, + "num_input_tokens_seen": 51492890, + "step": 2405 + }, + { + "epoch": 0.28930439487765286, + "flos": 21913346774400.0, + "grad_norm": 6.095422759693109, + "learning_rate": 3.335325572507782e-06, + "loss": 0.8323, + "num_input_tokens_seen": 51513390, + "step": 2406 + }, + { + "epoch": 0.28942463776829197, + "flos": 19254472646880.0, + "grad_norm": 1.7569042656902678, + "learning_rate": 3.3347455550174537e-06, + "loss": 0.737, + "num_input_tokens_seen": 51532730, + "step": 2407 + }, + { + "epoch": 0.289544880658931, + "flos": 14624919927360.0, + "grad_norm": 2.331126486444007, + "learning_rate": 3.3341653350472864e-06, + "loss": 0.6796, + "num_input_tokens_seen": 51549560, + "step": 2408 + }, + { + "epoch": 0.28966512354957014, + "flos": 28580829648960.0, + "grad_norm": 2.716553864350696, + "learning_rate": 3.333584912685298e-06, + "loss": 0.6968, + "num_input_tokens_seen": 51568180, + "step": 2409 + }, + { + "epoch": 0.28978536644020925, + "flos": 64619714694240.0, + "grad_norm": 0.9129165900361101, + "learning_rate": 3.3330042880195385e-06, + "loss": 0.5549, + "num_input_tokens_seen": 51622530, + "step": 2410 + }, + { + "epoch": 0.2899056093308483, + "flos": 18598137329280.0, + "grad_norm": 2.9084992267620073, + "learning_rate": 3.3324234611380888e-06, + "loss": 0.7824, + "num_input_tokens_seen": 51641260, + "step": 2411 + }, + { + "epoch": 0.2900258522214874, + "flos": 22861027369920.0, + "grad_norm": 1.7029541349882014, + "learning_rate": 3.3318424321290596e-06, + "loss": 0.815, + "num_input_tokens_seen": 51660975, + "step": 2412 + }, + { + "epoch": 0.2901460951121265, + "flos": 71003663674080.0, + "grad_norm": 0.8421522483454583, + "learning_rate": 3.3312612010805917e-06, + "loss": 0.5987, + "num_input_tokens_seen": 51720550, + "step": 2413 + }, + { + "epoch": 0.2902663380027656, + "flos": 32114780363520.0, + "grad_norm": 2.358245983685305, + "learning_rate": 3.330679768080858e-06, + "loss": 0.7079, + "num_input_tokens_seen": 51744435, + "step": 2414 + }, + { + "epoch": 0.2903865808934047, + "flos": 29309545956480.0, + "grad_norm": 5.016971015358302, + "learning_rate": 3.3300981332180627e-06, + "loss": 0.8345, + "num_input_tokens_seen": 51763640, + "step": 2415 + }, + { + "epoch": 0.29050682378404374, + "flos": 17064643918080.0, + "grad_norm": 2.162919606859794, + "learning_rate": 3.3295162965804373e-06, + "loss": 0.8033, + "num_input_tokens_seen": 51782135, + "step": 2416 + }, + { + "epoch": 0.29062706667468285, + "flos": 17832840244320.0, + "grad_norm": 12.111326077698989, + "learning_rate": 3.328934258256247e-06, + "loss": 0.7865, + "num_input_tokens_seen": 51800200, + "step": 2417 + }, + { + "epoch": 0.29074730956532197, + "flos": 24245075411520.0, + "grad_norm": 2.5811925036513186, + "learning_rate": 3.3283520183337856e-06, + "loss": 0.6689, + "num_input_tokens_seen": 51819905, + "step": 2418 + }, + { + "epoch": 0.290867552455961, + "flos": 22308858920640.0, + "grad_norm": 2.4244192234145103, + "learning_rate": 3.3277695769013797e-06, + "loss": 0.6919, + "num_input_tokens_seen": 51839350, + "step": 2419 + }, + { + "epoch": 0.29098779534660013, + "flos": 23153041703520.0, + "grad_norm": 2.1868596910942215, + "learning_rate": 3.327186934047385e-06, + "loss": 0.7765, + "num_input_tokens_seen": 51858045, + "step": 2420 + }, + { + "epoch": 0.29110803823723924, + "flos": 15282519016800.0, + "grad_norm": 2.42482395909579, + "learning_rate": 3.3266040898601877e-06, + "loss": 0.6539, + "num_input_tokens_seen": 51875880, + "step": 2421 + }, + { + "epoch": 0.2912282811278783, + "flos": 22562842856160.0, + "grad_norm": 2.7177601908869145, + "learning_rate": 3.3260210444282045e-06, + "loss": 0.781, + "num_input_tokens_seen": 51893835, + "step": 2422 + }, + { + "epoch": 0.2913485240185174, + "flos": 24462813097440.0, + "grad_norm": 2.189050288578148, + "learning_rate": 3.325437797839883e-06, + "loss": 0.736, + "num_input_tokens_seen": 51912205, + "step": 2423 + }, + { + "epoch": 0.2914687669091565, + "flos": 17905518592320.0, + "grad_norm": 3.1787200880724664, + "learning_rate": 3.3248543501837015e-06, + "loss": 0.7542, + "num_input_tokens_seen": 51929410, + "step": 2424 + }, + { + "epoch": 0.2915890097997956, + "flos": 22497152423040.0, + "grad_norm": 2.169385402503937, + "learning_rate": 3.3242707015481684e-06, + "loss": 0.7766, + "num_input_tokens_seen": 51949345, + "step": 2425 + }, + { + "epoch": 0.2917092526904347, + "flos": 13825681117920.0, + "grad_norm": 2.027748493813421, + "learning_rate": 3.323686852021823e-06, + "loss": 0.8076, + "num_input_tokens_seen": 51966855, + "step": 2426 + }, + { + "epoch": 0.2918294955810738, + "flos": 22642471949280.0, + "grad_norm": 2.6539689985015507, + "learning_rate": 3.323102801693235e-06, + "loss": 0.7979, + "num_input_tokens_seen": 51985620, + "step": 2427 + }, + { + "epoch": 0.29194973847171285, + "flos": 23404683944160.0, + "grad_norm": 3.108713050809835, + "learning_rate": 3.322518550651003e-06, + "loss": 0.8024, + "num_input_tokens_seen": 52004930, + "step": 2428 + }, + { + "epoch": 0.29206998136235196, + "flos": 21877806750240.0, + "grad_norm": 2.6338912610235528, + "learning_rate": 3.3219340989837586e-06, + "loss": 0.8115, + "num_input_tokens_seen": 52024920, + "step": 2429 + }, + { + "epoch": 0.292190224252991, + "flos": 23182523056800.0, + "grad_norm": 2.1066585427542064, + "learning_rate": 3.3213494467801625e-06, + "loss": 0.8025, + "num_input_tokens_seen": 52044095, + "step": 2430 + }, + { + "epoch": 0.2923104671436301, + "flos": 20711236206240.0, + "grad_norm": 2.669201768431327, + "learning_rate": 3.3207645941289063e-06, + "loss": 0.7173, + "num_input_tokens_seen": 52062440, + "step": 2431 + }, + { + "epoch": 0.29243071003426924, + "flos": 35757581336160.0, + "grad_norm": 2.1609068630471238, + "learning_rate": 3.320179541118711e-06, + "loss": 0.8002, + "num_input_tokens_seen": 52084940, + "step": 2432 + }, + { + "epoch": 0.2925509529249083, + "flos": 57999554657760.0, + "grad_norm": 1.0475479354451323, + "learning_rate": 3.3195942878383293e-06, + "loss": 0.6029, + "num_input_tokens_seen": 52141040, + "step": 2433 + }, + { + "epoch": 0.2926711958155474, + "flos": 21366567940320.0, + "grad_norm": 2.071696729168264, + "learning_rate": 3.319008834376543e-06, + "loss": 0.7774, + "num_input_tokens_seen": 52160730, + "step": 2434 + }, + { + "epoch": 0.2927914387061865, + "flos": 23152670005920.0, + "grad_norm": 3.0552088299939575, + "learning_rate": 3.3184231808221654e-06, + "loss": 0.8875, + "num_input_tokens_seen": 52175055, + "step": 2435 + }, + { + "epoch": 0.29291168159682557, + "flos": 22423470491520.0, + "grad_norm": 2.370633779841889, + "learning_rate": 3.3178373272640394e-06, + "loss": 0.6267, + "num_input_tokens_seen": 52194150, + "step": 2436 + }, + { + "epoch": 0.2930319244874647, + "flos": 21140132530560.0, + "grad_norm": 2.4448481597246254, + "learning_rate": 3.3172512737910387e-06, + "loss": 0.8488, + "num_input_tokens_seen": 52211660, + "step": 2437 + }, + { + "epoch": 0.2931521673781038, + "flos": 31312790991840.0, + "grad_norm": 12.613838927745874, + "learning_rate": 3.3166650204920674e-06, + "loss": 0.8834, + "num_input_tokens_seen": 52232830, + "step": 2438 + }, + { + "epoch": 0.29327241026874284, + "flos": 24166152543840.0, + "grad_norm": 4.377598889779067, + "learning_rate": 3.316078567456059e-06, + "loss": 0.8144, + "num_input_tokens_seen": 52250670, + "step": 2439 + }, + { + "epoch": 0.29339265315938196, + "flos": 24207899917920.0, + "grad_norm": 1.9124308579855422, + "learning_rate": 3.3154919147719786e-06, + "loss": 0.7605, + "num_input_tokens_seen": 52271685, + "step": 2440 + }, + { + "epoch": 0.29351289605002107, + "flos": 16922595330720.0, + "grad_norm": 2.1386049861115293, + "learning_rate": 3.31490506252882e-06, + "loss": 0.8631, + "num_input_tokens_seen": 52291065, + "step": 2441 + }, + { + "epoch": 0.2936331389406601, + "flos": 19800842613600.0, + "grad_norm": 2.1508850582524106, + "learning_rate": 3.31431801081561e-06, + "loss": 0.8391, + "num_input_tokens_seen": 52310000, + "step": 2442 + }, + { + "epoch": 0.29375338183129923, + "flos": 71314708924800.0, + "grad_norm": 0.9100399679796844, + "learning_rate": 3.313730759721402e-06, + "loss": 0.6792, + "num_input_tokens_seen": 52372930, + "step": 2443 + }, + { + "epoch": 0.29387362472193834, + "flos": 22023014767200.0, + "grad_norm": 2.1419149257388312, + "learning_rate": 3.313143309335282e-06, + "loss": 0.862, + "num_input_tokens_seen": 52391420, + "step": 2444 + }, + { + "epoch": 0.2939938676125774, + "flos": 22933594208640.0, + "grad_norm": 2.851956776573645, + "learning_rate": 3.3125556597463665e-06, + "loss": 0.8507, + "num_input_tokens_seen": 52410725, + "step": 2445 + }, + { + "epoch": 0.2941141105032165, + "flos": 31313683066080.0, + "grad_norm": 1.8689612738045047, + "learning_rate": 3.311967811043801e-06, + "loss": 0.6647, + "num_input_tokens_seen": 52432645, + "step": 2446 + }, + { + "epoch": 0.29423435339385556, + "flos": 23188953425280.0, + "grad_norm": 2.5967663610252734, + "learning_rate": 3.3113797633167617e-06, + "loss": 0.8195, + "num_input_tokens_seen": 52450940, + "step": 2447 + }, + { + "epoch": 0.2943545962844947, + "flos": 26825324029440.0, + "grad_norm": 2.611450740243722, + "learning_rate": 3.310791516654455e-06, + "loss": 0.687, + "num_input_tokens_seen": 52468000, + "step": 2448 + }, + { + "epoch": 0.2944748391751338, + "flos": 20202896637600.0, + "grad_norm": 4.371571532617867, + "learning_rate": 3.3102030711461177e-06, + "loss": 0.7944, + "num_input_tokens_seen": 52487575, + "step": 2449 + }, + { + "epoch": 0.29459508206577284, + "flos": 15938333957760.0, + "grad_norm": 2.132642466715559, + "learning_rate": 3.3096144268810156e-06, + "loss": 0.6821, + "num_input_tokens_seen": 52506335, + "step": 2450 + }, + { + "epoch": 0.29471532495641195, + "flos": 20383793357760.0, + "grad_norm": 2.3859656311655093, + "learning_rate": 3.3090255839484462e-06, + "loss": 0.7301, + "num_input_tokens_seen": 52524330, + "step": 2451 + }, + { + "epoch": 0.29483556784705106, + "flos": 20347547108160.0, + "grad_norm": 2.219020128596341, + "learning_rate": 3.3084365424377366e-06, + "loss": 0.854, + "num_input_tokens_seen": 52543095, + "step": 2452 + }, + { + "epoch": 0.2949558107376901, + "flos": 68457133762080.0, + "grad_norm": 0.7330651065462319, + "learning_rate": 3.307847302438245e-06, + "loss": 0.5592, + "num_input_tokens_seen": 52597075, + "step": 2453 + }, + { + "epoch": 0.2950760536283292, + "flos": 16084248200160.0, + "grad_norm": 2.998661696424672, + "learning_rate": 3.3072578640393562e-06, + "loss": 0.7813, + "num_input_tokens_seen": 52614410, + "step": 2454 + }, + { + "epoch": 0.29519629651896834, + "flos": 20454353029440.0, + "grad_norm": 2.1349181795187717, + "learning_rate": 3.3066682273304886e-06, + "loss": 0.7936, + "num_input_tokens_seen": 52632055, + "step": 2455 + }, + { + "epoch": 0.2953165394096074, + "flos": 18889556946720.0, + "grad_norm": 2.139218055970735, + "learning_rate": 3.3060783924010904e-06, + "loss": 0.7913, + "num_input_tokens_seen": 52649980, + "step": 2456 + }, + { + "epoch": 0.2954367823002465, + "flos": 20594803317120.0, + "grad_norm": 2.5688146615533354, + "learning_rate": 3.3054883593406387e-06, + "loss": 0.8474, + "num_input_tokens_seen": 52664730, + "step": 2457 + }, + { + "epoch": 0.2955570251908856, + "flos": 31130742009120.0, + "grad_norm": 2.1583423138091655, + "learning_rate": 3.3048981282386404e-06, + "loss": 0.6532, + "num_input_tokens_seen": 52686040, + "step": 2458 + }, + { + "epoch": 0.29567726808152467, + "flos": 21619325273760.0, + "grad_norm": 2.1207614589221047, + "learning_rate": 3.304307699184634e-06, + "loss": 0.8256, + "num_input_tokens_seen": 52704630, + "step": 2459 + }, + { + "epoch": 0.2957975109721638, + "flos": 24209200859520.0, + "grad_norm": 2.3909415241208087, + "learning_rate": 3.3037170722681866e-06, + "loss": 0.7852, + "num_input_tokens_seen": 52725465, + "step": 2460 + }, + { + "epoch": 0.29591775386280283, + "flos": 13460691078240.0, + "grad_norm": 2.0263463619035265, + "learning_rate": 3.3031262475788956e-06, + "loss": 0.6857, + "num_input_tokens_seen": 52742405, + "step": 2461 + }, + { + "epoch": 0.29603799675344195, + "flos": 17724584702400.0, + "grad_norm": 2.153037068602028, + "learning_rate": 3.3025352252063897e-06, + "loss": 0.7317, + "num_input_tokens_seen": 52761100, + "step": 2462 + }, + { + "epoch": 0.29615823964408106, + "flos": 22751359377120.0, + "grad_norm": 2.8487506537376013, + "learning_rate": 3.3019440052403252e-06, + "loss": 0.7504, + "num_input_tokens_seen": 52780965, + "step": 2463 + }, + { + "epoch": 0.2962784825347201, + "flos": 23480596061280.0, + "grad_norm": 1.7849483577620164, + "learning_rate": 3.30135258777039e-06, + "loss": 0.7092, + "num_input_tokens_seen": 52800335, + "step": 2464 + }, + { + "epoch": 0.2963987254253592, + "flos": 16339570247040.0, + "grad_norm": 5.0694899163609115, + "learning_rate": 3.3007609728863024e-06, + "loss": 0.7029, + "num_input_tokens_seen": 52818425, + "step": 2465 + }, + { + "epoch": 0.29651896831599833, + "flos": 33424068550560.0, + "grad_norm": 2.2384525253513763, + "learning_rate": 3.300169160677809e-06, + "loss": 0.7295, + "num_input_tokens_seen": 52842340, + "step": 2466 + }, + { + "epoch": 0.2966392112066374, + "flos": 23771309453280.0, + "grad_norm": 2.949124002513653, + "learning_rate": 3.2995771512346878e-06, + "loss": 0.7694, + "num_input_tokens_seen": 52860690, + "step": 2467 + }, + { + "epoch": 0.2967594540972765, + "flos": 19910064569280.0, + "grad_norm": 3.9406206389606493, + "learning_rate": 3.298984944646746e-06, + "loss": 0.7346, + "num_input_tokens_seen": 52879370, + "step": 2468 + }, + { + "epoch": 0.2968796969879156, + "flos": 23698891293600.0, + "grad_norm": 2.4477976885079262, + "learning_rate": 3.298392541003822e-06, + "loss": 0.8163, + "num_input_tokens_seen": 52898775, + "step": 2469 + }, + { + "epoch": 0.29699993987855466, + "flos": 22860990200160.0, + "grad_norm": 4.32951558493503, + "learning_rate": 3.2977999403957806e-06, + "loss": 0.9011, + "num_input_tokens_seen": 52917535, + "step": 2470 + }, + { + "epoch": 0.2971201827691938, + "flos": 33783668975040.0, + "grad_norm": 2.9042504107434026, + "learning_rate": 3.2972071429125207e-06, + "loss": 0.6675, + "num_input_tokens_seen": 52938755, + "step": 2471 + }, + { + "epoch": 0.2972404256598329, + "flos": 22022866088160.0, + "grad_norm": 2.4639863035700724, + "learning_rate": 3.2966141486439682e-06, + "loss": 0.8837, + "num_input_tokens_seen": 52957945, + "step": 2472 + }, + { + "epoch": 0.29736066855047194, + "flos": 31932657041280.0, + "grad_norm": 3.0210778148419277, + "learning_rate": 3.29602095768008e-06, + "loss": 0.641, + "num_input_tokens_seen": 52978460, + "step": 2473 + }, + { + "epoch": 0.29748091144111105, + "flos": 33462210457920.0, + "grad_norm": 3.4515764034987626, + "learning_rate": 3.2954275701108437e-06, + "loss": 0.6393, + "num_input_tokens_seen": 52999640, + "step": 2474 + }, + { + "epoch": 0.29760115433175016, + "flos": 41223920056320.0, + "grad_norm": 2.0814517361582165, + "learning_rate": 3.294833986026275e-06, + "loss": 0.69, + "num_input_tokens_seen": 53022880, + "step": 2475 + }, + { + "epoch": 0.2977213972223892, + "flos": 24458092537920.0, + "grad_norm": 2.417046834925149, + "learning_rate": 3.29424020551642e-06, + "loss": 0.8529, + "num_input_tokens_seen": 53041515, + "step": 2476 + }, + { + "epoch": 0.2978416401130283, + "flos": 21254632592160.0, + "grad_norm": 2.4966699356071285, + "learning_rate": 3.2936462286713546e-06, + "loss": 0.7212, + "num_input_tokens_seen": 53059865, + "step": 2477 + }, + { + "epoch": 0.2979618830036674, + "flos": 25735111639680.0, + "grad_norm": 2.876458005649337, + "learning_rate": 3.2930520555811846e-06, + "loss": 0.7724, + "num_input_tokens_seen": 53079490, + "step": 2478 + }, + { + "epoch": 0.2980821258943065, + "flos": 23443420567680.0, + "grad_norm": 2.2141497565865516, + "learning_rate": 3.292457686336046e-06, + "loss": 0.7972, + "num_input_tokens_seen": 53098810, + "step": 2479 + }, + { + "epoch": 0.2982023687849456, + "flos": 69653105586240.0, + "grad_norm": 0.87231134413585, + "learning_rate": 3.291863121026105e-06, + "loss": 0.6125, + "num_input_tokens_seen": 53162590, + "step": 2480 + }, + { + "epoch": 0.29832261167558466, + "flos": 29783126065920.0, + "grad_norm": 2.6082869219891114, + "learning_rate": 3.2912683597415547e-06, + "loss": 0.7623, + "num_input_tokens_seen": 53186675, + "step": 2481 + }, + { + "epoch": 0.29844285456622377, + "flos": 33861328070880.0, + "grad_norm": 2.393089370601787, + "learning_rate": 3.2906734025726213e-06, + "loss": 0.7834, + "num_input_tokens_seen": 53205940, + "step": 2482 + }, + { + "epoch": 0.2985630974568629, + "flos": 23842724029440.0, + "grad_norm": 7.878408511794177, + "learning_rate": 3.290078249609559e-06, + "loss": 0.8823, + "num_input_tokens_seen": 53225360, + "step": 2483 + }, + { + "epoch": 0.29868334034750194, + "flos": 21767990078400.0, + "grad_norm": 2.3862189286188875, + "learning_rate": 3.2894829009426514e-06, + "loss": 0.8807, + "num_input_tokens_seen": 53243195, + "step": 2484 + }, + { + "epoch": 0.29880358323814105, + "flos": 25629904018080.0, + "grad_norm": 11.04257632173258, + "learning_rate": 3.288887356662213e-06, + "loss": 0.7786, + "num_input_tokens_seen": 53264530, + "step": 2485 + }, + { + "epoch": 0.29892382612878016, + "flos": 70903957176960.0, + "grad_norm": 0.7785094066166723, + "learning_rate": 3.288291616858588e-06, + "loss": 0.5969, + "num_input_tokens_seen": 53319920, + "step": 2486 + }, + { + "epoch": 0.2990440690194192, + "flos": 25444546926720.0, + "grad_norm": 2.0406749678331484, + "learning_rate": 3.287695681622149e-06, + "loss": 0.7686, + "num_input_tokens_seen": 53339270, + "step": 2487 + }, + { + "epoch": 0.2991643119100583, + "flos": 23698408086720.0, + "grad_norm": 2.9442658120276777, + "learning_rate": 3.2870995510432982e-06, + "loss": 0.8129, + "num_input_tokens_seen": 53357750, + "step": 2488 + }, + { + "epoch": 0.29928455480069743, + "flos": 27378235873920.0, + "grad_norm": 2.768529885801607, + "learning_rate": 3.2865032252124697e-06, + "loss": 0.7685, + "num_input_tokens_seen": 53378265, + "step": 2489 + }, + { + "epoch": 0.2994047976913365, + "flos": 33644482459200.0, + "grad_norm": 2.7848015729253723, + "learning_rate": 3.2859067042201243e-06, + "loss": 0.7757, + "num_input_tokens_seen": 53400305, + "step": 2490 + }, + { + "epoch": 0.2995250405819756, + "flos": 16740174650400.0, + "grad_norm": 2.373617171216619, + "learning_rate": 3.2853099881567544e-06, + "loss": 0.7767, + "num_input_tokens_seen": 53418705, + "step": 2491 + }, + { + "epoch": 0.29964528347261465, + "flos": 22931066664960.0, + "grad_norm": 2.5219772412682184, + "learning_rate": 3.284713077112881e-06, + "loss": 0.7968, + "num_input_tokens_seen": 53438135, + "step": 2492 + }, + { + "epoch": 0.29976552636325376, + "flos": 16913674588320.0, + "grad_norm": 4.88320310937625, + "learning_rate": 3.284115971179056e-06, + "loss": 0.8643, + "num_input_tokens_seen": 53452165, + "step": 2493 + }, + { + "epoch": 0.2998857692538929, + "flos": 17032263323520.0, + "grad_norm": 2.1429726751154585, + "learning_rate": 3.283518670445859e-06, + "loss": 0.7815, + "num_input_tokens_seen": 53470075, + "step": 2494 + }, + { + "epoch": 0.30000601214453193, + "flos": 68733345214080.0, + "grad_norm": 0.7002043357703985, + "learning_rate": 3.2829211750038995e-06, + "loss": 0.543, + "num_input_tokens_seen": 53538705, + "step": 2495 + }, + { + "epoch": 0.30012625503517104, + "flos": 17578261592640.0, + "grad_norm": 2.070557086819939, + "learning_rate": 3.2823234849438183e-06, + "loss": 0.8865, + "num_input_tokens_seen": 53556740, + "step": 2496 + }, + { + "epoch": 0.30024649792581015, + "flos": 21222289167360.0, + "grad_norm": 2.5068688444348743, + "learning_rate": 3.2817256003562836e-06, + "loss": 0.7573, + "num_input_tokens_seen": 53577115, + "step": 2497 + }, + { + "epoch": 0.3003667408164492, + "flos": 22970398004640.0, + "grad_norm": 2.333100295918332, + "learning_rate": 3.281127521331995e-06, + "loss": 0.6599, + "num_input_tokens_seen": 53598295, + "step": 2498 + }, + { + "epoch": 0.3004869837070883, + "flos": 64140782139360.0, + "grad_norm": 0.8863351134843895, + "learning_rate": 3.2805292479616798e-06, + "loss": 0.6069, + "num_input_tokens_seen": 53657160, + "step": 2499 + }, + { + "epoch": 0.30060722659772743, + "flos": 26210996274240.0, + "grad_norm": 3.070308462048076, + "learning_rate": 3.2799307803360955e-06, + "loss": 0.9169, + "num_input_tokens_seen": 53673090, + "step": 2500 + }, + { + "epoch": 0.3007274694883665, + "flos": 24935055095520.0, + "grad_norm": 2.841044889498923, + "learning_rate": 3.27933211854603e-06, + "loss": 0.8158, + "num_input_tokens_seen": 53692145, + "step": 2501 + }, + { + "epoch": 0.3008477123790056, + "flos": 17031631437600.0, + "grad_norm": 1.7508003584479874, + "learning_rate": 3.278733262682299e-06, + "loss": 0.8726, + "num_input_tokens_seen": 53710000, + "step": 2502 + }, + { + "epoch": 0.3009679552696447, + "flos": 21475678386720.0, + "grad_norm": 3.035022079663977, + "learning_rate": 3.2781342128357484e-06, + "loss": 0.8288, + "num_input_tokens_seen": 53729355, + "step": 2503 + }, + { + "epoch": 0.30108819816028376, + "flos": 21103551753120.0, + "grad_norm": 2.3425506574920125, + "learning_rate": 3.2775349690972547e-06, + "loss": 0.8041, + "num_input_tokens_seen": 53743505, + "step": 2504 + }, + { + "epoch": 0.30120844105092287, + "flos": 71024776097760.0, + "grad_norm": 0.7610424462861013, + "learning_rate": 3.276935531557722e-06, + "loss": 0.5179, + "num_input_tokens_seen": 53808725, + "step": 2505 + }, + { + "epoch": 0.301328683941562, + "flos": 20235611760000.0, + "grad_norm": 2.5051957498235704, + "learning_rate": 3.2763359003080837e-06, + "loss": 0.7927, + "num_input_tokens_seen": 53825680, + "step": 2506 + }, + { + "epoch": 0.30144892683220104, + "flos": 70547107314720.0, + "grad_norm": 0.8756963819750746, + "learning_rate": 3.2757360754393047e-06, + "loss": 0.6244, + "num_input_tokens_seen": 53889750, + "step": 2507 + }, + { + "epoch": 0.30156916972284015, + "flos": 22787940154560.0, + "grad_norm": 10.053381537904531, + "learning_rate": 3.2751360570423767e-06, + "loss": 0.6464, + "num_input_tokens_seen": 53908135, + "step": 2508 + }, + { + "epoch": 0.3016894126134792, + "flos": 29856287620800.0, + "grad_norm": 2.5244014068227254, + "learning_rate": 3.2745358452083236e-06, + "loss": 0.7601, + "num_input_tokens_seen": 53931035, + "step": 2509 + }, + { + "epoch": 0.3018096555041183, + "flos": 21515678782080.0, + "grad_norm": 1.592003856845957, + "learning_rate": 3.2739354400281955e-06, + "loss": 0.8194, + "num_input_tokens_seen": 53952255, + "step": 2510 + }, + { + "epoch": 0.3019298983947574, + "flos": 59052591553920.0, + "grad_norm": 0.8623126711103205, + "learning_rate": 3.2733348415930744e-06, + "loss": 0.6364, + "num_input_tokens_seen": 54014125, + "step": 2511 + }, + { + "epoch": 0.3020501412853965, + "flos": 34373570464320.0, + "grad_norm": 2.817374096200772, + "learning_rate": 3.27273404999407e-06, + "loss": 0.8103, + "num_input_tokens_seen": 54036345, + "step": 2512 + }, + { + "epoch": 0.3021703841760356, + "flos": 70906484720640.0, + "grad_norm": 0.8076087370018706, + "learning_rate": 3.272133065322322e-06, + "loss": 0.6044, + "num_input_tokens_seen": 54094615, + "step": 2513 + }, + { + "epoch": 0.3022906270666747, + "flos": 21479952909120.0, + "grad_norm": 2.0407375894655386, + "learning_rate": 3.271531887669e-06, + "loss": 0.7934, + "num_input_tokens_seen": 54114675, + "step": 2514 + }, + { + "epoch": 0.30241086995731375, + "flos": 31087024637760.0, + "grad_norm": 3.6187174056039653, + "learning_rate": 3.2709305171253015e-06, + "loss": 0.6358, + "num_input_tokens_seen": 54134795, + "step": 2515 + }, + { + "epoch": 0.30253111284795287, + "flos": 23477845499040.0, + "grad_norm": 2.1547529026697516, + "learning_rate": 3.2703289537824536e-06, + "loss": 0.773, + "num_input_tokens_seen": 54154595, + "step": 2516 + }, + { + "epoch": 0.302651355738592, + "flos": 18698401372800.0, + "grad_norm": 3.0869376963960033, + "learning_rate": 3.269727197731714e-06, + "loss": 0.79, + "num_input_tokens_seen": 54167360, + "step": 2517 + }, + { + "epoch": 0.30277159862923103, + "flos": 22386852544320.0, + "grad_norm": 7.3437019508900185, + "learning_rate": 3.269125249064367e-06, + "loss": 0.7799, + "num_input_tokens_seen": 54187055, + "step": 2518 + }, + { + "epoch": 0.30289184151987014, + "flos": 22241309999520.0, + "grad_norm": 2.5092916434423014, + "learning_rate": 3.2685231078717297e-06, + "loss": 0.8311, + "num_input_tokens_seen": 54207245, + "step": 2519 + }, + { + "epoch": 0.30301208441050925, + "flos": 25189373558880.0, + "grad_norm": 2.147926408490925, + "learning_rate": 3.267920774245145e-06, + "loss": 0.7539, + "num_input_tokens_seen": 54226650, + "step": 2520 + }, + { + "epoch": 0.3031323273011483, + "flos": 23006309726400.0, + "grad_norm": 1.972314483950073, + "learning_rate": 3.2673182482759876e-06, + "loss": 0.845, + "num_input_tokens_seen": 54245765, + "step": 2521 + }, + { + "epoch": 0.3032525701917874, + "flos": 18849705230400.0, + "grad_norm": 3.670864274540088, + "learning_rate": 3.266715530055659e-06, + "loss": 0.6696, + "num_input_tokens_seen": 54263915, + "step": 2522 + }, + { + "epoch": 0.30337281308242653, + "flos": 17757262655040.0, + "grad_norm": 1.7527885018873877, + "learning_rate": 3.2661126196755927e-06, + "loss": 0.8048, + "num_input_tokens_seen": 54283025, + "step": 2523 + }, + { + "epoch": 0.3034930559730656, + "flos": 57742559971680.0, + "grad_norm": 0.7907737833185747, + "learning_rate": 3.265509517227248e-06, + "loss": 0.5593, + "num_input_tokens_seen": 54339840, + "step": 2524 + }, + { + "epoch": 0.3036132988637047, + "flos": 14734662259680.0, + "grad_norm": 2.2980849005318515, + "learning_rate": 3.264906222802115e-06, + "loss": 0.81, + "num_input_tokens_seen": 54357690, + "step": 2525 + }, + { + "epoch": 0.30373354175434375, + "flos": 21003213370080.0, + "grad_norm": 2.510286151031112, + "learning_rate": 3.264302736491715e-06, + "loss": 0.783, + "num_input_tokens_seen": 54377810, + "step": 2526 + }, + { + "epoch": 0.30385378464498286, + "flos": 21113104381440.0, + "grad_norm": 2.6714673539997316, + "learning_rate": 3.263699058387594e-06, + "loss": 0.8761, + "num_input_tokens_seen": 54395245, + "step": 2527 + }, + { + "epoch": 0.30397402753562197, + "flos": 20599746895200.0, + "grad_norm": 3.7153847477963455, + "learning_rate": 3.2630951885813315e-06, + "loss": 0.902, + "num_input_tokens_seen": 54412640, + "step": 2528 + }, + { + "epoch": 0.304094270426261, + "flos": 15064335293760.0, + "grad_norm": 2.5510800523986576, + "learning_rate": 3.262491127164533e-06, + "loss": 0.781, + "num_input_tokens_seen": 54431455, + "step": 2529 + }, + { + "epoch": 0.30421451331690014, + "flos": 13825681117920.0, + "grad_norm": 3.13263015026115, + "learning_rate": 3.2618868742288337e-06, + "loss": 0.8013, + "num_input_tokens_seen": 54448380, + "step": 2530 + }, + { + "epoch": 0.30433475620753925, + "flos": 17359520323200.0, + "grad_norm": 2.1221078376945615, + "learning_rate": 3.261282429865899e-06, + "loss": 0.7252, + "num_input_tokens_seen": 54466385, + "step": 2531 + }, + { + "epoch": 0.3044549990981783, + "flos": 18890040153600.0, + "grad_norm": 2.0142934970067645, + "learning_rate": 3.2606777941674225e-06, + "loss": 0.723, + "num_input_tokens_seen": 54485040, + "step": 2532 + }, + { + "epoch": 0.3045752419888174, + "flos": 21038344526880.0, + "grad_norm": 3.1674116329159907, + "learning_rate": 3.2600729672251276e-06, + "loss": 0.845, + "num_input_tokens_seen": 54502755, + "step": 2533 + }, + { + "epoch": 0.3046954848794565, + "flos": 29054669946720.0, + "grad_norm": 2.1632662523603226, + "learning_rate": 3.259467949130765e-06, + "loss": 0.6589, + "num_input_tokens_seen": 54524165, + "step": 2534 + }, + { + "epoch": 0.3048157277700956, + "flos": 20266133866560.0, + "grad_norm": 2.5358665658522193, + "learning_rate": 3.2588627399761164e-06, + "loss": 0.826, + "num_input_tokens_seen": 54540160, + "step": 2535 + }, + { + "epoch": 0.3049359706607347, + "flos": 22707158798880.0, + "grad_norm": 2.1292868737654387, + "learning_rate": 3.2582573398529903e-06, + "loss": 0.7048, + "num_input_tokens_seen": 54557515, + "step": 2536 + }, + { + "epoch": 0.3050562135513738, + "flos": 18408134017920.0, + "grad_norm": 2.7300314273352755, + "learning_rate": 3.2576517488532265e-06, + "loss": 0.7387, + "num_input_tokens_seen": 54573865, + "step": 2537 + }, + { + "epoch": 0.30517645644201286, + "flos": 20341079569920.0, + "grad_norm": 3.1374989604568277, + "learning_rate": 3.257045967068692e-06, + "loss": 0.8755, + "num_input_tokens_seen": 54593120, + "step": 2538 + }, + { + "epoch": 0.30529669933265197, + "flos": 21913792811520.0, + "grad_norm": 6.669573277402139, + "learning_rate": 3.2564399945912848e-06, + "loss": 0.8222, + "num_input_tokens_seen": 54612910, + "step": 2539 + }, + { + "epoch": 0.305416942223291, + "flos": 21804236328000.0, + "grad_norm": 2.785695498616233, + "learning_rate": 3.2558338315129287e-06, + "loss": 0.8242, + "num_input_tokens_seen": 54631055, + "step": 2540 + }, + { + "epoch": 0.30553718511393013, + "flos": 33863037879840.0, + "grad_norm": 2.2618312247710626, + "learning_rate": 3.2552274779255785e-06, + "loss": 0.7609, + "num_input_tokens_seen": 54651985, + "step": 2541 + }, + { + "epoch": 0.30565742800456924, + "flos": 22236552270240.0, + "grad_norm": 2.220415806363622, + "learning_rate": 3.2546209339212184e-06, + "loss": 0.771, + "num_input_tokens_seen": 54671005, + "step": 2542 + }, + { + "epoch": 0.3057776708952083, + "flos": 22532803956480.0, + "grad_norm": 2.4407101546706746, + "learning_rate": 3.25401419959186e-06, + "loss": 0.7742, + "num_input_tokens_seen": 54691615, + "step": 2543 + }, + { + "epoch": 0.3058979137858474, + "flos": 21768101587680.0, + "grad_norm": 4.136521865150617, + "learning_rate": 3.253407275029545e-06, + "loss": 0.762, + "num_input_tokens_seen": 54710520, + "step": 2544 + }, + { + "epoch": 0.3060181566764865, + "flos": 26940753335040.0, + "grad_norm": 1.9358761658393135, + "learning_rate": 3.2528001603263425e-06, + "loss": 0.806, + "num_input_tokens_seen": 54732590, + "step": 2545 + }, + { + "epoch": 0.3061383995671256, + "flos": 19836791505120.0, + "grad_norm": 2.7920550365652668, + "learning_rate": 3.2521928555743514e-06, + "loss": 0.8167, + "num_input_tokens_seen": 54749935, + "step": 2546 + }, + { + "epoch": 0.3062586424577647, + "flos": 22095916133760.0, + "grad_norm": 2.0554875313898693, + "learning_rate": 3.2515853608657e-06, + "loss": 0.6784, + "num_input_tokens_seen": 54767935, + "step": 2547 + }, + { + "epoch": 0.3063788853484038, + "flos": 20815588923360.0, + "grad_norm": 2.9807192795877397, + "learning_rate": 3.250977676292545e-06, + "loss": 0.7526, + "num_input_tokens_seen": 54786175, + "step": 2548 + }, + { + "epoch": 0.30649912823904285, + "flos": 16186482240960.0, + "grad_norm": 2.3945244793695064, + "learning_rate": 3.2503698019470712e-06, + "loss": 0.7934, + "num_input_tokens_seen": 54801945, + "step": 2549 + }, + { + "epoch": 0.30661937112968196, + "flos": 18591075074880.0, + "grad_norm": 2.802612557239422, + "learning_rate": 3.249761737921492e-06, + "loss": 0.7858, + "num_input_tokens_seen": 54818475, + "step": 2550 + }, + { + "epoch": 0.30673961402032107, + "flos": 31345580453760.0, + "grad_norm": 2.3523335254235405, + "learning_rate": 3.249153484308051e-06, + "loss": 0.7449, + "num_input_tokens_seen": 54837090, + "step": 2551 + }, + { + "epoch": 0.3068598569109601, + "flos": 20200963810080.0, + "grad_norm": 2.8441017687867323, + "learning_rate": 3.2485450411990194e-06, + "loss": 0.7779, + "num_input_tokens_seen": 54856445, + "step": 2552 + }, + { + "epoch": 0.30698009980159924, + "flos": 29559329709120.0, + "grad_norm": 2.2946000023615016, + "learning_rate": 3.2479364086866983e-06, + "loss": 0.8272, + "num_input_tokens_seen": 54876700, + "step": 2553 + }, + { + "epoch": 0.30710034269223835, + "flos": 23808336267840.0, + "grad_norm": 2.0777033327856516, + "learning_rate": 3.247327586863416e-06, + "loss": 0.8161, + "num_input_tokens_seen": 54897580, + "step": 2554 + }, + { + "epoch": 0.3072205855828774, + "flos": 25847901892320.0, + "grad_norm": 2.305151010229516, + "learning_rate": 3.2467185758215304e-06, + "loss": 0.7725, + "num_input_tokens_seen": 54917320, + "step": 2555 + }, + { + "epoch": 0.3073408284735165, + "flos": 22204617712800.0, + "grad_norm": 2.4528446887768376, + "learning_rate": 3.246109375653428e-06, + "loss": 0.8498, + "num_input_tokens_seen": 54934085, + "step": 2556 + }, + { + "epoch": 0.30746107136415557, + "flos": 19472507690880.0, + "grad_norm": 2.428707599829331, + "learning_rate": 3.2454999864515243e-06, + "loss": 0.7848, + "num_input_tokens_seen": 54953460, + "step": 2557 + }, + { + "epoch": 0.3075813142547947, + "flos": 21693750600480.0, + "grad_norm": 1.9968995858411651, + "learning_rate": 3.244890408308263e-06, + "loss": 0.6943, + "num_input_tokens_seen": 54974165, + "step": 2558 + }, + { + "epoch": 0.3077015571454338, + "flos": 24063509635680.0, + "grad_norm": 2.524057412547975, + "learning_rate": 3.2442806413161165e-06, + "loss": 0.6142, + "num_input_tokens_seen": 54993810, + "step": 2559 + }, + { + "epoch": 0.30782180003607285, + "flos": 18379284550560.0, + "grad_norm": 2.754280486955527, + "learning_rate": 3.243670685567586e-06, + "loss": 0.759, + "num_input_tokens_seen": 55011210, + "step": 2560 + }, + { + "epoch": 0.30794204292671196, + "flos": 23844396668640.0, + "grad_norm": 3.296135263482204, + "learning_rate": 3.2430605411552012e-06, + "loss": 0.8041, + "num_input_tokens_seen": 55031325, + "step": 2561 + }, + { + "epoch": 0.30806228581735107, + "flos": 67831240477920.0, + "grad_norm": 0.8948610189352528, + "learning_rate": 3.2424502081715205e-06, + "loss": 0.7061, + "num_input_tokens_seen": 55094440, + "step": 2562 + }, + { + "epoch": 0.3081825287079901, + "flos": 23809154002560.0, + "grad_norm": 2.4358534509082763, + "learning_rate": 3.241839686709132e-06, + "loss": 0.7784, + "num_input_tokens_seen": 55115375, + "step": 2563 + }, + { + "epoch": 0.30830277159862923, + "flos": 16186742429280.0, + "grad_norm": 2.914800719161067, + "learning_rate": 3.2412289768606495e-06, + "loss": 0.8268, + "num_input_tokens_seen": 55131945, + "step": 2564 + }, + { + "epoch": 0.30842301448926834, + "flos": 29307315770880.0, + "grad_norm": 2.0876631438869033, + "learning_rate": 3.240618078718718e-06, + "loss": 0.8301, + "num_input_tokens_seen": 55153405, + "step": 2565 + }, + { + "epoch": 0.3085432573799074, + "flos": 21913867151040.0, + "grad_norm": 2.7673661090886, + "learning_rate": 3.240006992376011e-06, + "loss": 0.7411, + "num_input_tokens_seen": 55173030, + "step": 2566 + }, + { + "epoch": 0.3086635002705465, + "flos": 22023237785760.0, + "grad_norm": 3.5501688312626416, + "learning_rate": 3.2393957179252284e-06, + "loss": 0.762, + "num_input_tokens_seen": 55189565, + "step": 2567 + }, + { + "epoch": 0.3087837431611856, + "flos": 32618771070240.0, + "grad_norm": 5.159691049656488, + "learning_rate": 3.2387842554591016e-06, + "loss": 0.8051, + "num_input_tokens_seen": 55210380, + "step": 2568 + }, + { + "epoch": 0.3089039860518247, + "flos": 17574098579520.0, + "grad_norm": 7.574121042755555, + "learning_rate": 3.238172605070388e-06, + "loss": 0.8742, + "num_input_tokens_seen": 55225795, + "step": 2569 + }, + { + "epoch": 0.3090242289424638, + "flos": 14363204681760.0, + "grad_norm": 3.43994871833192, + "learning_rate": 3.2375607668518745e-06, + "loss": 0.7849, + "num_input_tokens_seen": 55238830, + "step": 2570 + }, + { + "epoch": 0.30914447183310284, + "flos": 16045771764960.0, + "grad_norm": 3.256424935413889, + "learning_rate": 3.236948740896377e-06, + "loss": 0.9014, + "num_input_tokens_seen": 55253630, + "step": 2571 + }, + { + "epoch": 0.30926471472374195, + "flos": 32183778905280.0, + "grad_norm": 1.5255040945912752, + "learning_rate": 3.2363365272967384e-06, + "loss": 0.8398, + "num_input_tokens_seen": 55276200, + "step": 2572 + }, + { + "epoch": 0.30938495761438106, + "flos": 20341488437280.0, + "grad_norm": 2.1692820926906804, + "learning_rate": 3.235724126145832e-06, + "loss": 0.8165, + "num_input_tokens_seen": 55292235, + "step": 2573 + }, + { + "epoch": 0.3095052005050201, + "flos": 24026743009440.0, + "grad_norm": 3.5593719134095485, + "learning_rate": 3.235111537536558e-06, + "loss": 0.7749, + "num_input_tokens_seen": 55313050, + "step": 2574 + }, + { + "epoch": 0.30962544339565923, + "flos": 23367991657440.0, + "grad_norm": 2.447052942580173, + "learning_rate": 3.2344987615618456e-06, + "loss": 0.8289, + "num_input_tokens_seen": 55332885, + "step": 2575 + }, + { + "epoch": 0.30974568628629834, + "flos": 33751288380480.0, + "grad_norm": 1.8928740814253107, + "learning_rate": 3.2338857983146533e-06, + "loss": 0.7879, + "num_input_tokens_seen": 55354385, + "step": 2576 + }, + { + "epoch": 0.3098659291769374, + "flos": 20200740791520.0, + "grad_norm": 2.4521738139927973, + "learning_rate": 3.233272647887966e-06, + "loss": 0.768, + "num_input_tokens_seen": 55373275, + "step": 2577 + }, + { + "epoch": 0.3099861720675765, + "flos": 24712931377920.0, + "grad_norm": 2.1559642657713978, + "learning_rate": 3.2326593103747985e-06, + "loss": 0.8997, + "num_input_tokens_seen": 55392290, + "step": 2578 + }, + { + "epoch": 0.3101064149582156, + "flos": 11748977169600.0, + "grad_norm": 2.197840202652705, + "learning_rate": 3.2320457858681936e-06, + "loss": 0.852, + "num_input_tokens_seen": 55410680, + "step": 2579 + }, + { + "epoch": 0.31022665784885467, + "flos": 23000102376480.0, + "grad_norm": 3.4388199092812695, + "learning_rate": 3.2314320744612228e-06, + "loss": 0.8548, + "num_input_tokens_seen": 55425980, + "step": 2580 + }, + { + "epoch": 0.3103469007394938, + "flos": 16266074164320.0, + "grad_norm": 1.9823260805634393, + "learning_rate": 3.2308181762469854e-06, + "loss": 0.7682, + "num_input_tokens_seen": 55443925, + "step": 2581 + }, + { + "epoch": 0.3104671436301329, + "flos": 30471321601440.0, + "grad_norm": 2.487127159202456, + "learning_rate": 3.230204091318609e-06, + "loss": 0.7848, + "num_input_tokens_seen": 55464465, + "step": 2582 + }, + { + "epoch": 0.31058738652077195, + "flos": 20018654639040.0, + "grad_norm": 2.150283852402873, + "learning_rate": 3.2295898197692503e-06, + "loss": 0.8493, + "num_input_tokens_seen": 55483425, + "step": 2583 + }, + { + "epoch": 0.31070762941141106, + "flos": 28034385342720.0, + "grad_norm": 2.310700412269767, + "learning_rate": 3.228975361692094e-06, + "loss": 0.7917, + "num_input_tokens_seen": 55504445, + "step": 2584 + }, + { + "epoch": 0.31082787230205017, + "flos": 20492643615840.0, + "grad_norm": 2.9313149776455183, + "learning_rate": 3.228360717180352e-06, + "loss": 0.801, + "num_input_tokens_seen": 55521970, + "step": 2585 + }, + { + "epoch": 0.3109481151926892, + "flos": 62357021768640.0, + "grad_norm": 0.8127507043070984, + "learning_rate": 3.227745886327266e-06, + "loss": 0.5936, + "num_input_tokens_seen": 55580665, + "step": 2586 + }, + { + "epoch": 0.31106835808332833, + "flos": 44683476880320.0, + "grad_norm": 0.8114062797359797, + "learning_rate": 3.227130869226105e-06, + "loss": 0.5578, + "num_input_tokens_seen": 55637825, + "step": 2587 + }, + { + "epoch": 0.3111886009739674, + "flos": 23369552787360.0, + "grad_norm": 2.732167775443856, + "learning_rate": 3.226515665970167e-06, + "loss": 0.8272, + "num_input_tokens_seen": 55655365, + "step": 2588 + }, + { + "epoch": 0.3113088438646065, + "flos": 17505843432960.0, + "grad_norm": 2.470045787834761, + "learning_rate": 3.225900276652777e-06, + "loss": 0.8599, + "num_input_tokens_seen": 55672140, + "step": 2589 + }, + { + "epoch": 0.3114290867552456, + "flos": 28325098734720.0, + "grad_norm": 1.7986533611282427, + "learning_rate": 3.2252847013672906e-06, + "loss": 0.7595, + "num_input_tokens_seen": 55694800, + "step": 2590 + }, + { + "epoch": 0.31154932964588467, + "flos": 27339833778240.0, + "grad_norm": 2.792056744400557, + "learning_rate": 3.224668940207089e-06, + "loss": 0.7536, + "num_input_tokens_seen": 55713785, + "step": 2591 + }, + { + "epoch": 0.3116695725365238, + "flos": 26503828342560.0, + "grad_norm": 2.2340573171709424, + "learning_rate": 3.2240529932655828e-06, + "loss": 0.8659, + "num_input_tokens_seen": 55733050, + "step": 2592 + }, + { + "epoch": 0.3117898154271629, + "flos": 21147343464000.0, + "grad_norm": 4.363880741255582, + "learning_rate": 3.223436860636211e-06, + "loss": 0.8828, + "num_input_tokens_seen": 55748685, + "step": 2593 + }, + { + "epoch": 0.31191005831780194, + "flos": 27233362384800.0, + "grad_norm": 2.4609118263479752, + "learning_rate": 3.2228205424124403e-06, + "loss": 0.7439, + "num_input_tokens_seen": 55771840, + "step": 2594 + }, + { + "epoch": 0.31203030120844105, + "flos": 12945029066880.0, + "grad_norm": 2.9693757050324097, + "learning_rate": 3.222204038687765e-06, + "loss": 0.7483, + "num_input_tokens_seen": 55784975, + "step": 2595 + }, + { + "epoch": 0.31215054409908016, + "flos": 27522440307360.0, + "grad_norm": 1.9002944592631237, + "learning_rate": 3.221587349555709e-06, + "loss": 0.8804, + "num_input_tokens_seen": 55805235, + "step": 2596 + }, + { + "epoch": 0.3122707869897192, + "flos": 21475604047200.0, + "grad_norm": 1.8231940648777811, + "learning_rate": 3.2209704751098236e-06, + "loss": 0.6947, + "num_input_tokens_seen": 55824265, + "step": 2597 + }, + { + "epoch": 0.31239102988035833, + "flos": 15165231223200.0, + "grad_norm": 2.8727590700722914, + "learning_rate": 3.2203534154436875e-06, + "loss": 0.8239, + "num_input_tokens_seen": 55838620, + "step": 2598 + }, + { + "epoch": 0.31251127277099744, + "flos": 22022197032480.0, + "grad_norm": 5.069368765395296, + "learning_rate": 3.2197361706509084e-06, + "loss": 0.7587, + "num_input_tokens_seen": 55859375, + "step": 2599 + }, + { + "epoch": 0.3126315156616365, + "flos": 15172442156640.0, + "grad_norm": 3.1826709743939916, + "learning_rate": 3.2191187408251228e-06, + "loss": 0.8383, + "num_input_tokens_seen": 55876535, + "step": 2600 + }, + { + "epoch": 0.3127517585522756, + "flos": 18119353453440.0, + "grad_norm": 3.520149811677803, + "learning_rate": 3.218501126059993e-06, + "loss": 0.7884, + "num_input_tokens_seen": 55891930, + "step": 2601 + }, + { + "epoch": 0.31287200144291466, + "flos": 21877360713120.0, + "grad_norm": 2.439844031430099, + "learning_rate": 3.2178833264492116e-06, + "loss": 0.8167, + "num_input_tokens_seen": 55910635, + "step": 2602 + }, + { + "epoch": 0.31299224433355377, + "flos": 29854020265440.0, + "grad_norm": 2.022767180535217, + "learning_rate": 3.217265342086498e-06, + "loss": 0.7624, + "num_input_tokens_seen": 55931125, + "step": 2603 + }, + { + "epoch": 0.3131124872241929, + "flos": 11639532195360.0, + "grad_norm": 2.4935551540959655, + "learning_rate": 3.216647173065599e-06, + "loss": 0.7298, + "num_input_tokens_seen": 55946590, + "step": 2604 + }, + { + "epoch": 0.31323273011483194, + "flos": 49776542437920.0, + "grad_norm": 1.8987029090250302, + "learning_rate": 3.216028819480292e-06, + "loss": 0.7393, + "num_input_tokens_seen": 55967930, + "step": 2605 + }, + { + "epoch": 0.31335297300547105, + "flos": 22568975866560.0, + "grad_norm": 7.531877293718144, + "learning_rate": 3.2154102814243793e-06, + "loss": 0.7548, + "num_input_tokens_seen": 55987070, + "step": 2606 + }, + { + "epoch": 0.31347321589611016, + "flos": 34660789898880.0, + "grad_norm": 2.459612389707976, + "learning_rate": 3.2147915589916937e-06, + "loss": 0.6736, + "num_input_tokens_seen": 56008630, + "step": 2607 + }, + { + "epoch": 0.3135934587867492, + "flos": 19727643888960.0, + "grad_norm": 3.0832626918914112, + "learning_rate": 3.2141726522760938e-06, + "loss": 0.8302, + "num_input_tokens_seen": 56026690, + "step": 2608 + }, + { + "epoch": 0.3137137016773883, + "flos": 65722193104800.0, + "grad_norm": 0.7191676709229614, + "learning_rate": 3.213553561371469e-06, + "loss": 0.5261, + "num_input_tokens_seen": 56090425, + "step": 2609 + }, + { + "epoch": 0.31383394456802743, + "flos": 16229679235680.0, + "grad_norm": 2.8741735846249203, + "learning_rate": 3.212934286371733e-06, + "loss": 0.9577, + "num_input_tokens_seen": 56107135, + "step": 2610 + }, + { + "epoch": 0.3139541874586665, + "flos": 38739437940960.0, + "grad_norm": 2.6796052881202614, + "learning_rate": 3.2123148273708304e-06, + "loss": 0.8331, + "num_input_tokens_seen": 56127325, + "step": 2611 + }, + { + "epoch": 0.3140744303493056, + "flos": 25010484005760.0, + "grad_norm": 2.0071540673179955, + "learning_rate": 3.211695184462733e-06, + "loss": 0.7684, + "num_input_tokens_seen": 56148500, + "step": 2612 + }, + { + "epoch": 0.3141946732399447, + "flos": 72400275094560.0, + "grad_norm": 0.8982187087641608, + "learning_rate": 3.2110753577414383e-06, + "loss": 0.6045, + "num_input_tokens_seen": 56210080, + "step": 2613 + }, + { + "epoch": 0.31431491613058377, + "flos": 19212204896160.0, + "grad_norm": 2.3214734188400916, + "learning_rate": 3.2104553473009757e-06, + "loss": 0.7886, + "num_input_tokens_seen": 56228200, + "step": 2614 + }, + { + "epoch": 0.3144351590212229, + "flos": 36158483097600.0, + "grad_norm": 3.222119095858424, + "learning_rate": 3.209835153235399e-06, + "loss": 0.6785, + "num_input_tokens_seen": 56249755, + "step": 2615 + }, + { + "epoch": 0.314555401911862, + "flos": 18525347472000.0, + "grad_norm": 2.132823790356163, + "learning_rate": 3.2092147756387916e-06, + "loss": 0.6749, + "num_input_tokens_seen": 56270080, + "step": 2616 + }, + { + "epoch": 0.31467564480250104, + "flos": 16339681756320.0, + "grad_norm": 2.139256506314937, + "learning_rate": 3.208594214605264e-06, + "loss": 0.8313, + "num_input_tokens_seen": 56288625, + "step": 2617 + }, + { + "epoch": 0.31479588769314015, + "flos": 21621443950080.0, + "grad_norm": 3.0429085761059373, + "learning_rate": 3.2079734702289553e-06, + "loss": 0.7732, + "num_input_tokens_seen": 56307520, + "step": 2618 + }, + { + "epoch": 0.3149161305837792, + "flos": 65956662916320.0, + "grad_norm": 0.8051352875325758, + "learning_rate": 3.207352542604031e-06, + "loss": 0.604, + "num_input_tokens_seen": 56369855, + "step": 2619 + }, + { + "epoch": 0.3150363734744183, + "flos": 28945633839840.0, + "grad_norm": 2.0060516829826103, + "learning_rate": 3.2067314318246864e-06, + "loss": 0.7858, + "num_input_tokens_seen": 56390570, + "step": 2620 + }, + { + "epoch": 0.31515661636505743, + "flos": 27596977143360.0, + "grad_norm": 2.9041074012463164, + "learning_rate": 3.206110137985143e-06, + "loss": 0.778, + "num_input_tokens_seen": 56410775, + "step": 2621 + }, + { + "epoch": 0.3152768592556965, + "flos": 24570325244160.0, + "grad_norm": 2.5968497264747703, + "learning_rate": 3.2054886611796505e-06, + "loss": 0.9231, + "num_input_tokens_seen": 56429770, + "step": 2622 + }, + { + "epoch": 0.3153971021463356, + "flos": 68378285233920.0, + "grad_norm": 0.8923154147990875, + "learning_rate": 3.204867001502487e-06, + "loss": 0.6346, + "num_input_tokens_seen": 56488425, + "step": 2623 + }, + { + "epoch": 0.3155173450369747, + "flos": 25557076991040.0, + "grad_norm": 1.9650587567190958, + "learning_rate": 3.2042451590479567e-06, + "loss": 0.8035, + "num_input_tokens_seen": 56508940, + "step": 2624 + }, + { + "epoch": 0.31563758792761376, + "flos": 24274296576480.0, + "grad_norm": 2.872209936624401, + "learning_rate": 3.203623133910394e-06, + "loss": 0.872, + "num_input_tokens_seen": 56527245, + "step": 2625 + }, + { + "epoch": 0.31575783081825287, + "flos": 31857971526240.0, + "grad_norm": 3.7743300493964944, + "learning_rate": 3.203000926184158e-06, + "loss": 0.775, + "num_input_tokens_seen": 56546170, + "step": 2626 + }, + { + "epoch": 0.315878073708892, + "flos": 30766718383200.0, + "grad_norm": 2.1824085383351677, + "learning_rate": 3.202378535963639e-06, + "loss": 0.7727, + "num_input_tokens_seen": 56567085, + "step": 2627 + }, + { + "epoch": 0.31599831659953104, + "flos": 22168557312000.0, + "grad_norm": 2.200963901554786, + "learning_rate": 3.2017559633432516e-06, + "loss": 0.8381, + "num_input_tokens_seen": 56586715, + "step": 2628 + }, + { + "epoch": 0.31611855949017015, + "flos": 25556816802720.0, + "grad_norm": 2.8460298558784936, + "learning_rate": 3.2011332084174398e-06, + "loss": 0.6609, + "num_input_tokens_seen": 56607585, + "step": 2629 + }, + { + "epoch": 0.31623880238080926, + "flos": 20564690077920.0, + "grad_norm": 4.19147678397665, + "learning_rate": 3.2005102712806756e-06, + "loss": 0.8893, + "num_input_tokens_seen": 56626415, + "step": 2630 + }, + { + "epoch": 0.3163590452714483, + "flos": 12766845739200.0, + "grad_norm": 3.9990382230343275, + "learning_rate": 3.1998871520274575e-06, + "loss": 0.7279, + "num_input_tokens_seen": 56641750, + "step": 2631 + }, + { + "epoch": 0.3164792881620874, + "flos": 23008428402720.0, + "grad_norm": 2.152348513173892, + "learning_rate": 3.199263850752312e-06, + "loss": 0.852, + "num_input_tokens_seen": 56662585, + "step": 2632 + }, + { + "epoch": 0.31659953105272653, + "flos": 18270248443680.0, + "grad_norm": 2.409256976016289, + "learning_rate": 3.198640367549795e-06, + "loss": 0.8598, + "num_input_tokens_seen": 56681480, + "step": 2633 + }, + { + "epoch": 0.3167197739433656, + "flos": 25666633474560.0, + "grad_norm": 1.7996963244813171, + "learning_rate": 3.198016702514487e-06, + "loss": 0.8578, + "num_input_tokens_seen": 56701760, + "step": 2634 + }, + { + "epoch": 0.3168400168340047, + "flos": 23512196090880.0, + "grad_norm": 1.9632711316905622, + "learning_rate": 3.1973928557409972e-06, + "loss": 0.8455, + "num_input_tokens_seen": 56719800, + "step": 2635 + }, + { + "epoch": 0.31696025972464376, + "flos": 28325767790400.0, + "grad_norm": 2.7291869728167715, + "learning_rate": 3.1967688273239636e-06, + "loss": 0.7112, + "num_input_tokens_seen": 56739605, + "step": 2636 + }, + { + "epoch": 0.31708050261528287, + "flos": 16375853666400.0, + "grad_norm": 1.9422493836540258, + "learning_rate": 3.1961446173580503e-06, + "loss": 0.8165, + "num_input_tokens_seen": 56756545, + "step": 2637 + }, + { + "epoch": 0.317200745505922, + "flos": 26174750024640.0, + "grad_norm": 2.3449225989217815, + "learning_rate": 3.1955202259379502e-06, + "loss": 0.7734, + "num_input_tokens_seen": 56778275, + "step": 2638 + }, + { + "epoch": 0.31732098839656103, + "flos": 31305580058400.0, + "grad_norm": 1.975868314285633, + "learning_rate": 3.194895653158381e-06, + "loss": 0.8275, + "num_input_tokens_seen": 56797215, + "step": 2639 + }, + { + "epoch": 0.31744123128720014, + "flos": 58905487879200.0, + "grad_norm": 0.773339352015664, + "learning_rate": 3.194270899114093e-06, + "loss": 0.5544, + "num_input_tokens_seen": 56863010, + "step": 2640 + }, + { + "epoch": 0.31756147417783925, + "flos": 17392718652480.0, + "grad_norm": 2.22483882105954, + "learning_rate": 3.193645963899858e-06, + "loss": 0.8249, + "num_input_tokens_seen": 56880625, + "step": 2641 + }, + { + "epoch": 0.3176817170684783, + "flos": 25445104473120.0, + "grad_norm": 2.2222800313529896, + "learning_rate": 3.193020847610479e-06, + "loss": 0.8382, + "num_input_tokens_seen": 56900945, + "step": 2642 + }, + { + "epoch": 0.3178019599591174, + "flos": 24936839244000.0, + "grad_norm": 2.470875078934725, + "learning_rate": 3.192395550340787e-06, + "loss": 0.712, + "num_input_tokens_seen": 56918895, + "step": 2643 + }, + { + "epoch": 0.31792220284975653, + "flos": 12404754940800.0, + "grad_norm": 2.22621446532945, + "learning_rate": 3.191770072185638e-06, + "loss": 0.7685, + "num_input_tokens_seen": 56935890, + "step": 2644 + }, + { + "epoch": 0.3180424457403956, + "flos": 15463787434560.0, + "grad_norm": 3.0175722782609213, + "learning_rate": 3.191144413239916e-06, + "loss": 0.7301, + "num_input_tokens_seen": 56952460, + "step": 2645 + }, + { + "epoch": 0.3181626886310347, + "flos": 26137388682240.0, + "grad_norm": 2.461225414085518, + "learning_rate": 3.190518573598534e-06, + "loss": 0.8808, + "num_input_tokens_seen": 56969065, + "step": 2646 + }, + { + "epoch": 0.3182829315216738, + "flos": 25446591263520.0, + "grad_norm": 3.892387144281673, + "learning_rate": 3.1898925533564308e-06, + "loss": 0.7738, + "num_input_tokens_seen": 56990535, + "step": 2647 + }, + { + "epoch": 0.31840317441231286, + "flos": 18087939272640.0, + "grad_norm": 2.650005315379444, + "learning_rate": 3.1892663526085733e-06, + "loss": 0.6413, + "num_input_tokens_seen": 57008910, + "step": 2648 + }, + { + "epoch": 0.31852341730295197, + "flos": 64649604914880.0, + "grad_norm": 0.7576436906758176, + "learning_rate": 3.188639971449956e-06, + "loss": 0.5691, + "num_input_tokens_seen": 57074960, + "step": 2649 + }, + { + "epoch": 0.318643660193591, + "flos": 20638669367520.0, + "grad_norm": 2.4431502559953944, + "learning_rate": 3.1880134099756e-06, + "loss": 0.7241, + "num_input_tokens_seen": 57094595, + "step": 2650 + }, + { + "epoch": 0.31876390308423014, + "flos": 26904730104000.0, + "grad_norm": 2.739337276727297, + "learning_rate": 3.1873866682805535e-06, + "loss": 0.695, + "num_input_tokens_seen": 57115290, + "step": 2651 + }, + { + "epoch": 0.31888414597486925, + "flos": 18016301677920.0, + "grad_norm": 2.2179519937546757, + "learning_rate": 3.186759746459894e-06, + "loss": 0.8834, + "num_input_tokens_seen": 57134400, + "step": 2652 + }, + { + "epoch": 0.3190043888655083, + "flos": 25113089744160.0, + "grad_norm": 1.967551759865451, + "learning_rate": 3.1861326446087246e-06, + "loss": 0.7959, + "num_input_tokens_seen": 57153140, + "step": 2653 + }, + { + "epoch": 0.3191246317561474, + "flos": 22022382881280.0, + "grad_norm": 2.2507076210338024, + "learning_rate": 3.1855053628221763e-06, + "loss": 0.7171, + "num_input_tokens_seen": 57172395, + "step": 2654 + }, + { + "epoch": 0.3192448746467865, + "flos": 14880688011360.0, + "grad_norm": 4.522583770566555, + "learning_rate": 3.184877901195407e-06, + "loss": 0.9007, + "num_input_tokens_seen": 57189090, + "step": 2655 + }, + { + "epoch": 0.3193651175374256, + "flos": 67139290796640.0, + "grad_norm": 0.8005878866320388, + "learning_rate": 3.184250259823602e-06, + "loss": 0.6284, + "num_input_tokens_seen": 57251620, + "step": 2656 + }, + { + "epoch": 0.3194853604280647, + "flos": 12215643703680.0, + "grad_norm": 2.4802593953850867, + "learning_rate": 3.183622438801974e-06, + "loss": 0.817, + "num_input_tokens_seen": 57266910, + "step": 2657 + }, + { + "epoch": 0.3196056033187038, + "flos": 14918792748960.0, + "grad_norm": 3.2104092277095755, + "learning_rate": 3.1829944382257637e-06, + "loss": 0.7518, + "num_input_tokens_seen": 57285040, + "step": 2658 + }, + { + "epoch": 0.31972584620934286, + "flos": 23734877354880.0, + "grad_norm": 2.670297703665715, + "learning_rate": 3.1823662581902373e-06, + "loss": 0.8126, + "num_input_tokens_seen": 57302725, + "step": 2659 + }, + { + "epoch": 0.31984608909998197, + "flos": 21221211244320.0, + "grad_norm": 20.504561498269798, + "learning_rate": 3.1817378987906896e-06, + "loss": 0.7454, + "num_input_tokens_seen": 57322230, + "step": 2660 + }, + { + "epoch": 0.3199663319906211, + "flos": 18270211273920.0, + "grad_norm": 2.3747727134767715, + "learning_rate": 3.181109360122442e-06, + "loss": 0.8008, + "num_input_tokens_seen": 57339820, + "step": 2661 + }, + { + "epoch": 0.32008657488126013, + "flos": 18706950417600.0, + "grad_norm": 2.5319003929686, + "learning_rate": 3.1804806422808445e-06, + "loss": 0.7831, + "num_input_tokens_seen": 57356955, + "step": 2662 + }, + { + "epoch": 0.32020681777189924, + "flos": 20565991019520.0, + "grad_norm": 1.7557711128226705, + "learning_rate": 3.1798517453612714e-06, + "loss": 0.7325, + "num_input_tokens_seen": 57376760, + "step": 2663 + }, + { + "epoch": 0.32032706066253835, + "flos": 35210951181120.0, + "grad_norm": 2.7863186715084796, + "learning_rate": 3.1792226694591265e-06, + "loss": 0.7512, + "num_input_tokens_seen": 57398145, + "step": 2664 + }, + { + "epoch": 0.3204473035531774, + "flos": 15282556186560.0, + "grad_norm": 2.5644330432858045, + "learning_rate": 3.178593414669841e-06, + "loss": 0.8032, + "num_input_tokens_seen": 57416730, + "step": 2665 + }, + { + "epoch": 0.3205675464438165, + "flos": 24427570431360.0, + "grad_norm": 3.870362983654652, + "learning_rate": 3.1779639810888707e-06, + "loss": 0.7027, + "num_input_tokens_seen": 57436325, + "step": 2666 + }, + { + "epoch": 0.3206877893344556, + "flos": 22424176716960.0, + "grad_norm": 2.2577235854437636, + "learning_rate": 3.1773343688117013e-06, + "loss": 0.761, + "num_input_tokens_seen": 57457235, + "step": 2667 + }, + { + "epoch": 0.3208080322250947, + "flos": 20383124302080.0, + "grad_norm": 2.412995334037383, + "learning_rate": 3.1767045779338445e-06, + "loss": 0.8381, + "num_input_tokens_seen": 57474895, + "step": 2668 + }, + { + "epoch": 0.3209282751157338, + "flos": 21731037603360.0, + "grad_norm": 2.23615849946897, + "learning_rate": 3.176074608550839e-06, + "loss": 0.9167, + "num_input_tokens_seen": 57490715, + "step": 2669 + }, + { + "epoch": 0.32104851800637285, + "flos": 22023572313600.0, + "grad_norm": 3.716214253099378, + "learning_rate": 3.17544446075825e-06, + "loss": 0.8246, + "num_input_tokens_seen": 57509880, + "step": 2670 + }, + { + "epoch": 0.32116876089701196, + "flos": 36958985678880.0, + "grad_norm": 1.9298851745918395, + "learning_rate": 3.174814134651671e-06, + "loss": 0.7094, + "num_input_tokens_seen": 57532430, + "step": 2671 + }, + { + "epoch": 0.3212890037876511, + "flos": 21947771705760.0, + "grad_norm": 2.1120938333850012, + "learning_rate": 3.1741836303267215e-06, + "loss": 0.8036, + "num_input_tokens_seen": 57551965, + "step": 2672 + }, + { + "epoch": 0.32140924667829013, + "flos": 10327902313440.0, + "grad_norm": 2.06443972840379, + "learning_rate": 3.1735529478790496e-06, + "loss": 0.7508, + "num_input_tokens_seen": 57569955, + "step": 2673 + }, + { + "epoch": 0.32152948956892924, + "flos": 50725338126240.0, + "grad_norm": 2.167347752065032, + "learning_rate": 3.172922087404328e-06, + "loss": 0.7963, + "num_input_tokens_seen": 57592215, + "step": 2674 + }, + { + "epoch": 0.32164973245956835, + "flos": 63772075123680.0, + "grad_norm": 0.7775830231279168, + "learning_rate": 3.1722910489982586e-06, + "loss": 0.5522, + "num_input_tokens_seen": 57652575, + "step": 2675 + }, + { + "epoch": 0.3217699753502074, + "flos": 23480335872960.0, + "grad_norm": 2.123305149739701, + "learning_rate": 3.1716598327565694e-06, + "loss": 0.7991, + "num_input_tokens_seen": 57672215, + "step": 2676 + }, + { + "epoch": 0.3218902182408465, + "flos": 19035396849600.0, + "grad_norm": 6.364168726959168, + "learning_rate": 3.171028438775015e-06, + "loss": 0.8377, + "num_input_tokens_seen": 57691410, + "step": 2677 + }, + { + "epoch": 0.3220104611314856, + "flos": 20346506354880.0, + "grad_norm": 2.6318360280902118, + "learning_rate": 3.170396867149377e-06, + "loss": 0.8404, + "num_input_tokens_seen": 57709415, + "step": 2678 + }, + { + "epoch": 0.3221307040221247, + "flos": 20088210727200.0, + "grad_norm": 46.07260372990792, + "learning_rate": 3.1697651179754653e-06, + "loss": 0.8647, + "num_input_tokens_seen": 57728075, + "step": 2679 + }, + { + "epoch": 0.3222509469127638, + "flos": 23953469945280.0, + "grad_norm": 2.762910965950192, + "learning_rate": 3.1691331913491153e-06, + "loss": 0.732, + "num_input_tokens_seen": 57750645, + "step": 2680 + }, + { + "epoch": 0.32237118980340285, + "flos": 17650605412800.0, + "grad_norm": 3.406052329951088, + "learning_rate": 3.1685010873661898e-06, + "loss": 0.8469, + "num_input_tokens_seen": 57769820, + "step": 2681 + }, + { + "epoch": 0.32249143269404196, + "flos": 23114862626400.0, + "grad_norm": 3.555388329136421, + "learning_rate": 3.167868806122578e-06, + "loss": 0.8013, + "num_input_tokens_seen": 57788870, + "step": 2682 + }, + { + "epoch": 0.32261167558468107, + "flos": 24387235508160.0, + "grad_norm": 2.8767744476295287, + "learning_rate": 3.1672363477141968e-06, + "loss": 0.6625, + "num_input_tokens_seen": 57808165, + "step": 2683 + }, + { + "epoch": 0.3227319184753201, + "flos": 30323363022240.0, + "grad_norm": 2.0175982139155977, + "learning_rate": 3.1666037122369903e-06, + "loss": 0.8463, + "num_input_tokens_seen": 57828825, + "step": 2684 + }, + { + "epoch": 0.32285216136595923, + "flos": 16922112123840.0, + "grad_norm": 2.6011493343322427, + "learning_rate": 3.165970899786928e-06, + "loss": 0.8697, + "num_input_tokens_seen": 57846740, + "step": 2685 + }, + { + "epoch": 0.32297240425659834, + "flos": 21950150570400.0, + "grad_norm": 2.9732864253302647, + "learning_rate": 3.1653379104600067e-06, + "loss": 0.7563, + "num_input_tokens_seen": 57866885, + "step": 2686 + }, + { + "epoch": 0.3230926471472374, + "flos": 22715559164640.0, + "grad_norm": 2.179302550747953, + "learning_rate": 3.164704744352251e-06, + "loss": 0.6949, + "num_input_tokens_seen": 57887690, + "step": 2687 + }, + { + "epoch": 0.3232128900378765, + "flos": 16918804015200.0, + "grad_norm": 3.890238424805227, + "learning_rate": 3.164071401559713e-06, + "loss": 0.8084, + "num_input_tokens_seen": 57905090, + "step": 2688 + }, + { + "epoch": 0.3233331329285156, + "flos": 23989344497280.0, + "grad_norm": 2.2740515440078943, + "learning_rate": 3.1634378821784674e-06, + "loss": 0.7095, + "num_input_tokens_seen": 57925305, + "step": 2689 + }, + { + "epoch": 0.3234533758191547, + "flos": 18088087951680.0, + "grad_norm": 2.4207407602632007, + "learning_rate": 3.1628041863046208e-06, + "loss": 0.7421, + "num_input_tokens_seen": 57942720, + "step": 2690 + }, + { + "epoch": 0.3235736187097938, + "flos": 16922074954080.0, + "grad_norm": 2.568576864454724, + "learning_rate": 3.162170314034304e-06, + "loss": 0.9141, + "num_input_tokens_seen": 57960135, + "step": 2691 + }, + { + "epoch": 0.3236938616004329, + "flos": 22095544436160.0, + "grad_norm": 4.536118621311225, + "learning_rate": 3.1615362654636738e-06, + "loss": 0.8096, + "num_input_tokens_seen": 57980875, + "step": 2692 + }, + { + "epoch": 0.32381410449107195, + "flos": 17140221507360.0, + "grad_norm": 1.9833997597776858, + "learning_rate": 3.1609020406889163e-06, + "loss": 0.871, + "num_input_tokens_seen": 57998310, + "step": 2693 + }, + { + "epoch": 0.32393434738171106, + "flos": 16554483031200.0, + "grad_norm": 1.97466733038095, + "learning_rate": 3.1602676398062416e-06, + "loss": 0.8513, + "num_input_tokens_seen": 58017220, + "step": 2694 + }, + { + "epoch": 0.3240545902723502, + "flos": 25447111640160.0, + "grad_norm": 2.3490907979917406, + "learning_rate": 3.1596330629118886e-06, + "loss": 0.6159, + "num_input_tokens_seen": 58037190, + "step": 2695 + }, + { + "epoch": 0.32417483316298923, + "flos": 35794273622880.0, + "grad_norm": 12.899726308832038, + "learning_rate": 3.1589983101021223e-06, + "loss": 0.7318, + "num_input_tokens_seen": 58055820, + "step": 2696 + }, + { + "epoch": 0.32429507605362834, + "flos": 30037295850240.0, + "grad_norm": 2.2495377383622435, + "learning_rate": 3.1583633814732337e-06, + "loss": 0.8459, + "num_input_tokens_seen": 58075340, + "step": 2697 + }, + { + "epoch": 0.3244153189442674, + "flos": 18197347077120.0, + "grad_norm": 3.7056959538994265, + "learning_rate": 3.157728277121541e-06, + "loss": 0.7198, + "num_input_tokens_seen": 58093515, + "step": 2698 + }, + { + "epoch": 0.3245355618349065, + "flos": 17685067513920.0, + "grad_norm": 5.912222612580071, + "learning_rate": 3.1570929971433897e-06, + "loss": 0.7847, + "num_input_tokens_seen": 58109300, + "step": 2699 + }, + { + "epoch": 0.3246558047255456, + "flos": 23407285827360.0, + "grad_norm": 3.0119635139879644, + "learning_rate": 3.1564575416351504e-06, + "loss": 0.8383, + "num_input_tokens_seen": 58128000, + "step": 2700 + }, + { + "epoch": 0.32477604761618467, + "flos": 21729736661760.0, + "grad_norm": 2.257338107736987, + "learning_rate": 3.155821910693221e-06, + "loss": 0.7417, + "num_input_tokens_seen": 58147415, + "step": 2701 + }, + { + "epoch": 0.3248962905068238, + "flos": 19800062048640.0, + "grad_norm": 2.0272629561890527, + "learning_rate": 3.1551861044140275e-06, + "loss": 0.8598, + "num_input_tokens_seen": 58167050, + "step": 2702 + }, + { + "epoch": 0.3250165333974629, + "flos": 23913841247520.0, + "grad_norm": 1.954405830025458, + "learning_rate": 3.15455012289402e-06, + "loss": 0.7794, + "num_input_tokens_seen": 58186695, + "step": 2703 + }, + { + "epoch": 0.32513677628810195, + "flos": 23955068244960.0, + "grad_norm": 2.8061443147532823, + "learning_rate": 3.153913966229677e-06, + "loss": 0.8444, + "num_input_tokens_seen": 58205815, + "step": 2704 + }, + { + "epoch": 0.32525701917874106, + "flos": 70554950134080.0, + "grad_norm": 0.6497435458157382, + "learning_rate": 3.1532776345175027e-06, + "loss": 0.502, + "num_input_tokens_seen": 58270960, + "step": 2705 + }, + { + "epoch": 0.32537726206938017, + "flos": 19654593843360.0, + "grad_norm": 4.051569862815089, + "learning_rate": 3.1526411278540285e-06, + "loss": 0.7876, + "num_input_tokens_seen": 58289390, + "step": 2706 + }, + { + "epoch": 0.3254975049600192, + "flos": 28719495788160.0, + "grad_norm": 2.1669327743657867, + "learning_rate": 3.1520044463358116e-06, + "loss": 0.8106, + "num_input_tokens_seen": 58306120, + "step": 2707 + }, + { + "epoch": 0.32561774785065833, + "flos": 18850411455840.0, + "grad_norm": 1.6603427198816483, + "learning_rate": 3.151367590059436e-06, + "loss": 0.8029, + "num_input_tokens_seen": 58324545, + "step": 2708 + }, + { + "epoch": 0.32573799074129745, + "flos": 23079062413920.0, + "grad_norm": 2.7788675798326765, + "learning_rate": 3.1507305591215117e-06, + "loss": 0.8679, + "num_input_tokens_seen": 58342455, + "step": 2709 + }, + { + "epoch": 0.3258582336319365, + "flos": 71135893711200.0, + "grad_norm": 0.6774886507582843, + "learning_rate": 3.150093353618677e-06, + "loss": 0.5566, + "num_input_tokens_seen": 58407185, + "step": 2710 + }, + { + "epoch": 0.3259784765225756, + "flos": 22423953698400.0, + "grad_norm": 2.6784321594363067, + "learning_rate": 3.149455973647596e-06, + "loss": 0.8921, + "num_input_tokens_seen": 58425245, + "step": 2711 + }, + { + "epoch": 0.32609871941321467, + "flos": 20455579631520.0, + "grad_norm": 2.1536681709530305, + "learning_rate": 3.1488184193049563e-06, + "loss": 0.7681, + "num_input_tokens_seen": 58444780, + "step": 2712 + }, + { + "epoch": 0.3262189623038538, + "flos": 22384696698240.0, + "grad_norm": 2.6919180462337926, + "learning_rate": 3.1481806906874767e-06, + "loss": 0.7251, + "num_input_tokens_seen": 58466090, + "step": 2713 + }, + { + "epoch": 0.3263392051944929, + "flos": 20893656886560.0, + "grad_norm": 1.8095142501982493, + "learning_rate": 3.147542787891899e-06, + "loss": 0.8806, + "num_input_tokens_seen": 58485435, + "step": 2714 + }, + { + "epoch": 0.32645944808513194, + "flos": 23990831287680.0, + "grad_norm": 2.350202705893486, + "learning_rate": 3.1469047110149926e-06, + "loss": 0.7527, + "num_input_tokens_seen": 58504175, + "step": 2715 + }, + { + "epoch": 0.32657969097577105, + "flos": 21002395635360.0, + "grad_norm": 2.2508299674336496, + "learning_rate": 3.146266460153554e-06, + "loss": 0.8562, + "num_input_tokens_seen": 58523885, + "step": 2716 + }, + { + "epoch": 0.32669993386641016, + "flos": 22677788954880.0, + "grad_norm": 1.9274474203852123, + "learning_rate": 3.145628035404404e-06, + "loss": 0.8044, + "num_input_tokens_seen": 58543085, + "step": 2717 + }, + { + "epoch": 0.3268201767570492, + "flos": 72002421253440.0, + "grad_norm": 0.8736632899753133, + "learning_rate": 3.1449894368643922e-06, + "loss": 0.5738, + "num_input_tokens_seen": 58605400, + "step": 2718 + }, + { + "epoch": 0.32694041964768833, + "flos": 24500062930560.0, + "grad_norm": 1.6367681309208686, + "learning_rate": 3.1443506646303934e-06, + "loss": 0.7177, + "num_input_tokens_seen": 58626700, + "step": 2719 + }, + { + "epoch": 0.32706066253832744, + "flos": 33135325155840.0, + "grad_norm": 2.33749490829158, + "learning_rate": 3.1437117187993086e-06, + "loss": 0.6703, + "num_input_tokens_seen": 58649020, + "step": 2720 + }, + { + "epoch": 0.3271809054289665, + "flos": 24027226216320.0, + "grad_norm": 1.7622630426728112, + "learning_rate": 3.143072599468065e-06, + "loss": 0.8018, + "num_input_tokens_seen": 58668845, + "step": 2721 + }, + { + "epoch": 0.3273011483196056, + "flos": 38198680608000.0, + "grad_norm": 2.0334273186313423, + "learning_rate": 3.1424333067336174e-06, + "loss": 0.7603, + "num_input_tokens_seen": 58691610, + "step": 2722 + }, + { + "epoch": 0.3274213912102447, + "flos": 29012513705280.0, + "grad_norm": 2.168713361007377, + "learning_rate": 3.141793840692945e-06, + "loss": 0.7817, + "num_input_tokens_seen": 58712360, + "step": 2723 + }, + { + "epoch": 0.32754163410088377, + "flos": 29091473742720.0, + "grad_norm": 2.9374727288461493, + "learning_rate": 3.1411542014430553e-06, + "loss": 0.6129, + "num_input_tokens_seen": 58733440, + "step": 2724 + }, + { + "epoch": 0.3276618769915229, + "flos": 20601642552960.0, + "grad_norm": 3.7590529204771452, + "learning_rate": 3.1405143890809804e-06, + "loss": 0.8198, + "num_input_tokens_seen": 58752735, + "step": 2725 + }, + { + "epoch": 0.327782119882162, + "flos": 18630740942400.0, + "grad_norm": 2.096201130906426, + "learning_rate": 3.1398744037037796e-06, + "loss": 0.7025, + "num_input_tokens_seen": 58772070, + "step": 2726 + }, + { + "epoch": 0.32790236277280105, + "flos": 21764570460480.0, + "grad_norm": 2.209914762829591, + "learning_rate": 3.139234245408538e-06, + "loss": 0.8447, + "num_input_tokens_seen": 58790950, + "step": 2727 + }, + { + "epoch": 0.32802260566344016, + "flos": 23298063871680.0, + "grad_norm": 1.5858486507689666, + "learning_rate": 3.1385939142923666e-06, + "loss": 0.7632, + "num_input_tokens_seen": 58813340, + "step": 2728 + }, + { + "epoch": 0.3281428485540792, + "flos": 24172359893760.0, + "grad_norm": 2.5424949567926705, + "learning_rate": 3.137953410452405e-06, + "loss": 0.7816, + "num_input_tokens_seen": 58833610, + "step": 2729 + }, + { + "epoch": 0.3282630914447183, + "flos": 34079809152000.0, + "grad_norm": 2.352975009226186, + "learning_rate": 3.1373127339858146e-06, + "loss": 0.7424, + "num_input_tokens_seen": 58856210, + "step": 2730 + }, + { + "epoch": 0.32838333433535744, + "flos": 27561176930880.0, + "grad_norm": 2.179581178102453, + "learning_rate": 3.136671884989787e-06, + "loss": 0.7453, + "num_input_tokens_seen": 58876120, + "step": 2731 + }, + { + "epoch": 0.3285035772259965, + "flos": 12332150932320.0, + "grad_norm": 2.860399239957717, + "learning_rate": 3.1360308635615383e-06, + "loss": 0.8729, + "num_input_tokens_seen": 58894660, + "step": 2732 + }, + { + "epoch": 0.3286238201166356, + "flos": 24281284491360.0, + "grad_norm": 2.2656546755195537, + "learning_rate": 3.135389669798311e-06, + "loss": 0.7891, + "num_input_tokens_seen": 58912720, + "step": 2733 + }, + { + "epoch": 0.3287440630072747, + "flos": 21362107569120.0, + "grad_norm": 2.0660431223723763, + "learning_rate": 3.134748303797373e-06, + "loss": 0.8018, + "num_input_tokens_seen": 58930090, + "step": 2734 + }, + { + "epoch": 0.32886430589791377, + "flos": 23698593935520.0, + "grad_norm": 2.1107284159990733, + "learning_rate": 3.1341067656560203e-06, + "loss": 0.8126, + "num_input_tokens_seen": 58947935, + "step": 2735 + }, + { + "epoch": 0.3289845487885529, + "flos": 22386666695520.0, + "grad_norm": 3.5798873892180496, + "learning_rate": 3.133465055471572e-06, + "loss": 0.8632, + "num_input_tokens_seen": 58967720, + "step": 2736 + }, + { + "epoch": 0.329104791679192, + "flos": 19654556673600.0, + "grad_norm": 3.1387005883584718, + "learning_rate": 3.1328231733413767e-06, + "loss": 0.6633, + "num_input_tokens_seen": 58984360, + "step": 2737 + }, + { + "epoch": 0.32922503456983104, + "flos": 15974543037600.0, + "grad_norm": 3.6911329057195363, + "learning_rate": 3.1321811193628067e-06, + "loss": 0.9116, + "num_input_tokens_seen": 59002505, + "step": 2738 + }, + { + "epoch": 0.32934527746047015, + "flos": 26795619657600.0, + "grad_norm": 2.054440675712188, + "learning_rate": 3.131538893633261e-06, + "loss": 0.6982, + "num_input_tokens_seen": 59022065, + "step": 2739 + }, + { + "epoch": 0.32946552035110926, + "flos": 23370296182560.0, + "grad_norm": 3.1697189444729457, + "learning_rate": 3.130896496250165e-06, + "loss": 0.7784, + "num_input_tokens_seen": 59041890, + "step": 2740 + }, + { + "epoch": 0.3295857632417483, + "flos": 14152417740960.0, + "grad_norm": 2.3091682971336898, + "learning_rate": 3.1302539273109693e-06, + "loss": 0.8717, + "num_input_tokens_seen": 59058715, + "step": 2741 + }, + { + "epoch": 0.32970600613238743, + "flos": 22164877505760.0, + "grad_norm": 2.2155940216438337, + "learning_rate": 3.1296111869131513e-06, + "loss": 0.8091, + "num_input_tokens_seen": 59076140, + "step": 2742 + }, + { + "epoch": 0.32982624902302654, + "flos": 22022271372000.0, + "grad_norm": 14.311232864725472, + "learning_rate": 3.1289682751542153e-06, + "loss": 0.8573, + "num_input_tokens_seen": 59095700, + "step": 2743 + }, + { + "epoch": 0.3299464919136656, + "flos": 18635089804320.0, + "grad_norm": 3.2905249889278747, + "learning_rate": 3.1283251921316883e-06, + "loss": 0.7125, + "num_input_tokens_seen": 59113445, + "step": 2744 + }, + { + "epoch": 0.3300667348043047, + "flos": 13388458767360.0, + "grad_norm": 4.625159595931993, + "learning_rate": 3.1276819379431277e-06, + "loss": 0.8133, + "num_input_tokens_seen": 59131535, + "step": 2745 + }, + { + "epoch": 0.33018697769494376, + "flos": 15720261744000.0, + "grad_norm": 2.4514781502192093, + "learning_rate": 3.1270385126861134e-06, + "loss": 0.7503, + "num_input_tokens_seen": 59149520, + "step": 2746 + }, + { + "epoch": 0.3303072205855829, + "flos": 18232589743200.0, + "grad_norm": 2.050385122642467, + "learning_rate": 3.1263949164582533e-06, + "loss": 0.8232, + "num_input_tokens_seen": 59169170, + "step": 2747 + }, + { + "epoch": 0.330427463476222, + "flos": 17724213004800.0, + "grad_norm": 2.149032787734987, + "learning_rate": 3.1257511493571797e-06, + "loss": 0.7821, + "num_input_tokens_seen": 59187675, + "step": 2748 + }, + { + "epoch": 0.33054770636686104, + "flos": 27123657222240.0, + "grad_norm": 1.9939242702363653, + "learning_rate": 3.125107211480552e-06, + "loss": 0.7872, + "num_input_tokens_seen": 59207610, + "step": 2749 + }, + { + "epoch": 0.33066794925750015, + "flos": 20088396576000.0, + "grad_norm": 1.81600396408371, + "learning_rate": 3.124463102926054e-06, + "loss": 0.798, + "num_input_tokens_seen": 59226945, + "step": 2750 + }, + { + "epoch": 0.33078819214813926, + "flos": 70541308832160.0, + "grad_norm": 0.7743231586532812, + "learning_rate": 3.1238188237913984e-06, + "loss": 0.6151, + "num_input_tokens_seen": 59291485, + "step": 2751 + }, + { + "epoch": 0.3309084350387783, + "flos": 21111766270080.0, + "grad_norm": 2.233593036962384, + "learning_rate": 3.1231743741743202e-06, + "loss": 0.7633, + "num_input_tokens_seen": 59310990, + "step": 2752 + }, + { + "epoch": 0.3310286779294174, + "flos": 14189258706720.0, + "grad_norm": 3.288624931459855, + "learning_rate": 3.122529754172582e-06, + "loss": 0.84, + "num_input_tokens_seen": 59327875, + "step": 2753 + }, + { + "epoch": 0.33114892082005654, + "flos": 20748523209120.0, + "grad_norm": 2.2839114197558135, + "learning_rate": 3.1218849638839736e-06, + "loss": 0.7276, + "num_input_tokens_seen": 59347135, + "step": 2754 + }, + { + "epoch": 0.3312691637106956, + "flos": 17066428066560.0, + "grad_norm": 2.098619747698107, + "learning_rate": 3.121240003406307e-06, + "loss": 0.7853, + "num_input_tokens_seen": 59365985, + "step": 2755 + }, + { + "epoch": 0.3313894066013347, + "flos": 29413787164320.0, + "grad_norm": 2.439864454569381, + "learning_rate": 3.120594872837425e-06, + "loss": 0.7301, + "num_input_tokens_seen": 59384655, + "step": 2756 + }, + { + "epoch": 0.3315096494919738, + "flos": 61331644907520.0, + "grad_norm": 0.8341624585736889, + "learning_rate": 3.1199495722751906e-06, + "loss": 0.6236, + "num_input_tokens_seen": 59444185, + "step": 2757 + }, + { + "epoch": 0.33162989238261287, + "flos": 21622150175520.0, + "grad_norm": 1.9769821023164673, + "learning_rate": 3.1193041018174972e-06, + "loss": 0.8419, + "num_input_tokens_seen": 59464900, + "step": 2758 + }, + { + "epoch": 0.331750135273252, + "flos": 22642397609760.0, + "grad_norm": 3.0404101524232705, + "learning_rate": 3.118658461562261e-06, + "loss": 0.9459, + "num_input_tokens_seen": 59481000, + "step": 2759 + }, + { + "epoch": 0.33187037816389103, + "flos": 22714369732320.0, + "grad_norm": 1.9480441726595454, + "learning_rate": 3.118012651607426e-06, + "loss": 0.8513, + "num_input_tokens_seen": 59502605, + "step": 2760 + }, + { + "epoch": 0.33199062105453014, + "flos": 19176256004640.0, + "grad_norm": 2.707054501388239, + "learning_rate": 3.1173666720509603e-06, + "loss": 0.8331, + "num_input_tokens_seen": 59519540, + "step": 2761 + }, + { + "epoch": 0.33211086394516925, + "flos": 31531643770560.0, + "grad_norm": 2.6073036424212197, + "learning_rate": 3.116720522990859e-06, + "loss": 0.6818, + "num_input_tokens_seen": 59540415, + "step": 2762 + }, + { + "epoch": 0.3322311068358083, + "flos": 17907042552480.0, + "grad_norm": 4.788331814606467, + "learning_rate": 3.116074204525142e-06, + "loss": 0.6199, + "num_input_tokens_seen": 59558590, + "step": 2763 + }, + { + "epoch": 0.3323513497264474, + "flos": 32223444772800.0, + "grad_norm": 1.868973137537474, + "learning_rate": 3.1154277167518553e-06, + "loss": 0.8367, + "num_input_tokens_seen": 59580205, + "step": 2764 + }, + { + "epoch": 0.33247159261708653, + "flos": 52594037132160.0, + "grad_norm": 0.7868548566256393, + "learning_rate": 3.114781059769072e-06, + "loss": 0.5944, + "num_input_tokens_seen": 59631530, + "step": 2765 + }, + { + "epoch": 0.3325918355077256, + "flos": 27086333049600.0, + "grad_norm": 2.753038461734516, + "learning_rate": 3.1141342336748874e-06, + "loss": 0.6783, + "num_input_tokens_seen": 59651090, + "step": 2766 + }, + { + "epoch": 0.3327120783983647, + "flos": 23630747656320.0, + "grad_norm": 1.9424849339984687, + "learning_rate": 3.1134872385674253e-06, + "loss": 0.822, + "num_input_tokens_seen": 59675245, + "step": 2767 + }, + { + "epoch": 0.3328323212890038, + "flos": 19144061258880.0, + "grad_norm": 1.9682609313948127, + "learning_rate": 3.1128400745448353e-06, + "loss": 0.8575, + "num_input_tokens_seen": 59694625, + "step": 2768 + }, + { + "epoch": 0.33295256417964286, + "flos": 37652050452960.0, + "grad_norm": 3.2725597288131967, + "learning_rate": 3.11219274170529e-06, + "loss": 0.6332, + "num_input_tokens_seen": 59716435, + "step": 2769 + }, + { + "epoch": 0.333072807070282, + "flos": 26468028130080.0, + "grad_norm": 2.175103621371265, + "learning_rate": 3.1115452401469903e-06, + "loss": 0.8158, + "num_input_tokens_seen": 59736835, + "step": 2770 + }, + { + "epoch": 0.3331930499609211, + "flos": 21400212306720.0, + "grad_norm": 2.093117006183531, + "learning_rate": 3.1108975699681613e-06, + "loss": 0.8652, + "num_input_tokens_seen": 59754230, + "step": 2771 + }, + { + "epoch": 0.33331329285156014, + "flos": 20630194662240.0, + "grad_norm": 2.2637876229237763, + "learning_rate": 3.1102497312670542e-06, + "loss": 0.7156, + "num_input_tokens_seen": 59772235, + "step": 2772 + }, + { + "epoch": 0.33343353574219925, + "flos": 27961446806400.0, + "grad_norm": 2.1657504443102464, + "learning_rate": 3.109601724141946e-06, + "loss": 0.8053, + "num_input_tokens_seen": 59791230, + "step": 2773 + }, + { + "epoch": 0.33355377863283836, + "flos": 23730825851040.0, + "grad_norm": 2.2119724830171372, + "learning_rate": 3.108953548691138e-06, + "loss": 0.6843, + "num_input_tokens_seen": 59811315, + "step": 2774 + }, + { + "epoch": 0.3336740215234774, + "flos": 37725286347360.0, + "grad_norm": 2.8061464388287227, + "learning_rate": 3.108305205012959e-06, + "loss": 0.7236, + "num_input_tokens_seen": 59832010, + "step": 2775 + }, + { + "epoch": 0.3337942644141165, + "flos": 25482465815520.0, + "grad_norm": 2.166888543530333, + "learning_rate": 3.107656693205761e-06, + "loss": 0.8794, + "num_input_tokens_seen": 59851450, + "step": 2776 + }, + { + "epoch": 0.3339145073047556, + "flos": 25952440458240.0, + "grad_norm": 4.152596610436509, + "learning_rate": 3.107008013367924e-06, + "loss": 0.7032, + "num_input_tokens_seen": 59870245, + "step": 2777 + }, + { + "epoch": 0.3340347501953947, + "flos": 19035434019360.0, + "grad_norm": 2.416654411372538, + "learning_rate": 3.1063591655978507e-06, + "loss": 0.866, + "num_input_tokens_seen": 59886195, + "step": 2778 + }, + { + "epoch": 0.3341549930860338, + "flos": 18083813429280.0, + "grad_norm": 2.5054739152398913, + "learning_rate": 3.105710149993972e-06, + "loss": 0.7988, + "num_input_tokens_seen": 59905525, + "step": 2779 + }, + { + "epoch": 0.33427523597667286, + "flos": 22642583458560.0, + "grad_norm": 1.8734094867409492, + "learning_rate": 3.1050609666547427e-06, + "loss": 0.852, + "num_input_tokens_seen": 59925085, + "step": 2780 + }, + { + "epoch": 0.33439547886731197, + "flos": 22606002681120.0, + "grad_norm": 2.0625649644272808, + "learning_rate": 3.104411615678644e-06, + "loss": 0.7717, + "num_input_tokens_seen": 59943935, + "step": 2781 + }, + { + "epoch": 0.3345157217579511, + "flos": 24061874166240.0, + "grad_norm": 2.6085366369931715, + "learning_rate": 3.1037620971641803e-06, + "loss": 0.7327, + "num_input_tokens_seen": 59963725, + "step": 2782 + }, + { + "epoch": 0.33463596464859013, + "flos": 18962458313280.0, + "grad_norm": 12.413098191582439, + "learning_rate": 3.1031124112098844e-06, + "loss": 0.6493, + "num_input_tokens_seen": 59981935, + "step": 2783 + }, + { + "epoch": 0.33475620753922924, + "flos": 20346543524640.0, + "grad_norm": 2.6340331012960227, + "learning_rate": 3.1024625579143127e-06, + "loss": 0.7197, + "num_input_tokens_seen": 59999455, + "step": 2784 + }, + { + "epoch": 0.33487645042986836, + "flos": 18156566116800.0, + "grad_norm": 2.3137170836478824, + "learning_rate": 3.101812537376048e-06, + "loss": 0.7267, + "num_input_tokens_seen": 60018475, + "step": 2785 + }, + { + "epoch": 0.3349966933205074, + "flos": 25811804321760.0, + "grad_norm": 2.655806380981897, + "learning_rate": 3.1011623496936973e-06, + "loss": 0.8458, + "num_input_tokens_seen": 60036770, + "step": 2786 + }, + { + "epoch": 0.3351169362111465, + "flos": 28071151968960.0, + "grad_norm": 2.095551811507975, + "learning_rate": 3.100511994965893e-06, + "loss": 0.6994, + "num_input_tokens_seen": 60056725, + "step": 2787 + }, + { + "epoch": 0.33523717910178563, + "flos": 22640873649600.0, + "grad_norm": 1.7963388057622316, + "learning_rate": 3.0998614732912947e-06, + "loss": 0.8432, + "num_input_tokens_seen": 60076100, + "step": 2788 + }, + { + "epoch": 0.3353574219924247, + "flos": 15647620565760.0, + "grad_norm": 2.3108513600026557, + "learning_rate": 3.0992107847685855e-06, + "loss": 0.6786, + "num_input_tokens_seen": 60093815, + "step": 2789 + }, + { + "epoch": 0.3354776648830638, + "flos": 24755161958880.0, + "grad_norm": 2.544383134709442, + "learning_rate": 3.0985599294964736e-06, + "loss": 0.7962, + "num_input_tokens_seen": 60113370, + "step": 2790 + }, + { + "epoch": 0.33559790777370285, + "flos": 28653693845760.0, + "grad_norm": 2.600975909844947, + "learning_rate": 3.097908907573695e-06, + "loss": 0.6993, + "num_input_tokens_seen": 60131350, + "step": 2791 + }, + { + "epoch": 0.33571815066434196, + "flos": 22203985826880.0, + "grad_norm": 2.3189406529934553, + "learning_rate": 3.0972577190990067e-06, + "loss": 0.8939, + "num_input_tokens_seen": 60149830, + "step": 2792 + }, + { + "epoch": 0.3358383935549811, + "flos": 23808187588800.0, + "grad_norm": 2.432977246223164, + "learning_rate": 3.096606364171196e-06, + "loss": 0.8044, + "num_input_tokens_seen": 60169620, + "step": 2793 + }, + { + "epoch": 0.33595863644562013, + "flos": 22235288498400.0, + "grad_norm": 2.220775700358708, + "learning_rate": 3.0959548428890703e-06, + "loss": 0.8487, + "num_input_tokens_seen": 60188490, + "step": 2794 + }, + { + "epoch": 0.33607887933625924, + "flos": 20091072798720.0, + "grad_norm": 1.757123210123701, + "learning_rate": 3.095303155351468e-06, + "loss": 0.8386, + "num_input_tokens_seen": 60208695, + "step": 2795 + }, + { + "epoch": 0.33619912222689835, + "flos": 19290607387200.0, + "grad_norm": 2.4385481317090343, + "learning_rate": 3.0946513016572464e-06, + "loss": 0.7849, + "num_input_tokens_seen": 60227600, + "step": 2796 + }, + { + "epoch": 0.3363193651175374, + "flos": 16776829767360.0, + "grad_norm": 2.359563352632832, + "learning_rate": 3.0939992819052938e-06, + "loss": 0.7744, + "num_input_tokens_seen": 60245110, + "step": 2797 + }, + { + "epoch": 0.3364396080081765, + "flos": 23516619292320.0, + "grad_norm": 2.254500233149141, + "learning_rate": 3.0933470961945193e-06, + "loss": 0.814, + "num_input_tokens_seen": 60263405, + "step": 2798 + }, + { + "epoch": 0.3365598508988156, + "flos": 27997693056000.0, + "grad_norm": 2.1984660349433, + "learning_rate": 3.0926947446238597e-06, + "loss": 0.681, + "num_input_tokens_seen": 60282975, + "step": 2799 + }, + { + "epoch": 0.3366800937894547, + "flos": 16958655731520.0, + "grad_norm": 3.5705050827607723, + "learning_rate": 3.092042227292276e-06, + "loss": 0.8283, + "num_input_tokens_seen": 60299810, + "step": 2800 + }, + { + "epoch": 0.3368003366800938, + "flos": 23881126125120.0, + "grad_norm": 2.3142153589631915, + "learning_rate": 3.0913895442987557e-06, + "loss": 0.8823, + "num_input_tokens_seen": 60320775, + "step": 2801 + }, + { + "epoch": 0.3369205795707329, + "flos": 24755756675040.0, + "grad_norm": 1.8235642363789764, + "learning_rate": 3.090736695742308e-06, + "loss": 0.8559, + "num_input_tokens_seen": 60341905, + "step": 2802 + }, + { + "epoch": 0.33704082246137196, + "flos": 17905592931840.0, + "grad_norm": 3.07455438626027, + "learning_rate": 3.0900836817219713e-06, + "loss": 0.5168, + "num_input_tokens_seen": 60358335, + "step": 2803 + }, + { + "epoch": 0.33716106535201107, + "flos": 21256082212800.0, + "grad_norm": 2.0689888807599948, + "learning_rate": 3.089430502336807e-06, + "loss": 0.8361, + "num_input_tokens_seen": 60378305, + "step": 2804 + }, + { + "epoch": 0.3372813082426502, + "flos": 18376682667360.0, + "grad_norm": 2.9110107805792196, + "learning_rate": 3.088777157685902e-06, + "loss": 0.914, + "num_input_tokens_seen": 60393895, + "step": 2805 + }, + { + "epoch": 0.33740155113328923, + "flos": 17176950963840.0, + "grad_norm": 2.2130452276297725, + "learning_rate": 3.088123647868367e-06, + "loss": 0.8587, + "num_input_tokens_seen": 60410445, + "step": 2806 + }, + { + "epoch": 0.33752179402392835, + "flos": 29017234264800.0, + "grad_norm": 4.056052623113971, + "learning_rate": 3.0874699729833405e-06, + "loss": 0.8132, + "num_input_tokens_seen": 60431815, + "step": 2807 + }, + { + "epoch": 0.3376420369145674, + "flos": 25044760258080.0, + "grad_norm": 1.823584370716859, + "learning_rate": 3.086816133129983e-06, + "loss": 0.7966, + "num_input_tokens_seen": 60452075, + "step": 2808 + }, + { + "epoch": 0.3377622798052065, + "flos": 27451211580000.0, + "grad_norm": 2.0174483694208405, + "learning_rate": 3.0861621284074826e-06, + "loss": 0.7623, + "num_input_tokens_seen": 60472600, + "step": 2809 + }, + { + "epoch": 0.3378825226958456, + "flos": 21943608692640.0, + "grad_norm": 1.5690685298388234, + "learning_rate": 3.085507958915051e-06, + "loss": 0.7265, + "num_input_tokens_seen": 60491030, + "step": 2810 + }, + { + "epoch": 0.3380027655864847, + "flos": 42461905176480.0, + "grad_norm": 2.047503425865428, + "learning_rate": 3.084853624751925e-06, + "loss": 0.7131, + "num_input_tokens_seen": 60512615, + "step": 2811 + }, + { + "epoch": 0.3381230084771238, + "flos": 26686806569280.0, + "grad_norm": 2.7565356532572225, + "learning_rate": 3.0841991260173668e-06, + "loss": 0.8559, + "num_input_tokens_seen": 60532520, + "step": 2812 + }, + { + "epoch": 0.3382432513677629, + "flos": 22677900464160.0, + "grad_norm": 2.131445366191697, + "learning_rate": 3.0835444628106634e-06, + "loss": 0.8088, + "num_input_tokens_seen": 60551500, + "step": 2813 + }, + { + "epoch": 0.33836349425840195, + "flos": 22090638027840.0, + "grad_norm": 2.008629421182973, + "learning_rate": 3.082889635231126e-06, + "loss": 0.8282, + "num_input_tokens_seen": 60570160, + "step": 2814 + }, + { + "epoch": 0.33848373714904106, + "flos": 27268939578720.0, + "grad_norm": 2.698565921698494, + "learning_rate": 3.0822346433780925e-06, + "loss": 0.7659, + "num_input_tokens_seen": 60590885, + "step": 2815 + }, + { + "epoch": 0.3386039800396802, + "flos": 25811692812480.0, + "grad_norm": 2.4530529813460555, + "learning_rate": 3.0815794873509237e-06, + "loss": 0.8759, + "num_input_tokens_seen": 60609170, + "step": 2816 + }, + { + "epoch": 0.33872422293031923, + "flos": 18853793904000.0, + "grad_norm": 1.9728268389072738, + "learning_rate": 3.0809241672490066e-06, + "loss": 0.729, + "num_input_tokens_seen": 60627340, + "step": 2817 + }, + { + "epoch": 0.33884446582095834, + "flos": 23113821873120.0, + "grad_norm": 1.8912086531713845, + "learning_rate": 3.080268683171753e-06, + "loss": 0.8475, + "num_input_tokens_seen": 60647630, + "step": 2818 + }, + { + "epoch": 0.33896470871159745, + "flos": 15974654546880.0, + "grad_norm": 2.719474871236516, + "learning_rate": 3.0796130352185985e-06, + "loss": 0.8921, + "num_input_tokens_seen": 60664485, + "step": 2819 + }, + { + "epoch": 0.3390849516022365, + "flos": 34446137303040.0, + "grad_norm": 2.1414356784962556, + "learning_rate": 3.0789572234890057e-06, + "loss": 0.6663, + "num_input_tokens_seen": 60686125, + "step": 2820 + }, + { + "epoch": 0.3392051944928756, + "flos": 16157372585280.0, + "grad_norm": 2.0149679369777425, + "learning_rate": 3.0783012480824596e-06, + "loss": 0.7746, + "num_input_tokens_seen": 60705270, + "step": 2821 + }, + { + "epoch": 0.33932543738351467, + "flos": 17062190713920.0, + "grad_norm": 2.5878881375282856, + "learning_rate": 3.077645109098471e-06, + "loss": 0.7469, + "num_input_tokens_seen": 60721540, + "step": 2822 + }, + { + "epoch": 0.3394456802741538, + "flos": 22095358587360.0, + "grad_norm": 3.8427615251562504, + "learning_rate": 3.076988806636577e-06, + "loss": 0.7223, + "num_input_tokens_seen": 60739300, + "step": 2823 + }, + { + "epoch": 0.3395659231647929, + "flos": 25190711670240.0, + "grad_norm": 2.1385579396734453, + "learning_rate": 3.0763323407963377e-06, + "loss": 0.8913, + "num_input_tokens_seen": 60758910, + "step": 2824 + }, + { + "epoch": 0.33968616605543195, + "flos": 29054744286240.0, + "grad_norm": 3.3066471142600973, + "learning_rate": 3.075675711677337e-06, + "loss": 0.7987, + "num_input_tokens_seen": 60779005, + "step": 2825 + }, + { + "epoch": 0.33980640894607106, + "flos": 21403371736320.0, + "grad_norm": 2.1442674151444234, + "learning_rate": 3.0750189193791865e-06, + "loss": 0.7799, + "num_input_tokens_seen": 60798590, + "step": 2826 + }, + { + "epoch": 0.33992665183671017, + "flos": 32443375474560.0, + "grad_norm": 4.742963900160694, + "learning_rate": 3.0743619640015203e-06, + "loss": 0.7041, + "num_input_tokens_seen": 60818840, + "step": 2827 + }, + { + "epoch": 0.3400468947273492, + "flos": 17031185400480.0, + "grad_norm": 2.1537474331920734, + "learning_rate": 3.073704845643999e-06, + "loss": 0.925, + "num_input_tokens_seen": 60835965, + "step": 2828 + }, + { + "epoch": 0.34016713761798834, + "flos": 16849024908480.0, + "grad_norm": 3.380788431023781, + "learning_rate": 3.0730475644063063e-06, + "loss": 0.7754, + "num_input_tokens_seen": 60851065, + "step": 2829 + }, + { + "epoch": 0.34028738050862745, + "flos": 21875836752960.0, + "grad_norm": 1.8164552087235932, + "learning_rate": 3.072390120388151e-06, + "loss": 0.6492, + "num_input_tokens_seen": 60869390, + "step": 2830 + }, + { + "epoch": 0.3404076233992665, + "flos": 22713477658080.0, + "grad_norm": 2.366732109758942, + "learning_rate": 3.071732513689267e-06, + "loss": 0.7128, + "num_input_tokens_seen": 60887925, + "step": 2831 + }, + { + "epoch": 0.3405278662899056, + "flos": 17028025970880.0, + "grad_norm": 2.854825929856795, + "learning_rate": 3.0710747444094134e-06, + "loss": 0.6761, + "num_input_tokens_seen": 60905955, + "step": 2832 + }, + { + "epoch": 0.3406481091805447, + "flos": 42751354796640.0, + "grad_norm": 2.1607690531737918, + "learning_rate": 3.070416812648372e-06, + "loss": 0.6488, + "num_input_tokens_seen": 60929405, + "step": 2833 + }, + { + "epoch": 0.3407683520711838, + "flos": 26723127158400.0, + "grad_norm": 11.048479581463312, + "learning_rate": 3.069758718505951e-06, + "loss": 0.6552, + "num_input_tokens_seen": 60951145, + "step": 2834 + }, + { + "epoch": 0.3408885949618229, + "flos": 28726372193760.0, + "grad_norm": 1.9885591863543508, + "learning_rate": 3.0691004620819836e-06, + "loss": 0.8042, + "num_input_tokens_seen": 60971275, + "step": 2835 + }, + { + "epoch": 0.341008837852462, + "flos": 63485673423840.0, + "grad_norm": 0.8128941386933761, + "learning_rate": 3.0684420434763254e-06, + "loss": 0.6013, + "num_input_tokens_seen": 61037460, + "step": 2836 + }, + { + "epoch": 0.34112908074310105, + "flos": 20782427763840.0, + "grad_norm": 2.8848909158430023, + "learning_rate": 3.06778346278886e-06, + "loss": 0.7714, + "num_input_tokens_seen": 61056295, + "step": 2837 + }, + { + "epoch": 0.34124932363374016, + "flos": 24940816408320.0, + "grad_norm": 2.0653885929717988, + "learning_rate": 3.0671247201194906e-06, + "loss": 0.7893, + "num_input_tokens_seen": 61078430, + "step": 2838 + }, + { + "epoch": 0.3413695665243792, + "flos": 28362237058560.0, + "grad_norm": 4.1588539628339545, + "learning_rate": 3.066465815568151e-06, + "loss": 0.7553, + "num_input_tokens_seen": 61099260, + "step": 2839 + }, + { + "epoch": 0.34148980941501833, + "flos": 25265806052640.0, + "grad_norm": 3.1296714961327896, + "learning_rate": 3.0658067492347947e-06, + "loss": 0.6906, + "num_input_tokens_seen": 61121900, + "step": 2840 + }, + { + "epoch": 0.34161005230565744, + "flos": 17505174377280.0, + "grad_norm": 3.9465653951212754, + "learning_rate": 3.065147521219402e-06, + "loss": 0.6709, + "num_input_tokens_seen": 61141155, + "step": 2841 + }, + { + "epoch": 0.3417302951962965, + "flos": 43587025704480.0, + "grad_norm": 2.0920925127365053, + "learning_rate": 3.064488131621977e-06, + "loss": 0.7441, + "num_input_tokens_seen": 61164480, + "step": 2842 + }, + { + "epoch": 0.3418505380869356, + "flos": 30837798431520.0, + "grad_norm": 2.052744712041349, + "learning_rate": 3.063828580542549e-06, + "loss": 0.7416, + "num_input_tokens_seen": 61185675, + "step": 2843 + }, + { + "epoch": 0.3419707809775747, + "flos": 19436149932000.0, + "grad_norm": 2.38275716748639, + "learning_rate": 3.0631688680811706e-06, + "loss": 0.733, + "num_input_tokens_seen": 61205300, + "step": 2844 + }, + { + "epoch": 0.3420910238682138, + "flos": 28686483307680.0, + "grad_norm": 7.098426117867966, + "learning_rate": 3.062508994337921e-06, + "loss": 0.7552, + "num_input_tokens_seen": 61224905, + "step": 2845 + }, + { + "epoch": 0.3422112667588529, + "flos": 21366716619360.0, + "grad_norm": 2.8202933947478757, + "learning_rate": 3.0618489594129013e-06, + "loss": 0.7949, + "num_input_tokens_seen": 61243045, + "step": 2846 + }, + { + "epoch": 0.342331509649492, + "flos": 13861853028000.0, + "grad_norm": 2.3129224649076146, + "learning_rate": 3.061188763406239e-06, + "loss": 0.7137, + "num_input_tokens_seen": 61259190, + "step": 2847 + }, + { + "epoch": 0.34245175254013105, + "flos": 28580383611840.0, + "grad_norm": 2.2744436233453595, + "learning_rate": 3.060528406418085e-06, + "loss": 0.8172, + "num_input_tokens_seen": 61279040, + "step": 2848 + }, + { + "epoch": 0.34257199543077016, + "flos": 34077839154720.0, + "grad_norm": 1.7848453042238046, + "learning_rate": 3.0598678885486145e-06, + "loss": 0.6171, + "num_input_tokens_seen": 61301580, + "step": 2849 + }, + { + "epoch": 0.34269223832140927, + "flos": 19945530253920.0, + "grad_norm": 2.3478831151434028, + "learning_rate": 3.0592072098980282e-06, + "loss": 0.7466, + "num_input_tokens_seen": 61321240, + "step": 2850 + }, + { + "epoch": 0.3428124812120483, + "flos": 27196632928320.0, + "grad_norm": 3.6564936108412724, + "learning_rate": 3.0585463705665514e-06, + "loss": 0.7327, + "num_input_tokens_seen": 61341615, + "step": 2851 + }, + { + "epoch": 0.34293272410268744, + "flos": 24535677294240.0, + "grad_norm": 2.777618695048294, + "learning_rate": 3.0578853706544304e-06, + "loss": 0.7028, + "num_input_tokens_seen": 61359005, + "step": 2852 + }, + { + "epoch": 0.34305296699332655, + "flos": 21476161593600.0, + "grad_norm": 2.3239330468767743, + "learning_rate": 3.0572242102619404e-06, + "loss": 0.6613, + "num_input_tokens_seen": 61376160, + "step": 2853 + }, + { + "epoch": 0.3431732098839656, + "flos": 24022096789440.0, + "grad_norm": 2.1267096430024592, + "learning_rate": 3.0565628894893784e-06, + "loss": 0.8063, + "num_input_tokens_seen": 61396795, + "step": 2854 + }, + { + "epoch": 0.3432934527746047, + "flos": 16777052785920.0, + "grad_norm": 1.9119049368213934, + "learning_rate": 3.0559014084370655e-06, + "loss": 0.7459, + "num_input_tokens_seen": 61415320, + "step": 2855 + }, + { + "epoch": 0.34341369566524377, + "flos": 23406170734560.0, + "grad_norm": 2.0411226235410393, + "learning_rate": 3.055239767205349e-06, + "loss": 0.7876, + "num_input_tokens_seen": 61437070, + "step": 2856 + }, + { + "epoch": 0.3435339385558829, + "flos": 17242678566720.0, + "grad_norm": 2.0535335418807734, + "learning_rate": 3.054577965894599e-06, + "loss": 0.7818, + "num_input_tokens_seen": 61453215, + "step": 2857 + }, + { + "epoch": 0.343654181446522, + "flos": 22167628068000.0, + "grad_norm": 1.7536813298632152, + "learning_rate": 3.0539160046052094e-06, + "loss": 0.7017, + "num_input_tokens_seen": 61472915, + "step": 2858 + }, + { + "epoch": 0.34377442433716104, + "flos": 19873446622080.0, + "grad_norm": 2.5930173827389353, + "learning_rate": 3.0532538834376003e-06, + "loss": 0.7038, + "num_input_tokens_seen": 61492475, + "step": 2859 + }, + { + "epoch": 0.34389466722780015, + "flos": 22165881089280.0, + "grad_norm": 2.498644365880182, + "learning_rate": 3.0525916024922143e-06, + "loss": 0.7827, + "num_input_tokens_seen": 61511660, + "step": 2860 + }, + { + "epoch": 0.34401491011843927, + "flos": 18598062989760.0, + "grad_norm": 4.5450123120971035, + "learning_rate": 3.0519291618695193e-06, + "loss": 0.8461, + "num_input_tokens_seen": 61529060, + "step": 2861 + }, + { + "epoch": 0.3441351530090783, + "flos": 17832803074560.0, + "grad_norm": 1.731896012067584, + "learning_rate": 3.0512665616700065e-06, + "loss": 0.7593, + "num_input_tokens_seen": 61548125, + "step": 2862 + }, + { + "epoch": 0.34425539589971743, + "flos": 23079731469600.0, + "grad_norm": 2.2213269165323424, + "learning_rate": 3.0506038019941933e-06, + "loss": 0.8915, + "num_input_tokens_seen": 61568135, + "step": 2863 + }, + { + "epoch": 0.34437563879035654, + "flos": 21876394299360.0, + "grad_norm": 2.9722592142562125, + "learning_rate": 3.049940882942617e-06, + "loss": 0.6742, + "num_input_tokens_seen": 61586920, + "step": 2864 + }, + { + "epoch": 0.3444958816809956, + "flos": 23043968426880.0, + "grad_norm": 2.187368770477791, + "learning_rate": 3.0492778046158448e-06, + "loss": 0.8042, + "num_input_tokens_seen": 61608340, + "step": 2865 + }, + { + "epoch": 0.3446161245716347, + "flos": 21877360713120.0, + "grad_norm": 2.163731408526911, + "learning_rate": 3.0486145671144633e-06, + "loss": 0.7713, + "num_input_tokens_seen": 61628130, + "step": 2866 + }, + { + "epoch": 0.3447363674622738, + "flos": 25076657645760.0, + "grad_norm": 3.2564420676376806, + "learning_rate": 3.047951170539086e-06, + "loss": 0.7661, + "num_input_tokens_seen": 61647755, + "step": 2867 + }, + { + "epoch": 0.3448566103529129, + "flos": 11967867118080.0, + "grad_norm": 2.2496822980589295, + "learning_rate": 3.047287614990349e-06, + "loss": 0.8402, + "num_input_tokens_seen": 61665675, + "step": 2868 + }, + { + "epoch": 0.344976853243552, + "flos": 40129990690560.0, + "grad_norm": 2.6904816167288814, + "learning_rate": 3.046623900568914e-06, + "loss": 0.6163, + "num_input_tokens_seen": 61688240, + "step": 2869 + }, + { + "epoch": 0.34509709613419104, + "flos": 28682171615520.0, + "grad_norm": 4.181452013276137, + "learning_rate": 3.045960027375465e-06, + "loss": 0.6938, + "num_input_tokens_seen": 61706075, + "step": 2870 + }, + { + "epoch": 0.34521733902483015, + "flos": 29924431258080.0, + "grad_norm": 6.8084021326306345, + "learning_rate": 3.045295995510711e-06, + "loss": 0.8296, + "num_input_tokens_seen": 61723045, + "step": 2871 + }, + { + "epoch": 0.34533758191546926, + "flos": 27882821296800.0, + "grad_norm": 1.8833142636889386, + "learning_rate": 3.0446318050753865e-06, + "loss": 0.7333, + "num_input_tokens_seen": 61743525, + "step": 2872 + }, + { + "epoch": 0.3454578248061083, + "flos": 27087002105280.0, + "grad_norm": 2.9804271530108393, + "learning_rate": 3.0439674561702474e-06, + "loss": 0.7751, + "num_input_tokens_seen": 61763025, + "step": 2873 + }, + { + "epoch": 0.3455780676967474, + "flos": 18997775318880.0, + "grad_norm": 2.5724031705584984, + "learning_rate": 3.043302948896076e-06, + "loss": 0.8811, + "num_input_tokens_seen": 61781630, + "step": 2874 + }, + { + "epoch": 0.34569831058738654, + "flos": 34446397491360.0, + "grad_norm": 2.018936040000292, + "learning_rate": 3.0426382833536756e-06, + "loss": 0.6012, + "num_input_tokens_seen": 61804985, + "step": 2875 + }, + { + "epoch": 0.3458185534780256, + "flos": 31093938213120.0, + "grad_norm": 2.8753860877817075, + "learning_rate": 3.041973459643877e-06, + "loss": 0.7761, + "num_input_tokens_seen": 61824440, + "step": 2876 + }, + { + "epoch": 0.3459387963686647, + "flos": 32405865453120.0, + "grad_norm": 13.169096195430125, + "learning_rate": 3.0413084778675334e-06, + "loss": 0.6703, + "num_input_tokens_seen": 61845690, + "step": 2877 + }, + { + "epoch": 0.3460590392593038, + "flos": 24640141520640.0, + "grad_norm": 2.1003795825681464, + "learning_rate": 3.0406433381255214e-06, + "loss": 0.8414, + "num_input_tokens_seen": 61863725, + "step": 2878 + }, + { + "epoch": 0.34617928214994287, + "flos": 18780558009600.0, + "grad_norm": 3.2845767640661054, + "learning_rate": 3.0399780405187425e-06, + "loss": 0.8249, + "num_input_tokens_seen": 61882720, + "step": 2879 + }, + { + "epoch": 0.346299525040582, + "flos": 24821855975520.0, + "grad_norm": 2.000429584143907, + "learning_rate": 3.0393125851481216e-06, + "loss": 0.7904, + "num_input_tokens_seen": 61902595, + "step": 2880 + }, + { + "epoch": 0.3464197679312211, + "flos": 16411579539360.0, + "grad_norm": 3.5306297365151087, + "learning_rate": 3.038646972114608e-06, + "loss": 0.8655, + "num_input_tokens_seen": 61918920, + "step": 2881 + }, + { + "epoch": 0.34654001082186014, + "flos": 22350494785440.0, + "grad_norm": 2.0702132291205846, + "learning_rate": 3.037981201519174e-06, + "loss": 0.6779, + "num_input_tokens_seen": 61939520, + "step": 2882 + }, + { + "epoch": 0.34666025371249926, + "flos": 19546040943360.0, + "grad_norm": 2.30871530074071, + "learning_rate": 3.0373152734628175e-06, + "loss": 0.7164, + "num_input_tokens_seen": 61957800, + "step": 2883 + }, + { + "epoch": 0.34678049660313837, + "flos": 15246012578880.0, + "grad_norm": 2.3531342889974773, + "learning_rate": 3.0366491880465584e-06, + "loss": 0.7618, + "num_input_tokens_seen": 61975820, + "step": 2884 + }, + { + "epoch": 0.3469007394937774, + "flos": 21151432137600.0, + "grad_norm": 1.8870639524716097, + "learning_rate": 3.035982945371443e-06, + "loss": 0.8214, + "num_input_tokens_seen": 61997515, + "step": 2885 + }, + { + "epoch": 0.34702098238441653, + "flos": 22343246682240.0, + "grad_norm": 2.7774519114619065, + "learning_rate": 3.035316545538537e-06, + "loss": 0.8523, + "num_input_tokens_seen": 62016310, + "step": 2886 + }, + { + "epoch": 0.3471412252750556, + "flos": 22896641733600.0, + "grad_norm": 2.145424911987953, + "learning_rate": 3.034649988648935e-06, + "loss": 0.7967, + "num_input_tokens_seen": 62036075, + "step": 2887 + }, + { + "epoch": 0.3472614681656947, + "flos": 21293815252800.0, + "grad_norm": 6.3828887001074195, + "learning_rate": 3.033983274803752e-06, + "loss": 0.8023, + "num_input_tokens_seen": 62055865, + "step": 2888 + }, + { + "epoch": 0.3473817110563338, + "flos": 23442119626080.0, + "grad_norm": 2.631792816205181, + "learning_rate": 3.0333164041041283e-06, + "loss": 0.7263, + "num_input_tokens_seen": 62072180, + "step": 2889 + }, + { + "epoch": 0.34750195394697286, + "flos": 22314099856800.0, + "grad_norm": 2.0706703884355773, + "learning_rate": 3.032649376651228e-06, + "loss": 0.7226, + "num_input_tokens_seen": 62091600, + "step": 2890 + }, + { + "epoch": 0.347622196837612, + "flos": 29053220326080.0, + "grad_norm": 2.44942433347709, + "learning_rate": 3.031982192546238e-06, + "loss": 0.758, + "num_input_tokens_seen": 62114695, + "step": 2891 + }, + { + "epoch": 0.3477424397282511, + "flos": 22423656340320.0, + "grad_norm": 2.45236806672124, + "learning_rate": 3.0313148518903696e-06, + "loss": 0.9426, + "num_input_tokens_seen": 62134520, + "step": 2892 + }, + { + "epoch": 0.34786268261889014, + "flos": 15756433654080.0, + "grad_norm": 2.63284011962168, + "learning_rate": 3.030647354784859e-06, + "loss": 0.8142, + "num_input_tokens_seen": 62151560, + "step": 2893 + }, + { + "epoch": 0.34798292550952925, + "flos": 20748002832480.0, + "grad_norm": 2.3305940784255705, + "learning_rate": 3.029979701330964e-06, + "loss": 0.774, + "num_input_tokens_seen": 62170985, + "step": 2894 + }, + { + "epoch": 0.34810316840016836, + "flos": 19909395513600.0, + "grad_norm": 2.5406690004862376, + "learning_rate": 3.029311891629966e-06, + "loss": 0.8048, + "num_input_tokens_seen": 62189440, + "step": 2895 + }, + { + "epoch": 0.3482234112908074, + "flos": 23589520658880.0, + "grad_norm": 2.1368365769615196, + "learning_rate": 3.0286439257831744e-06, + "loss": 0.7466, + "num_input_tokens_seen": 62208910, + "step": 2896 + }, + { + "epoch": 0.3483436541814465, + "flos": 23952354852480.0, + "grad_norm": 2.8564474826380644, + "learning_rate": 3.0279758038919156e-06, + "loss": 0.7145, + "num_input_tokens_seen": 62227135, + "step": 2897 + }, + { + "epoch": 0.34846389707208564, + "flos": 22606002681120.0, + "grad_norm": 2.351116657765333, + "learning_rate": 3.0273075260575455e-06, + "loss": 0.7878, + "num_input_tokens_seen": 62246035, + "step": 2898 + }, + { + "epoch": 0.3485841399627247, + "flos": 21761522540160.0, + "grad_norm": 1.9753712317291934, + "learning_rate": 3.0266390923814396e-06, + "loss": 0.8087, + "num_input_tokens_seen": 62264095, + "step": 2899 + }, + { + "epoch": 0.3487043828533638, + "flos": 16994864811360.0, + "grad_norm": 2.0778160152328113, + "learning_rate": 3.0259705029650008e-06, + "loss": 0.8194, + "num_input_tokens_seen": 62282025, + "step": 2900 + }, + { + "epoch": 0.34882462574400286, + "flos": 22568827187520.0, + "grad_norm": 1.9194387485820186, + "learning_rate": 3.025301757909652e-06, + "loss": 0.7296, + "num_input_tokens_seen": 62302220, + "step": 2901 + }, + { + "epoch": 0.34894486863464197, + "flos": 29818628920320.0, + "grad_norm": 1.773626645122316, + "learning_rate": 3.024632857316842e-06, + "loss": 0.8067, + "num_input_tokens_seen": 62323535, + "step": 2902 + }, + { + "epoch": 0.3490651115252811, + "flos": 22090489348800.0, + "grad_norm": 2.126710651963278, + "learning_rate": 3.0239638012880412e-06, + "loss": 0.7725, + "num_input_tokens_seen": 62343430, + "step": 2903 + }, + { + "epoch": 0.34918535441592014, + "flos": 12658813215840.0, + "grad_norm": 3.8041352609519765, + "learning_rate": 3.0232945899247466e-06, + "loss": 0.8117, + "num_input_tokens_seen": 62360365, + "step": 2904 + }, + { + "epoch": 0.34930559730655925, + "flos": 23152484157120.0, + "grad_norm": 2.3926396077878356, + "learning_rate": 3.022625223328476e-06, + "loss": 0.7777, + "num_input_tokens_seen": 62382105, + "step": 2905 + }, + { + "epoch": 0.34942584019719836, + "flos": 22822290746400.0, + "grad_norm": 1.5902605939637275, + "learning_rate": 3.0219557016007723e-06, + "loss": 0.6934, + "num_input_tokens_seen": 62402235, + "step": 2906 + }, + { + "epoch": 0.3495460830878374, + "flos": 24391101163200.0, + "grad_norm": 2.1062841305103976, + "learning_rate": 3.021286024843202e-06, + "loss": 0.7001, + "num_input_tokens_seen": 62420430, + "step": 2907 + }, + { + "epoch": 0.3496663259784765, + "flos": 69907875820320.0, + "grad_norm": 1.082341834708582, + "learning_rate": 3.0206161931573526e-06, + "loss": 0.648, + "num_input_tokens_seen": 62472980, + "step": 2908 + }, + { + "epoch": 0.34978656886911563, + "flos": 28651575169440.0, + "grad_norm": 2.067421560955957, + "learning_rate": 3.0199462066448388e-06, + "loss": 0.9302, + "num_input_tokens_seen": 62495175, + "step": 2909 + }, + { + "epoch": 0.3499068117597547, + "flos": 21111989288640.0, + "grad_norm": 2.0059796587900616, + "learning_rate": 3.019276065407296e-06, + "loss": 0.6918, + "num_input_tokens_seen": 62514295, + "step": 2910 + }, + { + "epoch": 0.3500270546503938, + "flos": 22750095605280.0, + "grad_norm": 4.437970848355642, + "learning_rate": 3.018605769546385e-06, + "loss": 0.8081, + "num_input_tokens_seen": 62534850, + "step": 2911 + }, + { + "epoch": 0.3501472975410329, + "flos": 22861361897760.0, + "grad_norm": 2.3435830629024843, + "learning_rate": 3.017935319163788e-06, + "loss": 0.7958, + "num_input_tokens_seen": 62553810, + "step": 2912 + }, + { + "epoch": 0.35026754043167196, + "flos": 25410121995360.0, + "grad_norm": 2.263652284683889, + "learning_rate": 3.017264714361213e-06, + "loss": 0.7055, + "num_input_tokens_seen": 62573460, + "step": 2913 + }, + { + "epoch": 0.3503877833223111, + "flos": 19545892264320.0, + "grad_norm": 2.216019145304471, + "learning_rate": 3.016593955240389e-06, + "loss": 0.824, + "num_input_tokens_seen": 62592150, + "step": 2914 + }, + { + "epoch": 0.3505080262129502, + "flos": 64979315118720.0, + "grad_norm": 0.8295314649278769, + "learning_rate": 3.015923041903071e-06, + "loss": 0.6367, + "num_input_tokens_seen": 62658275, + "step": 2915 + }, + { + "epoch": 0.35062826910358924, + "flos": 29272481972160.0, + "grad_norm": 2.5327115960704307, + "learning_rate": 3.0152519744510347e-06, + "loss": 0.8364, + "num_input_tokens_seen": 62678075, + "step": 2916 + }, + { + "epoch": 0.35074851199422835, + "flos": 23953395605760.0, + "grad_norm": 2.8248018583504986, + "learning_rate": 3.014580752986081e-06, + "loss": 0.8312, + "num_input_tokens_seen": 62696950, + "step": 2917 + }, + { + "epoch": 0.3508687548848674, + "flos": 15202629735360.0, + "grad_norm": 2.187217694185984, + "learning_rate": 3.0139093776100345e-06, + "loss": 0.7865, + "num_input_tokens_seen": 62713540, + "step": 2918 + }, + { + "epoch": 0.3509889977755065, + "flos": 21331288104480.0, + "grad_norm": 1.9085323241097152, + "learning_rate": 3.013237848424741e-06, + "loss": 0.7539, + "num_input_tokens_seen": 62732925, + "step": 2919 + }, + { + "epoch": 0.35110924066614563, + "flos": 19108112367360.0, + "grad_norm": 2.454681485834652, + "learning_rate": 3.012566165532072e-06, + "loss": 0.75, + "num_input_tokens_seen": 62751715, + "step": 2920 + }, + { + "epoch": 0.3512294835567847, + "flos": 21949481514720.0, + "grad_norm": 2.5847691780793456, + "learning_rate": 3.0118943290339207e-06, + "loss": 0.764, + "num_input_tokens_seen": 62771925, + "step": 2921 + }, + { + "epoch": 0.3513497264474238, + "flos": 17791799095680.0, + "grad_norm": 2.034129874232558, + "learning_rate": 3.011222339032204e-06, + "loss": 0.6826, + "num_input_tokens_seen": 62790075, + "step": 2922 + }, + { + "epoch": 0.3514699693380629, + "flos": 26904804443520.0, + "grad_norm": 2.0097370349566916, + "learning_rate": 3.0105501956288626e-06, + "loss": 0.6954, + "num_input_tokens_seen": 62810545, + "step": 2923 + }, + { + "epoch": 0.35159021222870196, + "flos": 15245938239360.0, + "grad_norm": 2.7352001580007617, + "learning_rate": 3.0098778989258602e-06, + "loss": 0.7229, + "num_input_tokens_seen": 62829155, + "step": 2924 + }, + { + "epoch": 0.35171045511934107, + "flos": 13964161408320.0, + "grad_norm": 3.2084858020042097, + "learning_rate": 3.009205449025183e-06, + "loss": 0.8818, + "num_input_tokens_seen": 62845350, + "step": 2925 + }, + { + "epoch": 0.3518306980099802, + "flos": 14263275166080.0, + "grad_norm": 2.4260492591416982, + "learning_rate": 3.008532846028842e-06, + "loss": 0.6281, + "num_input_tokens_seen": 62862110, + "step": 2926 + }, + { + "epoch": 0.35195094090061924, + "flos": 27014100738720.0, + "grad_norm": 7.061874767073548, + "learning_rate": 3.0078600900388694e-06, + "loss": 0.7222, + "num_input_tokens_seen": 62882415, + "step": 2927 + }, + { + "epoch": 0.35207118379125835, + "flos": 25592542675680.0, + "grad_norm": 2.4360615324946324, + "learning_rate": 3.007187181157323e-06, + "loss": 0.7373, + "num_input_tokens_seen": 62902425, + "step": 2928 + }, + { + "epoch": 0.35219142668189746, + "flos": 17979126184320.0, + "grad_norm": 3.043264587673234, + "learning_rate": 3.006514119486282e-06, + "loss": 0.684, + "num_input_tokens_seen": 62919255, + "step": 2929 + }, + { + "epoch": 0.3523116695725365, + "flos": 14008324816800.0, + "grad_norm": 1.9520316764106185, + "learning_rate": 3.005840905127849e-06, + "loss": 0.6958, + "num_input_tokens_seen": 62936160, + "step": 2930 + }, + { + "epoch": 0.3524319124631756, + "flos": 21403371736320.0, + "grad_norm": 2.592688101820725, + "learning_rate": 3.0051675381841516e-06, + "loss": 0.8667, + "num_input_tokens_seen": 62953470, + "step": 2931 + }, + { + "epoch": 0.3525521553538147, + "flos": 26285049903360.0, + "grad_norm": 1.7933684258885905, + "learning_rate": 3.0044940187573363e-06, + "loss": 0.7721, + "num_input_tokens_seen": 62974480, + "step": 2932 + }, + { + "epoch": 0.3526723982444538, + "flos": 21512556522240.0, + "grad_norm": 2.4430546193147875, + "learning_rate": 3.003820346949578e-06, + "loss": 0.6535, + "num_input_tokens_seen": 62992560, + "step": 2933 + }, + { + "epoch": 0.3527926411350929, + "flos": 23699746198080.0, + "grad_norm": 2.5201700982708983, + "learning_rate": 3.003146522863071e-06, + "loss": 0.7923, + "num_input_tokens_seen": 63013445, + "step": 2934 + }, + { + "epoch": 0.35291288402573195, + "flos": 30401728343520.0, + "grad_norm": 2.9490640157566537, + "learning_rate": 3.0024725466000345e-06, + "loss": 0.8608, + "num_input_tokens_seen": 63033400, + "step": 2935 + }, + { + "epoch": 0.35303312691637107, + "flos": 23079136753440.0, + "grad_norm": 1.8917286701655947, + "learning_rate": 3.0017984182627087e-06, + "loss": 0.7876, + "num_input_tokens_seen": 63052725, + "step": 2936 + }, + { + "epoch": 0.3531533698070102, + "flos": 21804533686080.0, + "grad_norm": 2.6065925793798175, + "learning_rate": 3.00112413795336e-06, + "loss": 0.8242, + "num_input_tokens_seen": 63072560, + "step": 2937 + }, + { + "epoch": 0.35327361269764923, + "flos": 15756619502880.0, + "grad_norm": 2.4104831665502093, + "learning_rate": 3.000449705774275e-06, + "loss": 0.8028, + "num_input_tokens_seen": 63090800, + "step": 2938 + }, + { + "epoch": 0.35339385558828834, + "flos": 22058852149440.0, + "grad_norm": 2.2668307548855884, + "learning_rate": 2.9997751218277654e-06, + "loss": 0.7159, + "num_input_tokens_seen": 63108955, + "step": 2939 + }, + { + "epoch": 0.35351409847892745, + "flos": 24130240822080.0, + "grad_norm": 2.4805250747173004, + "learning_rate": 2.999100386216166e-06, + "loss": 0.7805, + "num_input_tokens_seen": 63127895, + "step": 2940 + }, + { + "epoch": 0.3536343413695665, + "flos": 27013506022560.0, + "grad_norm": 2.061120699128742, + "learning_rate": 2.998425499041831e-06, + "loss": 0.7418, + "num_input_tokens_seen": 63148410, + "step": 2941 + }, + { + "epoch": 0.3537545842602056, + "flos": 65897217002880.0, + "grad_norm": 1.231052272430525, + "learning_rate": 2.997750460407142e-06, + "loss": 0.6452, + "num_input_tokens_seen": 63209765, + "step": 2942 + }, + { + "epoch": 0.35387482715084473, + "flos": 18410475712800.0, + "grad_norm": 2.2994381881870023, + "learning_rate": 2.997075270414501e-06, + "loss": 0.7039, + "num_input_tokens_seen": 63226980, + "step": 2943 + }, + { + "epoch": 0.3539950700414838, + "flos": 65495274488160.0, + "grad_norm": 0.7156951503436382, + "learning_rate": 2.9963999291663347e-06, + "loss": 0.5773, + "num_input_tokens_seen": 63292760, + "step": 2944 + }, + { + "epoch": 0.3541153129321229, + "flos": 20492383427520.0, + "grad_norm": 3.698813846265277, + "learning_rate": 2.9957244367650915e-06, + "loss": 0.7353, + "num_input_tokens_seen": 63309005, + "step": 2945 + }, + { + "epoch": 0.354235555822762, + "flos": 19545706415520.0, + "grad_norm": 1.9936112645588722, + "learning_rate": 2.9950487933132425e-06, + "loss": 0.8378, + "num_input_tokens_seen": 63326420, + "step": 2946 + }, + { + "epoch": 0.35435579871340106, + "flos": 20747631134880.0, + "grad_norm": 2.414911805905954, + "learning_rate": 2.994372998913283e-06, + "loss": 0.7154, + "num_input_tokens_seen": 63344925, + "step": 2947 + }, + { + "epoch": 0.35447604160404017, + "flos": 23917558223520.0, + "grad_norm": 2.62462467961846, + "learning_rate": 2.99369705366773e-06, + "loss": 0.6256, + "num_input_tokens_seen": 63365170, + "step": 2948 + }, + { + "epoch": 0.3545962844946792, + "flos": 23401970551680.0, + "grad_norm": 2.964111398505708, + "learning_rate": 2.9930209576791244e-06, + "loss": 0.8192, + "num_input_tokens_seen": 63383645, + "step": 2949 + }, + { + "epoch": 0.35471652738531834, + "flos": 22053388194720.0, + "grad_norm": 2.077800639372979, + "learning_rate": 2.9923447110500285e-06, + "loss": 0.6371, + "num_input_tokens_seen": 63402390, + "step": 2950 + }, + { + "epoch": 0.35483677027595745, + "flos": 27302583945120.0, + "grad_norm": 2.2608819180967643, + "learning_rate": 2.9916683138830295e-06, + "loss": 0.7551, + "num_input_tokens_seen": 63423055, + "step": 2951 + }, + { + "epoch": 0.3549570131665965, + "flos": 13497086006880.0, + "grad_norm": 2.6952969315673565, + "learning_rate": 2.9909917662807353e-06, + "loss": 0.8135, + "num_input_tokens_seen": 63440855, + "step": 2952 + }, + { + "epoch": 0.3550772560572356, + "flos": 20857782334560.0, + "grad_norm": 2.572815021767105, + "learning_rate": 2.9903150683457783e-06, + "loss": 0.6955, + "num_input_tokens_seen": 63459400, + "step": 2953 + }, + { + "epoch": 0.3551974989478747, + "flos": 20165275106880.0, + "grad_norm": 5.015073942055801, + "learning_rate": 2.9896382201808126e-06, + "loss": 0.6517, + "num_input_tokens_seen": 63476800, + "step": 2954 + }, + { + "epoch": 0.3553177418385138, + "flos": 19800582425280.0, + "grad_norm": 6.411713740459692, + "learning_rate": 2.988961221888516e-06, + "loss": 0.8137, + "num_input_tokens_seen": 63493475, + "step": 2955 + }, + { + "epoch": 0.3554379847291529, + "flos": 14808121172640.0, + "grad_norm": 3.8679509189651613, + "learning_rate": 2.988284073571589e-06, + "loss": 0.7877, + "num_input_tokens_seen": 63509560, + "step": 2956 + }, + { + "epoch": 0.355558227619792, + "flos": 20456137177920.0, + "grad_norm": 3.169666743044589, + "learning_rate": 2.9876067753327528e-06, + "loss": 0.7342, + "num_input_tokens_seen": 63527290, + "step": 2957 + }, + { + "epoch": 0.35567847051043106, + "flos": 37609150816320.0, + "grad_norm": 4.297592491588155, + "learning_rate": 2.986929327274754e-06, + "loss": 0.8094, + "num_input_tokens_seen": 63547870, + "step": 2958 + }, + { + "epoch": 0.35579871340107017, + "flos": 26904432745920.0, + "grad_norm": 1.8517628492937248, + "learning_rate": 2.9862517295003617e-06, + "loss": 0.7891, + "num_input_tokens_seen": 63568285, + "step": 2959 + }, + { + "epoch": 0.3559189562917093, + "flos": 28252754914560.0, + "grad_norm": 1.6393829003455271, + "learning_rate": 2.9855739821123654e-06, + "loss": 0.728, + "num_input_tokens_seen": 63589865, + "step": 2960 + }, + { + "epoch": 0.35603919918234833, + "flos": 25627227795360.0, + "grad_norm": 1.759753025022525, + "learning_rate": 2.98489608521358e-06, + "loss": 0.8205, + "num_input_tokens_seen": 63609805, + "step": 2961 + }, + { + "epoch": 0.35615944207298744, + "flos": 22967907630720.0, + "grad_norm": 2.4397586237226028, + "learning_rate": 2.9842180389068425e-06, + "loss": 0.7981, + "num_input_tokens_seen": 63627115, + "step": 2962 + }, + { + "epoch": 0.35627968496362655, + "flos": 68154074276160.0, + "grad_norm": 0.7699507136486459, + "learning_rate": 2.98353984329501e-06, + "loss": 0.5924, + "num_input_tokens_seen": 63691460, + "step": 2963 + }, + { + "epoch": 0.3563999278542656, + "flos": 22609496638560.0, + "grad_norm": 1.8856251708458307, + "learning_rate": 2.982861498480965e-06, + "loss": 0.7045, + "num_input_tokens_seen": 63713840, + "step": 2964 + }, + { + "epoch": 0.3565201707449047, + "flos": 25915153455360.0, + "grad_norm": 3.7484037180718386, + "learning_rate": 2.9821830045676122e-06, + "loss": 0.8275, + "num_input_tokens_seen": 63733540, + "step": 2965 + }, + { + "epoch": 0.3566404136355438, + "flos": 28434878236800.0, + "grad_norm": 1.8806186415012383, + "learning_rate": 2.9815043616578793e-06, + "loss": 0.7312, + "num_input_tokens_seen": 63754335, + "step": 2966 + }, + { + "epoch": 0.3567606565261829, + "flos": 38307865393920.0, + "grad_norm": 2.2826483634203036, + "learning_rate": 2.9808255698547145e-06, + "loss": 0.7699, + "num_input_tokens_seen": 63777135, + "step": 2967 + }, + { + "epoch": 0.356880899416822, + "flos": 21949630193760.0, + "grad_norm": 3.093971629825812, + "learning_rate": 2.9801466292610913e-06, + "loss": 0.7987, + "num_input_tokens_seen": 63797020, + "step": 2968 + }, + { + "epoch": 0.35700114230746105, + "flos": 18962086615680.0, + "grad_norm": 3.889472514969369, + "learning_rate": 2.979467539980003e-06, + "loss": 0.8107, + "num_input_tokens_seen": 63813810, + "step": 2969 + }, + { + "epoch": 0.35712138519810016, + "flos": 19727978416800.0, + "grad_norm": 3.373858506053873, + "learning_rate": 2.978788302114468e-06, + "loss": 0.769, + "num_input_tokens_seen": 63830925, + "step": 2970 + }, + { + "epoch": 0.35724162808873927, + "flos": 35132176992480.0, + "grad_norm": 2.3305741299401923, + "learning_rate": 2.9781089157675255e-06, + "loss": 0.8374, + "num_input_tokens_seen": 63849385, + "step": 2971 + }, + { + "epoch": 0.3573618709793783, + "flos": 25520124516000.0, + "grad_norm": 1.838136695012643, + "learning_rate": 2.977429381042238e-06, + "loss": 0.8858, + "num_input_tokens_seen": 63870060, + "step": 2972 + }, + { + "epoch": 0.35748211387001744, + "flos": 29090544498720.0, + "grad_norm": 2.3884722422711153, + "learning_rate": 2.9767496980416913e-06, + "loss": 0.8894, + "num_input_tokens_seen": 63889355, + "step": 2973 + }, + { + "epoch": 0.35760235676065655, + "flos": 13934457036480.0, + "grad_norm": 3.4107434792934765, + "learning_rate": 2.9760698668689914e-06, + "loss": 0.8111, + "num_input_tokens_seen": 63905580, + "step": 2974 + }, + { + "epoch": 0.3577225996512956, + "flos": 43955212343520.0, + "grad_norm": 2.651045783638646, + "learning_rate": 2.975389887627269e-06, + "loss": 0.7117, + "num_input_tokens_seen": 63928180, + "step": 2975 + }, + { + "epoch": 0.3578428425419347, + "flos": 17030962381920.0, + "grad_norm": 2.501549431351521, + "learning_rate": 2.9747097604196764e-06, + "loss": 0.8983, + "num_input_tokens_seen": 63944825, + "step": 2976 + }, + { + "epoch": 0.3579630854325738, + "flos": 71574193984800.0, + "grad_norm": 0.6802002299255951, + "learning_rate": 2.9740294853493875e-06, + "loss": 0.5666, + "num_input_tokens_seen": 64016385, + "step": 2977 + }, + { + "epoch": 0.3580833283232129, + "flos": 25010818533600.0, + "grad_norm": 2.2163337122839923, + "learning_rate": 2.9733490625196008e-06, + "loss": 0.6753, + "num_input_tokens_seen": 64036245, + "step": 2978 + }, + { + "epoch": 0.358203571213852, + "flos": 13933379113440.0, + "grad_norm": 2.83480904987323, + "learning_rate": 2.9726684920335353e-06, + "loss": 0.7607, + "num_input_tokens_seen": 64054110, + "step": 2979 + }, + { + "epoch": 0.35832381410449105, + "flos": 20273493479040.0, + "grad_norm": 2.436133857726215, + "learning_rate": 2.971987773994432e-06, + "loss": 0.8195, + "num_input_tokens_seen": 64070295, + "step": 2980 + }, + { + "epoch": 0.35844405699513016, + "flos": 16959138938400.0, + "grad_norm": 2.294494848901017, + "learning_rate": 2.9713069085055566e-06, + "loss": 0.8299, + "num_input_tokens_seen": 64088605, + "step": 2981 + }, + { + "epoch": 0.35856429988576927, + "flos": 23183489470560.0, + "grad_norm": 2.2768135304180976, + "learning_rate": 2.9706258956701958e-06, + "loss": 0.7945, + "num_input_tokens_seen": 64108595, + "step": 2982 + }, + { + "epoch": 0.3586845427764083, + "flos": 23001477657600.0, + "grad_norm": 2.9593066553040686, + "learning_rate": 2.9699447355916575e-06, + "loss": 0.7734, + "num_input_tokens_seen": 64127405, + "step": 2983 + }, + { + "epoch": 0.35880478566704743, + "flos": 19982148201120.0, + "grad_norm": 2.106654921189091, + "learning_rate": 2.969263428373275e-06, + "loss": 0.7366, + "num_input_tokens_seen": 64145755, + "step": 2984 + }, + { + "epoch": 0.35892502855768654, + "flos": 13679766875520.0, + "grad_norm": 2.795286014915448, + "learning_rate": 2.9685819741184007e-06, + "loss": 0.7954, + "num_input_tokens_seen": 64164195, + "step": 2985 + }, + { + "epoch": 0.3590452714483256, + "flos": 18088348140000.0, + "grad_norm": 2.630105639174551, + "learning_rate": 2.967900372930411e-06, + "loss": 0.6899, + "num_input_tokens_seen": 64182705, + "step": 2986 + }, + { + "epoch": 0.3591655143389647, + "flos": 17724138665280.0, + "grad_norm": 2.4800711726158657, + "learning_rate": 2.9672186249127046e-06, + "loss": 0.7964, + "num_input_tokens_seen": 64202170, + "step": 2987 + }, + { + "epoch": 0.3592857572296038, + "flos": 25188146956800.0, + "grad_norm": 2.6672324732702406, + "learning_rate": 2.9665367301687014e-06, + "loss": 0.795, + "num_input_tokens_seen": 64220415, + "step": 2988 + }, + { + "epoch": 0.3594060001202429, + "flos": 29342112399840.0, + "grad_norm": 2.608080879969953, + "learning_rate": 2.965854688801845e-06, + "loss": 0.7696, + "num_input_tokens_seen": 64242475, + "step": 2989 + }, + { + "epoch": 0.359526243010882, + "flos": 17028583517280.0, + "grad_norm": 2.418185863700085, + "learning_rate": 2.9651725009156005e-06, + "loss": 0.7665, + "num_input_tokens_seen": 64260220, + "step": 2990 + }, + { + "epoch": 0.3596464859015211, + "flos": 22932218927520.0, + "grad_norm": 2.4431200039808356, + "learning_rate": 2.964490166613454e-06, + "loss": 0.7462, + "num_input_tokens_seen": 64280145, + "step": 2991 + }, + { + "epoch": 0.35976672879216015, + "flos": 54661485810240.0, + "grad_norm": 0.7684664645326812, + "learning_rate": 2.963807685998917e-06, + "loss": 0.5771, + "num_input_tokens_seen": 64336010, + "step": 2992 + }, + { + "epoch": 0.35988697168279926, + "flos": 43076678968800.0, + "grad_norm": 1.6905906772478774, + "learning_rate": 2.9631250591755196e-06, + "loss": 0.7824, + "num_input_tokens_seen": 64358725, + "step": 2993 + }, + { + "epoch": 0.36000721457343837, + "flos": 35794013434560.0, + "grad_norm": 1.979674652226525, + "learning_rate": 2.962442286246817e-06, + "loss": 0.576, + "num_input_tokens_seen": 64381555, + "step": 2994 + }, + { + "epoch": 0.3601274574640774, + "flos": 18264895998240.0, + "grad_norm": 1.8433101582004454, + "learning_rate": 2.9617593673163853e-06, + "loss": 0.6968, + "num_input_tokens_seen": 64400375, + "step": 2995 + }, + { + "epoch": 0.36024770035471654, + "flos": 13315148533440.0, + "grad_norm": 2.297702873662299, + "learning_rate": 2.9610763024878216e-06, + "loss": 0.7721, + "num_input_tokens_seen": 64415880, + "step": 2996 + }, + { + "epoch": 0.3603679432453556, + "flos": 20238139303680.0, + "grad_norm": 1.8401421630921357, + "learning_rate": 2.960393091864747e-06, + "loss": 0.9167, + "num_input_tokens_seen": 64434260, + "step": 2997 + }, + { + "epoch": 0.3604881861359947, + "flos": 22418898611040.0, + "grad_norm": 2.13216940997245, + "learning_rate": 2.959709735550804e-06, + "loss": 0.7454, + "num_input_tokens_seen": 64453855, + "step": 2998 + }, + { + "epoch": 0.3606084290266338, + "flos": 22022754578880.0, + "grad_norm": 2.3440391870744763, + "learning_rate": 2.9590262336496575e-06, + "loss": 0.7608, + "num_input_tokens_seen": 64473585, + "step": 2999 + }, + { + "epoch": 0.36072867191727287, + "flos": 15610779600000.0, + "grad_norm": 2.7569295101791558, + "learning_rate": 2.9583425862649936e-06, + "loss": 0.8555, + "num_input_tokens_seen": 64491720, + "step": 3000 + }, + { + "epoch": 0.360848914807912, + "flos": 19648758191040.0, + "grad_norm": 2.5082463634871055, + "learning_rate": 2.9576587935005215e-06, + "loss": 0.7383, + "num_input_tokens_seen": 64509800, + "step": 3001 + }, + { + "epoch": 0.3609691576985511, + "flos": 18850857492960.0, + "grad_norm": 3.5441698690137304, + "learning_rate": 2.9569748554599713e-06, + "loss": 0.7256, + "num_input_tokens_seen": 64525410, + "step": 3002 + }, + { + "epoch": 0.36108940058919015, + "flos": 42162977267520.0, + "grad_norm": 2.1082453210526126, + "learning_rate": 2.956290772247097e-06, + "loss": 0.7314, + "num_input_tokens_seen": 64544085, + "step": 3003 + }, + { + "epoch": 0.36120964347982926, + "flos": 23152000950240.0, + "grad_norm": 1.7959010926434968, + "learning_rate": 2.9556065439656724e-06, + "loss": 0.7313, + "num_input_tokens_seen": 64563135, + "step": 3004 + }, + { + "epoch": 0.36132988637046837, + "flos": 18087270216960.0, + "grad_norm": 1.863901478926537, + "learning_rate": 2.9549221707194952e-06, + "loss": 0.8171, + "num_input_tokens_seen": 64581985, + "step": 3005 + }, + { + "epoch": 0.3614501292611074, + "flos": 27773004624960.0, + "grad_norm": 4.3277764025060135, + "learning_rate": 2.954237652612384e-06, + "loss": 0.7329, + "num_input_tokens_seen": 64601035, + "step": 3006 + }, + { + "epoch": 0.36157037215174653, + "flos": 22602471553920.0, + "grad_norm": 2.086129740453922, + "learning_rate": 2.9535529897481796e-06, + "loss": 0.8458, + "num_input_tokens_seen": 64618580, + "step": 3007 + }, + { + "epoch": 0.36169061504238564, + "flos": 12582975438240.0, + "grad_norm": 5.759753041271435, + "learning_rate": 2.9528681822307446e-06, + "loss": 0.7649, + "num_input_tokens_seen": 64635430, + "step": 3008 + }, + { + "epoch": 0.3618108579330247, + "flos": 26646285797280.0, + "grad_norm": 2.0895215508051828, + "learning_rate": 2.952183230163964e-06, + "loss": 0.8223, + "num_input_tokens_seen": 64655005, + "step": 3009 + }, + { + "epoch": 0.3619311008236638, + "flos": 22784297518080.0, + "grad_norm": 2.253472826983255, + "learning_rate": 2.9514981336517448e-06, + "loss": 0.7323, + "num_input_tokens_seen": 64674775, + "step": 3010 + }, + { + "epoch": 0.36205134371430286, + "flos": 25882401163200.0, + "grad_norm": 2.016392051767985, + "learning_rate": 2.950812892798015e-06, + "loss": 0.818, + "num_input_tokens_seen": 64695670, + "step": 3011 + }, + { + "epoch": 0.362171586604942, + "flos": 26102108846400.0, + "grad_norm": 2.139450865015296, + "learning_rate": 2.9501275077067256e-06, + "loss": 0.8718, + "num_input_tokens_seen": 64716290, + "step": 3012 + }, + { + "epoch": 0.3622918294955811, + "flos": 28033939305600.0, + "grad_norm": 1.955368645392202, + "learning_rate": 2.949441978481848e-06, + "loss": 0.8877, + "num_input_tokens_seen": 64737190, + "step": 3013 + }, + { + "epoch": 0.36241207238622014, + "flos": 19799727520800.0, + "grad_norm": 2.287998407816833, + "learning_rate": 2.9487563052273778e-06, + "loss": 0.7994, + "num_input_tokens_seen": 64753950, + "step": 3014 + }, + { + "epoch": 0.36253231527685925, + "flos": 21366605110080.0, + "grad_norm": 2.054684831160636, + "learning_rate": 2.94807048804733e-06, + "loss": 0.86, + "num_input_tokens_seen": 64773570, + "step": 3015 + }, + { + "epoch": 0.36265255816749836, + "flos": 18336533592960.0, + "grad_norm": 2.1603563433336372, + "learning_rate": 2.9473845270457434e-06, + "loss": 0.9051, + "num_input_tokens_seen": 64790385, + "step": 3016 + }, + { + "epoch": 0.3627728010581374, + "flos": 18633454334880.0, + "grad_norm": 2.8043117430872395, + "learning_rate": 2.946698422326677e-06, + "loss": 0.7028, + "num_input_tokens_seen": 64807845, + "step": 3017 + }, + { + "epoch": 0.36289304394877653, + "flos": 27484707267360.0, + "grad_norm": 3.020331944039439, + "learning_rate": 2.946012173994213e-06, + "loss": 0.7963, + "num_input_tokens_seen": 64827630, + "step": 3018 + }, + { + "epoch": 0.36301328683941564, + "flos": 34483610154720.0, + "grad_norm": 1.5008455913042413, + "learning_rate": 2.945325782152454e-06, + "loss": 0.6779, + "num_input_tokens_seen": 64850665, + "step": 3019 + }, + { + "epoch": 0.3631335297300547, + "flos": 18998890411680.0, + "grad_norm": 4.413450377565097, + "learning_rate": 2.9446392469055257e-06, + "loss": 0.7866, + "num_input_tokens_seen": 64868700, + "step": 3020 + }, + { + "epoch": 0.3632537726206938, + "flos": 19508939789280.0, + "grad_norm": 2.0250431009109104, + "learning_rate": 2.9439525683575745e-06, + "loss": 0.7984, + "num_input_tokens_seen": 64887620, + "step": 3021 + }, + { + "epoch": 0.3633740155113329, + "flos": 21038530375680.0, + "grad_norm": 3.5632933686896533, + "learning_rate": 2.9432657466127694e-06, + "loss": 0.7501, + "num_input_tokens_seen": 64907190, + "step": 3022 + }, + { + "epoch": 0.36349425840197197, + "flos": 20269627824000.0, + "grad_norm": 1.9343710213714258, + "learning_rate": 2.9425787817753007e-06, + "loss": 0.7695, + "num_input_tokens_seen": 64925850, + "step": 3023 + }, + { + "epoch": 0.3636145012926111, + "flos": 29674127128800.0, + "grad_norm": 1.8684121429310105, + "learning_rate": 2.94189167394938e-06, + "loss": 0.7108, + "num_input_tokens_seen": 64948285, + "step": 3024 + }, + { + "epoch": 0.3637347441832502, + "flos": 21401178720480.0, + "grad_norm": 5.643732541079464, + "learning_rate": 2.941204423239241e-06, + "loss": 0.8141, + "num_input_tokens_seen": 64967160, + "step": 3025 + }, + { + "epoch": 0.36385498707388925, + "flos": 29491557769440.0, + "grad_norm": 1.8703968127885133, + "learning_rate": 2.9405170297491395e-06, + "loss": 0.7599, + "num_input_tokens_seen": 64987155, + "step": 3026 + }, + { + "epoch": 0.36397522996452836, + "flos": 22204803561600.0, + "grad_norm": 3.870383188759484, + "learning_rate": 2.939829493583353e-06, + "loss": 0.8048, + "num_input_tokens_seen": 65003800, + "step": 3027 + }, + { + "epoch": 0.3640954728551674, + "flos": 21475789896000.0, + "grad_norm": 2.7444266282231333, + "learning_rate": 2.939141814846179e-06, + "loss": 0.8292, + "num_input_tokens_seen": 65021215, + "step": 3028 + }, + { + "epoch": 0.3642157157458065, + "flos": 17687706566880.0, + "grad_norm": 1.7117384337943575, + "learning_rate": 2.938453993641938e-06, + "loss": 0.8285, + "num_input_tokens_seen": 65039590, + "step": 3029 + }, + { + "epoch": 0.36433595863644563, + "flos": 17614470672480.0, + "grad_norm": 2.552486019413814, + "learning_rate": 2.937766030074973e-06, + "loss": 0.6975, + "num_input_tokens_seen": 65056845, + "step": 3030 + }, + { + "epoch": 0.3644562015270847, + "flos": 26794801922880.0, + "grad_norm": 2.1903785294822886, + "learning_rate": 2.937077924249646e-06, + "loss": 0.8288, + "num_input_tokens_seen": 65079435, + "step": 3031 + }, + { + "epoch": 0.3645764444177238, + "flos": 14262531770880.0, + "grad_norm": 3.0288113882004883, + "learning_rate": 2.9363896762703443e-06, + "loss": 0.7568, + "num_input_tokens_seen": 65096500, + "step": 3032 + }, + { + "epoch": 0.3646966873083629, + "flos": 20638000311840.0, + "grad_norm": 1.9755027855322684, + "learning_rate": 2.9357012862414725e-06, + "loss": 0.8444, + "num_input_tokens_seen": 65115860, + "step": 3033 + }, + { + "epoch": 0.36481693019900197, + "flos": 27742779876480.0, + "grad_norm": 2.274580625509269, + "learning_rate": 2.9350127542674593e-06, + "loss": 0.7156, + "num_input_tokens_seen": 65138070, + "step": 3034 + }, + { + "epoch": 0.3649371730896411, + "flos": 19684261045440.0, + "grad_norm": 2.6613338513510794, + "learning_rate": 2.934324080452755e-06, + "loss": 0.7639, + "num_input_tokens_seen": 65155860, + "step": 3035 + }, + { + "epoch": 0.3650574159802802, + "flos": 24714938544960.0, + "grad_norm": 1.9171185444538295, + "learning_rate": 2.9336352649018307e-06, + "loss": 0.7821, + "num_input_tokens_seen": 65175930, + "step": 3036 + }, + { + "epoch": 0.36517765887091924, + "flos": 32806246837920.0, + "grad_norm": 1.889255816063151, + "learning_rate": 2.9329463077191783e-06, + "loss": 0.7044, + "num_input_tokens_seen": 65199305, + "step": 3037 + }, + { + "epoch": 0.36529790176155835, + "flos": 20091630345120.0, + "grad_norm": 2.5100709247067314, + "learning_rate": 2.9322572090093135e-06, + "loss": 0.642, + "num_input_tokens_seen": 65218010, + "step": 3038 + }, + { + "epoch": 0.36541814465219746, + "flos": 17614507842240.0, + "grad_norm": 3.6324919528438855, + "learning_rate": 2.9315679688767713e-06, + "loss": 0.7686, + "num_input_tokens_seen": 65236020, + "step": 3039 + }, + { + "epoch": 0.3655383875428365, + "flos": 22641951572640.0, + "grad_norm": 2.2594560495193248, + "learning_rate": 2.9308785874261085e-06, + "loss": 0.6661, + "num_input_tokens_seen": 65256210, + "step": 3040 + }, + { + "epoch": 0.36565863043347563, + "flos": 21949778872800.0, + "grad_norm": 2.1786308325143855, + "learning_rate": 2.9301890647619045e-06, + "loss": 0.8168, + "num_input_tokens_seen": 65275505, + "step": 3041 + }, + { + "epoch": 0.36577887332411474, + "flos": 24792114433920.0, + "grad_norm": 3.168435322261987, + "learning_rate": 2.929499400988759e-06, + "loss": 0.8017, + "num_input_tokens_seen": 65296665, + "step": 3042 + }, + { + "epoch": 0.3658991162147538, + "flos": 28252606235520.0, + "grad_norm": 1.9809099116091515, + "learning_rate": 2.9288095962112927e-06, + "loss": 0.6546, + "num_input_tokens_seen": 65317370, + "step": 3043 + }, + { + "epoch": 0.3660193591053929, + "flos": 17760273405600.0, + "grad_norm": 2.0121527046362315, + "learning_rate": 2.9281196505341503e-06, + "loss": 0.8536, + "num_input_tokens_seen": 65334540, + "step": 3044 + }, + { + "epoch": 0.36613960199603196, + "flos": 10328348350560.0, + "grad_norm": 2.308498115946939, + "learning_rate": 2.9274295640619946e-06, + "loss": 0.7831, + "num_input_tokens_seen": 65351670, + "step": 3045 + }, + { + "epoch": 0.36625984488667107, + "flos": 19727458040160.0, + "grad_norm": 2.008975512243782, + "learning_rate": 2.9267393368995103e-06, + "loss": 0.7842, + "num_input_tokens_seen": 65370900, + "step": 3046 + }, + { + "epoch": 0.3663800877773102, + "flos": 17649081452640.0, + "grad_norm": 3.7171256741308487, + "learning_rate": 2.926048969151407e-06, + "loss": 0.7454, + "num_input_tokens_seen": 65388185, + "step": 3047 + }, + { + "epoch": 0.36650033066794924, + "flos": 20274682911360.0, + "grad_norm": 2.3802583607119234, + "learning_rate": 2.92535846092241e-06, + "loss": 0.6873, + "num_input_tokens_seen": 65407200, + "step": 3048 + }, + { + "epoch": 0.36662057355858835, + "flos": 24680810971680.0, + "grad_norm": 3.8585950549830756, + "learning_rate": 2.9246678123172704e-06, + "loss": 0.8291, + "num_input_tokens_seen": 65427290, + "step": 3049 + }, + { + "epoch": 0.36674081644922746, + "flos": 12367839635520.0, + "grad_norm": 2.4787070438666134, + "learning_rate": 2.9239770234407596e-06, + "loss": 0.7451, + "num_input_tokens_seen": 65445595, + "step": 3050 + }, + { + "epoch": 0.3668610593398665, + "flos": 21075817378560.0, + "grad_norm": 1.9748461361597585, + "learning_rate": 2.9232860943976686e-06, + "loss": 0.684, + "num_input_tokens_seen": 65466330, + "step": 3051 + }, + { + "epoch": 0.3669813022305056, + "flos": 26719633200960.0, + "grad_norm": 1.911939666043512, + "learning_rate": 2.9225950252928115e-06, + "loss": 0.8427, + "num_input_tokens_seen": 65487180, + "step": 3052 + }, + { + "epoch": 0.36710154512114473, + "flos": 19072349324640.0, + "grad_norm": 2.6686139646070854, + "learning_rate": 2.9219038162310217e-06, + "loss": 0.8189, + "num_input_tokens_seen": 65505380, + "step": 3053 + }, + { + "epoch": 0.3672217880117838, + "flos": 20782279084800.0, + "grad_norm": 2.4877914833012005, + "learning_rate": 2.921212467317157e-06, + "loss": 0.8237, + "num_input_tokens_seen": 65524825, + "step": 3054 + }, + { + "epoch": 0.3673420309024229, + "flos": 13571251145280.0, + "grad_norm": 3.28587262737615, + "learning_rate": 2.920520978656093e-06, + "loss": 0.7999, + "num_input_tokens_seen": 65541790, + "step": 3055 + }, + { + "epoch": 0.367462273793062, + "flos": 28945299312000.0, + "grad_norm": 2.1542978710403977, + "learning_rate": 2.919829350352729e-06, + "loss": 0.7667, + "num_input_tokens_seen": 65563395, + "step": 3056 + }, + { + "epoch": 0.36758251668370107, + "flos": 62553163824000.0, + "grad_norm": 0.7573677803069419, + "learning_rate": 2.919137582511983e-06, + "loss": 0.599, + "num_input_tokens_seen": 65619835, + "step": 3057 + }, + { + "epoch": 0.3677027595743402, + "flos": 12695988709440.0, + "grad_norm": 2.6627558303209025, + "learning_rate": 2.918445675238797e-06, + "loss": 0.6415, + "num_input_tokens_seen": 65636760, + "step": 3058 + }, + { + "epoch": 0.36782300246497923, + "flos": 25010744194080.0, + "grad_norm": 3.5141012322787635, + "learning_rate": 2.917753628638132e-06, + "loss": 0.7034, + "num_input_tokens_seen": 65657065, + "step": 3059 + }, + { + "epoch": 0.36794324535561834, + "flos": 17394725819520.0, + "grad_norm": 2.705425628485897, + "learning_rate": 2.9170614428149716e-06, + "loss": 0.7064, + "num_input_tokens_seen": 65675400, + "step": 3060 + }, + { + "epoch": 0.36806348824625745, + "flos": 24054365874720.0, + "grad_norm": 2.6034403487453095, + "learning_rate": 2.9163691178743195e-06, + "loss": 0.8695, + "num_input_tokens_seen": 65694050, + "step": 3061 + }, + { + "epoch": 0.3681837311368965, + "flos": 20492606446080.0, + "grad_norm": 2.4710834919043916, + "learning_rate": 2.9156766539212006e-06, + "loss": 0.772, + "num_input_tokens_seen": 65711695, + "step": 3062 + }, + { + "epoch": 0.3683039740275356, + "flos": 21435863840160.0, + "grad_norm": 2.879378923099231, + "learning_rate": 2.9149840510606614e-06, + "loss": 0.721, + "num_input_tokens_seen": 65730350, + "step": 3063 + }, + { + "epoch": 0.36842421691817473, + "flos": 70279853774880.0, + "grad_norm": 1.0292536746374825, + "learning_rate": 2.914291309397769e-06, + "loss": 0.6412, + "num_input_tokens_seen": 65787820, + "step": 3064 + }, + { + "epoch": 0.3685444598088138, + "flos": 23298026701920.0, + "grad_norm": 2.415949013361489, + "learning_rate": 2.9135984290376117e-06, + "loss": 0.7854, + "num_input_tokens_seen": 65806685, + "step": 3065 + }, + { + "epoch": 0.3686647026994529, + "flos": 23037092021280.0, + "grad_norm": 3.2897106266015106, + "learning_rate": 2.9129054100853e-06, + "loss": 0.8307, + "num_input_tokens_seen": 65825525, + "step": 3066 + }, + { + "epoch": 0.368784945590092, + "flos": 25083980088480.0, + "grad_norm": 1.9879234006926636, + "learning_rate": 2.912212252645963e-06, + "loss": 0.7593, + "num_input_tokens_seen": 65848110, + "step": 3067 + }, + { + "epoch": 0.36890518848073106, + "flos": 18416051176800.0, + "grad_norm": 3.1751529686791025, + "learning_rate": 2.9115189568247523e-06, + "loss": 0.7669, + "num_input_tokens_seen": 65865670, + "step": 3068 + }, + { + "epoch": 0.36902543137137017, + "flos": 16339235719200.0, + "grad_norm": 2.1946033133234675, + "learning_rate": 2.910825522726841e-06, + "loss": 0.9226, + "num_input_tokens_seen": 65883195, + "step": 3069 + }, + { + "epoch": 0.3691456742620093, + "flos": 12258320321760.0, + "grad_norm": 3.23347040325002, + "learning_rate": 2.9101319504574215e-06, + "loss": 0.7757, + "num_input_tokens_seen": 65899635, + "step": 3070 + }, + { + "epoch": 0.36926591715264834, + "flos": 17761165479840.0, + "grad_norm": 12.177791230069765, + "learning_rate": 2.909438240121709e-06, + "loss": 0.7634, + "num_input_tokens_seen": 65919910, + "step": 3071 + }, + { + "epoch": 0.36938616004328745, + "flos": 28906711367520.0, + "grad_norm": 1.9357657515702649, + "learning_rate": 2.908744391824939e-06, + "loss": 0.7053, + "num_input_tokens_seen": 65940930, + "step": 3072 + }, + { + "epoch": 0.36950640293392656, + "flos": 29163483035040.0, + "grad_norm": 2.059432783459417, + "learning_rate": 2.908050405672367e-06, + "loss": 0.7933, + "num_input_tokens_seen": 65960475, + "step": 3073 + }, + { + "epoch": 0.3696266458245656, + "flos": 24792263112960.0, + "grad_norm": 2.054913526511622, + "learning_rate": 2.9073562817692703e-06, + "loss": 0.793, + "num_input_tokens_seen": 65979160, + "step": 3074 + }, + { + "epoch": 0.3697468887152047, + "flos": 59801905642080.0, + "grad_norm": 0.7299996159826863, + "learning_rate": 2.9066620202209468e-06, + "loss": 0.5649, + "num_input_tokens_seen": 66041650, + "step": 3075 + }, + { + "epoch": 0.3698671316058438, + "flos": 26100027339840.0, + "grad_norm": 2.119195077785422, + "learning_rate": 2.905967621132716e-06, + "loss": 0.7798, + "num_input_tokens_seen": 66059260, + "step": 3076 + }, + { + "epoch": 0.3699873744964829, + "flos": 24572220901920.0, + "grad_norm": 2.376889430926859, + "learning_rate": 2.9052730846099172e-06, + "loss": 0.7526, + "num_input_tokens_seen": 66080605, + "step": 3077 + }, + { + "epoch": 0.370107617387122, + "flos": 64792694255520.0, + "grad_norm": 0.8780355251789402, + "learning_rate": 2.9045784107579123e-06, + "loss": 0.6081, + "num_input_tokens_seen": 66140710, + "step": 3078 + }, + { + "epoch": 0.37022786027776106, + "flos": 15939077352960.0, + "grad_norm": 1.9446253041517698, + "learning_rate": 2.9038835996820807e-06, + "loss": 0.6708, + "num_input_tokens_seen": 66158320, + "step": 3079 + }, + { + "epoch": 0.37034810316840017, + "flos": 18520366724160.0, + "grad_norm": 6.164943076162522, + "learning_rate": 2.903188651487826e-06, + "loss": 0.793, + "num_input_tokens_seen": 66176875, + "step": 3080 + }, + { + "epoch": 0.3704683460590393, + "flos": 17796593994720.0, + "grad_norm": 2.710322769340871, + "learning_rate": 2.902493566280571e-06, + "loss": 0.8654, + "num_input_tokens_seen": 66194980, + "step": 3081 + }, + { + "epoch": 0.37058858894967833, + "flos": 14114015645280.0, + "grad_norm": 7.906223221684518, + "learning_rate": 2.9017983441657595e-06, + "loss": 0.8178, + "num_input_tokens_seen": 66212310, + "step": 3082 + }, + { + "epoch": 0.37070883184031744, + "flos": 13934791564320.0, + "grad_norm": 2.514489338110171, + "learning_rate": 2.9011029852488564e-06, + "loss": 0.7514, + "num_input_tokens_seen": 66229545, + "step": 3083 + }, + { + "epoch": 0.37082907473095655, + "flos": 52240389602880.0, + "grad_norm": 1.0471019214255837, + "learning_rate": 2.9004074896353465e-06, + "loss": 0.6248, + "num_input_tokens_seen": 66283545, + "step": 3084 + }, + { + "epoch": 0.3709493176215956, + "flos": 15975435111840.0, + "grad_norm": 2.082784147992515, + "learning_rate": 2.8997118574307362e-06, + "loss": 0.8165, + "num_input_tokens_seen": 66300500, + "step": 3085 + }, + { + "epoch": 0.3710695605122347, + "flos": 20929940305920.0, + "grad_norm": 2.4103891022165844, + "learning_rate": 2.899016088740553e-06, + "loss": 0.7485, + "num_input_tokens_seen": 66318610, + "step": 3086 + }, + { + "epoch": 0.37118980340287383, + "flos": 14335284458400.0, + "grad_norm": 2.4535888427731947, + "learning_rate": 2.898320183670344e-06, + "loss": 0.7931, + "num_input_tokens_seen": 66336025, + "step": 3087 + }, + { + "epoch": 0.3713100462935129, + "flos": 25848310759680.0, + "grad_norm": 2.7659711890495795, + "learning_rate": 2.8976241423256767e-06, + "loss": 0.8879, + "num_input_tokens_seen": 66356120, + "step": 3088 + }, + { + "epoch": 0.371430289184152, + "flos": 30475112916960.0, + "grad_norm": 3.1240765542334947, + "learning_rate": 2.896927964812142e-06, + "loss": 0.6828, + "num_input_tokens_seen": 66376685, + "step": 3089 + }, + { + "epoch": 0.37155053207479105, + "flos": 15719927216160.0, + "grad_norm": 4.179326977734143, + "learning_rate": 2.8962316512353465e-06, + "loss": 0.7544, + "num_input_tokens_seen": 66394230, + "step": 3090 + }, + { + "epoch": 0.37167077496543016, + "flos": 23370593540640.0, + "grad_norm": 1.8925509140845707, + "learning_rate": 2.8955352017009233e-06, + "loss": 0.7506, + "num_input_tokens_seen": 66413475, + "step": 3091 + }, + { + "epoch": 0.3717910178560693, + "flos": 22057030831200.0, + "grad_norm": 4.092075850051321, + "learning_rate": 2.8948386163145212e-06, + "loss": 0.7739, + "num_input_tokens_seen": 66432685, + "step": 3092 + }, + { + "epoch": 0.3719112607467083, + "flos": 26901236146560.0, + "grad_norm": 2.6448253203258285, + "learning_rate": 2.8941418951818135e-06, + "loss": 0.7974, + "num_input_tokens_seen": 66452245, + "step": 3093 + }, + { + "epoch": 0.37203150363734744, + "flos": 12149395724160.0, + "grad_norm": 2.3430571485673664, + "learning_rate": 2.8934450384084903e-06, + "loss": 0.7131, + "num_input_tokens_seen": 66469760, + "step": 3094 + }, + { + "epoch": 0.37215174652798655, + "flos": 23663760136800.0, + "grad_norm": 3.0108750642203255, + "learning_rate": 2.8927480461002653e-06, + "loss": 0.6964, + "num_input_tokens_seen": 66489730, + "step": 3095 + }, + { + "epoch": 0.3722719894186256, + "flos": 17862061409280.0, + "grad_norm": 4.35130607904177, + "learning_rate": 2.892050918362872e-06, + "loss": 0.8581, + "num_input_tokens_seen": 66504785, + "step": 3096 + }, + { + "epoch": 0.3723922323092647, + "flos": 62330073692640.0, + "grad_norm": 0.8537243341299391, + "learning_rate": 2.8913536553020626e-06, + "loss": 0.5582, + "num_input_tokens_seen": 66558840, + "step": 3097 + }, + { + "epoch": 0.3725124751999038, + "flos": 23006830103040.0, + "grad_norm": 2.0529423834236025, + "learning_rate": 2.8906562570236137e-06, + "loss": 0.85, + "num_input_tokens_seen": 66576750, + "step": 3098 + }, + { + "epoch": 0.3726327180905429, + "flos": 20890608966240.0, + "grad_norm": 1.88142704938485, + "learning_rate": 2.889958723633318e-06, + "loss": 0.7679, + "num_input_tokens_seen": 66595690, + "step": 3099 + }, + { + "epoch": 0.372752960981182, + "flos": 30548237302080.0, + "grad_norm": 1.834772137830747, + "learning_rate": 2.889261055236992e-06, + "loss": 0.7423, + "num_input_tokens_seen": 66617905, + "step": 3100 + }, + { + "epoch": 0.3728732038718211, + "flos": 25080709149600.0, + "grad_norm": 2.134668122880587, + "learning_rate": 2.8885632519404704e-06, + "loss": 0.8306, + "num_input_tokens_seen": 66638175, + "step": 3101 + }, + { + "epoch": 0.37299344676246016, + "flos": 25265843222400.0, + "grad_norm": 2.338825392698073, + "learning_rate": 2.8878653138496107e-06, + "loss": 0.759, + "num_input_tokens_seen": 66658670, + "step": 3102 + }, + { + "epoch": 0.37311368965309927, + "flos": 23808447777120.0, + "grad_norm": 2.4828248375981614, + "learning_rate": 2.8871672410702878e-06, + "loss": 0.7638, + "num_input_tokens_seen": 66676030, + "step": 3103 + }, + { + "epoch": 0.3732339325437384, + "flos": 25774963356000.0, + "grad_norm": 1.9233740566025992, + "learning_rate": 2.8864690337084008e-06, + "loss": 0.8259, + "num_input_tokens_seen": 66695305, + "step": 3104 + }, + { + "epoch": 0.37335417543437743, + "flos": 26170698520800.0, + "grad_norm": 1.929573580819448, + "learning_rate": 2.885770691869866e-06, + "loss": 0.783, + "num_input_tokens_seen": 66716185, + "step": 3105 + }, + { + "epoch": 0.37347441832501654, + "flos": 23989493176320.0, + "grad_norm": 3.7609383317211895, + "learning_rate": 2.8850722156606207e-06, + "loss": 0.7448, + "num_input_tokens_seen": 66734575, + "step": 3106 + }, + { + "epoch": 0.3735946612156556, + "flos": 19686677079840.0, + "grad_norm": 3.5743425525006955, + "learning_rate": 2.8843736051866252e-06, + "loss": 0.676, + "num_input_tokens_seen": 66753500, + "step": 3107 + }, + { + "epoch": 0.3737149041062947, + "flos": 23006681424000.0, + "grad_norm": 1.9984147226379227, + "learning_rate": 2.8836748605538557e-06, + "loss": 0.6938, + "num_input_tokens_seen": 66775140, + "step": 3108 + }, + { + "epoch": 0.3738351469969338, + "flos": 34628223455520.0, + "grad_norm": 2.4999220835941562, + "learning_rate": 2.882975981868313e-06, + "loss": 0.6333, + "num_input_tokens_seen": 66795525, + "step": 3109 + }, + { + "epoch": 0.3739553898875729, + "flos": 43445237305440.0, + "grad_norm": 8.365381895515142, + "learning_rate": 2.882276969236016e-06, + "loss": 0.6889, + "num_input_tokens_seen": 66816885, + "step": 3110 + }, + { + "epoch": 0.374075632778212, + "flos": 12838260315360.0, + "grad_norm": 2.6076263233282213, + "learning_rate": 2.881577822763005e-06, + "loss": 0.7631, + "num_input_tokens_seen": 66834835, + "step": 3111 + }, + { + "epoch": 0.3741958756688511, + "flos": 25986493692000.0, + "grad_norm": 2.0863028788507734, + "learning_rate": 2.880878542555338e-06, + "loss": 0.8741, + "num_input_tokens_seen": 66852600, + "step": 3112 + }, + { + "epoch": 0.37431611855949015, + "flos": 21403297396800.0, + "grad_norm": 3.537220456566089, + "learning_rate": 2.8801791287190976e-06, + "loss": 0.8067, + "num_input_tokens_seen": 66870595, + "step": 3113 + }, + { + "epoch": 0.37443636145012926, + "flos": 24172471403040.0, + "grad_norm": 2.9038126579680865, + "learning_rate": 2.8794795813603817e-06, + "loss": 0.8591, + "num_input_tokens_seen": 66886060, + "step": 3114 + }, + { + "epoch": 0.3745566043407684, + "flos": 15356870004000.0, + "grad_norm": 2.020871689684784, + "learning_rate": 2.878779900585314e-06, + "loss": 0.8147, + "num_input_tokens_seen": 66903900, + "step": 3115 + }, + { + "epoch": 0.37467684723140743, + "flos": 24573001466880.0, + "grad_norm": 1.830517369079293, + "learning_rate": 2.8780800865000336e-06, + "loss": 0.7512, + "num_input_tokens_seen": 66925210, + "step": 3116 + }, + { + "epoch": 0.37479709012204654, + "flos": 64285692798240.0, + "grad_norm": 0.9811329510605113, + "learning_rate": 2.877380139210702e-06, + "loss": 0.5918, + "num_input_tokens_seen": 66983880, + "step": 3117 + }, + { + "epoch": 0.37491733301268565, + "flos": 23769859832640.0, + "grad_norm": 1.820187740389178, + "learning_rate": 2.876680058823501e-06, + "loss": 0.764, + "num_input_tokens_seen": 67004280, + "step": 3118 + }, + { + "epoch": 0.3750375759033247, + "flos": 32114743193760.0, + "grad_norm": 2.248559808751258, + "learning_rate": 2.8759798454446314e-06, + "loss": 0.6584, + "num_input_tokens_seen": 67027445, + "step": 3119 + }, + { + "epoch": 0.3751578187939638, + "flos": 23334384460800.0, + "grad_norm": 2.230502313738822, + "learning_rate": 2.8752794991803173e-06, + "loss": 0.8166, + "num_input_tokens_seen": 67045130, + "step": 3120 + }, + { + "epoch": 0.37527806168460287, + "flos": 14584882362240.0, + "grad_norm": 2.4202837238608197, + "learning_rate": 2.8745790201367976e-06, + "loss": 0.7552, + "num_input_tokens_seen": 67060885, + "step": 3121 + }, + { + "epoch": 0.375398304575242, + "flos": 26352821843040.0, + "grad_norm": 2.2662721959167014, + "learning_rate": 2.8738784084203373e-06, + "loss": 0.842, + "num_input_tokens_seen": 67080960, + "step": 3122 + }, + { + "epoch": 0.3755185474658811, + "flos": 22204840731360.0, + "grad_norm": 1.8900358685112388, + "learning_rate": 2.873177664137216e-06, + "loss": 0.7895, + "num_input_tokens_seen": 67101890, + "step": 3123 + }, + { + "epoch": 0.37563879035652015, + "flos": 30768762720000.0, + "grad_norm": 1.8923133058566102, + "learning_rate": 2.8724767873937384e-06, + "loss": 0.6913, + "num_input_tokens_seen": 67126010, + "step": 3124 + }, + { + "epoch": 0.37575903324715926, + "flos": 20743765479840.0, + "grad_norm": 4.30874713752636, + "learning_rate": 2.871775778296225e-06, + "loss": 0.8771, + "num_input_tokens_seen": 67143100, + "step": 3125 + }, + { + "epoch": 0.37587927613779837, + "flos": 18671075865600.0, + "grad_norm": 7.303303992477854, + "learning_rate": 2.8710746369510196e-06, + "loss": 0.7845, + "num_input_tokens_seen": 67161085, + "step": 3126 + }, + { + "epoch": 0.3759995190284374, + "flos": 13605527397600.0, + "grad_norm": 3.1733945016281644, + "learning_rate": 2.8703733634644846e-06, + "loss": 0.8337, + "num_input_tokens_seen": 67175840, + "step": 3127 + }, + { + "epoch": 0.37611976191907653, + "flos": 20455616801280.0, + "grad_norm": 1.8962223017259516, + "learning_rate": 2.869671957943002e-06, + "loss": 0.7939, + "num_input_tokens_seen": 67194995, + "step": 3128 + }, + { + "epoch": 0.37624000480971564, + "flos": 21111617591040.0, + "grad_norm": 2.338310133949673, + "learning_rate": 2.8689704204929747e-06, + "loss": 0.745, + "num_input_tokens_seen": 67214175, + "step": 3129 + }, + { + "epoch": 0.3763602477003547, + "flos": 22532209240320.0, + "grad_norm": 2.6728348820325785, + "learning_rate": 2.8682687512208253e-06, + "loss": 0.8094, + "num_input_tokens_seen": 67233185, + "step": 3130 + }, + { + "epoch": 0.3764804905909938, + "flos": 27487383490080.0, + "grad_norm": 2.2041502386489293, + "learning_rate": 2.8675669502329972e-06, + "loss": 0.8093, + "num_input_tokens_seen": 67254715, + "step": 3131 + }, + { + "epoch": 0.3766007334816329, + "flos": 22495888651200.0, + "grad_norm": 4.8389203019600044, + "learning_rate": 2.866865017635952e-06, + "loss": 0.8573, + "num_input_tokens_seen": 67272535, + "step": 3132 + }, + { + "epoch": 0.376720976372272, + "flos": 25919948354400.0, + "grad_norm": 1.9284860677851312, + "learning_rate": 2.866162953536174e-06, + "loss": 0.7923, + "num_input_tokens_seen": 67293505, + "step": 3133 + }, + { + "epoch": 0.3768412192629111, + "flos": 18015521112960.0, + "grad_norm": 2.0360033026108253, + "learning_rate": 2.8654607580401634e-06, + "loss": 0.7492, + "num_input_tokens_seen": 67313240, + "step": 3134 + }, + { + "epoch": 0.3769614621535502, + "flos": 62899637589600.0, + "grad_norm": 0.9020966691109897, + "learning_rate": 2.8647584312544446e-06, + "loss": 0.6519, + "num_input_tokens_seen": 67378445, + "step": 3135 + }, + { + "epoch": 0.37708170504418925, + "flos": 23627922754560.0, + "grad_norm": 2.0290884401045988, + "learning_rate": 2.864055973285559e-06, + "loss": 0.8498, + "num_input_tokens_seen": 67400445, + "step": 3136 + }, + { + "epoch": 0.37720194793482836, + "flos": 24388127582400.0, + "grad_norm": 2.332449155132711, + "learning_rate": 2.8633533842400698e-06, + "loss": 0.8664, + "num_input_tokens_seen": 67420645, + "step": 3137 + }, + { + "epoch": 0.3773221908254674, + "flos": 20966558253120.0, + "grad_norm": 2.5589762394507467, + "learning_rate": 2.862650664224558e-06, + "loss": 0.7719, + "num_input_tokens_seen": 67439495, + "step": 3138 + }, + { + "epoch": 0.37744243371610653, + "flos": 37577476447200.0, + "grad_norm": 1.700225321479135, + "learning_rate": 2.861947813345627e-06, + "loss": 0.6983, + "num_input_tokens_seen": 67462195, + "step": 3139 + }, + { + "epoch": 0.37756267660674564, + "flos": 26103075260160.0, + "grad_norm": 1.9491431578998484, + "learning_rate": 2.8612448317098974e-06, + "loss": 0.7244, + "num_input_tokens_seen": 67482530, + "step": 3140 + }, + { + "epoch": 0.3776829194973847, + "flos": 19400238210240.0, + "grad_norm": 2.5758130428815407, + "learning_rate": 2.8605417194240114e-06, + "loss": 0.8291, + "num_input_tokens_seen": 67500890, + "step": 3141 + }, + { + "epoch": 0.3778031623880238, + "flos": 17357550325920.0, + "grad_norm": 2.4736735825881677, + "learning_rate": 2.8598384765946315e-06, + "loss": 0.792, + "num_input_tokens_seen": 67519545, + "step": 3142 + }, + { + "epoch": 0.3779234052786629, + "flos": 27086964935520.0, + "grad_norm": 2.34056566720891, + "learning_rate": 2.8591351033284377e-06, + "loss": 0.7209, + "num_input_tokens_seen": 67539275, + "step": 3143 + }, + { + "epoch": 0.37804364816930197, + "flos": 19654593843360.0, + "grad_norm": 3.038289655858695, + "learning_rate": 2.8584315997321325e-06, + "loss": 0.8375, + "num_input_tokens_seen": 67558280, + "step": 3144 + }, + { + "epoch": 0.3781638910599411, + "flos": 22670169154080.0, + "grad_norm": 2.9231817413949988, + "learning_rate": 2.8577279659124356e-06, + "loss": 0.7836, + "num_input_tokens_seen": 67575955, + "step": 3145 + }, + { + "epoch": 0.3782841339505802, + "flos": 14626889924640.0, + "grad_norm": 3.047365902210558, + "learning_rate": 2.857024201976089e-06, + "loss": 0.8352, + "num_input_tokens_seen": 67593515, + "step": 3146 + }, + { + "epoch": 0.37840437684121925, + "flos": 32771078511360.0, + "grad_norm": 5.035051698390866, + "learning_rate": 2.8563203080298516e-06, + "loss": 0.7273, + "num_input_tokens_seen": 67614290, + "step": 3147 + }, + { + "epoch": 0.37852461973185836, + "flos": 18343075470720.0, + "grad_norm": 2.6756814534309528, + "learning_rate": 2.855616284180505e-06, + "loss": 0.8973, + "num_input_tokens_seen": 67631340, + "step": 3148 + }, + { + "epoch": 0.37864486262249747, + "flos": 59416429331040.0, + "grad_norm": 0.8834090791171377, + "learning_rate": 2.8549121305348477e-06, + "loss": 0.6598, + "num_input_tokens_seen": 67691125, + "step": 3149 + }, + { + "epoch": 0.3787651055131365, + "flos": 23329440882720.0, + "grad_norm": 2.774538493227807, + "learning_rate": 2.8542078471997006e-06, + "loss": 0.8323, + "num_input_tokens_seen": 67708740, + "step": 3150 + }, + { + "epoch": 0.37888534840377563, + "flos": 24565939212480.0, + "grad_norm": 1.847288066802714, + "learning_rate": 2.8535034342819013e-06, + "loss": 0.7616, + "num_input_tokens_seen": 67726150, + "step": 3151 + }, + { + "epoch": 0.37900559129441475, + "flos": 23952540701280.0, + "grad_norm": 3.1855799558080773, + "learning_rate": 2.85279889188831e-06, + "loss": 0.7268, + "num_input_tokens_seen": 67746525, + "step": 3152 + }, + { + "epoch": 0.3791258341850538, + "flos": 24609396395520.0, + "grad_norm": 1.8601436789198802, + "learning_rate": 2.852094220125805e-06, + "loss": 0.8087, + "num_input_tokens_seen": 67766035, + "step": 3153 + }, + { + "epoch": 0.3792460770756929, + "flos": 17396101100640.0, + "grad_norm": 4.133915370036147, + "learning_rate": 2.8513894191012846e-06, + "loss": 0.7071, + "num_input_tokens_seen": 67785015, + "step": 3154 + }, + { + "epoch": 0.37936631996633197, + "flos": 24171913856640.0, + "grad_norm": 1.7623693106605673, + "learning_rate": 2.8506844889216664e-06, + "loss": 0.7888, + "num_input_tokens_seen": 67804400, + "step": 3155 + }, + { + "epoch": 0.3794865628569711, + "flos": 70197022348800.0, + "grad_norm": 0.8782675902236353, + "learning_rate": 2.849979429693887e-06, + "loss": 0.6288, + "num_input_tokens_seen": 67865385, + "step": 3156 + }, + { + "epoch": 0.3796068057476102, + "flos": 15757028370240.0, + "grad_norm": 2.1909799416136204, + "learning_rate": 2.8492742415249042e-06, + "loss": 0.7411, + "num_input_tokens_seen": 67883030, + "step": 3157 + }, + { + "epoch": 0.37972704863824924, + "flos": 25155617683200.0, + "grad_norm": 2.258611095656111, + "learning_rate": 2.848568924521694e-06, + "loss": 0.7663, + "num_input_tokens_seen": 67903810, + "step": 3158 + }, + { + "epoch": 0.37984729152888835, + "flos": 26172705687840.0, + "grad_norm": 3.0781713182091743, + "learning_rate": 2.8478634787912526e-06, + "loss": 0.7379, + "num_input_tokens_seen": 67921865, + "step": 3159 + }, + { + "epoch": 0.37996753441952746, + "flos": 25592579845440.0, + "grad_norm": 2.8933359991081837, + "learning_rate": 2.847157904440596e-06, + "loss": 0.7656, + "num_input_tokens_seen": 67941595, + "step": 3160 + }, + { + "epoch": 0.3800877773101665, + "flos": 20089660347840.0, + "grad_norm": 1.9117226760007067, + "learning_rate": 2.846452201576759e-06, + "loss": 0.739, + "num_input_tokens_seen": 67960970, + "step": 3161 + }, + { + "epoch": 0.38020802020080563, + "flos": 62963358025440.0, + "grad_norm": 0.8621701992133871, + "learning_rate": 2.845746370306795e-06, + "loss": 0.6268, + "num_input_tokens_seen": 68026160, + "step": 3162 + }, + { + "epoch": 0.38032826309144474, + "flos": 21258052210080.0, + "grad_norm": 2.871241559226716, + "learning_rate": 2.84504041073778e-06, + "loss": 0.7864, + "num_input_tokens_seen": 68044575, + "step": 3163 + }, + { + "epoch": 0.3804485059820838, + "flos": 18927178477440.0, + "grad_norm": 2.1924076070424223, + "learning_rate": 2.844334322976806e-06, + "loss": 0.7927, + "num_input_tokens_seen": 68064790, + "step": 3164 + }, + { + "epoch": 0.3805687488727229, + "flos": 21802526519040.0, + "grad_norm": 1.9158847153609506, + "learning_rate": 2.8436281071309866e-06, + "loss": 0.8348, + "num_input_tokens_seen": 68083330, + "step": 3165 + }, + { + "epoch": 0.380688991763362, + "flos": 58463136101760.0, + "grad_norm": 0.7304205241067901, + "learning_rate": 2.842921763307455e-06, + "loss": 0.5291, + "num_input_tokens_seen": 68146140, + "step": 3166 + }, + { + "epoch": 0.38080923465400107, + "flos": 23765250782400.0, + "grad_norm": 1.960707388822807, + "learning_rate": 2.842215291613361e-06, + "loss": 0.8288, + "num_input_tokens_seen": 68164520, + "step": 3167 + }, + { + "epoch": 0.3809294775446402, + "flos": 54891972723840.0, + "grad_norm": 0.8538245508421259, + "learning_rate": 2.8415086921558774e-06, + "loss": 0.5918, + "num_input_tokens_seen": 68227945, + "step": 3168 + }, + { + "epoch": 0.38104972043527924, + "flos": 24608392812000.0, + "grad_norm": 1.807230293338511, + "learning_rate": 2.840801965042194e-06, + "loss": 0.7875, + "num_input_tokens_seen": 68247405, + "step": 3169 + }, + { + "epoch": 0.38116996332591835, + "flos": 22824112064640.0, + "grad_norm": 2.485085343827688, + "learning_rate": 2.840095110379521e-06, + "loss": 0.8427, + "num_input_tokens_seen": 68266925, + "step": 3170 + }, + { + "epoch": 0.38129020621655746, + "flos": 60749437558560.0, + "grad_norm": 0.7443863670704731, + "learning_rate": 2.8393881282750884e-06, + "loss": 0.5387, + "num_input_tokens_seen": 68329755, + "step": 3171 + }, + { + "epoch": 0.3814104491071965, + "flos": 21617020748640.0, + "grad_norm": 2.1826069193690234, + "learning_rate": 2.838681018836144e-06, + "loss": 0.7841, + "num_input_tokens_seen": 68347075, + "step": 3172 + }, + { + "epoch": 0.3815306919978356, + "flos": 19071903287520.0, + "grad_norm": 2.030839731892224, + "learning_rate": 2.837973782169955e-06, + "loss": 0.785, + "num_input_tokens_seen": 68365450, + "step": 3173 + }, + { + "epoch": 0.38165093488847474, + "flos": 66971855263200.0, + "grad_norm": 0.8558989187948443, + "learning_rate": 2.8372664183838096e-06, + "loss": 0.5914, + "num_input_tokens_seen": 68428470, + "step": 3174 + }, + { + "epoch": 0.3817711777791138, + "flos": 22309007599680.0, + "grad_norm": 18.164086021490146, + "learning_rate": 2.836558927585015e-06, + "loss": 0.6835, + "num_input_tokens_seen": 68445440, + "step": 3175 + }, + { + "epoch": 0.3818914206697529, + "flos": 22788014494080.0, + "grad_norm": 2.387551619685702, + "learning_rate": 2.8358513098808957e-06, + "loss": 0.8227, + "num_input_tokens_seen": 68465755, + "step": 3176 + }, + { + "epoch": 0.382011663560392, + "flos": 24351063598080.0, + "grad_norm": 2.5896515235505597, + "learning_rate": 2.835143565378798e-06, + "loss": 0.7688, + "num_input_tokens_seen": 68486220, + "step": 3177 + }, + { + "epoch": 0.38213190645103107, + "flos": 21950187740160.0, + "grad_norm": 2.4774187695594834, + "learning_rate": 2.8344356941860847e-06, + "loss": 0.7846, + "num_input_tokens_seen": 68505850, + "step": 3178 + }, + { + "epoch": 0.3822521493416702, + "flos": 35465827190880.0, + "grad_norm": 2.712260426431904, + "learning_rate": 2.8337276964101403e-06, + "loss": 0.6624, + "num_input_tokens_seen": 68526290, + "step": 3179 + }, + { + "epoch": 0.3823723922323093, + "flos": 21039868487040.0, + "grad_norm": 5.4830515424833735, + "learning_rate": 2.833019572158367e-06, + "loss": 0.7652, + "num_input_tokens_seen": 68544725, + "step": 3180 + }, + { + "epoch": 0.38249263512294834, + "flos": 19760842218240.0, + "grad_norm": 1.9109720621547075, + "learning_rate": 2.8323113215381872e-06, + "loss": 0.8032, + "num_input_tokens_seen": 68563070, + "step": 3181 + }, + { + "epoch": 0.38261287801358745, + "flos": 21403334566560.0, + "grad_norm": 2.0789047728615335, + "learning_rate": 2.831602944657042e-06, + "loss": 0.7619, + "num_input_tokens_seen": 68581150, + "step": 3182 + }, + { + "epoch": 0.38273312090422656, + "flos": 21950187740160.0, + "grad_norm": 4.2160622406044235, + "learning_rate": 2.830894441622391e-06, + "loss": 0.7412, + "num_input_tokens_seen": 68597800, + "step": 3183 + }, + { + "epoch": 0.3828533637948656, + "flos": 24755422147200.0, + "grad_norm": 2.4014692498135997, + "learning_rate": 2.8301858125417134e-06, + "loss": 0.7968, + "num_input_tokens_seen": 68617640, + "step": 3184 + }, + { + "epoch": 0.38297360668550473, + "flos": 22423358982240.0, + "grad_norm": 3.546610818174622, + "learning_rate": 2.8294770575225082e-06, + "loss": 0.74, + "num_input_tokens_seen": 68637770, + "step": 3185 + }, + { + "epoch": 0.3830938495761438, + "flos": 24861150145440.0, + "grad_norm": 2.0071190256159452, + "learning_rate": 2.828768176672293e-06, + "loss": 0.8411, + "num_input_tokens_seen": 68656885, + "step": 3186 + }, + { + "epoch": 0.3832140924667829, + "flos": 32988741857760.0, + "grad_norm": 2.126439336367539, + "learning_rate": 2.8280591700986044e-06, + "loss": 0.7129, + "num_input_tokens_seen": 68678750, + "step": 3187 + }, + { + "epoch": 0.383334335357422, + "flos": 31857822847200.0, + "grad_norm": 3.3774334646994424, + "learning_rate": 2.827350037908999e-06, + "loss": 0.7491, + "num_input_tokens_seen": 68698190, + "step": 3188 + }, + { + "epoch": 0.38345457824806106, + "flos": 19763704289760.0, + "grad_norm": 3.532794175908222, + "learning_rate": 2.8266407802110496e-06, + "loss": 0.7898, + "num_input_tokens_seen": 68716445, + "step": 3189 + }, + { + "epoch": 0.3835748211387002, + "flos": 22387372920960.0, + "grad_norm": 2.422702278848909, + "learning_rate": 2.8259313971123515e-06, + "loss": 0.7673, + "num_input_tokens_seen": 68737565, + "step": 3190 + }, + { + "epoch": 0.3836950640293393, + "flos": 25082456128320.0, + "grad_norm": 1.7742902411264256, + "learning_rate": 2.8252218887205166e-06, + "loss": 0.787, + "num_input_tokens_seen": 68758255, + "step": 3191 + }, + { + "epoch": 0.38381530691997834, + "flos": 21768101587680.0, + "grad_norm": 1.902949743861128, + "learning_rate": 2.824512255143178e-06, + "loss": 0.8053, + "num_input_tokens_seen": 68777490, + "step": 3192 + }, + { + "epoch": 0.38393554981061745, + "flos": 21222400676640.0, + "grad_norm": 1.925423795940995, + "learning_rate": 2.8238024964879855e-06, + "loss": 0.7932, + "num_input_tokens_seen": 68798385, + "step": 3193 + }, + { + "epoch": 0.38405579270125656, + "flos": 16995199339200.0, + "grad_norm": 3.0420478338108623, + "learning_rate": 2.8230926128626095e-06, + "loss": 0.7719, + "num_input_tokens_seen": 68816880, + "step": 3194 + }, + { + "epoch": 0.3841760355918956, + "flos": 21804199158240.0, + "grad_norm": 2.385302733875234, + "learning_rate": 2.822382604374738e-06, + "loss": 0.7946, + "num_input_tokens_seen": 68835805, + "step": 3195 + }, + { + "epoch": 0.3842962784825347, + "flos": 25877754943200.0, + "grad_norm": 7.238026959830989, + "learning_rate": 2.8216724711320793e-06, + "loss": 0.6593, + "num_input_tokens_seen": 68854930, + "step": 3196 + }, + { + "epoch": 0.38441652137317384, + "flos": 25301197397760.0, + "grad_norm": 1.7017011377552607, + "learning_rate": 2.820962213242361e-06, + "loss": 0.7973, + "num_input_tokens_seen": 68874260, + "step": 3197 + }, + { + "epoch": 0.3845367642638129, + "flos": 18087716254080.0, + "grad_norm": 2.5144058969805387, + "learning_rate": 2.8202518308133264e-06, + "loss": 0.8438, + "num_input_tokens_seen": 68891535, + "step": 3198 + }, + { + "epoch": 0.384657007154452, + "flos": 25192272800160.0, + "grad_norm": 2.0166576017486038, + "learning_rate": 2.8195413239527426e-06, + "loss": 0.7369, + "num_input_tokens_seen": 68911275, + "step": 3199 + }, + { + "epoch": 0.38477725004509106, + "flos": 19837572070080.0, + "grad_norm": 2.398621592461646, + "learning_rate": 2.8188306927683906e-06, + "loss": 0.8047, + "num_input_tokens_seen": 68930745, + "step": 3200 + }, + { + "epoch": 0.38489749293573017, + "flos": 18233407477920.0, + "grad_norm": 2.2752458125190023, + "learning_rate": 2.818119937368074e-06, + "loss": 0.7464, + "num_input_tokens_seen": 68949380, + "step": 3201 + }, + { + "epoch": 0.3850177358263693, + "flos": 24354260197440.0, + "grad_norm": 2.080333826529643, + "learning_rate": 2.817409057859613e-06, + "loss": 0.6537, + "num_input_tokens_seen": 68968370, + "step": 3202 + }, + { + "epoch": 0.38513797871700833, + "flos": 17646256550880.0, + "grad_norm": 2.4919827655177906, + "learning_rate": 2.8166980543508482e-06, + "loss": 0.7891, + "num_input_tokens_seen": 68984260, + "step": 3203 + }, + { + "epoch": 0.38525822160764744, + "flos": 25702731045120.0, + "grad_norm": 2.06638963544833, + "learning_rate": 2.815986926949638e-06, + "loss": 0.7978, + "num_input_tokens_seen": 69002760, + "step": 3204 + }, + { + "epoch": 0.38537846449828655, + "flos": 20164866239520.0, + "grad_norm": 2.163517191244457, + "learning_rate": 2.8152756757638597e-06, + "loss": 0.8039, + "num_input_tokens_seen": 69021260, + "step": 3205 + }, + { + "epoch": 0.3854987073889256, + "flos": 23006421235680.0, + "grad_norm": 5.796265932562647, + "learning_rate": 2.8145643009014093e-06, + "loss": 0.8468, + "num_input_tokens_seen": 69039755, + "step": 3206 + }, + { + "epoch": 0.3856189502795647, + "flos": 20161743979680.0, + "grad_norm": 2.0291404920870306, + "learning_rate": 2.813852802470202e-06, + "loss": 0.7935, + "num_input_tokens_seen": 69057650, + "step": 3207 + }, + { + "epoch": 0.38573919317020383, + "flos": 25665815739840.0, + "grad_norm": 3.193490775201774, + "learning_rate": 2.8131411805781717e-06, + "loss": 0.7241, + "num_input_tokens_seen": 69077775, + "step": 3208 + }, + { + "epoch": 0.3858594360608429, + "flos": 29778442676160.0, + "grad_norm": 5.684458831286391, + "learning_rate": 2.8124294353332707e-06, + "loss": 0.6373, + "num_input_tokens_seen": 69096450, + "step": 3209 + }, + { + "epoch": 0.385979678951482, + "flos": 24754938940320.0, + "grad_norm": 2.1516221735594483, + "learning_rate": 2.8117175668434713e-06, + "loss": 0.7727, + "num_input_tokens_seen": 69116110, + "step": 3210 + }, + { + "epoch": 0.3860999218421211, + "flos": 21257494663680.0, + "grad_norm": 3.1491126006398673, + "learning_rate": 2.811005575216762e-06, + "loss": 0.7018, + "num_input_tokens_seen": 69134825, + "step": 3211 + }, + { + "epoch": 0.38622016473276016, + "flos": 24501661230240.0, + "grad_norm": 1.7377195672366987, + "learning_rate": 2.8102934605611513e-06, + "loss": 0.7873, + "num_input_tokens_seen": 69156460, + "step": 3212 + }, + { + "epoch": 0.3863404076233993, + "flos": 20528889865440.0, + "grad_norm": 5.289905204861049, + "learning_rate": 2.8095812229846665e-06, + "loss": 0.6703, + "num_input_tokens_seen": 69176420, + "step": 3213 + }, + { + "epoch": 0.3864606505140384, + "flos": 22314360045120.0, + "grad_norm": 3.2005791199697025, + "learning_rate": 2.808868862595355e-06, + "loss": 0.6927, + "num_input_tokens_seen": 69196665, + "step": 3214 + }, + { + "epoch": 0.38658089340467744, + "flos": 25666819323360.0, + "grad_norm": 3.7678728794190266, + "learning_rate": 2.8081563795012795e-06, + "loss": 0.7935, + "num_input_tokens_seen": 69216290, + "step": 3215 + }, + { + "epoch": 0.38670113629531655, + "flos": 33753927433440.0, + "grad_norm": 2.1406668777267717, + "learning_rate": 2.807443773810524e-06, + "loss": 0.7371, + "num_input_tokens_seen": 69237550, + "step": 3216 + }, + { + "epoch": 0.3868213791859556, + "flos": 23298175380960.0, + "grad_norm": 5.488324691220424, + "learning_rate": 2.80673104563119e-06, + "loss": 0.8952, + "num_input_tokens_seen": 69258175, + "step": 3217 + }, + { + "epoch": 0.3869416220765947, + "flos": 18415270611840.0, + "grad_norm": 2.4295522768716555, + "learning_rate": 2.8060181950713976e-06, + "loss": 0.7918, + "num_input_tokens_seen": 69274965, + "step": 3218 + }, + { + "epoch": 0.3870618649672338, + "flos": 15610853939520.0, + "grad_norm": 9.63398145279575, + "learning_rate": 2.805305222239286e-06, + "loss": 0.8089, + "num_input_tokens_seen": 69292220, + "step": 3219 + }, + { + "epoch": 0.3871821078578729, + "flos": 23480187193920.0, + "grad_norm": 1.9587576250951542, + "learning_rate": 2.8045921272430118e-06, + "loss": 0.7369, + "num_input_tokens_seen": 69311900, + "step": 3220 + }, + { + "epoch": 0.387302350748512, + "flos": 17752727944320.0, + "grad_norm": 2.590365212072476, + "learning_rate": 2.803878910190753e-06, + "loss": 0.7644, + "num_input_tokens_seen": 69328820, + "step": 3221 + }, + { + "epoch": 0.3874225936391511, + "flos": 11487038905440.0, + "grad_norm": 3.333193362794066, + "learning_rate": 2.8031655711907017e-06, + "loss": 0.8237, + "num_input_tokens_seen": 69342525, + "step": 3222 + }, + { + "epoch": 0.38754283652979016, + "flos": 21913829981280.0, + "grad_norm": 2.5909800618625454, + "learning_rate": 2.8024521103510723e-06, + "loss": 0.8083, + "num_input_tokens_seen": 69359295, + "step": 3223 + }, + { + "epoch": 0.38766307942042927, + "flos": 21145670824800.0, + "grad_norm": 1.8639755395012916, + "learning_rate": 2.8017385277800952e-06, + "loss": 0.7528, + "num_input_tokens_seen": 69377650, + "step": 3224 + }, + { + "epoch": 0.3877833223110684, + "flos": 27377975685600.0, + "grad_norm": 2.526690202551391, + "learning_rate": 2.8010248235860213e-06, + "loss": 0.7517, + "num_input_tokens_seen": 69397765, + "step": 3225 + }, + { + "epoch": 0.38790356520170743, + "flos": 64408518886080.0, + "grad_norm": 0.8362773311465154, + "learning_rate": 2.8003109978771192e-06, + "loss": 0.6278, + "num_input_tokens_seen": 69458930, + "step": 3226 + }, + { + "epoch": 0.38802380809234654, + "flos": 22313467970880.0, + "grad_norm": 2.7016174875623644, + "learning_rate": 2.799597050761674e-06, + "loss": 0.7901, + "num_input_tokens_seen": 69475135, + "step": 3227 + }, + { + "epoch": 0.38814405098298566, + "flos": 25225062262080.0, + "grad_norm": 2.3238672840219508, + "learning_rate": 2.7988829823479924e-06, + "loss": 0.7911, + "num_input_tokens_seen": 69493525, + "step": 3228 + }, + { + "epoch": 0.3882642938736247, + "flos": 18814276715520.0, + "grad_norm": 2.051512030324264, + "learning_rate": 2.7981687927443976e-06, + "loss": 0.6397, + "num_input_tokens_seen": 69510325, + "step": 3229 + }, + { + "epoch": 0.3883845367642638, + "flos": 21621889987200.0, + "grad_norm": 2.1256233871709576, + "learning_rate": 2.797454482059231e-06, + "loss": 0.8547, + "num_input_tokens_seen": 69530480, + "step": 3230 + }, + { + "epoch": 0.3885047796549029, + "flos": 20528183640000.0, + "grad_norm": 2.2790102480180487, + "learning_rate": 2.7967400504008537e-06, + "loss": 0.849, + "num_input_tokens_seen": 69549780, + "step": 3231 + }, + { + "epoch": 0.388625022545542, + "flos": 64233203363520.0, + "grad_norm": 0.7983592926362807, + "learning_rate": 2.7960254978776456e-06, + "loss": 0.5745, + "num_input_tokens_seen": 69611870, + "step": 3232 + }, + { + "epoch": 0.3887452654361811, + "flos": 18088125121440.0, + "grad_norm": 3.162997144140015, + "learning_rate": 2.7953108245980006e-06, + "loss": 0.82, + "num_input_tokens_seen": 69630385, + "step": 3233 + }, + { + "epoch": 0.38886550832682015, + "flos": 24939589806240.0, + "grad_norm": 1.8821327596979374, + "learning_rate": 2.7945960306703365e-06, + "loss": 0.7389, + "num_input_tokens_seen": 69653370, + "step": 3234 + }, + { + "epoch": 0.38898575121745926, + "flos": 27160535357760.0, + "grad_norm": 1.8779536824471332, + "learning_rate": 2.7938811162030865e-06, + "loss": 0.6557, + "num_input_tokens_seen": 69673635, + "step": 3235 + }, + { + "epoch": 0.3891059941080984, + "flos": 28722320689920.0, + "grad_norm": 1.7966734399945055, + "learning_rate": 2.793166081304702e-06, + "loss": 0.8226, + "num_input_tokens_seen": 69694130, + "step": 3236 + }, + { + "epoch": 0.38922623699873743, + "flos": 22860730011840.0, + "grad_norm": 2.1073685887682303, + "learning_rate": 2.7924509260836543e-06, + "loss": 0.8246, + "num_input_tokens_seen": 69713255, + "step": 3237 + }, + { + "epoch": 0.38934647988937654, + "flos": 19764224666400.0, + "grad_norm": 2.304588082965634, + "learning_rate": 2.791735650648431e-06, + "loss": 0.6862, + "num_input_tokens_seen": 69732375, + "step": 3238 + }, + { + "epoch": 0.38946672278001565, + "flos": 19174509025920.0, + "grad_norm": 3.956218850734836, + "learning_rate": 2.791020255107538e-06, + "loss": 0.7447, + "num_input_tokens_seen": 69749745, + "step": 3239 + }, + { + "epoch": 0.3895869656706547, + "flos": 24900778843200.0, + "grad_norm": 1.736033840579148, + "learning_rate": 2.7903047395695023e-06, + "loss": 0.8094, + "num_input_tokens_seen": 69769645, + "step": 3240 + }, + { + "epoch": 0.3897072085612938, + "flos": 24099086829600.0, + "grad_norm": 2.4403408808309015, + "learning_rate": 2.789589104142865e-06, + "loss": 0.901, + "num_input_tokens_seen": 69787270, + "step": 3241 + }, + { + "epoch": 0.3898274514519329, + "flos": 17142005655840.0, + "grad_norm": 2.4037851366955487, + "learning_rate": 2.7888733489361895e-06, + "loss": 0.7676, + "num_input_tokens_seen": 69805685, + "step": 3242 + }, + { + "epoch": 0.389947694342572, + "flos": 65979899749920.0, + "grad_norm": 0.7312921381489835, + "learning_rate": 2.788157474058054e-06, + "loss": 0.5868, + "num_input_tokens_seen": 69867960, + "step": 3243 + }, + { + "epoch": 0.3900679372332111, + "flos": 25704106326240.0, + "grad_norm": 2.0361914756612207, + "learning_rate": 2.7874414796170555e-06, + "loss": 0.6987, + "num_input_tokens_seen": 69889450, + "step": 3244 + }, + { + "epoch": 0.3901881801238502, + "flos": 11785037570400.0, + "grad_norm": 5.401111961601327, + "learning_rate": 2.7867253657218113e-06, + "loss": 0.8395, + "num_input_tokens_seen": 69903340, + "step": 3245 + }, + { + "epoch": 0.39030842301448926, + "flos": 27269831652960.0, + "grad_norm": 2.1861890952087757, + "learning_rate": 2.7860091324809544e-06, + "loss": 0.7314, + "num_input_tokens_seen": 69924400, + "step": 3246 + }, + { + "epoch": 0.39042866590512837, + "flos": 27124437787200.0, + "grad_norm": 1.8652498414362038, + "learning_rate": 2.7852927800031377e-06, + "loss": 0.8155, + "num_input_tokens_seen": 69944565, + "step": 3247 + }, + { + "epoch": 0.3905489087957674, + "flos": 29673458073120.0, + "grad_norm": 2.463969467042029, + "learning_rate": 2.7845763083970298e-06, + "loss": 0.8311, + "num_input_tokens_seen": 69964965, + "step": 3248 + }, + { + "epoch": 0.39066915168640653, + "flos": 24463556492640.0, + "grad_norm": 2.22413969231041, + "learning_rate": 2.7838597177713205e-06, + "loss": 0.8265, + "num_input_tokens_seen": 69986055, + "step": 3249 + }, + { + "epoch": 0.39078939457704565, + "flos": 20529261563040.0, + "grad_norm": 2.0448598478833095, + "learning_rate": 2.7831430082347143e-06, + "loss": 0.737, + "num_input_tokens_seen": 70005260, + "step": 3250 + }, + { + "epoch": 0.3909096374676847, + "flos": 22750839000480.0, + "grad_norm": 2.5229601334708023, + "learning_rate": 2.7824261798959373e-06, + "loss": 0.8226, + "num_input_tokens_seen": 70023440, + "step": 3251 + }, + { + "epoch": 0.3910298803583238, + "flos": 22970249325600.0, + "grad_norm": 2.235784446228289, + "learning_rate": 2.78170923286373e-06, + "loss": 0.7994, + "num_input_tokens_seen": 70043040, + "step": 3252 + }, + { + "epoch": 0.3911501232489629, + "flos": 24281470340160.0, + "grad_norm": 2.641181841944397, + "learning_rate": 2.780992167246854e-06, + "loss": 0.8443, + "num_input_tokens_seen": 70060565, + "step": 3253 + }, + { + "epoch": 0.391270366139602, + "flos": 60782152680960.0, + "grad_norm": 0.9879255898752553, + "learning_rate": 2.7802749831540883e-06, + "loss": 0.7213, + "num_input_tokens_seen": 70119465, + "step": 3254 + }, + { + "epoch": 0.3913906090302411, + "flos": 21512742371040.0, + "grad_norm": 2.1648564724116417, + "learning_rate": 2.7795576806942268e-06, + "loss": 0.8215, + "num_input_tokens_seen": 70138270, + "step": 3255 + }, + { + "epoch": 0.3915108519208802, + "flos": 49768731054720.0, + "grad_norm": 0.7680351314009672, + "learning_rate": 2.778840259976085e-06, + "loss": 0.5479, + "num_input_tokens_seen": 70193500, + "step": 3256 + }, + { + "epoch": 0.39163109481151925, + "flos": 16482956945760.0, + "grad_norm": 2.2462820488326294, + "learning_rate": 2.778122721108495e-06, + "loss": 0.7704, + "num_input_tokens_seen": 70211730, + "step": 3257 + }, + { + "epoch": 0.39175133770215836, + "flos": 26030136723840.0, + "grad_norm": 2.1604098676416967, + "learning_rate": 2.7774050642003076e-06, + "loss": 0.8815, + "num_input_tokens_seen": 70232540, + "step": 3258 + }, + { + "epoch": 0.3918715805927975, + "flos": 21840854275200.0, + "grad_norm": 2.7961939829713662, + "learning_rate": 2.7766872893603896e-06, + "loss": 0.9333, + "num_input_tokens_seen": 70252515, + "step": 3259 + }, + { + "epoch": 0.39199182348343653, + "flos": 20347138240800.0, + "grad_norm": 2.0370061451403028, + "learning_rate": 2.7759693966976275e-06, + "loss": 0.7326, + "num_input_tokens_seen": 70271020, + "step": 3260 + }, + { + "epoch": 0.39211206637407564, + "flos": 21652374924000.0, + "grad_norm": 2.1430420977382183, + "learning_rate": 2.7752513863209242e-06, + "loss": 0.8552, + "num_input_tokens_seen": 70289600, + "step": 3261 + }, + { + "epoch": 0.39223230926471475, + "flos": 21036114341280.0, + "grad_norm": 1.741926028212881, + "learning_rate": 2.774533258339203e-06, + "loss": 0.8436, + "num_input_tokens_seen": 70307060, + "step": 3262 + }, + { + "epoch": 0.3923525521553538, + "flos": 17578001404320.0, + "grad_norm": 2.354125492332276, + "learning_rate": 2.7738150128614014e-06, + "loss": 0.7974, + "num_input_tokens_seen": 70324465, + "step": 3263 + }, + { + "epoch": 0.3924727950459929, + "flos": 20528964204960.0, + "grad_norm": 2.077167605951458, + "learning_rate": 2.7730966499964777e-06, + "loss": 0.8968, + "num_input_tokens_seen": 70341415, + "step": 3264 + }, + { + "epoch": 0.39259303793663197, + "flos": 16193098458240.0, + "grad_norm": 7.782323801488904, + "learning_rate": 2.772378169853408e-06, + "loss": 0.8049, + "num_input_tokens_seen": 70358985, + "step": 3265 + }, + { + "epoch": 0.3927132808272711, + "flos": 16773038451840.0, + "grad_norm": 2.1827559117687274, + "learning_rate": 2.771659572541183e-06, + "loss": 0.7452, + "num_input_tokens_seen": 70376915, + "step": 3266 + }, + { + "epoch": 0.3928335237179102, + "flos": 20238362322240.0, + "grad_norm": 2.145433357763074, + "learning_rate": 2.7709408581688143e-06, + "loss": 0.8677, + "num_input_tokens_seen": 70396900, + "step": 3267 + }, + { + "epoch": 0.39295376660854925, + "flos": 24937508299680.0, + "grad_norm": 1.6995263533764113, + "learning_rate": 2.7702220268453307e-06, + "loss": 0.878, + "num_input_tokens_seen": 70417260, + "step": 3268 + }, + { + "epoch": 0.39307400949918836, + "flos": 18671187374880.0, + "grad_norm": 2.568747557107548, + "learning_rate": 2.7695030786797785e-06, + "loss": 0.8493, + "num_input_tokens_seen": 70433155, + "step": 3269 + }, + { + "epoch": 0.39319425238982747, + "flos": 22383321417120.0, + "grad_norm": 2.2937588475859676, + "learning_rate": 2.7687840137812206e-06, + "loss": 0.7497, + "num_input_tokens_seen": 70451240, + "step": 3270 + }, + { + "epoch": 0.3933144952804665, + "flos": 66098191127040.0, + "grad_norm": 0.8078145866440878, + "learning_rate": 2.7680648322587395e-06, + "loss": 0.6207, + "num_input_tokens_seen": 70516115, + "step": 3271 + }, + { + "epoch": 0.39343473817110564, + "flos": 15465237055200.0, + "grad_norm": 3.6260320575473184, + "learning_rate": 2.7673455342214334e-06, + "loss": 0.809, + "num_input_tokens_seen": 70533105, + "step": 3272 + }, + { + "epoch": 0.39355498106174475, + "flos": 21294484308480.0, + "grad_norm": 3.417320948656343, + "learning_rate": 2.7666261197784198e-06, + "loss": 0.7581, + "num_input_tokens_seen": 70551920, + "step": 3273 + }, + { + "epoch": 0.3936752239523838, + "flos": 13277564172480.0, + "grad_norm": 2.6402166829628975, + "learning_rate": 2.7659065890388336e-06, + "loss": 0.7617, + "num_input_tokens_seen": 70567920, + "step": 3274 + }, + { + "epoch": 0.3937954668430229, + "flos": 16776309390720.0, + "grad_norm": 2.112386291874631, + "learning_rate": 2.7651869421118266e-06, + "loss": 0.8518, + "num_input_tokens_seen": 70584530, + "step": 3275 + }, + { + "epoch": 0.393915709733662, + "flos": 21033995664960.0, + "grad_norm": 2.2991120739168363, + "learning_rate": 2.76446717910657e-06, + "loss": 0.8299, + "num_input_tokens_seen": 70605235, + "step": 3276 + }, + { + "epoch": 0.3940359526243011, + "flos": 17140890563040.0, + "grad_norm": 2.980898604886578, + "learning_rate": 2.763747300132249e-06, + "loss": 0.7731, + "num_input_tokens_seen": 70622675, + "step": 3277 + }, + { + "epoch": 0.3941561955149402, + "flos": 20965889197440.0, + "grad_norm": 2.1735866032018722, + "learning_rate": 2.7630273052980704e-06, + "loss": 0.8668, + "num_input_tokens_seen": 70643265, + "step": 3278 + }, + { + "epoch": 0.39427643840557924, + "flos": 18816581240640.0, + "grad_norm": 2.334919787546485, + "learning_rate": 2.7623071947132554e-06, + "loss": 0.6687, + "num_input_tokens_seen": 70660295, + "step": 3279 + }, + { + "epoch": 0.39439668129621835, + "flos": 23225422693440.0, + "grad_norm": 2.3080911621786617, + "learning_rate": 2.7615869684870458e-06, + "loss": 0.791, + "num_input_tokens_seen": 70679605, + "step": 3280 + }, + { + "epoch": 0.39451692418685746, + "flos": 26614128221280.0, + "grad_norm": 1.8808145012657114, + "learning_rate": 2.7608666267286986e-06, + "loss": 0.8466, + "num_input_tokens_seen": 70699835, + "step": 3281 + }, + { + "epoch": 0.3946371670774965, + "flos": 18232701252480.0, + "grad_norm": 2.708139344139328, + "learning_rate": 2.760146169547489e-06, + "loss": 0.8631, + "num_input_tokens_seen": 70716600, + "step": 3282 + }, + { + "epoch": 0.39475740996813563, + "flos": 24171467819520.0, + "grad_norm": 1.9391088881311687, + "learning_rate": 2.75942559705271e-06, + "loss": 0.7626, + "num_input_tokens_seen": 70736335, + "step": 3283 + }, + { + "epoch": 0.39487765285877474, + "flos": 19290347198880.0, + "grad_norm": 2.2404165684413435, + "learning_rate": 2.7587049093536713e-06, + "loss": 0.8916, + "num_input_tokens_seen": 70752665, + "step": 3284 + }, + { + "epoch": 0.3949978957494138, + "flos": 17287176503040.0, + "grad_norm": 2.151030040891202, + "learning_rate": 2.757984106559701e-06, + "loss": 0.8117, + "num_input_tokens_seen": 70771650, + "step": 3285 + }, + { + "epoch": 0.3951181386400529, + "flos": 36265512037440.0, + "grad_norm": 2.9604660906995925, + "learning_rate": 2.7572631887801446e-06, + "loss": 0.7177, + "num_input_tokens_seen": 70793275, + "step": 3286 + }, + { + "epoch": 0.395238381530692, + "flos": 23077241095680.0, + "grad_norm": 1.8701281514854846, + "learning_rate": 2.7565421561243654e-06, + "loss": 0.7685, + "num_input_tokens_seen": 70811440, + "step": 3287 + }, + { + "epoch": 0.3953586244213311, + "flos": 24312587162880.0, + "grad_norm": 2.6350720950547966, + "learning_rate": 2.7558210087017413e-06, + "loss": 0.8269, + "num_input_tokens_seen": 70831375, + "step": 3288 + }, + { + "epoch": 0.3954788673119702, + "flos": 23406542432160.0, + "grad_norm": 2.67951043881367, + "learning_rate": 2.7550997466216724e-06, + "loss": 0.7367, + "num_input_tokens_seen": 70849170, + "step": 3289 + }, + { + "epoch": 0.3955991102026093, + "flos": 17469225485760.0, + "grad_norm": 2.1693573774557815, + "learning_rate": 2.7543783699935714e-06, + "loss": 0.8104, + "num_input_tokens_seen": 70867200, + "step": 3290 + }, + { + "epoch": 0.39571935309324835, + "flos": 18195339910080.0, + "grad_norm": 2.8345612688695048, + "learning_rate": 2.753656878926872e-06, + "loss": 0.8572, + "num_input_tokens_seen": 70883080, + "step": 3291 + }, + { + "epoch": 0.39583959598388746, + "flos": 17723581118880.0, + "grad_norm": 1.963194016121199, + "learning_rate": 2.752935273531023e-06, + "loss": 0.7403, + "num_input_tokens_seen": 70901230, + "step": 3292 + }, + { + "epoch": 0.39595983887452657, + "flos": 19324734960480.0, + "grad_norm": 2.910422170893699, + "learning_rate": 2.752213553915492e-06, + "loss": 0.7876, + "num_input_tokens_seen": 70919585, + "step": 3293 + }, + { + "epoch": 0.3960800817651656, + "flos": 60595494648000.0, + "grad_norm": 0.9204280942590546, + "learning_rate": 2.751491720189762e-06, + "loss": 0.6605, + "num_input_tokens_seen": 70977695, + "step": 3294 + }, + { + "epoch": 0.39620032465580474, + "flos": 16812815828640.0, + "grad_norm": 2.232295849597861, + "learning_rate": 2.7507697724633364e-06, + "loss": 0.9169, + "num_input_tokens_seen": 70994455, + "step": 3295 + }, + { + "epoch": 0.3963205675464438, + "flos": 68972349736320.0, + "grad_norm": 0.7828477162493693, + "learning_rate": 2.7500477108457327e-06, + "loss": 0.546, + "num_input_tokens_seen": 71046465, + "step": 3296 + }, + { + "epoch": 0.3964408104370829, + "flos": 25630721752800.0, + "grad_norm": 2.9143398449516074, + "learning_rate": 2.7493255354464877e-06, + "loss": 0.8082, + "num_input_tokens_seen": 71064275, + "step": 3297 + }, + { + "epoch": 0.396561053327722, + "flos": 24242659377120.0, + "grad_norm": 2.0276391433402217, + "learning_rate": 2.748603246375156e-06, + "loss": 0.7616, + "num_input_tokens_seen": 71082290, + "step": 3298 + }, + { + "epoch": 0.39668129621836107, + "flos": 20493052483200.0, + "grad_norm": 2.443425434539899, + "learning_rate": 2.7478808437413055e-06, + "loss": 0.6983, + "num_input_tokens_seen": 71101700, + "step": 3299 + }, + { + "epoch": 0.3968015391090002, + "flos": 27013803380640.0, + "grad_norm": 2.1263424615866873, + "learning_rate": 2.7471583276545263e-06, + "loss": 0.6618, + "num_input_tokens_seen": 71122360, + "step": 3300 + }, + { + "epoch": 0.3969217819996393, + "flos": 12514199915040.0, + "grad_norm": 2.5014903394690595, + "learning_rate": 2.7464356982244224e-06, + "loss": 0.702, + "num_input_tokens_seen": 71140360, + "step": 3301 + }, + { + "epoch": 0.39704202489027834, + "flos": 66147000755520.0, + "grad_norm": 2.126143091957653, + "learning_rate": 2.745712955560617e-06, + "loss": 0.6166, + "num_input_tokens_seen": 71196565, + "step": 3302 + }, + { + "epoch": 0.39716226778091746, + "flos": 16958878750080.0, + "grad_norm": 3.1028447419350056, + "learning_rate": 2.7449900997727496e-06, + "loss": 0.7753, + "num_input_tokens_seen": 71214675, + "step": 3303 + }, + { + "epoch": 0.39728251067155657, + "flos": 23443569246720.0, + "grad_norm": 2.1288020957176674, + "learning_rate": 2.744267130970476e-06, + "loss": 0.8425, + "num_input_tokens_seen": 71234400, + "step": 3304 + }, + { + "epoch": 0.3974027535621956, + "flos": 20674915617120.0, + "grad_norm": 3.927679257833688, + "learning_rate": 2.7435440492634697e-06, + "loss": 0.7757, + "num_input_tokens_seen": 71253725, + "step": 3305 + }, + { + "epoch": 0.39752299645283473, + "flos": 21512519352480.0, + "grad_norm": 2.9557820633287557, + "learning_rate": 2.7428208547614228e-06, + "loss": 0.6755, + "num_input_tokens_seen": 71273220, + "step": 3306 + }, + { + "epoch": 0.39764323934347384, + "flos": 19180939394400.0, + "grad_norm": 4.35052423058792, + "learning_rate": 2.742097547574043e-06, + "loss": 0.7671, + "num_input_tokens_seen": 71291445, + "step": 3307 + }, + { + "epoch": 0.3977634822341129, + "flos": 20820458161920.0, + "grad_norm": 2.9394469606455296, + "learning_rate": 2.7413741278110544e-06, + "loss": 0.7807, + "num_input_tokens_seen": 71310135, + "step": 3308 + }, + { + "epoch": 0.397883725124752, + "flos": 39728717231520.0, + "grad_norm": 2.563040420487356, + "learning_rate": 2.7406505955822016e-06, + "loss": 0.6853, + "num_input_tokens_seen": 71331160, + "step": 3309 + }, + { + "epoch": 0.39800396801539106, + "flos": 17354651084640.0, + "grad_norm": 6.89896585205934, + "learning_rate": 2.7399269509972415e-06, + "loss": 0.6657, + "num_input_tokens_seen": 71345315, + "step": 3310 + }, + { + "epoch": 0.3981242109060302, + "flos": 19175215251360.0, + "grad_norm": 2.2728749001847084, + "learning_rate": 2.7392031941659514e-06, + "loss": 0.8563, + "num_input_tokens_seen": 71363160, + "step": 3311 + }, + { + "epoch": 0.3982444537966693, + "flos": 24529730132640.0, + "grad_norm": 1.998705266054705, + "learning_rate": 2.7384793251981244e-06, + "loss": 0.8629, + "num_input_tokens_seen": 71382145, + "step": 3312 + }, + { + "epoch": 0.39836469668730834, + "flos": 26176311154560.0, + "grad_norm": 2.646309957352097, + "learning_rate": 2.737755344203571e-06, + "loss": 0.8055, + "num_input_tokens_seen": 71402095, + "step": 3313 + }, + { + "epoch": 0.39848493957794745, + "flos": 27596493936480.0, + "grad_norm": 1.807089326142142, + "learning_rate": 2.7370312512921186e-06, + "loss": 0.8016, + "num_input_tokens_seen": 71423875, + "step": 3314 + }, + { + "epoch": 0.39860518246858656, + "flos": 12222111241920.0, + "grad_norm": 2.5684161858022954, + "learning_rate": 2.736307046573611e-06, + "loss": 0.7701, + "num_input_tokens_seen": 71439745, + "step": 3315 + }, + { + "epoch": 0.3987254253592256, + "flos": 22347037997760.0, + "grad_norm": 1.8354548692425994, + "learning_rate": 2.73558273015791e-06, + "loss": 0.8204, + "num_input_tokens_seen": 71459575, + "step": 3316 + }, + { + "epoch": 0.3988456682498647, + "flos": 23480410212480.0, + "grad_norm": 4.799161384757575, + "learning_rate": 2.734858302154894e-06, + "loss": 0.7094, + "num_input_tokens_seen": 71481075, + "step": 3317 + }, + { + "epoch": 0.39896591114050384, + "flos": 19181422601280.0, + "grad_norm": 2.5034519205903027, + "learning_rate": 2.734133762674457e-06, + "loss": 0.7651, + "num_input_tokens_seen": 71500665, + "step": 3318 + }, + { + "epoch": 0.3990861540311429, + "flos": 28360229891520.0, + "grad_norm": 4.226135570801411, + "learning_rate": 2.7334091118265124e-06, + "loss": 0.7078, + "num_input_tokens_seen": 71522560, + "step": 3319 + }, + { + "epoch": 0.399206396921782, + "flos": 61670870569920.0, + "grad_norm": 0.6901695700308554, + "learning_rate": 2.732684349720989e-06, + "loss": 0.5781, + "num_input_tokens_seen": 71590520, + "step": 3320 + }, + { + "epoch": 0.3993266398124211, + "flos": 28034311003200.0, + "grad_norm": 1.8953737253739407, + "learning_rate": 2.7319594764678318e-06, + "loss": 0.7553, + "num_input_tokens_seen": 71612620, + "step": 3321 + }, + { + "epoch": 0.39944688270306017, + "flos": 23192001345600.0, + "grad_norm": 1.7436007254637302, + "learning_rate": 2.7312344921770044e-06, + "loss": 0.8299, + "num_input_tokens_seen": 71634320, + "step": 3322 + }, + { + "epoch": 0.3995671255936993, + "flos": 19362802528320.0, + "grad_norm": 2.1062708867789137, + "learning_rate": 2.7305093969584857e-06, + "loss": 0.7844, + "num_input_tokens_seen": 71653145, + "step": 3323 + }, + { + "epoch": 0.3996873684843384, + "flos": 23808559286400.0, + "grad_norm": 2.8510111236515825, + "learning_rate": 2.729784190922272e-06, + "loss": 0.7968, + "num_input_tokens_seen": 71674580, + "step": 3324 + }, + { + "epoch": 0.39980761137497745, + "flos": 66482249253600.0, + "grad_norm": 0.9782789480916145, + "learning_rate": 2.729058874178378e-06, + "loss": 0.5714, + "num_input_tokens_seen": 71745260, + "step": 3325 + }, + { + "epoch": 0.39992785426561656, + "flos": 28508225640480.0, + "grad_norm": 4.966852749575119, + "learning_rate": 2.7283334468368315e-06, + "loss": 0.6898, + "num_input_tokens_seen": 71766540, + "step": 3326 + }, + { + "epoch": 0.4000480971562556, + "flos": 15028126213920.0, + "grad_norm": 2.7465457028847466, + "learning_rate": 2.72760790900768e-06, + "loss": 0.7307, + "num_input_tokens_seen": 71783125, + "step": 3327 + }, + { + "epoch": 0.4001683400468947, + "flos": 23881423483200.0, + "grad_norm": 2.5565001796675344, + "learning_rate": 2.7268822608009875e-06, + "loss": 0.7909, + "num_input_tokens_seen": 71802500, + "step": 3328 + }, + { + "epoch": 0.40028858293753383, + "flos": 24317828099040.0, + "grad_norm": 2.085239822353711, + "learning_rate": 2.726156502326834e-06, + "loss": 0.783, + "num_input_tokens_seen": 71823800, + "step": 3329 + }, + { + "epoch": 0.4004088258281729, + "flos": 66691772422560.0, + "grad_norm": 0.7028418076622224, + "learning_rate": 2.725430633695316e-06, + "loss": 0.6026, + "num_input_tokens_seen": 71886880, + "step": 3330 + }, + { + "epoch": 0.400529068718812, + "flos": 58514479007520.0, + "grad_norm": 0.9067660210905537, + "learning_rate": 2.7247046550165485e-06, + "loss": 0.5793, + "num_input_tokens_seen": 71939005, + "step": 3331 + }, + { + "epoch": 0.4006493116094511, + "flos": 25341234962880.0, + "grad_norm": 1.6494397194709285, + "learning_rate": 2.7239785664006606e-06, + "loss": 0.7602, + "num_input_tokens_seen": 71962545, + "step": 3332 + }, + { + "epoch": 0.40076955450009016, + "flos": 60194518547040.0, + "grad_norm": 0.7719596461165514, + "learning_rate": 2.7232523679578002e-06, + "loss": 0.6174, + "num_input_tokens_seen": 72025625, + "step": 3333 + }, + { + "epoch": 0.4008897973907293, + "flos": 16593665691840.0, + "grad_norm": 2.629590697838578, + "learning_rate": 2.7225260597981295e-06, + "loss": 0.7978, + "num_input_tokens_seen": 72043330, + "step": 3334 + }, + { + "epoch": 0.4010100402813684, + "flos": 15355569062400.0, + "grad_norm": 3.31096123996953, + "learning_rate": 2.721799642031831e-06, + "loss": 0.7813, + "num_input_tokens_seen": 72059500, + "step": 3335 + }, + { + "epoch": 0.40113028317200744, + "flos": 13280128885920.0, + "grad_norm": 2.318691625143056, + "learning_rate": 2.721073114769101e-06, + "loss": 0.7753, + "num_input_tokens_seen": 72077175, + "step": 3336 + }, + { + "epoch": 0.40125052606264655, + "flos": 20638557858240.0, + "grad_norm": 2.5050286263602413, + "learning_rate": 2.7203464781201523e-06, + "loss": 0.7518, + "num_input_tokens_seen": 72096490, + "step": 3337 + }, + { + "epoch": 0.40137076895328566, + "flos": 24572443920480.0, + "grad_norm": 2.2603200959233973, + "learning_rate": 2.719619732195215e-06, + "loss": 0.7792, + "num_input_tokens_seen": 72118130, + "step": 3338 + }, + { + "epoch": 0.4014910118439247, + "flos": 24171951026400.0, + "grad_norm": 1.8551745372209205, + "learning_rate": 2.7188928771045377e-06, + "loss": 0.7271, + "num_input_tokens_seen": 72139450, + "step": 3339 + }, + { + "epoch": 0.4016112547345638, + "flos": 26686806569280.0, + "grad_norm": 2.4296706220525115, + "learning_rate": 2.7181659129583815e-06, + "loss": 0.8032, + "num_input_tokens_seen": 72159040, + "step": 3340 + }, + { + "epoch": 0.4017314976252029, + "flos": 21257643342720.0, + "grad_norm": 1.9598793123028582, + "learning_rate": 2.7174388398670276e-06, + "loss": 0.7596, + "num_input_tokens_seen": 72178220, + "step": 3341 + }, + { + "epoch": 0.401851740515842, + "flos": 25448226732960.0, + "grad_norm": 2.815846296244649, + "learning_rate": 2.716711657940773e-06, + "loss": 0.9226, + "num_input_tokens_seen": 72199230, + "step": 3342 + }, + { + "epoch": 0.4019719834064811, + "flos": 55316371507200.0, + "grad_norm": 0.8200296913563332, + "learning_rate": 2.7159843672899284e-06, + "loss": 0.5641, + "num_input_tokens_seen": 72263430, + "step": 3343 + }, + { + "epoch": 0.40209222629712016, + "flos": 18154001403360.0, + "grad_norm": 3.2857604497426864, + "learning_rate": 2.715256968024825e-06, + "loss": 0.8178, + "num_input_tokens_seen": 72280505, + "step": 3344 + }, + { + "epoch": 0.40221246918775927, + "flos": 25923925518720.0, + "grad_norm": 1.6636800550379287, + "learning_rate": 2.7145294602558083e-06, + "loss": 0.8257, + "num_input_tokens_seen": 72301215, + "step": 3345 + }, + { + "epoch": 0.4023327120783984, + "flos": 33790248022560.0, + "grad_norm": 2.8030132088542197, + "learning_rate": 2.713801844093241e-06, + "loss": 0.71, + "num_input_tokens_seen": 72322365, + "step": 3346 + }, + { + "epoch": 0.40245295496903744, + "flos": 26861458769760.0, + "grad_norm": 2.900267647855835, + "learning_rate": 2.7130741196475014e-06, + "loss": 0.8842, + "num_input_tokens_seen": 72340495, + "step": 3347 + }, + { + "epoch": 0.40257319785967655, + "flos": 36849577874400.0, + "grad_norm": 2.7461155419821233, + "learning_rate": 2.7123462870289848e-06, + "loss": 0.7889, + "num_input_tokens_seen": 72360545, + "step": 3348 + }, + { + "epoch": 0.40269344075031566, + "flos": 24318348475680.0, + "grad_norm": 1.8483413813180454, + "learning_rate": 2.711618346348102e-06, + "loss": 0.8135, + "num_input_tokens_seen": 72381070, + "step": 3349 + }, + { + "epoch": 0.4028136836409547, + "flos": 14369226182880.0, + "grad_norm": 2.1801017996441185, + "learning_rate": 2.7108902977152825e-06, + "loss": 0.6365, + "num_input_tokens_seen": 72398970, + "step": 3350 + }, + { + "epoch": 0.4029339265315938, + "flos": 26098429040160.0, + "grad_norm": 2.31311408472462, + "learning_rate": 2.7101621412409704e-06, + "loss": 0.755, + "num_input_tokens_seen": 72418455, + "step": 3351 + }, + { + "epoch": 0.40305416942223293, + "flos": 23223006659040.0, + "grad_norm": 5.773161307246853, + "learning_rate": 2.7094338770356256e-06, + "loss": 0.8568, + "num_input_tokens_seen": 72437540, + "step": 3352 + }, + { + "epoch": 0.403174412312872, + "flos": 27050830195200.0, + "grad_norm": 2.56745338850789, + "learning_rate": 2.708705505209726e-06, + "loss": 0.6472, + "num_input_tokens_seen": 72458315, + "step": 3353 + }, + { + "epoch": 0.4032946552035111, + "flos": 21725053272000.0, + "grad_norm": 2.8661072902362723, + "learning_rate": 2.7079770258737646e-06, + "loss": 0.9188, + "num_input_tokens_seen": 72476225, + "step": 3354 + }, + { + "epoch": 0.4034148980941502, + "flos": 17319185400000.0, + "grad_norm": 2.367663445797548, + "learning_rate": 2.707248439138251e-06, + "loss": 0.7559, + "num_input_tokens_seen": 72492460, + "step": 3355 + }, + { + "epoch": 0.40353514098478926, + "flos": 21986396820000.0, + "grad_norm": 21.774595188096352, + "learning_rate": 2.7065197451137114e-06, + "loss": 0.6567, + "num_input_tokens_seen": 72513840, + "step": 3356 + }, + { + "epoch": 0.4036553838754284, + "flos": 14226322691040.0, + "grad_norm": 3.346306652320547, + "learning_rate": 2.7057909439106894e-06, + "loss": 0.6742, + "num_input_tokens_seen": 72531695, + "step": 3357 + }, + { + "epoch": 0.40377562676606743, + "flos": 24755161958880.0, + "grad_norm": 2.6274655888098017, + "learning_rate": 2.7050620356397417e-06, + "loss": 0.7864, + "num_input_tokens_seen": 72550645, + "step": 3358 + }, + { + "epoch": 0.40389586965670654, + "flos": 24026817348960.0, + "grad_norm": 2.0740690877768713, + "learning_rate": 2.7043330204114437e-06, + "loss": 0.7246, + "num_input_tokens_seen": 72569355, + "step": 3359 + }, + { + "epoch": 0.40401611254734565, + "flos": 16376374043040.0, + "grad_norm": 2.759750420163352, + "learning_rate": 2.7036038983363862e-06, + "loss": 0.8577, + "num_input_tokens_seen": 72585960, + "step": 3360 + }, + { + "epoch": 0.4041363554379847, + "flos": 23953915982400.0, + "grad_norm": 1.8751792822002558, + "learning_rate": 2.702874669525177e-06, + "loss": 0.8435, + "num_input_tokens_seen": 72604440, + "step": 3361 + }, + { + "epoch": 0.4042565983286238, + "flos": 28361196305280.0, + "grad_norm": 2.379539067328611, + "learning_rate": 2.7021453340884394e-06, + "loss": 0.7006, + "num_input_tokens_seen": 72622165, + "step": 3362 + }, + { + "epoch": 0.40437684121926293, + "flos": 17685736569600.0, + "grad_norm": 3.7199787065450303, + "learning_rate": 2.7014158921368125e-06, + "loss": 0.729, + "num_input_tokens_seen": 72640490, + "step": 3363 + }, + { + "epoch": 0.404497084109902, + "flos": 23984400919200.0, + "grad_norm": 2.1900104830765326, + "learning_rate": 2.700686343780953e-06, + "loss": 0.8552, + "num_input_tokens_seen": 72660360, + "step": 3364 + }, + { + "epoch": 0.4046173270005411, + "flos": 22897087770720.0, + "grad_norm": 2.4331173744346732, + "learning_rate": 2.699956689131532e-06, + "loss": 0.8794, + "num_input_tokens_seen": 72680345, + "step": 3365 + }, + { + "epoch": 0.4047375698911802, + "flos": 20638669367520.0, + "grad_norm": 2.164482654314337, + "learning_rate": 2.699226928299238e-06, + "loss": 0.8453, + "num_input_tokens_seen": 72698365, + "step": 3366 + }, + { + "epoch": 0.40485781278181926, + "flos": 28871357192160.0, + "grad_norm": 2.7591487251726883, + "learning_rate": 2.698497061394774e-06, + "loss": 0.79, + "num_input_tokens_seen": 72716850, + "step": 3367 + }, + { + "epoch": 0.40497805567245837, + "flos": 23115271493760.0, + "grad_norm": 1.9339995603787732, + "learning_rate": 2.6977670885288627e-06, + "loss": 0.8104, + "num_input_tokens_seen": 72738250, + "step": 3368 + }, + { + "epoch": 0.4050982985630975, + "flos": 16266297182880.0, + "grad_norm": 2.5064009737504174, + "learning_rate": 2.6970370098122378e-06, + "loss": 0.7545, + "num_input_tokens_seen": 72755235, + "step": 3369 + }, + { + "epoch": 0.40521854145373654, + "flos": 34409928223200.0, + "grad_norm": 1.6890146028148272, + "learning_rate": 2.6963068253556535e-06, + "loss": 0.8672, + "num_input_tokens_seen": 72776620, + "step": 3370 + }, + { + "epoch": 0.40533878434437565, + "flos": 25294581180480.0, + "grad_norm": 2.116996165283363, + "learning_rate": 2.6955765352698763e-06, + "loss": 0.8555, + "num_input_tokens_seen": 72796765, + "step": 3371 + }, + { + "epoch": 0.40545902723501476, + "flos": 14990802041280.0, + "grad_norm": 2.9233812335913463, + "learning_rate": 2.6948461396656923e-06, + "loss": 0.7331, + "num_input_tokens_seen": 72814175, + "step": 3372 + }, + { + "epoch": 0.4055792701256538, + "flos": 25484584491840.0, + "grad_norm": 2.7320557267857284, + "learning_rate": 2.6941156386539013e-06, + "loss": 0.7458, + "num_input_tokens_seen": 72834125, + "step": 3373 + }, + { + "epoch": 0.4056995130162929, + "flos": 19546747168800.0, + "grad_norm": 4.876745253942488, + "learning_rate": 2.6933850323453203e-06, + "loss": 0.8131, + "num_input_tokens_seen": 72852570, + "step": 3374 + }, + { + "epoch": 0.405819755906932, + "flos": 15391963991040.0, + "grad_norm": 2.064243620001594, + "learning_rate": 2.6926543208507806e-06, + "loss": 0.75, + "num_input_tokens_seen": 72871250, + "step": 3375 + }, + { + "epoch": 0.4059399987975711, + "flos": 21403334566560.0, + "grad_norm": 2.3364587922987967, + "learning_rate": 2.6919235042811316e-06, + "loss": 0.8026, + "num_input_tokens_seen": 72890035, + "step": 3376 + }, + { + "epoch": 0.4060602416882102, + "flos": 25520347534560.0, + "grad_norm": 2.837026843982369, + "learning_rate": 2.691192582747237e-06, + "loss": 0.7636, + "num_input_tokens_seen": 72909665, + "step": 3377 + }, + { + "epoch": 0.40618048457884925, + "flos": 23732684339040.0, + "grad_norm": 3.556357056627075, + "learning_rate": 2.6904615563599765e-06, + "loss": 0.7415, + "num_input_tokens_seen": 72929465, + "step": 3378 + }, + { + "epoch": 0.40630072746948837, + "flos": 17614619351520.0, + "grad_norm": 2.4640086553347276, + "learning_rate": 2.6897304252302477e-06, + "loss": 0.8351, + "num_input_tokens_seen": 72946785, + "step": 3379 + }, + { + "epoch": 0.4064209703601275, + "flos": 60749326049280.0, + "grad_norm": 0.8902480683020545, + "learning_rate": 2.688999189468962e-06, + "loss": 0.5476, + "num_input_tokens_seen": 73003215, + "step": 3380 + }, + { + "epoch": 0.40654121325076653, + "flos": 23989976383200.0, + "grad_norm": 2.932876682138101, + "learning_rate": 2.6882678491870464e-06, + "loss": 0.7635, + "num_input_tokens_seen": 73023650, + "step": 3381 + }, + { + "epoch": 0.40666145614140564, + "flos": 27305037149280.0, + "grad_norm": 2.0077205889256104, + "learning_rate": 2.6875364044954453e-06, + "loss": 0.7173, + "num_input_tokens_seen": 73043880, + "step": 3382 + }, + { + "epoch": 0.40678169903204475, + "flos": 26139210000480.0, + "grad_norm": 1.6971313985375374, + "learning_rate": 2.6868048555051185e-06, + "loss": 0.8262, + "num_input_tokens_seen": 73065410, + "step": 3383 + }, + { + "epoch": 0.4069019419226838, + "flos": 28581387195360.0, + "grad_norm": 3.972527455930614, + "learning_rate": 2.686073202327041e-06, + "loss": 0.8543, + "num_input_tokens_seen": 73084410, + "step": 3384 + }, + { + "epoch": 0.4070221848133229, + "flos": 25192904686080.0, + "grad_norm": 1.894441079422295, + "learning_rate": 2.6853414450722043e-06, + "loss": 0.7318, + "num_input_tokens_seen": 73104075, + "step": 3385 + }, + { + "epoch": 0.40714242770396203, + "flos": 18379024362240.0, + "grad_norm": 2.178285537561179, + "learning_rate": 2.684609583851616e-06, + "loss": 0.853, + "num_input_tokens_seen": 73122250, + "step": 3386 + }, + { + "epoch": 0.4072626705946011, + "flos": 30185588957280.0, + "grad_norm": 4.735640655648779, + "learning_rate": 2.683877618776297e-06, + "loss": 0.8069, + "num_input_tokens_seen": 73144505, + "step": 3387 + }, + { + "epoch": 0.4073829134852402, + "flos": 21803344253760.0, + "grad_norm": 2.378755411211074, + "learning_rate": 2.6831455499572876e-06, + "loss": 0.7418, + "num_input_tokens_seen": 73162440, + "step": 3388 + }, + { + "epoch": 0.40750315637587925, + "flos": 25223872829760.0, + "grad_norm": 2.5658169190229807, + "learning_rate": 2.682413377505641e-06, + "loss": 0.7783, + "num_input_tokens_seen": 73181245, + "step": 3389 + }, + { + "epoch": 0.40762339926651836, + "flos": 19684446894240.0, + "grad_norm": 1.973152143637877, + "learning_rate": 2.6816811015324284e-06, + "loss": 0.768, + "num_input_tokens_seen": 73199295, + "step": 3390 + }, + { + "epoch": 0.40774364215715747, + "flos": 71346978010080.0, + "grad_norm": 0.7962514644351829, + "learning_rate": 2.6809487221487343e-06, + "loss": 0.5659, + "num_input_tokens_seen": 73258780, + "step": 3391 + }, + { + "epoch": 0.4078638850477965, + "flos": 15062736994080.0, + "grad_norm": 3.958232731126325, + "learning_rate": 2.6802162394656605e-06, + "loss": 0.8219, + "num_input_tokens_seen": 73275490, + "step": 3392 + }, + { + "epoch": 0.40798412793843564, + "flos": 23808819474720.0, + "grad_norm": 2.1647780574659237, + "learning_rate": 2.679483653594324e-06, + "loss": 0.7184, + "num_input_tokens_seen": 73297260, + "step": 3393 + }, + { + "epoch": 0.40810437082907475, + "flos": 21035593964640.0, + "grad_norm": 2.638209756385751, + "learning_rate": 2.678750964645857e-06, + "loss": 0.762, + "num_input_tokens_seen": 73316340, + "step": 3394 + }, + { + "epoch": 0.4082246137197138, + "flos": 11305287280800.0, + "grad_norm": 3.459176814769642, + "learning_rate": 2.6780181727314094e-06, + "loss": 0.833, + "num_input_tokens_seen": 73331245, + "step": 3395 + }, + { + "epoch": 0.4083448566103529, + "flos": 19035434019360.0, + "grad_norm": 2.091670241739123, + "learning_rate": 2.6772852779621435e-06, + "loss": 0.7776, + "num_input_tokens_seen": 73349105, + "step": 3396 + }, + { + "epoch": 0.408465099500992, + "flos": 23516582122560.0, + "grad_norm": 2.411844364079438, + "learning_rate": 2.676552280449239e-06, + "loss": 0.8671, + "num_input_tokens_seen": 73368990, + "step": 3397 + }, + { + "epoch": 0.4085853423916311, + "flos": 12732049110240.0, + "grad_norm": 3.2660027244449052, + "learning_rate": 2.6758191803038917e-06, + "loss": 0.7571, + "num_input_tokens_seen": 73387485, + "step": 3398 + }, + { + "epoch": 0.4087055852822702, + "flos": 24318236966400.0, + "grad_norm": 2.0030112247619125, + "learning_rate": 2.6750859776373125e-06, + "loss": 0.8259, + "num_input_tokens_seen": 73406940, + "step": 3399 + }, + { + "epoch": 0.4088258281729093, + "flos": 66293361035040.0, + "grad_norm": 0.7811605402271685, + "learning_rate": 2.674352672560727e-06, + "loss": 0.6037, + "num_input_tokens_seen": 73468385, + "step": 3400 + }, + { + "epoch": 0.40894607106354836, + "flos": 20420113946880.0, + "grad_norm": 4.497194899366785, + "learning_rate": 2.673619265185377e-06, + "loss": 0.7682, + "num_input_tokens_seen": 73488225, + "step": 3401 + }, + { + "epoch": 0.40906631395418747, + "flos": 27014881303680.0, + "grad_norm": 2.1887373558215146, + "learning_rate": 2.672885755622521e-06, + "loss": 0.7816, + "num_input_tokens_seen": 73510640, + "step": 3402 + }, + { + "epoch": 0.4091865568448266, + "flos": 25447669186560.0, + "grad_norm": 2.948418474565588, + "learning_rate": 2.67215214398343e-06, + "loss": 0.7006, + "num_input_tokens_seen": 73530655, + "step": 3403 + }, + { + "epoch": 0.40930679973546563, + "flos": 28616741370720.0, + "grad_norm": 3.079291263136298, + "learning_rate": 2.671418430379393e-06, + "loss": 0.7826, + "num_input_tokens_seen": 73549340, + "step": 3404 + }, + { + "epoch": 0.40942704262610474, + "flos": 20856481392960.0, + "grad_norm": 3.4557090515302535, + "learning_rate": 2.670684614921715e-06, + "loss": 0.8332, + "num_input_tokens_seen": 73568315, + "step": 3405 + }, + { + "epoch": 0.4095472855167438, + "flos": 21587279207040.0, + "grad_norm": 2.654647343664705, + "learning_rate": 2.6699506977217128e-06, + "loss": 0.6913, + "num_input_tokens_seen": 73588490, + "step": 3406 + }, + { + "epoch": 0.4096675284073829, + "flos": 27885980726400.0, + "grad_norm": 2.49724159652819, + "learning_rate": 2.6692166788907233e-06, + "loss": 0.702, + "num_input_tokens_seen": 73608685, + "step": 3407 + }, + { + "epoch": 0.409787771298022, + "flos": 19179192415680.0, + "grad_norm": 2.4428039944156503, + "learning_rate": 2.6684825585400957e-06, + "loss": 0.7711, + "num_input_tokens_seen": 73627630, + "step": 3408 + }, + { + "epoch": 0.4099080141886611, + "flos": 59184864494400.0, + "grad_norm": 0.81757172984371, + "learning_rate": 2.6677483367811947e-06, + "loss": 0.651, + "num_input_tokens_seen": 73687670, + "step": 3409 + }, + { + "epoch": 0.4100282570793002, + "flos": 21875093357760.0, + "grad_norm": 2.0221949705912317, + "learning_rate": 2.6670140137254028e-06, + "loss": 0.7599, + "num_input_tokens_seen": 73707145, + "step": 3410 + }, + { + "epoch": 0.4101484999699393, + "flos": 18525198792960.0, + "grad_norm": 4.243337907291733, + "learning_rate": 2.666279589484115e-06, + "loss": 0.9073, + "num_input_tokens_seen": 73725045, + "step": 3411 + }, + { + "epoch": 0.41026874286057835, + "flos": 19067182728000.0, + "grad_norm": 2.188823408620151, + "learning_rate": 2.6655450641687435e-06, + "loss": 0.8074, + "num_input_tokens_seen": 73742610, + "step": 3412 + }, + { + "epoch": 0.41038898575121746, + "flos": 31164646563840.0, + "grad_norm": 2.443815705569736, + "learning_rate": 2.664810437890715e-06, + "loss": 0.6904, + "num_input_tokens_seen": 73764280, + "step": 3413 + }, + { + "epoch": 0.41050922864185657, + "flos": 14335581816480.0, + "grad_norm": 1.933946488698351, + "learning_rate": 2.6640757107614714e-06, + "loss": 0.7987, + "num_input_tokens_seen": 73782375, + "step": 3414 + }, + { + "epoch": 0.4106294715324956, + "flos": 30911814890880.0, + "grad_norm": 2.324228905407173, + "learning_rate": 2.6633408828924697e-06, + "loss": 0.6921, + "num_input_tokens_seen": 73801240, + "step": 3415 + }, + { + "epoch": 0.41074971442313474, + "flos": 24422143646400.0, + "grad_norm": 1.811999679439087, + "learning_rate": 2.662605954395185e-06, + "loss": 0.7015, + "num_input_tokens_seen": 73821200, + "step": 3416 + }, + { + "epoch": 0.41086995731377385, + "flos": 21512221994400.0, + "grad_norm": 2.1319737164879426, + "learning_rate": 2.6618709253811027e-06, + "loss": 0.8385, + "num_input_tokens_seen": 73840655, + "step": 3417 + }, + { + "epoch": 0.4109902002044129, + "flos": 20673168638400.0, + "grad_norm": 4.722360808389619, + "learning_rate": 2.6611357959617277e-06, + "loss": 0.88, + "num_input_tokens_seen": 73860275, + "step": 3418 + }, + { + "epoch": 0.411110443095052, + "flos": 18153889894080.0, + "grad_norm": 4.5964424391384995, + "learning_rate": 2.660400566248578e-06, + "loss": 0.9103, + "num_input_tokens_seen": 73878400, + "step": 3419 + }, + { + "epoch": 0.41123068598569107, + "flos": 14554657613760.0, + "grad_norm": 4.531937119575484, + "learning_rate": 2.6596652363531876e-06, + "loss": 0.6714, + "num_input_tokens_seen": 73894275, + "step": 3420 + }, + { + "epoch": 0.4113509288763302, + "flos": 21147752331360.0, + "grad_norm": 6.3444956073527825, + "learning_rate": 2.6589298063871055e-06, + "loss": 0.7868, + "num_input_tokens_seen": 73914450, + "step": 3421 + }, + { + "epoch": 0.4114711717669693, + "flos": 18415976837280.0, + "grad_norm": 2.254859550821981, + "learning_rate": 2.658194276461895e-06, + "loss": 0.6977, + "num_input_tokens_seen": 73932375, + "step": 3422 + }, + { + "epoch": 0.41159141465760835, + "flos": 27194291233440.0, + "grad_norm": 2.4558963550169812, + "learning_rate": 2.6574586466891368e-06, + "loss": 0.6729, + "num_input_tokens_seen": 73952850, + "step": 3423 + }, + { + "epoch": 0.41171165754824746, + "flos": 19978208206560.0, + "grad_norm": 2.7076590554531403, + "learning_rate": 2.6567229171804247e-06, + "loss": 0.6496, + "num_input_tokens_seen": 73970735, + "step": 3424 + }, + { + "epoch": 0.41183190043888657, + "flos": 17978717316960.0, + "grad_norm": 2.6829989688969413, + "learning_rate": 2.655987088047368e-06, + "loss": 0.8768, + "num_input_tokens_seen": 73989080, + "step": 3425 + }, + { + "epoch": 0.4119521433295256, + "flos": 27124474956960.0, + "grad_norm": 2.116524004696681, + "learning_rate": 2.6552511594015912e-06, + "loss": 0.7848, + "num_input_tokens_seen": 74009470, + "step": 3426 + }, + { + "epoch": 0.41207238622016473, + "flos": 15100618713120.0, + "grad_norm": 2.352569465868296, + "learning_rate": 2.654515131354735e-06, + "loss": 0.8543, + "num_input_tokens_seen": 74027735, + "step": 3427 + }, + { + "epoch": 0.41219262911080384, + "flos": 27014100738720.0, + "grad_norm": 2.358469606198438, + "learning_rate": 2.653779004018453e-06, + "loss": 0.854, + "num_input_tokens_seen": 74048300, + "step": 3428 + }, + { + "epoch": 0.4123128720014429, + "flos": 24651218109120.0, + "grad_norm": 2.5841400261073373, + "learning_rate": 2.653042777504417e-06, + "loss": 0.8238, + "num_input_tokens_seen": 74070890, + "step": 3429 + }, + { + "epoch": 0.412433114892082, + "flos": 26206981940160.0, + "grad_norm": 2.2314335595286767, + "learning_rate": 2.6523064519243105e-06, + "loss": 0.8011, + "num_input_tokens_seen": 74090460, + "step": 3430 + }, + { + "epoch": 0.4125533577827211, + "flos": 21330544709280.0, + "grad_norm": 2.7160507482975866, + "learning_rate": 2.6515700273898333e-06, + "loss": 0.7893, + "num_input_tokens_seen": 74108335, + "step": 3431 + }, + { + "epoch": 0.4126736006733602, + "flos": 26030211063360.0, + "grad_norm": 3.0926886385201375, + "learning_rate": 2.6508335040127018e-06, + "loss": 0.6869, + "num_input_tokens_seen": 74128030, + "step": 3432 + }, + { + "epoch": 0.4127938435639993, + "flos": 25629383641440.0, + "grad_norm": 1.6708315150884436, + "learning_rate": 2.6500968819046446e-06, + "loss": 0.7706, + "num_input_tokens_seen": 74148330, + "step": 3433 + }, + { + "epoch": 0.4129140864546384, + "flos": 17970019593120.0, + "grad_norm": 2.715944618581832, + "learning_rate": 2.649360161177408e-06, + "loss": 0.5921, + "num_input_tokens_seen": 74163390, + "step": 3434 + }, + { + "epoch": 0.41303432934527745, + "flos": 23698965633120.0, + "grad_norm": 6.606569673279802, + "learning_rate": 2.6486233419427504e-06, + "loss": 0.7344, + "num_input_tokens_seen": 74183205, + "step": 3435 + }, + { + "epoch": 0.41315457223591656, + "flos": 19727123512320.0, + "grad_norm": 3.79852208961691, + "learning_rate": 2.6478864243124484e-06, + "loss": 0.7565, + "num_input_tokens_seen": 74202870, + "step": 3436 + }, + { + "epoch": 0.4132748151265556, + "flos": 20893471037760.0, + "grad_norm": 2.0220487682072115, + "learning_rate": 2.6471494083982903e-06, + "loss": 0.8505, + "num_input_tokens_seen": 74222255, + "step": 3437 + }, + { + "epoch": 0.4133950580171947, + "flos": 32186157769920.0, + "grad_norm": 2.2301755268143384, + "learning_rate": 2.6464122943120818e-06, + "loss": 0.7465, + "num_input_tokens_seen": 74242480, + "step": 3438 + }, + { + "epoch": 0.41351530090783384, + "flos": 23259104229600.0, + "grad_norm": 4.965988255364383, + "learning_rate": 2.645675082165642e-06, + "loss": 0.8199, + "num_input_tokens_seen": 74258690, + "step": 3439 + }, + { + "epoch": 0.4136355437984729, + "flos": 25556928312000.0, + "grad_norm": 2.4557479151107513, + "learning_rate": 2.644937772070806e-06, + "loss": 0.7511, + "num_input_tokens_seen": 74277135, + "step": 3440 + }, + { + "epoch": 0.413755786689112, + "flos": 19800359406720.0, + "grad_norm": 3.408219494274802, + "learning_rate": 2.6442003641394225e-06, + "loss": 0.8367, + "num_input_tokens_seen": 74294730, + "step": 3441 + }, + { + "epoch": 0.4138760295797511, + "flos": 26831940246720.0, + "grad_norm": 1.6132612250242135, + "learning_rate": 2.643462858483356e-06, + "loss": 0.8398, + "num_input_tokens_seen": 74315015, + "step": 3442 + }, + { + "epoch": 0.41399627247039017, + "flos": 16376002345440.0, + "grad_norm": 2.686967042818105, + "learning_rate": 2.6427252552144856e-06, + "loss": 0.7289, + "num_input_tokens_seen": 74333665, + "step": 3443 + }, + { + "epoch": 0.4141165153610293, + "flos": 22897459468320.0, + "grad_norm": 2.412821923158345, + "learning_rate": 2.6419875544447044e-06, + "loss": 0.7496, + "num_input_tokens_seen": 74355065, + "step": 3444 + }, + { + "epoch": 0.4142367582516684, + "flos": 25156509757440.0, + "grad_norm": 1.9270327449722573, + "learning_rate": 2.6412497562859218e-06, + "loss": 0.7167, + "num_input_tokens_seen": 74376345, + "step": 3445 + }, + { + "epoch": 0.41435700114230745, + "flos": 21659548687680.0, + "grad_norm": 6.277249556892869, + "learning_rate": 2.6405118608500617e-06, + "loss": 0.7589, + "num_input_tokens_seen": 74395170, + "step": 3446 + }, + { + "epoch": 0.41447724403294656, + "flos": 25957941582720.0, + "grad_norm": 2.183356256825709, + "learning_rate": 2.6397738682490613e-06, + "loss": 0.8131, + "num_input_tokens_seen": 74416070, + "step": 3447 + }, + { + "epoch": 0.41459748692358567, + "flos": 18233704836000.0, + "grad_norm": 2.0136156342549265, + "learning_rate": 2.6390357785948734e-06, + "loss": 0.751, + "num_input_tokens_seen": 74433185, + "step": 3448 + }, + { + "epoch": 0.4147177298142247, + "flos": 24133809119040.0, + "grad_norm": 1.9156866539746082, + "learning_rate": 2.6382975919994667e-06, + "loss": 0.8062, + "num_input_tokens_seen": 74453760, + "step": 3449 + }, + { + "epoch": 0.41483797270486383, + "flos": 20056053151200.0, + "grad_norm": 1.776314439457014, + "learning_rate": 2.637559308574822e-06, + "loss": 0.728, + "num_input_tokens_seen": 74473505, + "step": 3450 + }, + { + "epoch": 0.4149582155955029, + "flos": 30037853396640.0, + "grad_norm": 2.407534764119076, + "learning_rate": 2.6368209284329376e-06, + "loss": 0.7107, + "num_input_tokens_seen": 74494135, + "step": 3451 + }, + { + "epoch": 0.415078458486142, + "flos": 16740620687520.0, + "grad_norm": 2.3142654278453065, + "learning_rate": 2.636082451685825e-06, + "loss": 0.7529, + "num_input_tokens_seen": 74512335, + "step": 3452 + }, + { + "epoch": 0.4151987013767811, + "flos": 25996492357440.0, + "grad_norm": 1.9577147000447332, + "learning_rate": 2.6353438784455094e-06, + "loss": 0.8617, + "num_input_tokens_seen": 74535620, + "step": 3453 + }, + { + "epoch": 0.41531894426742016, + "flos": 24572778448320.0, + "grad_norm": 3.9550383389202564, + "learning_rate": 2.6346052088240326e-06, + "loss": 0.7137, + "num_input_tokens_seen": 74555140, + "step": 3454 + }, + { + "epoch": 0.4154391871580593, + "flos": 14954592961440.0, + "grad_norm": 2.284353888084033, + "learning_rate": 2.63386644293345e-06, + "loss": 0.7743, + "num_input_tokens_seen": 74571485, + "step": 3455 + }, + { + "epoch": 0.4155594300486984, + "flos": 14626443887520.0, + "grad_norm": 3.241536785227104, + "learning_rate": 2.633127580885833e-06, + "loss": 0.8313, + "num_input_tokens_seen": 74585985, + "step": 3456 + }, + { + "epoch": 0.41567967293933744, + "flos": 29454939822240.0, + "grad_norm": 6.463474057131121, + "learning_rate": 2.632388622793265e-06, + "loss": 0.6516, + "num_input_tokens_seen": 74605140, + "step": 3457 + }, + { + "epoch": 0.41579991582997655, + "flos": 19210755275520.0, + "grad_norm": 2.4649188095247334, + "learning_rate": 2.6316495687678457e-06, + "loss": 0.6835, + "num_input_tokens_seen": 74623790, + "step": 3458 + }, + { + "epoch": 0.41592015872061566, + "flos": 24427050054720.0, + "grad_norm": 5.110798582021805, + "learning_rate": 2.6309104189216887e-06, + "loss": 0.7686, + "num_input_tokens_seen": 74641355, + "step": 3459 + }, + { + "epoch": 0.4160404016112547, + "flos": 20745958495680.0, + "grad_norm": 2.2407559067966316, + "learning_rate": 2.630171173366923e-06, + "loss": 0.7443, + "num_input_tokens_seen": 74657155, + "step": 3460 + }, + { + "epoch": 0.41616064450189383, + "flos": 13896761166240.0, + "grad_norm": 2.7083339006472262, + "learning_rate": 2.629431832215691e-06, + "loss": 0.7513, + "num_input_tokens_seen": 74671960, + "step": 3461 + }, + { + "epoch": 0.41628088739253294, + "flos": 19982296880160.0, + "grad_norm": 1.9141586590960487, + "learning_rate": 2.628692395580151e-06, + "loss": 0.8701, + "num_input_tokens_seen": 74692050, + "step": 3462 + }, + { + "epoch": 0.416401130283172, + "flos": 29126939427360.0, + "grad_norm": 7.462219703304816, + "learning_rate": 2.6279528635724747e-06, + "loss": 0.7902, + "num_input_tokens_seen": 74712205, + "step": 3463 + }, + { + "epoch": 0.4165213731738111, + "flos": 16223546225280.0, + "grad_norm": 2.8421972038686394, + "learning_rate": 2.627213236304848e-06, + "loss": 0.7809, + "num_input_tokens_seen": 74729005, + "step": 3464 + }, + { + "epoch": 0.4166416160644502, + "flos": 33716863449120.0, + "grad_norm": 4.118141020469981, + "learning_rate": 2.626473513889472e-06, + "loss": 0.7131, + "num_input_tokens_seen": 74751385, + "step": 3465 + }, + { + "epoch": 0.41676185895508927, + "flos": 20887449536640.0, + "grad_norm": 5.230801128057172, + "learning_rate": 2.625733696438562e-06, + "loss": 0.8292, + "num_input_tokens_seen": 74768890, + "step": 3466 + }, + { + "epoch": 0.4168821018457284, + "flos": 18379953606240.0, + "grad_norm": 1.9352812830204242, + "learning_rate": 2.6249937840643476e-06, + "loss": 0.7529, + "num_input_tokens_seen": 74787195, + "step": 3467 + }, + { + "epoch": 0.41700234473636744, + "flos": 18671819260800.0, + "grad_norm": 1.9497812564133234, + "learning_rate": 2.6242537768790733e-06, + "loss": 0.668, + "num_input_tokens_seen": 74806350, + "step": 3468 + }, + { + "epoch": 0.41712258762700655, + "flos": 30988842100800.0, + "grad_norm": 2.1755137527371295, + "learning_rate": 2.6235136749949975e-06, + "loss": 0.6904, + "num_input_tokens_seen": 74829800, + "step": 3469 + }, + { + "epoch": 0.41724283051764566, + "flos": 35862937636800.0, + "grad_norm": 2.3920545106245625, + "learning_rate": 2.6227734785243924e-06, + "loss": 0.6133, + "num_input_tokens_seen": 74849760, + "step": 3470 + }, + { + "epoch": 0.4173630734082847, + "flos": 25297220233440.0, + "grad_norm": 2.3934958181063792, + "learning_rate": 2.6220331875795466e-06, + "loss": 0.79, + "num_input_tokens_seen": 74869110, + "step": 3471 + }, + { + "epoch": 0.4174833162989238, + "flos": 26647066362240.0, + "grad_norm": 2.2084052476029803, + "learning_rate": 2.62129280227276e-06, + "loss": 0.7524, + "num_input_tokens_seen": 74889110, + "step": 3472 + }, + { + "epoch": 0.41760355918956293, + "flos": 74631373724160.0, + "grad_norm": 2.959679812047876, + "learning_rate": 2.62055232271635e-06, + "loss": 0.6875, + "num_input_tokens_seen": 74916260, + "step": 3473 + }, + { + "epoch": 0.417723802080202, + "flos": 14297216890560.0, + "grad_norm": 2.5232797106844993, + "learning_rate": 2.619811749022646e-06, + "loss": 0.8813, + "num_input_tokens_seen": 74931570, + "step": 3474 + }, + { + "epoch": 0.4178440449708411, + "flos": 14622949930080.0, + "grad_norm": 2.7900673474622875, + "learning_rate": 2.6190710813039917e-06, + "loss": 0.7129, + "num_input_tokens_seen": 74944695, + "step": 3475 + }, + { + "epoch": 0.4179642878614802, + "flos": 21476607630720.0, + "grad_norm": 2.5299007446058774, + "learning_rate": 2.618330319672747e-06, + "loss": 0.8368, + "num_input_tokens_seen": 74964115, + "step": 3476 + }, + { + "epoch": 0.41808453075211927, + "flos": 18415753818720.0, + "grad_norm": 2.90786370217464, + "learning_rate": 2.617589464241284e-06, + "loss": 0.9169, + "num_input_tokens_seen": 74978515, + "step": 3477 + }, + { + "epoch": 0.4182047736427584, + "flos": 20272006688640.0, + "grad_norm": 2.372149714667864, + "learning_rate": 2.6168485151219914e-06, + "loss": 0.7487, + "num_input_tokens_seen": 74998135, + "step": 3478 + }, + { + "epoch": 0.4183250165333975, + "flos": 18849630890880.0, + "grad_norm": 5.462324454762522, + "learning_rate": 2.616107472427269e-06, + "loss": 0.7136, + "num_input_tokens_seen": 75012830, + "step": 3479 + }, + { + "epoch": 0.41844525942403654, + "flos": 17714511697440.0, + "grad_norm": 2.6514491959002235, + "learning_rate": 2.615366336269533e-06, + "loss": 0.763, + "num_input_tokens_seen": 75026495, + "step": 3480 + }, + { + "epoch": 0.41856550231467565, + "flos": 18334972463040.0, + "grad_norm": 2.496850031989281, + "learning_rate": 2.6146251067612126e-06, + "loss": 0.8069, + "num_input_tokens_seen": 75041970, + "step": 3481 + }, + { + "epoch": 0.41868574520531476, + "flos": 22749166361280.0, + "grad_norm": 1.8565208263179398, + "learning_rate": 2.6138837840147525e-06, + "loss": 0.8266, + "num_input_tokens_seen": 75061005, + "step": 3482 + }, + { + "epoch": 0.4188059880959538, + "flos": 13679841215040.0, + "grad_norm": 2.4538901526777215, + "learning_rate": 2.6131423681426103e-06, + "loss": 0.7649, + "num_input_tokens_seen": 75076920, + "step": 3483 + }, + { + "epoch": 0.41892623098659293, + "flos": 37763911461600.0, + "grad_norm": 12.625774699699743, + "learning_rate": 2.6124008592572587e-06, + "loss": 0.7293, + "num_input_tokens_seen": 75100905, + "step": 3484 + }, + { + "epoch": 0.419046473877232, + "flos": 23225645712000.0, + "grad_norm": 2.384432682724514, + "learning_rate": 2.6116592574711835e-06, + "loss": 0.8168, + "num_input_tokens_seen": 75119205, + "step": 3485 + }, + { + "epoch": 0.4191667167678711, + "flos": 20711459224800.0, + "grad_norm": 3.5282931481188604, + "learning_rate": 2.6109175628968853e-06, + "loss": 0.84, + "num_input_tokens_seen": 75138970, + "step": 3486 + }, + { + "epoch": 0.4192869596585102, + "flos": 23552679693120.0, + "grad_norm": 3.1326196473235246, + "learning_rate": 2.610175775646878e-06, + "loss": 0.8301, + "num_input_tokens_seen": 75157550, + "step": 3487 + }, + { + "epoch": 0.41940720254914926, + "flos": 25045317804480.0, + "grad_norm": 2.6363725901245374, + "learning_rate": 2.6094338958336907e-06, + "loss": 0.7327, + "num_input_tokens_seen": 75176220, + "step": 3488 + }, + { + "epoch": 0.41952744543978837, + "flos": 15531893902080.0, + "grad_norm": 3.9729450312166836, + "learning_rate": 2.608691923569867e-06, + "loss": 0.8229, + "num_input_tokens_seen": 75193095, + "step": 3489 + }, + { + "epoch": 0.4196476883304275, + "flos": 24609507904800.0, + "grad_norm": 1.767567806319156, + "learning_rate": 2.6079498589679616e-06, + "loss": 0.758, + "num_input_tokens_seen": 75214185, + "step": 3490 + }, + { + "epoch": 0.41976793122106654, + "flos": 24496085766240.0, + "grad_norm": 2.831519350287878, + "learning_rate": 2.6072077021405465e-06, + "loss": 0.7628, + "num_input_tokens_seen": 75233020, + "step": 3491 + }, + { + "epoch": 0.41988817411170565, + "flos": 21144964599360.0, + "grad_norm": 2.1516974113901575, + "learning_rate": 2.6064654532002054e-06, + "loss": 0.6944, + "num_input_tokens_seen": 75252030, + "step": 3492 + }, + { + "epoch": 0.42000841700234476, + "flos": 31603913251200.0, + "grad_norm": 1.8277881018447848, + "learning_rate": 2.6057231122595375e-06, + "loss": 0.7601, + "num_input_tokens_seen": 75273340, + "step": 3493 + }, + { + "epoch": 0.4201286598929838, + "flos": 21250915616160.0, + "grad_norm": 1.9828975914747506, + "learning_rate": 2.604980679431154e-06, + "loss": 0.7289, + "num_input_tokens_seen": 75291580, + "step": 3494 + }, + { + "epoch": 0.4202489027836229, + "flos": 18519957856800.0, + "grad_norm": 2.33637940172423, + "learning_rate": 2.604238154827684e-06, + "loss": 0.7429, + "num_input_tokens_seen": 75308205, + "step": 3495 + }, + { + "epoch": 0.42036914567426203, + "flos": 19290124180320.0, + "grad_norm": 2.2229649441692443, + "learning_rate": 2.6034955385617656e-06, + "loss": 0.7345, + "num_input_tokens_seen": 75326535, + "step": 3496 + }, + { + "epoch": 0.4204893885649011, + "flos": 67746073090560.0, + "grad_norm": 0.72791226435005, + "learning_rate": 2.6027528307460544e-06, + "loss": 0.6157, + "num_input_tokens_seen": 75390540, + "step": 3497 + }, + { + "epoch": 0.4206096314555402, + "flos": 21877509392160.0, + "grad_norm": 2.0817711758957285, + "learning_rate": 2.602010031493217e-06, + "loss": 0.8669, + "num_input_tokens_seen": 75408770, + "step": 3498 + }, + { + "epoch": 0.42072987434617926, + "flos": 29236161383040.0, + "grad_norm": 3.1377981948488465, + "learning_rate": 2.6012671409159367e-06, + "loss": 0.8641, + "num_input_tokens_seen": 75430105, + "step": 3499 + }, + { + "epoch": 0.42085011723681837, + "flos": 27561176930880.0, + "grad_norm": 2.1765722620978587, + "learning_rate": 2.6005241591269097e-06, + "loss": 0.8202, + "num_input_tokens_seen": 75449475, + "step": 3500 + }, + { + "epoch": 0.4209703601274575, + "flos": 27778728768000.0, + "grad_norm": 1.8589634800566488, + "learning_rate": 2.5997810862388454e-06, + "loss": 0.7971, + "num_input_tokens_seen": 75469315, + "step": 3501 + }, + { + "epoch": 0.42109060301809653, + "flos": 27486008208960.0, + "grad_norm": 7.08531675805759, + "learning_rate": 2.599037922364467e-06, + "loss": 0.7605, + "num_input_tokens_seen": 75488215, + "step": 3502 + }, + { + "epoch": 0.42121084590873564, + "flos": 29272184614080.0, + "grad_norm": 2.627924334727222, + "learning_rate": 2.5982946676165112e-06, + "loss": 0.7528, + "num_input_tokens_seen": 75507985, + "step": 3503 + }, + { + "epoch": 0.42133108879937475, + "flos": 67302383201760.0, + "grad_norm": 0.7379503265349121, + "learning_rate": 2.5975513221077313e-06, + "loss": 0.5753, + "num_input_tokens_seen": 75571955, + "step": 3504 + }, + { + "epoch": 0.4214513316900138, + "flos": 23072780724480.0, + "grad_norm": 2.436145210341609, + "learning_rate": 2.5968078859508897e-06, + "loss": 0.8892, + "num_input_tokens_seen": 75589155, + "step": 3505 + }, + { + "epoch": 0.4215715745806529, + "flos": 15314304895200.0, + "grad_norm": 2.3509348288280925, + "learning_rate": 2.5960643592587673e-06, + "loss": 0.7989, + "num_input_tokens_seen": 75606565, + "step": 3506 + }, + { + "epoch": 0.42169181747129203, + "flos": 22095655945440.0, + "grad_norm": 2.15689806943244, + "learning_rate": 2.5953207421441553e-06, + "loss": 0.8137, + "num_input_tokens_seen": 75625165, + "step": 3507 + }, + { + "epoch": 0.4218120603619311, + "flos": 22598122692000.0, + "grad_norm": 2.355448303640011, + "learning_rate": 2.5945770347198603e-06, + "loss": 0.7511, + "num_input_tokens_seen": 75643320, + "step": 3508 + }, + { + "epoch": 0.4219323032525702, + "flos": 19654779692160.0, + "grad_norm": 2.1410008678265258, + "learning_rate": 2.593833237098701e-06, + "loss": 0.8203, + "num_input_tokens_seen": 75660920, + "step": 3509 + }, + { + "epoch": 0.4220525461432093, + "flos": 30147409880160.0, + "grad_norm": 1.9829674482112114, + "learning_rate": 2.593089349393512e-06, + "loss": 0.629, + "num_input_tokens_seen": 75681410, + "step": 3510 + }, + { + "epoch": 0.42217278903384836, + "flos": 24281135812320.0, + "grad_norm": 2.78921260483242, + "learning_rate": 2.592345371717141e-06, + "loss": 0.835, + "num_input_tokens_seen": 75700940, + "step": 3511 + }, + { + "epoch": 0.42229303192448747, + "flos": 17068212215040.0, + "grad_norm": 2.452984070829977, + "learning_rate": 2.591601304182448e-06, + "loss": 0.722, + "num_input_tokens_seen": 75718910, + "step": 3512 + }, + { + "epoch": 0.4224132748151266, + "flos": 22751991263040.0, + "grad_norm": 2.070751445949832, + "learning_rate": 2.5908571469023067e-06, + "loss": 0.7918, + "num_input_tokens_seen": 75738395, + "step": 3513 + }, + { + "epoch": 0.42253351770576564, + "flos": 17793657583680.0, + "grad_norm": 2.3153179629242495, + "learning_rate": 2.5901128999896067e-06, + "loss": 0.7524, + "num_input_tokens_seen": 75753825, + "step": 3514 + }, + { + "epoch": 0.42265376059640475, + "flos": 28471496184000.0, + "grad_norm": 1.84577326565634, + "learning_rate": 2.5893685635572487e-06, + "loss": 0.6843, + "num_input_tokens_seen": 75774675, + "step": 3515 + }, + { + "epoch": 0.4227740034870438, + "flos": 16230311121600.0, + "grad_norm": 4.773381159523279, + "learning_rate": 2.5886241377181483e-06, + "loss": 0.6926, + "num_input_tokens_seen": 75793545, + "step": 3516 + }, + { + "epoch": 0.4228942463776829, + "flos": 25259264174880.0, + "grad_norm": 2.5508455557552687, + "learning_rate": 2.587879622585234e-06, + "loss": 0.8105, + "num_input_tokens_seen": 75812145, + "step": 3517 + }, + { + "epoch": 0.423014489268322, + "flos": 26357951269920.0, + "grad_norm": 2.8753690068153, + "learning_rate": 2.5871350182714486e-06, + "loss": 0.7625, + "num_input_tokens_seen": 75833025, + "step": 3518 + }, + { + "epoch": 0.4231347321589611, + "flos": 17249443463040.0, + "grad_norm": 3.0064356439256925, + "learning_rate": 2.586390324889748e-06, + "loss": 0.806, + "num_input_tokens_seen": 75848925, + "step": 3519 + }, + { + "epoch": 0.4232549750496002, + "flos": 22966643858880.0, + "grad_norm": 2.2680299819633056, + "learning_rate": 2.5856455425531003e-06, + "loss": 0.6773, + "num_input_tokens_seen": 75870400, + "step": 3520 + }, + { + "epoch": 0.4233752179402393, + "flos": 21217903135680.0, + "grad_norm": 2.6544633129672213, + "learning_rate": 2.5849006713744902e-06, + "loss": 0.8077, + "num_input_tokens_seen": 75889195, + "step": 3521 + }, + { + "epoch": 0.42349546083087836, + "flos": 20674766938080.0, + "grad_norm": 3.143025709561077, + "learning_rate": 2.5841557114669135e-06, + "loss": 0.7274, + "num_input_tokens_seen": 75906055, + "step": 3522 + }, + { + "epoch": 0.42361570372151747, + "flos": 18558508631520.0, + "grad_norm": 3.292786870347845, + "learning_rate": 2.58341066294338e-06, + "loss": 0.672, + "num_input_tokens_seen": 75922720, + "step": 3523 + }, + { + "epoch": 0.4237359466121566, + "flos": 20929568608320.0, + "grad_norm": 2.5791375223792317, + "learning_rate": 2.5826655259169124e-06, + "loss": 0.8558, + "num_input_tokens_seen": 75941375, + "step": 3524 + }, + { + "epoch": 0.42385618950279563, + "flos": 18012436022880.0, + "grad_norm": 2.2356727939239454, + "learning_rate": 2.5819203005005475e-06, + "loss": 0.9079, + "num_input_tokens_seen": 75958710, + "step": 3525 + }, + { + "epoch": 0.42397643239343474, + "flos": 23735137543200.0, + "grad_norm": 10.802737729670877, + "learning_rate": 2.581174986807336e-06, + "loss": 0.7877, + "num_input_tokens_seen": 75978945, + "step": 3526 + }, + { + "epoch": 0.42409667528407385, + "flos": 16521619229760.0, + "grad_norm": 2.6815678974911266, + "learning_rate": 2.580429584950341e-06, + "loss": 0.9134, + "num_input_tokens_seen": 75994695, + "step": 3527 + }, + { + "epoch": 0.4242169181747129, + "flos": 16011458342880.0, + "grad_norm": 2.2014882385854637, + "learning_rate": 2.5796840950426397e-06, + "loss": 0.668, + "num_input_tokens_seen": 76011780, + "step": 3528 + }, + { + "epoch": 0.424337161065352, + "flos": 20055384095520.0, + "grad_norm": 2.4919802779860802, + "learning_rate": 2.578938517197322e-06, + "loss": 0.6584, + "num_input_tokens_seen": 76029875, + "step": 3529 + }, + { + "epoch": 0.4244574039559911, + "flos": 23844210819840.0, + "grad_norm": 2.710970128107638, + "learning_rate": 2.5781928515274916e-06, + "loss": 0.6258, + "num_input_tokens_seen": 76048595, + "step": 3530 + }, + { + "epoch": 0.4245776468466302, + "flos": 17540937420000.0, + "grad_norm": 2.4001050489005804, + "learning_rate": 2.577447098146265e-06, + "loss": 0.6835, + "num_input_tokens_seen": 76065770, + "step": 3531 + }, + { + "epoch": 0.4246978897372693, + "flos": 27736200828960.0, + "grad_norm": 2.3515969714302742, + "learning_rate": 2.5767012571667724e-06, + "loss": 0.7904, + "num_input_tokens_seen": 76085250, + "step": 3532 + }, + { + "epoch": 0.42481813262790835, + "flos": 15574533350400.0, + "grad_norm": 2.6416968703924955, + "learning_rate": 2.5759553287021587e-06, + "loss": 0.6822, + "num_input_tokens_seen": 76103580, + "step": 3533 + }, + { + "epoch": 0.42493837551854746, + "flos": 23916814828320.0, + "grad_norm": 67.662662080676, + "learning_rate": 2.5752093128655786e-06, + "loss": 0.7727, + "num_input_tokens_seen": 76121825, + "step": 3534 + }, + { + "epoch": 0.4250586184091866, + "flos": 20783765875200.0, + "grad_norm": 1.9080143877162719, + "learning_rate": 2.574463209770204e-06, + "loss": 0.7351, + "num_input_tokens_seen": 76141140, + "step": 3535 + }, + { + "epoch": 0.42517886129982563, + "flos": 30327340186560.0, + "grad_norm": 1.9177688525539702, + "learning_rate": 2.5737170195292165e-06, + "loss": 0.7958, + "num_input_tokens_seen": 76164475, + "step": 3536 + }, + { + "epoch": 0.42529910419046474, + "flos": 20051890138080.0, + "grad_norm": 2.5413953188107885, + "learning_rate": 2.572970742255814e-06, + "loss": 0.785, + "num_input_tokens_seen": 76182965, + "step": 3537 + }, + { + "epoch": 0.42541934708110385, + "flos": 22599906840480.0, + "grad_norm": 2.1962054734121104, + "learning_rate": 2.5722243780632046e-06, + "loss": 0.8173, + "num_input_tokens_seen": 76201230, + "step": 3538 + }, + { + "epoch": 0.4255395899717429, + "flos": 66106182625440.0, + "grad_norm": 0.7777741740536839, + "learning_rate": 2.5714779270646125e-06, + "loss": 0.6049, + "num_input_tokens_seen": 76262000, + "step": 3539 + }, + { + "epoch": 0.425659832862382, + "flos": 17906336327040.0, + "grad_norm": 8.827129368635367, + "learning_rate": 2.5707313893732735e-06, + "loss": 0.7759, + "num_input_tokens_seen": 76280540, + "step": 3540 + }, + { + "epoch": 0.4257800757530211, + "flos": 23988117895200.0, + "grad_norm": 2.306800416842986, + "learning_rate": 2.5699847651024364e-06, + "loss": 0.7707, + "num_input_tokens_seen": 76301180, + "step": 3541 + }, + { + "epoch": 0.4259003186436602, + "flos": 23662124667360.0, + "grad_norm": 2.2612309432188824, + "learning_rate": 2.5692380543653627e-06, + "loss": 0.7631, + "num_input_tokens_seen": 76319335, + "step": 3542 + }, + { + "epoch": 0.4260205615342993, + "flos": 15237872401440.0, + "grad_norm": 2.123236990718677, + "learning_rate": 2.5684912572753293e-06, + "loss": 0.7037, + "num_input_tokens_seen": 76335010, + "step": 3543 + }, + { + "epoch": 0.4261408044249384, + "flos": 30621287347680.0, + "grad_norm": 2.03272405105034, + "learning_rate": 2.5677443739456245e-06, + "loss": 0.8468, + "num_input_tokens_seen": 76356670, + "step": 3544 + }, + { + "epoch": 0.42626104731557746, + "flos": 23224642128480.0, + "grad_norm": 2.654884076937277, + "learning_rate": 2.5669974044895495e-06, + "loss": 0.7931, + "num_input_tokens_seen": 76373065, + "step": 3545 + }, + { + "epoch": 0.42638129020621657, + "flos": 25847678873760.0, + "grad_norm": 2.659119167192552, + "learning_rate": 2.5662503490204187e-06, + "loss": 0.7943, + "num_input_tokens_seen": 76393230, + "step": 3546 + }, + { + "epoch": 0.4265015330968556, + "flos": 26464273984320.0, + "grad_norm": 2.2677081824042555, + "learning_rate": 2.5655032076515603e-06, + "loss": 0.7648, + "num_input_tokens_seen": 76412555, + "step": 3547 + }, + { + "epoch": 0.42662177598749473, + "flos": 24354631895040.0, + "grad_norm": 2.565595478440937, + "learning_rate": 2.5647559804963155e-06, + "loss": 0.8245, + "num_input_tokens_seen": 76432485, + "step": 3548 + }, + { + "epoch": 0.42674201887813384, + "flos": 23115531682080.0, + "grad_norm": 3.4291866692801305, + "learning_rate": 2.5640086676680364e-06, + "loss": 0.7822, + "num_input_tokens_seen": 76453980, + "step": 3549 + }, + { + "epoch": 0.4268622617687729, + "flos": 21658656613440.0, + "grad_norm": 2.9065492537462916, + "learning_rate": 2.5632612692800923e-06, + "loss": 0.8077, + "num_input_tokens_seen": 76473045, + "step": 3550 + }, + { + "epoch": 0.426982504659412, + "flos": 23406616771680.0, + "grad_norm": 2.32441979828981, + "learning_rate": 2.5625137854458603e-06, + "loss": 0.7521, + "num_input_tokens_seen": 76492060, + "step": 3551 + }, + { + "epoch": 0.4271027475500511, + "flos": 18889631286240.0, + "grad_norm": 2.276814577576862, + "learning_rate": 2.561766216278735e-06, + "loss": 0.8051, + "num_input_tokens_seen": 76509655, + "step": 3552 + }, + { + "epoch": 0.4272229904406902, + "flos": 26832051756000.0, + "grad_norm": 2.184470558532147, + "learning_rate": 2.561018561892121e-06, + "loss": 0.8203, + "num_input_tokens_seen": 76528795, + "step": 3553 + }, + { + "epoch": 0.4273432333313293, + "flos": 23917112186400.0, + "grad_norm": 1.8863227381247178, + "learning_rate": 2.5602708223994363e-06, + "loss": 0.7681, + "num_input_tokens_seen": 76550660, + "step": 3554 + }, + { + "epoch": 0.4274634762219684, + "flos": 29527915528320.0, + "grad_norm": 2.603563427039127, + "learning_rate": 2.559522997914115e-06, + "loss": 0.6826, + "num_input_tokens_seen": 76570240, + "step": 3555 + }, + { + "epoch": 0.42758371911260745, + "flos": 21403557585120.0, + "grad_norm": 2.809138174969643, + "learning_rate": 2.558775088549599e-06, + "loss": 0.8469, + "num_input_tokens_seen": 76589820, + "step": 3556 + }, + { + "epoch": 0.42770396200324656, + "flos": 14731800188160.0, + "grad_norm": 3.1017423716931174, + "learning_rate": 2.5580270944193467e-06, + "loss": 0.6683, + "num_input_tokens_seen": 76606640, + "step": 3557 + }, + { + "epoch": 0.4278242048938857, + "flos": 70553835041280.0, + "grad_norm": 0.7489942323643429, + "learning_rate": 2.557279015636827e-06, + "loss": 0.5547, + "num_input_tokens_seen": 76667050, + "step": 3558 + }, + { + "epoch": 0.42794444778452473, + "flos": 69266742934560.0, + "grad_norm": 0.7784902657504817, + "learning_rate": 2.5565308523155245e-06, + "loss": 0.6116, + "num_input_tokens_seen": 76726650, + "step": 3559 + }, + { + "epoch": 0.42806469067516384, + "flos": 18189132560160.0, + "grad_norm": 2.7855886843246087, + "learning_rate": 2.5557826045689336e-06, + "loss": 0.8227, + "num_input_tokens_seen": 76742890, + "step": 3560 + }, + { + "epoch": 0.4281849335658029, + "flos": 54458690367840.0, + "grad_norm": 0.8492065176131066, + "learning_rate": 2.5550342725105643e-06, + "loss": 0.5872, + "num_input_tokens_seen": 76801055, + "step": 3561 + }, + { + "epoch": 0.428305176456442, + "flos": 17250298367520.0, + "grad_norm": 2.083360086905042, + "learning_rate": 2.554285856253937e-06, + "loss": 0.815, + "num_input_tokens_seen": 76819890, + "step": 3562 + }, + { + "epoch": 0.4284254193470811, + "flos": 26321742190080.0, + "grad_norm": 2.1290172302996613, + "learning_rate": 2.5535373559125855e-06, + "loss": 0.7759, + "num_input_tokens_seen": 76840255, + "step": 3563 + }, + { + "epoch": 0.42854566223772017, + "flos": 29671525245600.0, + "grad_norm": 1.727467571566954, + "learning_rate": 2.552788771600057e-06, + "loss": 0.8196, + "num_input_tokens_seen": 76860565, + "step": 3564 + }, + { + "epoch": 0.4286659051283593, + "flos": 21986731347840.0, + "grad_norm": 1.8338177070001598, + "learning_rate": 2.5520401034299118e-06, + "loss": 0.8179, + "num_input_tokens_seen": 76878325, + "step": 3565 + }, + { + "epoch": 0.4287861480189984, + "flos": 13315631740320.0, + "grad_norm": 2.446228635730733, + "learning_rate": 2.551291351515722e-06, + "loss": 0.8753, + "num_input_tokens_seen": 76895340, + "step": 3566 + }, + { + "epoch": 0.42890639090963745, + "flos": 26613347656320.0, + "grad_norm": 1.8866582825253126, + "learning_rate": 2.5505425159710726e-06, + "loss": 0.8562, + "num_input_tokens_seen": 76915150, + "step": 3567 + }, + { + "epoch": 0.42902663380027656, + "flos": 24021167545440.0, + "grad_norm": 2.2808845958958384, + "learning_rate": 2.549793596909561e-06, + "loss": 0.8271, + "num_input_tokens_seen": 76933770, + "step": 3568 + }, + { + "epoch": 0.42914687669091567, + "flos": 15610407902400.0, + "grad_norm": 2.5082091020846473, + "learning_rate": 2.5490445944447976e-06, + "loss": 0.6613, + "num_input_tokens_seen": 76952265, + "step": 3569 + }, + { + "epoch": 0.4292671195815547, + "flos": 31422607663680.0, + "grad_norm": 3.0241974262499816, + "learning_rate": 2.548295508690406e-06, + "loss": 0.6544, + "num_input_tokens_seen": 76973560, + "step": 3570 + }, + { + "epoch": 0.42938736247219383, + "flos": 30214103896800.0, + "grad_norm": 2.1428834459416186, + "learning_rate": 2.5475463397600217e-06, + "loss": 0.7652, + "num_input_tokens_seen": 76993640, + "step": 3571 + }, + { + "epoch": 0.42950760536283294, + "flos": 29307501619680.0, + "grad_norm": 2.193262955812883, + "learning_rate": 2.546797087767293e-06, + "loss": 0.7708, + "num_input_tokens_seen": 77013765, + "step": 3572 + }, + { + "epoch": 0.429627848253472, + "flos": 26831419870080.0, + "grad_norm": 2.230915940011296, + "learning_rate": 2.546047752825881e-06, + "loss": 0.8707, + "num_input_tokens_seen": 77034370, + "step": 3573 + }, + { + "epoch": 0.4297480911441111, + "flos": 13861667179200.0, + "grad_norm": 2.374552031984159, + "learning_rate": 2.5452983350494595e-06, + "loss": 0.9298, + "num_input_tokens_seen": 77049925, + "step": 3574 + }, + { + "epoch": 0.4298683340347502, + "flos": 20711570734080.0, + "grad_norm": 2.4454173105221364, + "learning_rate": 2.544548834551713e-06, + "loss": 0.6517, + "num_input_tokens_seen": 77068930, + "step": 3575 + }, + { + "epoch": 0.4299885769253893, + "flos": 20852169700800.0, + "grad_norm": 2.707409060316755, + "learning_rate": 2.5437992514463424e-06, + "loss": 0.9464, + "num_input_tokens_seen": 77082255, + "step": 3576 + }, + { + "epoch": 0.4301088198160284, + "flos": 25448152393440.0, + "grad_norm": 1.8928914002765629, + "learning_rate": 2.5430495858470565e-06, + "loss": 0.8793, + "num_input_tokens_seen": 77101725, + "step": 3577 + }, + { + "epoch": 0.43022906270666744, + "flos": 18233444647680.0, + "grad_norm": 2.819584041787223, + "learning_rate": 2.54229983786758e-06, + "loss": 0.7768, + "num_input_tokens_seen": 77119670, + "step": 3578 + }, + { + "epoch": 0.43034930559730655, + "flos": 23365910150880.0, + "grad_norm": 2.5092788164795494, + "learning_rate": 2.541550007621651e-06, + "loss": 0.857, + "num_input_tokens_seen": 77138160, + "step": 3579 + }, + { + "epoch": 0.43046954848794566, + "flos": 28143978996000.0, + "grad_norm": 2.09432668003497, + "learning_rate": 2.5408000952230156e-06, + "loss": 0.8004, + "num_input_tokens_seen": 77156950, + "step": 3580 + }, + { + "epoch": 0.4305897913785847, + "flos": 28539305293440.0, + "grad_norm": 2.2446362883450774, + "learning_rate": 2.5400501007854357e-06, + "loss": 0.9059, + "num_input_tokens_seen": 77176395, + "step": 3581 + }, + { + "epoch": 0.43071003426922383, + "flos": 20419556400480.0, + "grad_norm": 2.1581929616559337, + "learning_rate": 2.539300024422685e-06, + "loss": 0.7568, + "num_input_tokens_seen": 77194340, + "step": 3582 + }, + { + "epoch": 0.43083027715986294, + "flos": 51923800324320.0, + "grad_norm": 0.7981571280062866, + "learning_rate": 2.538549866248549e-06, + "loss": 0.6088, + "num_input_tokens_seen": 77249320, + "step": 3583 + }, + { + "epoch": 0.430950520050502, + "flos": 16667198944320.0, + "grad_norm": 2.238483626185849, + "learning_rate": 2.5377996263768274e-06, + "loss": 0.8159, + "num_input_tokens_seen": 77267400, + "step": 3584 + }, + { + "epoch": 0.4310707629411411, + "flos": 24573224485440.0, + "grad_norm": 1.9230126216115164, + "learning_rate": 2.5370493049213293e-06, + "loss": 0.6852, + "num_input_tokens_seen": 77287045, + "step": 3585 + }, + { + "epoch": 0.4311910058317802, + "flos": 26395126763520.0, + "grad_norm": 2.66652478396938, + "learning_rate": 2.536298901995878e-06, + "loss": 0.8018, + "num_input_tokens_seen": 77306255, + "step": 3586 + }, + { + "epoch": 0.43131124872241927, + "flos": 25119891810240.0, + "grad_norm": 1.9590637860314473, + "learning_rate": 2.535548417714311e-06, + "loss": 0.8004, + "num_input_tokens_seen": 77325555, + "step": 3587 + }, + { + "epoch": 0.4314314916130584, + "flos": 21583896758880.0, + "grad_norm": 1.7235969863928997, + "learning_rate": 2.534797852190474e-06, + "loss": 0.8746, + "num_input_tokens_seen": 77345735, + "step": 3588 + }, + { + "epoch": 0.4315517345036975, + "flos": 19247336052960.0, + "grad_norm": 2.0764496438938047, + "learning_rate": 2.5340472055382283e-06, + "loss": 0.8155, + "num_input_tokens_seen": 77361765, + "step": 3589 + }, + { + "epoch": 0.43167197739433655, + "flos": 24238905231360.0, + "grad_norm": 3.789886951219963, + "learning_rate": 2.5332964778714468e-06, + "loss": 0.8102, + "num_input_tokens_seen": 77378950, + "step": 3590 + }, + { + "epoch": 0.43179222028497566, + "flos": 16843672463040.0, + "grad_norm": 1.9348217696283194, + "learning_rate": 2.5325456693040123e-06, + "loss": 0.6655, + "num_input_tokens_seen": 77396145, + "step": 3591 + }, + { + "epoch": 0.43191246317561477, + "flos": 17614136144640.0, + "grad_norm": 2.5609552332768617, + "learning_rate": 2.531794779949824e-06, + "loss": 0.7525, + "num_input_tokens_seen": 77414045, + "step": 3592 + }, + { + "epoch": 0.4320327060662538, + "flos": 23844656856960.0, + "grad_norm": 2.029944836848877, + "learning_rate": 2.5310438099227903e-06, + "loss": 0.8792, + "num_input_tokens_seen": 77431310, + "step": 3593 + }, + { + "epoch": 0.43215294895689293, + "flos": 66300571968480.0, + "grad_norm": 0.8228827038931218, + "learning_rate": 2.530292759336833e-06, + "loss": 0.5336, + "num_input_tokens_seen": 77492045, + "step": 3594 + }, + { + "epoch": 0.432273191847532, + "flos": 20564875926720.0, + "grad_norm": 3.535794869587636, + "learning_rate": 2.5295416283058855e-06, + "loss": 0.7004, + "num_input_tokens_seen": 77510910, + "step": 3595 + }, + { + "epoch": 0.4323934347381711, + "flos": 19254695665440.0, + "grad_norm": 1.79101256325563, + "learning_rate": 2.5287904169438943e-06, + "loss": 0.6579, + "num_input_tokens_seen": 77530115, + "step": 3596 + }, + { + "epoch": 0.4325136776288102, + "flos": 21695311730400.0, + "grad_norm": 3.1791679838951916, + "learning_rate": 2.528039125364817e-06, + "loss": 0.6378, + "num_input_tokens_seen": 77548920, + "step": 3597 + }, + { + "epoch": 0.43263392051944927, + "flos": 22308635902080.0, + "grad_norm": 6.11910509086395, + "learning_rate": 2.5272877536826246e-06, + "loss": 0.7577, + "num_input_tokens_seen": 77567715, + "step": 3598 + }, + { + "epoch": 0.4327541634100884, + "flos": 29126530560000.0, + "grad_norm": 3.247009074845481, + "learning_rate": 2.5265363020112986e-06, + "loss": 0.7023, + "num_input_tokens_seen": 77588035, + "step": 3599 + }, + { + "epoch": 0.4328744063007275, + "flos": 26030099554080.0, + "grad_norm": 2.241830192566932, + "learning_rate": 2.5257847704648344e-06, + "loss": 0.8381, + "num_input_tokens_seen": 77609265, + "step": 3600 + }, + { + "epoch": 0.43299464919136654, + "flos": 16557828309600.0, + "grad_norm": 2.0231285301180053, + "learning_rate": 2.525033159157239e-06, + "loss": 0.7505, + "num_input_tokens_seen": 77625580, + "step": 3601 + }, + { + "epoch": 0.43311489208200565, + "flos": 16084434048960.0, + "grad_norm": 2.0078117875272325, + "learning_rate": 2.52428146820253e-06, + "loss": 0.7721, + "num_input_tokens_seen": 77643835, + "step": 3602 + }, + { + "epoch": 0.43323513497264476, + "flos": 22897459468320.0, + "grad_norm": 2.585521284185091, + "learning_rate": 2.52352969771474e-06, + "loss": 0.8186, + "num_input_tokens_seen": 77663060, + "step": 3603 + }, + { + "epoch": 0.4333553778632838, + "flos": 25262720962560.0, + "grad_norm": 2.138448569427877, + "learning_rate": 2.5227778478079106e-06, + "loss": 0.8887, + "num_input_tokens_seen": 77682470, + "step": 3604 + }, + { + "epoch": 0.43347562075392293, + "flos": 19359420080160.0, + "grad_norm": 3.5379894389786846, + "learning_rate": 2.522025918596098e-06, + "loss": 0.7692, + "num_input_tokens_seen": 77700770, + "step": 3605 + }, + { + "epoch": 0.43359586364456204, + "flos": 26288358012000.0, + "grad_norm": 2.343415503474326, + "learning_rate": 2.521273910193368e-06, + "loss": 0.6551, + "num_input_tokens_seen": 77723950, + "step": 3606 + }, + { + "epoch": 0.4337161065352011, + "flos": 15966291350880.0, + "grad_norm": 2.2909715002677493, + "learning_rate": 2.5205218227138006e-06, + "loss": 0.8701, + "num_input_tokens_seen": 77736905, + "step": 3607 + }, + { + "epoch": 0.4338363494258402, + "flos": 20195685704160.0, + "grad_norm": 2.927194903183029, + "learning_rate": 2.519769656271486e-06, + "loss": 0.7908, + "num_input_tokens_seen": 77754325, + "step": 3608 + }, + { + "epoch": 0.43395659231647926, + "flos": 20055272586240.0, + "grad_norm": 3.35651037815171, + "learning_rate": 2.5190174109805285e-06, + "loss": 0.676, + "num_input_tokens_seen": 77774150, + "step": 3609 + }, + { + "epoch": 0.43407683520711837, + "flos": 19873446622080.0, + "grad_norm": 2.4596663810163566, + "learning_rate": 2.518265086955042e-06, + "loss": 0.6469, + "num_input_tokens_seen": 77791870, + "step": 3610 + }, + { + "epoch": 0.4341970780977575, + "flos": 23075568456480.0, + "grad_norm": 2.6875184524257514, + "learning_rate": 2.5175126843091534e-06, + "loss": 0.8355, + "num_input_tokens_seen": 77811240, + "step": 3611 + }, + { + "epoch": 0.43431732098839654, + "flos": 37354200467040.0, + "grad_norm": 2.314147259423989, + "learning_rate": 2.5167602031570034e-06, + "loss": 0.7566, + "num_input_tokens_seen": 77831425, + "step": 3612 + }, + { + "epoch": 0.43443756387903565, + "flos": 31822394332320.0, + "grad_norm": 1.9001757038589586, + "learning_rate": 2.51600764361274e-06, + "loss": 0.7369, + "num_input_tokens_seen": 77852950, + "step": 3613 + }, + { + "epoch": 0.43455780676967476, + "flos": 23443569246720.0, + "grad_norm": 2.5425774761087183, + "learning_rate": 2.5152550057905283e-06, + "loss": 0.7882, + "num_input_tokens_seen": 77872840, + "step": 3614 + }, + { + "epoch": 0.4346780496603138, + "flos": 24172434233280.0, + "grad_norm": 4.347950612648366, + "learning_rate": 2.5145022898045415e-06, + "loss": 0.7709, + "num_input_tokens_seen": 77892025, + "step": 3615 + }, + { + "epoch": 0.4347982925509529, + "flos": 17068100705760.0, + "grad_norm": 2.630638334168445, + "learning_rate": 2.5137494957689664e-06, + "loss": 0.9018, + "num_input_tokens_seen": 77907635, + "step": 3616 + }, + { + "epoch": 0.43491853544159204, + "flos": 60858659514240.0, + "grad_norm": 0.7778403837877421, + "learning_rate": 2.5129966237980016e-06, + "loss": 0.573, + "num_input_tokens_seen": 77970630, + "step": 3617 + }, + { + "epoch": 0.4350387783322311, + "flos": 21913309604640.0, + "grad_norm": 2.0685866738008265, + "learning_rate": 2.512243674005857e-06, + "loss": 0.7821, + "num_input_tokens_seen": 77989995, + "step": 3618 + }, + { + "epoch": 0.4351590212228702, + "flos": 25046804594880.0, + "grad_norm": 2.3029076896848193, + "learning_rate": 2.5114906465067537e-06, + "loss": 0.8616, + "num_input_tokens_seen": 78010980, + "step": 3619 + }, + { + "epoch": 0.4352792641135093, + "flos": 21475827065760.0, + "grad_norm": 2.180738407683641, + "learning_rate": 2.5107375414149264e-06, + "loss": 0.7535, + "num_input_tokens_seen": 78030225, + "step": 3620 + }, + { + "epoch": 0.43539950700414837, + "flos": 16230422630880.0, + "grad_norm": 3.386253527375078, + "learning_rate": 2.5099843588446197e-06, + "loss": 0.7153, + "num_input_tokens_seen": 78048545, + "step": 3621 + }, + { + "epoch": 0.4355197498947875, + "flos": 16668165358080.0, + "grad_norm": 1.8936838673541947, + "learning_rate": 2.509231098910091e-06, + "loss": 0.6189, + "num_input_tokens_seen": 78068415, + "step": 3622 + }, + { + "epoch": 0.4356399927854266, + "flos": 16339533077280.0, + "grad_norm": 3.507725364095575, + "learning_rate": 2.508477761725611e-06, + "loss": 0.7497, + "num_input_tokens_seen": 78086690, + "step": 3623 + }, + { + "epoch": 0.43576023567606564, + "flos": 17177620019520.0, + "grad_norm": 1.9388756329223866, + "learning_rate": 2.507724347405458e-06, + "loss": 0.8081, + "num_input_tokens_seen": 78104640, + "step": 3624 + }, + { + "epoch": 0.43588047856670475, + "flos": 15894839604960.0, + "grad_norm": 3.739356768163719, + "learning_rate": 2.5069708560639243e-06, + "loss": 0.8236, + "num_input_tokens_seen": 78122550, + "step": 3625 + }, + { + "epoch": 0.4360007214573438, + "flos": 23625729738720.0, + "grad_norm": 2.237420520756087, + "learning_rate": 2.5062172878153158e-06, + "loss": 0.6154, + "num_input_tokens_seen": 78141825, + "step": 3626 + }, + { + "epoch": 0.4361209643479829, + "flos": 21947288498880.0, + "grad_norm": 4.114585908655576, + "learning_rate": 2.505463642773947e-06, + "loss": 0.8733, + "num_input_tokens_seen": 78161790, + "step": 3627 + }, + { + "epoch": 0.43624120723862203, + "flos": 17395394875200.0, + "grad_norm": 3.328986974138952, + "learning_rate": 2.504709921054146e-06, + "loss": 0.7478, + "num_input_tokens_seen": 78178605, + "step": 3628 + }, + { + "epoch": 0.4363614501292611, + "flos": 17869867058880.0, + "grad_norm": 2.4768030669523564, + "learning_rate": 2.50395612277025e-06, + "loss": 0.8379, + "num_input_tokens_seen": 78194460, + "step": 3629 + }, + { + "epoch": 0.4364816930199002, + "flos": 20273939516160.0, + "grad_norm": 2.2938237066665836, + "learning_rate": 2.503202248036612e-06, + "loss": 0.731, + "num_input_tokens_seen": 78213315, + "step": 3630 + }, + { + "epoch": 0.4366019359105393, + "flos": 24026928858240.0, + "grad_norm": 9.552881127795786, + "learning_rate": 2.5024482969675927e-06, + "loss": 0.7311, + "num_input_tokens_seen": 78233270, + "step": 3631 + }, + { + "epoch": 0.43672217880117836, + "flos": 21722748746880.0, + "grad_norm": 12.186239167554037, + "learning_rate": 2.501694269677566e-06, + "loss": 0.8473, + "num_input_tokens_seen": 78251040, + "step": 3632 + }, + { + "epoch": 0.4368424216918175, + "flos": 18009350932800.0, + "grad_norm": 2.4167805896960712, + "learning_rate": 2.500940166280918e-06, + "loss": 0.8047, + "num_input_tokens_seen": 78269470, + "step": 3633 + }, + { + "epoch": 0.4369626645824566, + "flos": 25410679541760.0, + "grad_norm": 2.215016084186376, + "learning_rate": 2.500185986892045e-06, + "loss": 0.7894, + "num_input_tokens_seen": 78288955, + "step": 3634 + }, + { + "epoch": 0.43708290747309564, + "flos": 25266512278080.0, + "grad_norm": 2.9088428375287627, + "learning_rate": 2.499431731625355e-06, + "loss": 0.7713, + "num_input_tokens_seen": 78307980, + "step": 3635 + }, + { + "epoch": 0.43720315036373475, + "flos": 31529673773280.0, + "grad_norm": 2.583180874253227, + "learning_rate": 2.4986774005952686e-06, + "loss": 0.7902, + "num_input_tokens_seen": 78330355, + "step": 3636 + }, + { + "epoch": 0.43732339325437386, + "flos": 23078950904640.0, + "grad_norm": 2.7938097987933066, + "learning_rate": 2.4979229939162166e-06, + "loss": 0.8458, + "num_input_tokens_seen": 78349135, + "step": 3637 + }, + { + "epoch": 0.4374436361450129, + "flos": 27706496457120.0, + "grad_norm": 1.999086051970343, + "learning_rate": 2.4971685117026433e-06, + "loss": 0.804, + "num_input_tokens_seen": 78368900, + "step": 3638 + }, + { + "epoch": 0.437563879035652, + "flos": 24137860622880.0, + "grad_norm": 2.171806389253223, + "learning_rate": 2.4964139540690018e-06, + "loss": 0.7654, + "num_input_tokens_seen": 78392235, + "step": 3639 + }, + { + "epoch": 0.4376841219262911, + "flos": 23443643586240.0, + "grad_norm": 1.9748468605256653, + "learning_rate": 2.495659321129758e-06, + "loss": 0.7295, + "num_input_tokens_seen": 78409980, + "step": 3640 + }, + { + "epoch": 0.4378043648169302, + "flos": 25411385767200.0, + "grad_norm": 2.053726841986977, + "learning_rate": 2.494904612999389e-06, + "loss": 0.7542, + "num_input_tokens_seen": 78428690, + "step": 3641 + }, + { + "epoch": 0.4379246077075693, + "flos": 53837783565120.0, + "grad_norm": 0.7618136615799087, + "learning_rate": 2.4941498297923843e-06, + "loss": 0.5649, + "num_input_tokens_seen": 78489260, + "step": 3642 + }, + { + "epoch": 0.43804485059820836, + "flos": 20558705746560.0, + "grad_norm": 1.9402078963504894, + "learning_rate": 2.4933949716232424e-06, + "loss": 0.6991, + "num_input_tokens_seen": 78506785, + "step": 3643 + }, + { + "epoch": 0.43816509348884747, + "flos": 23841980634240.0, + "grad_norm": 2.8864646872087194, + "learning_rate": 2.492640038606476e-06, + "loss": 0.7352, + "num_input_tokens_seen": 78525150, + "step": 3644 + }, + { + "epoch": 0.4382853363794866, + "flos": 14663693720640.0, + "grad_norm": 3.2898271627739493, + "learning_rate": 2.491885030856608e-06, + "loss": 0.7827, + "num_input_tokens_seen": 78543245, + "step": 3645 + }, + { + "epoch": 0.43840557927012563, + "flos": 17140927732800.0, + "grad_norm": 2.8312899757240944, + "learning_rate": 2.4911299484881713e-06, + "loss": 0.8266, + "num_input_tokens_seen": 78560775, + "step": 3646 + }, + { + "epoch": 0.43852582216076474, + "flos": 19363062716640.0, + "grad_norm": 1.759919052971842, + "learning_rate": 2.490374791615712e-06, + "loss": 0.8113, + "num_input_tokens_seen": 78580800, + "step": 3647 + }, + { + "epoch": 0.43864606505140386, + "flos": 18049351328160.0, + "grad_norm": 4.181688061676291, + "learning_rate": 2.4896195603537867e-06, + "loss": 0.7796, + "num_input_tokens_seen": 78595410, + "step": 3648 + }, + { + "epoch": 0.4387663079420429, + "flos": 19616860803360.0, + "grad_norm": 3.135537653180347, + "learning_rate": 2.488864254816964e-06, + "loss": 0.7405, + "num_input_tokens_seen": 78614100, + "step": 3649 + }, + { + "epoch": 0.438886550832682, + "flos": 19691248960320.0, + "grad_norm": 4.6562145667040005, + "learning_rate": 2.4881088751198218e-06, + "loss": 0.6868, + "num_input_tokens_seen": 78632260, + "step": 3650 + }, + { + "epoch": 0.43900679372332113, + "flos": 14516069669280.0, + "grad_norm": 3.2591099339613425, + "learning_rate": 2.4873534213769517e-06, + "loss": 0.6469, + "num_input_tokens_seen": 78647490, + "step": 3651 + }, + { + "epoch": 0.4391270366139602, + "flos": 24021650752320.0, + "grad_norm": 1.9294165293601873, + "learning_rate": 2.4865978937029547e-06, + "loss": 0.7209, + "num_input_tokens_seen": 78666470, + "step": 3652 + }, + { + "epoch": 0.4392472795045993, + "flos": 31493130165600.0, + "grad_norm": 2.02395269838479, + "learning_rate": 2.485842292212445e-06, + "loss": 0.6625, + "num_input_tokens_seen": 78687685, + "step": 3653 + }, + { + "epoch": 0.4393675223952384, + "flos": 14845259496480.0, + "grad_norm": 2.0020090979217904, + "learning_rate": 2.485086617020045e-06, + "loss": 0.806, + "num_input_tokens_seen": 78706180, + "step": 3654 + }, + { + "epoch": 0.43948776528587746, + "flos": 14804775894240.0, + "grad_norm": 4.929485721962508, + "learning_rate": 2.4843308682403903e-06, + "loss": 0.8209, + "num_input_tokens_seen": 78723095, + "step": 3655 + }, + { + "epoch": 0.4396080081765166, + "flos": 13894753999200.0, + "grad_norm": 1.8750134785485595, + "learning_rate": 2.4835750459881294e-06, + "loss": 0.8276, + "num_input_tokens_seen": 78739720, + "step": 3656 + }, + { + "epoch": 0.43972825106715563, + "flos": 18197049719040.0, + "grad_norm": 2.0560911772359867, + "learning_rate": 2.4828191503779177e-06, + "loss": 0.8211, + "num_input_tokens_seen": 78757820, + "step": 3657 + }, + { + "epoch": 0.43984849395779474, + "flos": 16849285096800.0, + "grad_norm": 2.5776398173315576, + "learning_rate": 2.482063181524425e-06, + "loss": 0.8982, + "num_input_tokens_seen": 78773720, + "step": 3658 + }, + { + "epoch": 0.43996873684843385, + "flos": 18664682666880.0, + "grad_norm": 2.3181569783176994, + "learning_rate": 2.4813071395423307e-06, + "loss": 0.8119, + "num_input_tokens_seen": 78791285, + "step": 3659 + }, + { + "epoch": 0.4400889797390729, + "flos": 23619410879520.0, + "grad_norm": 2.2239896130353576, + "learning_rate": 2.4805510245463263e-06, + "loss": 0.6447, + "num_input_tokens_seen": 78811440, + "step": 3660 + }, + { + "epoch": 0.440209222629712, + "flos": 23116126398240.0, + "grad_norm": 2.216650493342929, + "learning_rate": 2.4797948366511137e-06, + "loss": 0.6044, + "num_input_tokens_seen": 78831150, + "step": 3661 + }, + { + "epoch": 0.4403294655203511, + "flos": 24787728402240.0, + "grad_norm": 2.599474736488056, + "learning_rate": 2.4790385759714055e-06, + "loss": 0.7675, + "num_input_tokens_seen": 78850215, + "step": 3662 + }, + { + "epoch": 0.4404497084109902, + "flos": 22532841126240.0, + "grad_norm": 2.0539858240011672, + "learning_rate": 2.478282242621926e-06, + "loss": 0.7084, + "num_input_tokens_seen": 78870675, + "step": 3663 + }, + { + "epoch": 0.4405699513016293, + "flos": 64874627873760.0, + "grad_norm": 0.8813590854738506, + "learning_rate": 2.477525836717411e-06, + "loss": 0.5947, + "num_input_tokens_seen": 78938440, + "step": 3664 + }, + { + "epoch": 0.4406901941922684, + "flos": 35610589170720.0, + "grad_norm": 8.031188727119135, + "learning_rate": 2.476769358372606e-06, + "loss": 0.7951, + "num_input_tokens_seen": 78960090, + "step": 3665 + }, + { + "epoch": 0.44081043708290746, + "flos": 18015112245600.0, + "grad_norm": 2.3458033404513587, + "learning_rate": 2.4760128077022683e-06, + "loss": 0.7509, + "num_input_tokens_seen": 78978225, + "step": 3666 + }, + { + "epoch": 0.44093067997354657, + "flos": 30110197216800.0, + "grad_norm": 1.8896951201436507, + "learning_rate": 2.4752561848211672e-06, + "loss": 0.686, + "num_input_tokens_seen": 79000625, + "step": 3667 + }, + { + "epoch": 0.4410509228641857, + "flos": 23221705717440.0, + "grad_norm": 2.417634364657546, + "learning_rate": 2.4744994898440797e-06, + "loss": 0.7105, + "num_input_tokens_seen": 79019415, + "step": 3668 + }, + { + "epoch": 0.44117116575482473, + "flos": 19472805048960.0, + "grad_norm": 2.3625538612967274, + "learning_rate": 2.473742722885797e-06, + "loss": 0.8355, + "num_input_tokens_seen": 79038435, + "step": 3669 + }, + { + "epoch": 0.44129140864546385, + "flos": 27014360927040.0, + "grad_norm": 6.2282149014107, + "learning_rate": 2.4729858840611197e-06, + "loss": 0.655, + "num_input_tokens_seen": 79057345, + "step": 3670 + }, + { + "epoch": 0.4414116515361029, + "flos": 26064635994720.0, + "grad_norm": 2.4461098245526904, + "learning_rate": 2.4722289734848605e-06, + "loss": 0.7247, + "num_input_tokens_seen": 79077810, + "step": 3671 + }, + { + "epoch": 0.441531894426742, + "flos": 21874721660160.0, + "grad_norm": 2.9050150883031645, + "learning_rate": 2.471471991271841e-06, + "loss": 0.7846, + "num_input_tokens_seen": 79094935, + "step": 3672 + }, + { + "epoch": 0.4416521373173811, + "flos": 23403457342080.0, + "grad_norm": 3.647408775354928, + "learning_rate": 2.470714937536896e-06, + "loss": 0.7978, + "num_input_tokens_seen": 79113660, + "step": 3673 + }, + { + "epoch": 0.4417723802080202, + "flos": 20305167848160.0, + "grad_norm": 3.0455123593084767, + "learning_rate": 2.469957812394868e-06, + "loss": 0.7071, + "num_input_tokens_seen": 79132470, + "step": 3674 + }, + { + "epoch": 0.4418926230986593, + "flos": 18853831073760.0, + "grad_norm": 2.487901020791987, + "learning_rate": 2.4692006159606148e-06, + "loss": 0.7629, + "num_input_tokens_seen": 79150000, + "step": 3675 + }, + { + "epoch": 0.4420128659892984, + "flos": 19436707478400.0, + "grad_norm": 2.1022533996968265, + "learning_rate": 2.468443348349e-06, + "loss": 0.7837, + "num_input_tokens_seen": 79167875, + "step": 3676 + }, + { + "epoch": 0.44213310887993745, + "flos": 17868900645120.0, + "grad_norm": 2.8729018142976495, + "learning_rate": 2.467686009674902e-06, + "loss": 0.8259, + "num_input_tokens_seen": 79182325, + "step": 3677 + }, + { + "epoch": 0.44225335177057656, + "flos": 19177185248640.0, + "grad_norm": 2.532157735187111, + "learning_rate": 2.466928600053209e-06, + "loss": 0.8547, + "num_input_tokens_seen": 79197630, + "step": 3678 + }, + { + "epoch": 0.4423735946612157, + "flos": 23437547745600.0, + "grad_norm": 2.0222090007480182, + "learning_rate": 2.466171119598818e-06, + "loss": 0.7124, + "num_input_tokens_seen": 79217600, + "step": 3679 + }, + { + "epoch": 0.44249383755185473, + "flos": 26647066362240.0, + "grad_norm": 2.0305167415241527, + "learning_rate": 2.465413568426639e-06, + "loss": 0.7731, + "num_input_tokens_seen": 79238550, + "step": 3680 + }, + { + "epoch": 0.44261408044249384, + "flos": 23114528098560.0, + "grad_norm": 1.883365534054205, + "learning_rate": 2.464655946651591e-06, + "loss": 0.8134, + "num_input_tokens_seen": 79257555, + "step": 3681 + }, + { + "epoch": 0.44273432333313295, + "flos": 24427570431360.0, + "grad_norm": 2.3381827800497423, + "learning_rate": 2.4638982543886065e-06, + "loss": 0.8072, + "num_input_tokens_seen": 79277595, + "step": 3682 + }, + { + "epoch": 0.442854566223772, + "flos": 17503984944960.0, + "grad_norm": 2.4118909234961814, + "learning_rate": 2.4631404917526254e-06, + "loss": 0.8685, + "num_input_tokens_seen": 79294550, + "step": 3683 + }, + { + "epoch": 0.4429748091144111, + "flos": 24860518259520.0, + "grad_norm": 2.050118591786086, + "learning_rate": 2.4623826588586e-06, + "loss": 0.7936, + "num_input_tokens_seen": 79314820, + "step": 3684 + }, + { + "epoch": 0.4430950520050502, + "flos": 21583785249600.0, + "grad_norm": 1.7805234531570855, + "learning_rate": 2.461624755821492e-06, + "loss": 0.831, + "num_input_tokens_seen": 79334300, + "step": 3685 + }, + { + "epoch": 0.4432152948956893, + "flos": 24537126914880.0, + "grad_norm": 1.8917885974905375, + "learning_rate": 2.4608667827562763e-06, + "loss": 0.7675, + "num_input_tokens_seen": 79353630, + "step": 3686 + }, + { + "epoch": 0.4433355377863284, + "flos": 21731186282400.0, + "grad_norm": 2.6678758105151803, + "learning_rate": 2.460108739777936e-06, + "loss": 0.8989, + "num_input_tokens_seen": 79372765, + "step": 3687 + }, + { + "epoch": 0.44345578067696745, + "flos": 20055867302400.0, + "grad_norm": 1.7119156641063173, + "learning_rate": 2.4593506270014656e-06, + "loss": 0.7636, + "num_input_tokens_seen": 79391735, + "step": 3688 + }, + { + "epoch": 0.44357602356760656, + "flos": 24134887042080.0, + "grad_norm": 2.302789145506668, + "learning_rate": 2.45859244454187e-06, + "loss": 0.8187, + "num_input_tokens_seen": 79411525, + "step": 3689 + }, + { + "epoch": 0.44369626645824567, + "flos": 22675149901920.0, + "grad_norm": 2.2582455124149017, + "learning_rate": 2.4578341925141655e-06, + "loss": 0.6643, + "num_input_tokens_seen": 79430740, + "step": 3690 + }, + { + "epoch": 0.4438165093488847, + "flos": 37975293118560.0, + "grad_norm": 3.2372677163707135, + "learning_rate": 2.457075871033378e-06, + "loss": 0.7231, + "num_input_tokens_seen": 79452615, + "step": 3691 + }, + { + "epoch": 0.44393675223952384, + "flos": 15501223116480.0, + "grad_norm": 2.363737103342815, + "learning_rate": 2.4563174802145445e-06, + "loss": 0.886, + "num_input_tokens_seen": 79469865, + "step": 3692 + }, + { + "epoch": 0.44405699513016295, + "flos": 64483353080160.0, + "grad_norm": 0.681713308049894, + "learning_rate": 2.455559020172712e-06, + "loss": 0.4856, + "num_input_tokens_seen": 79537220, + "step": 3693 + }, + { + "epoch": 0.444177238020802, + "flos": 23953061077920.0, + "grad_norm": 2.051613833038706, + "learning_rate": 2.4548004910229385e-06, + "loss": 0.8996, + "num_input_tokens_seen": 79554795, + "step": 3694 + }, + { + "epoch": 0.4442974809114411, + "flos": 22530759619680.0, + "grad_norm": 2.3221984802241122, + "learning_rate": 2.4540418928802913e-06, + "loss": 0.8694, + "num_input_tokens_seen": 79573530, + "step": 3695 + }, + { + "epoch": 0.4444177238020802, + "flos": 17650568243040.0, + "grad_norm": 3.0281749740901787, + "learning_rate": 2.4532832258598506e-06, + "loss": 0.6567, + "num_input_tokens_seen": 79591360, + "step": 3696 + }, + { + "epoch": 0.4445379666927193, + "flos": 28580532290880.0, + "grad_norm": 2.0571085659264696, + "learning_rate": 2.4525244900767047e-06, + "loss": 0.8088, + "num_input_tokens_seen": 79612175, + "step": 3697 + }, + { + "epoch": 0.4446582095833584, + "flos": 70387285848480.0, + "grad_norm": 0.7837874974554327, + "learning_rate": 2.4517656856459536e-06, + "loss": 0.6053, + "num_input_tokens_seen": 79678020, + "step": 3698 + }, + { + "epoch": 0.4447784524739975, + "flos": 26467210395360.0, + "grad_norm": 1.9931204014004673, + "learning_rate": 2.4510068126827073e-06, + "loss": 0.6789, + "num_input_tokens_seen": 79699020, + "step": 3699 + }, + { + "epoch": 0.44489869536463655, + "flos": 11639866723200.0, + "grad_norm": 2.6213916319356363, + "learning_rate": 2.450247871302086e-06, + "loss": 0.8234, + "num_input_tokens_seen": 79716795, + "step": 3700 + }, + { + "epoch": 0.44501893825527566, + "flos": 20419296212160.0, + "grad_norm": 2.953032154056753, + "learning_rate": 2.44948886161922e-06, + "loss": 0.8303, + "num_input_tokens_seen": 79735810, + "step": 3701 + }, + { + "epoch": 0.4451391811459148, + "flos": 18235154456640.0, + "grad_norm": 1.858248778274157, + "learning_rate": 2.4487297837492524e-06, + "loss": 0.8493, + "num_input_tokens_seen": 79754975, + "step": 3702 + }, + { + "epoch": 0.44525942403655383, + "flos": 16886163232320.0, + "grad_norm": 2.0925946820166454, + "learning_rate": 2.4479706378073323e-06, + "loss": 0.6169, + "num_input_tokens_seen": 79773515, + "step": 3703 + }, + { + "epoch": 0.44537966692719294, + "flos": 23225608542240.0, + "grad_norm": 3.053029891126728, + "learning_rate": 2.447211423908623e-06, + "loss": 0.8395, + "num_input_tokens_seen": 79793560, + "step": 3704 + }, + { + "epoch": 0.445499909817832, + "flos": 21693081544800.0, + "grad_norm": 2.284730737960802, + "learning_rate": 2.4464521421682966e-06, + "loss": 0.7521, + "num_input_tokens_seen": 79811150, + "step": 3705 + }, + { + "epoch": 0.4456201527084711, + "flos": 23953655794080.0, + "grad_norm": 1.4361935359257427, + "learning_rate": 2.4456927927015345e-06, + "loss": 0.8769, + "num_input_tokens_seen": 79832190, + "step": 3706 + }, + { + "epoch": 0.4457403955991102, + "flos": 18780297821280.0, + "grad_norm": 2.281344163597317, + "learning_rate": 2.4449333756235307e-06, + "loss": 0.768, + "num_input_tokens_seen": 79848905, + "step": 3707 + }, + { + "epoch": 0.4458606384897493, + "flos": 19180381848000.0, + "grad_norm": 2.671957628188272, + "learning_rate": 2.4441738910494876e-06, + "loss": 0.7844, + "num_input_tokens_seen": 79863640, + "step": 3708 + }, + { + "epoch": 0.4459808813803884, + "flos": 21330804897600.0, + "grad_norm": 2.062660442238515, + "learning_rate": 2.4434143390946176e-06, + "loss": 0.8237, + "num_input_tokens_seen": 79882450, + "step": 3709 + }, + { + "epoch": 0.4461011242710275, + "flos": 23255461593120.0, + "grad_norm": 2.3959937553324386, + "learning_rate": 2.4426547198741457e-06, + "loss": 0.8567, + "num_input_tokens_seen": 79900655, + "step": 3710 + }, + { + "epoch": 0.44622136716166655, + "flos": 20164197183840.0, + "grad_norm": 2.7813713818754975, + "learning_rate": 2.441895033503305e-06, + "loss": 0.7482, + "num_input_tokens_seen": 79918575, + "step": 3711 + }, + { + "epoch": 0.44634161005230566, + "flos": 21252476746080.0, + "grad_norm": 1.8360691753811096, + "learning_rate": 2.4411352800973375e-06, + "loss": 0.8231, + "num_input_tokens_seen": 79937010, + "step": 3712 + }, + { + "epoch": 0.44646185294294477, + "flos": 22896567394080.0, + "grad_norm": 2.9730736017184896, + "learning_rate": 2.4403754597715005e-06, + "loss": 0.7586, + "num_input_tokens_seen": 79956455, + "step": 3713 + }, + { + "epoch": 0.4465820958335838, + "flos": 22605333625440.0, + "grad_norm": 2.341498043283496, + "learning_rate": 2.4396155726410553e-06, + "loss": 0.9326, + "num_input_tokens_seen": 79975065, + "step": 3714 + }, + { + "epoch": 0.44670233872422294, + "flos": 22640055914880.0, + "grad_norm": 4.642945194457431, + "learning_rate": 2.438855618821278e-06, + "loss": 0.9113, + "num_input_tokens_seen": 79990865, + "step": 3715 + }, + { + "epoch": 0.44682258161486205, + "flos": 23548293661440.0, + "grad_norm": 7.063529420620867, + "learning_rate": 2.4380955984274517e-06, + "loss": 0.6743, + "num_input_tokens_seen": 80009520, + "step": 3716 + }, + { + "epoch": 0.4469428245055011, + "flos": 26463233231040.0, + "grad_norm": 5.611322856386754, + "learning_rate": 2.4373355115748716e-06, + "loss": 0.7669, + "num_input_tokens_seen": 80030350, + "step": 3717 + }, + { + "epoch": 0.4470630673961402, + "flos": 21474154426560.0, + "grad_norm": 2.2450649393115722, + "learning_rate": 2.436575358378842e-06, + "loss": 0.7243, + "num_input_tokens_seen": 80049840, + "step": 3718 + }, + { + "epoch": 0.44718331028677927, + "flos": 16150830707520.0, + "grad_norm": 3.097946249626706, + "learning_rate": 2.4358151389546782e-06, + "loss": 0.8253, + "num_input_tokens_seen": 80066525, + "step": 3719 + }, + { + "epoch": 0.4473035531774184, + "flos": 19653144222720.0, + "grad_norm": 4.163783080153698, + "learning_rate": 2.4350548534177035e-06, + "loss": 0.7589, + "num_input_tokens_seen": 80083355, + "step": 3720 + }, + { + "epoch": 0.4474237960680575, + "flos": 41367864301440.0, + "grad_norm": 2.037256369037435, + "learning_rate": 2.434294501883254e-06, + "loss": 0.6687, + "num_input_tokens_seen": 80106450, + "step": 3721 + }, + { + "epoch": 0.44754403895869654, + "flos": 22858536996000.0, + "grad_norm": 2.1758364768013467, + "learning_rate": 2.433534084466674e-06, + "loss": 0.6564, + "num_input_tokens_seen": 80125670, + "step": 3722 + }, + { + "epoch": 0.44766428184933565, + "flos": 25593583428960.0, + "grad_norm": 2.6571815203711897, + "learning_rate": 2.4327736012833178e-06, + "loss": 0.7074, + "num_input_tokens_seen": 80147390, + "step": 3723 + }, + { + "epoch": 0.44778452473997477, + "flos": 20419816588800.0, + "grad_norm": 2.89065056866859, + "learning_rate": 2.4320130524485506e-06, + "loss": 0.766, + "num_input_tokens_seen": 80166185, + "step": 3724 + }, + { + "epoch": 0.4479047676306138, + "flos": 21943868880960.0, + "grad_norm": 1.6887196971961458, + "learning_rate": 2.431252438077746e-06, + "loss": 0.7986, + "num_input_tokens_seen": 80185685, + "step": 3725 + }, + { + "epoch": 0.44802501052125293, + "flos": 21436458556320.0, + "grad_norm": 2.6609318480596813, + "learning_rate": 2.4304917582862906e-06, + "loss": 0.774, + "num_input_tokens_seen": 80204865, + "step": 3726 + }, + { + "epoch": 0.44814525341189204, + "flos": 22095358587360.0, + "grad_norm": 2.7242286990448705, + "learning_rate": 2.4297310131895774e-06, + "loss": 0.875, + "num_input_tokens_seen": 80223410, + "step": 3727 + }, + { + "epoch": 0.4482654963025311, + "flos": 16630283639040.0, + "grad_norm": 3.812674846704651, + "learning_rate": 2.4289702029030113e-06, + "loss": 0.7493, + "num_input_tokens_seen": 80240880, + "step": 3728 + }, + { + "epoch": 0.4483857391931702, + "flos": 18814351055040.0, + "grad_norm": 2.1062446254582414, + "learning_rate": 2.4282093275420057e-06, + "loss": 0.8333, + "num_input_tokens_seen": 80259910, + "step": 3729 + }, + { + "epoch": 0.4485059820838093, + "flos": 20343421264800.0, + "grad_norm": 2.277539933458585, + "learning_rate": 2.4274483872219863e-06, + "loss": 0.7094, + "num_input_tokens_seen": 80277270, + "step": 3730 + }, + { + "epoch": 0.4486262249744484, + "flos": 20019175015680.0, + "grad_norm": 2.054768607855842, + "learning_rate": 2.426687382058386e-06, + "loss": 0.9379, + "num_input_tokens_seen": 80295550, + "step": 3731 + }, + { + "epoch": 0.4487464678650875, + "flos": 64503541993440.0, + "grad_norm": 0.8932111903273326, + "learning_rate": 2.425926312166649e-06, + "loss": 0.5986, + "num_input_tokens_seen": 80348425, + "step": 3732 + }, + { + "epoch": 0.4488667107557266, + "flos": 20739899824800.0, + "grad_norm": 2.8505370922288247, + "learning_rate": 2.42516517766223e-06, + "loss": 0.7335, + "num_input_tokens_seen": 80366505, + "step": 3733 + }, + { + "epoch": 0.44898695364636565, + "flos": 23917483884000.0, + "grad_norm": 2.13056889275205, + "learning_rate": 2.4244039786605907e-06, + "loss": 0.6804, + "num_input_tokens_seen": 80387510, + "step": 3734 + }, + { + "epoch": 0.44910719653700476, + "flos": 18597839971200.0, + "grad_norm": 5.03365386969617, + "learning_rate": 2.4236427152772055e-06, + "loss": 0.8317, + "num_input_tokens_seen": 80405915, + "step": 3735 + }, + { + "epoch": 0.4492274394276438, + "flos": 56951950280160.0, + "grad_norm": 0.8567173524685499, + "learning_rate": 2.422881387627557e-06, + "loss": 0.5732, + "num_input_tokens_seen": 80458365, + "step": 3736 + }, + { + "epoch": 0.4493476823182829, + "flos": 23221519868640.0, + "grad_norm": 1.7573259824768865, + "learning_rate": 2.422119995827139e-06, + "loss": 0.7721, + "num_input_tokens_seen": 80478490, + "step": 3737 + }, + { + "epoch": 0.44946792520892204, + "flos": 15792717073440.0, + "grad_norm": 4.972550767857078, + "learning_rate": 2.4213585399914528e-06, + "loss": 0.7419, + "num_input_tokens_seen": 80495695, + "step": 3738 + }, + { + "epoch": 0.4495881680995611, + "flos": 19582138513920.0, + "grad_norm": 2.170706455186284, + "learning_rate": 2.4205970202360113e-06, + "loss": 0.8538, + "num_input_tokens_seen": 80515260, + "step": 3739 + }, + { + "epoch": 0.4497084109902002, + "flos": 25994187832320.0, + "grad_norm": 2.073083254256121, + "learning_rate": 2.4198354366763354e-06, + "loss": 0.7809, + "num_input_tokens_seen": 80533900, + "step": 3740 + }, + { + "epoch": 0.4498286538808393, + "flos": 14772766997280.0, + "grad_norm": 2.969890174761768, + "learning_rate": 2.4190737894279587e-06, + "loss": 0.7828, + "num_input_tokens_seen": 80551270, + "step": 3741 + }, + { + "epoch": 0.44994889677147837, + "flos": 15428284580160.0, + "grad_norm": 2.274394879192662, + "learning_rate": 2.4183120786064203e-06, + "loss": 0.8076, + "num_input_tokens_seen": 80568145, + "step": 3742 + }, + { + "epoch": 0.4500691396621175, + "flos": 21767172343680.0, + "grad_norm": 2.8138238122546215, + "learning_rate": 2.417550304327273e-06, + "loss": 0.8529, + "num_input_tokens_seen": 80586185, + "step": 3743 + }, + { + "epoch": 0.4501893825527566, + "flos": 32335565969760.0, + "grad_norm": 2.4447077286663683, + "learning_rate": 2.4167884667060763e-06, + "loss": 0.7578, + "num_input_tokens_seen": 80608895, + "step": 3744 + }, + { + "epoch": 0.45030962544339564, + "flos": 16521470550720.0, + "grad_norm": 10.312542169658098, + "learning_rate": 2.4160265658584e-06, + "loss": 0.879, + "num_input_tokens_seen": 80623785, + "step": 3745 + }, + { + "epoch": 0.45042986833403476, + "flos": 19545148869120.0, + "grad_norm": 2.3935940325649865, + "learning_rate": 2.4152646018998253e-06, + "loss": 0.6861, + "num_input_tokens_seen": 80642735, + "step": 3746 + }, + { + "epoch": 0.45055011122467387, + "flos": 23079434111520.0, + "grad_norm": 2.080407631435087, + "learning_rate": 2.4145025749459403e-06, + "loss": 0.7195, + "num_input_tokens_seen": 80663760, + "step": 3747 + }, + { + "epoch": 0.4506703541153129, + "flos": 19906087404960.0, + "grad_norm": 2.6152507846992643, + "learning_rate": 2.413740485112344e-06, + "loss": 0.6996, + "num_input_tokens_seen": 80682100, + "step": 3748 + }, + { + "epoch": 0.45079059700595203, + "flos": 19471727125920.0, + "grad_norm": 1.8494620340023378, + "learning_rate": 2.412978332514646e-06, + "loss": 0.8234, + "num_input_tokens_seen": 80700880, + "step": 3749 + }, + { + "epoch": 0.4509108398965911, + "flos": 27596679785280.0, + "grad_norm": 2.7394860157283327, + "learning_rate": 2.4122161172684623e-06, + "loss": 0.7204, + "num_input_tokens_seen": 80721675, + "step": 3750 + }, + { + "epoch": 0.4510310827872302, + "flos": 20965740518400.0, + "grad_norm": 2.695734670996453, + "learning_rate": 2.4114538394894216e-06, + "loss": 0.8421, + "num_input_tokens_seen": 80739315, + "step": 3751 + }, + { + "epoch": 0.4511513256778693, + "flos": 16193693174400.0, + "grad_norm": 2.0875746342457178, + "learning_rate": 2.410691499293161e-06, + "loss": 0.8304, + "num_input_tokens_seen": 80757470, + "step": 3752 + }, + { + "epoch": 0.45127156856850836, + "flos": 25150451086560.0, + "grad_norm": 1.740774359878177, + "learning_rate": 2.409929096795326e-06, + "loss": 0.742, + "num_input_tokens_seen": 80777840, + "step": 3753 + }, + { + "epoch": 0.4513918114591475, + "flos": 20383421660160.0, + "grad_norm": 2.3265930870612572, + "learning_rate": 2.409166632111573e-06, + "loss": 0.7927, + "num_input_tokens_seen": 80795975, + "step": 3754 + }, + { + "epoch": 0.4515120543497866, + "flos": 26610188226720.0, + "grad_norm": 4.887289744193725, + "learning_rate": 2.4084041053575674e-06, + "loss": 0.8076, + "num_input_tokens_seen": 80815395, + "step": 3755 + }, + { + "epoch": 0.45163229724042564, + "flos": 20565730831200.0, + "grad_norm": 2.82990659795503, + "learning_rate": 2.4076415166489834e-06, + "loss": 0.7255, + "num_input_tokens_seen": 80834805, + "step": 3756 + }, + { + "epoch": 0.45175254013106475, + "flos": 21658024727520.0, + "grad_norm": 1.9922730670579185, + "learning_rate": 2.406878866101506e-06, + "loss": 0.7887, + "num_input_tokens_seen": 80853770, + "step": 3757 + }, + { + "epoch": 0.45187278302170386, + "flos": 18851935416000.0, + "grad_norm": 2.672733637389919, + "learning_rate": 2.4061161538308273e-06, + "loss": 0.7862, + "num_input_tokens_seen": 80870615, + "step": 3758 + }, + { + "epoch": 0.4519930259123429, + "flos": 18555832408800.0, + "grad_norm": 2.1925837707174995, + "learning_rate": 2.4053533799526523e-06, + "loss": 0.8923, + "num_input_tokens_seen": 80886335, + "step": 3759 + }, + { + "epoch": 0.452113268802982, + "flos": 25156323908640.0, + "grad_norm": 2.147150160301456, + "learning_rate": 2.404590544582691e-06, + "loss": 0.8667, + "num_input_tokens_seen": 80904570, + "step": 3760 + }, + { + "epoch": 0.45223351169362114, + "flos": 39321979817760.0, + "grad_norm": 3.3705382347999757, + "learning_rate": 2.403827647836666e-06, + "loss": 0.8157, + "num_input_tokens_seen": 80925125, + "step": 3761 + }, + { + "epoch": 0.4523537545842602, + "flos": 21551441824800.0, + "grad_norm": 2.3170298979982484, + "learning_rate": 2.4030646898303075e-06, + "loss": 0.6952, + "num_input_tokens_seen": 80946615, + "step": 3762 + }, + { + "epoch": 0.4524739974748993, + "flos": 28398743496480.0, + "grad_norm": 2.9505914386181096, + "learning_rate": 2.4023016706793566e-06, + "loss": 0.8226, + "num_input_tokens_seen": 80966410, + "step": 3763 + }, + { + "epoch": 0.4525942403655384, + "flos": 61468378219200.0, + "grad_norm": 0.7696464181298608, + "learning_rate": 2.401538590499561e-06, + "loss": 0.5686, + "num_input_tokens_seen": 81026865, + "step": 3764 + }, + { + "epoch": 0.45271448325617747, + "flos": 27490468580160.0, + "grad_norm": 3.1915827017181835, + "learning_rate": 2.400775449406682e-06, + "loss": 0.7211, + "num_input_tokens_seen": 81050060, + "step": 3765 + }, + { + "epoch": 0.4528347261468166, + "flos": 22420422571200.0, + "grad_norm": 2.3236947118000377, + "learning_rate": 2.400012247516485e-06, + "loss": 0.7288, + "num_input_tokens_seen": 81070625, + "step": 3766 + }, + { + "epoch": 0.45295496903745563, + "flos": 21073698702240.0, + "grad_norm": 2.31807294970732, + "learning_rate": 2.3992489849447484e-06, + "loss": 0.9046, + "num_input_tokens_seen": 81089080, + "step": 3767 + }, + { + "epoch": 0.45307521192809475, + "flos": 23187949841760.0, + "grad_norm": 2.5824844031956946, + "learning_rate": 2.3984856618072584e-06, + "loss": 0.7874, + "num_input_tokens_seen": 81109115, + "step": 3768 + }, + { + "epoch": 0.45319545481873386, + "flos": 15538064082240.0, + "grad_norm": 2.1187749450475457, + "learning_rate": 2.3977222782198098e-06, + "loss": 0.7417, + "num_input_tokens_seen": 81127750, + "step": 3769 + }, + { + "epoch": 0.4533156977093729, + "flos": 21913421113920.0, + "grad_norm": 3.585940691379299, + "learning_rate": 2.3969588342982077e-06, + "loss": 0.7566, + "num_input_tokens_seen": 81147125, + "step": 3770 + }, + { + "epoch": 0.453435940600012, + "flos": 24207453880800.0, + "grad_norm": 1.9193045321179927, + "learning_rate": 2.396195330158267e-06, + "loss": 0.7308, + "num_input_tokens_seen": 81167015, + "step": 3771 + }, + { + "epoch": 0.45355618349065113, + "flos": 23406839790240.0, + "grad_norm": 2.4662323681084595, + "learning_rate": 2.3954317659158094e-06, + "loss": 0.7983, + "num_input_tokens_seen": 81187350, + "step": 3772 + }, + { + "epoch": 0.4536764263812902, + "flos": 66807982293120.0, + "grad_norm": 0.9062608520910389, + "learning_rate": 2.394668141686667e-06, + "loss": 0.5685, + "num_input_tokens_seen": 81249315, + "step": 3773 + }, + { + "epoch": 0.4537966692719293, + "flos": 42678081732480.0, + "grad_norm": 2.5611509167861795, + "learning_rate": 2.3939044575866813e-06, + "loss": 0.6959, + "num_input_tokens_seen": 81272065, + "step": 3774 + }, + { + "epoch": 0.4539169121625684, + "flos": 35498170615680.0, + "grad_norm": 2.4131616189189313, + "learning_rate": 2.3931407137317024e-06, + "loss": 0.7543, + "num_input_tokens_seen": 81292220, + "step": 3775 + }, + { + "epoch": 0.45403715505320746, + "flos": 18488357827200.0, + "grad_norm": 5.708493566932606, + "learning_rate": 2.3923769102375907e-06, + "loss": 0.8471, + "num_input_tokens_seen": 81311085, + "step": 3776 + }, + { + "epoch": 0.4541573979438466, + "flos": 25009963629120.0, + "grad_norm": 4.6173954899087315, + "learning_rate": 2.391613047220213e-06, + "loss": 0.789, + "num_input_tokens_seen": 81330985, + "step": 3777 + }, + { + "epoch": 0.4542776408344857, + "flos": 18306308844480.0, + "grad_norm": 2.6323274341040697, + "learning_rate": 2.390849124795447e-06, + "loss": 0.7892, + "num_input_tokens_seen": 81346985, + "step": 3778 + }, + { + "epoch": 0.45439788372512474, + "flos": 20671533168960.0, + "grad_norm": 2.173581228433366, + "learning_rate": 2.3900851430791804e-06, + "loss": 0.8408, + "num_input_tokens_seen": 81365005, + "step": 3779 + }, + { + "epoch": 0.45451812661576385, + "flos": 22277556249120.0, + "grad_norm": 2.659126878004586, + "learning_rate": 2.389321102187307e-06, + "loss": 0.8485, + "num_input_tokens_seen": 81383785, + "step": 3780 + }, + { + "epoch": 0.4546383695064029, + "flos": 21732264205440.0, + "grad_norm": 2.501178082409719, + "learning_rate": 2.3885570022357326e-06, + "loss": 0.8147, + "num_input_tokens_seen": 81402915, + "step": 3781 + }, + { + "epoch": 0.454758612397042, + "flos": 64151561369760.0, + "grad_norm": 0.8053184924126614, + "learning_rate": 2.38779284334037e-06, + "loss": 0.6081, + "num_input_tokens_seen": 81467890, + "step": 3782 + }, + { + "epoch": 0.4548788552876811, + "flos": 27264888074880.0, + "grad_norm": 2.3063537781065544, + "learning_rate": 2.387028625617141e-06, + "loss": 0.7894, + "num_input_tokens_seen": 81487105, + "step": 3783 + }, + { + "epoch": 0.4549990981783202, + "flos": 22824483762240.0, + "grad_norm": 3.7906752467461446, + "learning_rate": 2.3862643491819766e-06, + "loss": 0.8454, + "num_input_tokens_seen": 81505670, + "step": 3784 + }, + { + "epoch": 0.4551193410689593, + "flos": 23225348353920.0, + "grad_norm": 1.9502985530110704, + "learning_rate": 2.3855000141508186e-06, + "loss": 0.8459, + "num_input_tokens_seen": 81525825, + "step": 3785 + }, + { + "epoch": 0.4552395839595984, + "flos": 20747891323200.0, + "grad_norm": 3.625647651431946, + "learning_rate": 2.3847356206396143e-06, + "loss": 0.8378, + "num_input_tokens_seen": 81543135, + "step": 3786 + }, + { + "epoch": 0.45535982685023746, + "flos": 23224233261120.0, + "grad_norm": 1.743778887547038, + "learning_rate": 2.3839711687643227e-06, + "loss": 0.785, + "num_input_tokens_seen": 81564360, + "step": 3787 + }, + { + "epoch": 0.45548006974087657, + "flos": 19618124575200.0, + "grad_norm": 2.221926282674483, + "learning_rate": 2.38320665864091e-06, + "loss": 0.7441, + "num_input_tokens_seen": 81583710, + "step": 3788 + }, + { + "epoch": 0.4556003126315157, + "flos": 20019212185440.0, + "grad_norm": 2.0822009505917096, + "learning_rate": 2.3824420903853516e-06, + "loss": 0.8194, + "num_input_tokens_seen": 81602175, + "step": 3789 + }, + { + "epoch": 0.45572055552215474, + "flos": 22926680633280.0, + "grad_norm": 3.4870329841561256, + "learning_rate": 2.3816774641136324e-06, + "loss": 0.8225, + "num_input_tokens_seen": 81619430, + "step": 3790 + }, + { + "epoch": 0.45584079841279385, + "flos": 33062349449760.0, + "grad_norm": 1.9096717766158342, + "learning_rate": 2.380912779941745e-06, + "loss": 0.7152, + "num_input_tokens_seen": 81642550, + "step": 3791 + }, + { + "epoch": 0.45596104130343296, + "flos": 27233102196480.0, + "grad_norm": 2.6047537993727317, + "learning_rate": 2.3801480379856918e-06, + "loss": 0.8371, + "num_input_tokens_seen": 81664260, + "step": 3792 + }, + { + "epoch": 0.456081284194072, + "flos": 21548988620640.0, + "grad_norm": 2.0385280800355097, + "learning_rate": 2.379383238361484e-06, + "loss": 0.8372, + "num_input_tokens_seen": 81683620, + "step": 3793 + }, + { + "epoch": 0.4562015270847111, + "flos": 35867137819680.0, + "grad_norm": 2.2377919566022966, + "learning_rate": 2.3786183811851407e-06, + "loss": 0.7959, + "num_input_tokens_seen": 81704325, + "step": 3794 + }, + { + "epoch": 0.45632176997535023, + "flos": 13570619259360.0, + "grad_norm": 6.078919503423414, + "learning_rate": 2.3778534665726892e-06, + "loss": 0.8036, + "num_input_tokens_seen": 81721155, + "step": 3795 + }, + { + "epoch": 0.4564420128659893, + "flos": 32588509152000.0, + "grad_norm": 2.2267629432803138, + "learning_rate": 2.377088494640168e-06, + "loss": 0.7257, + "num_input_tokens_seen": 81742905, + "step": 3796 + }, + { + "epoch": 0.4565622557566284, + "flos": 20348699370720.0, + "grad_norm": 1.9450142873126515, + "learning_rate": 2.3763234655036216e-06, + "loss": 0.7835, + "num_input_tokens_seen": 81762105, + "step": 3797 + }, + { + "epoch": 0.45668249864726745, + "flos": 25338038363520.0, + "grad_norm": 2.3537552960822943, + "learning_rate": 2.3755583792791046e-06, + "loss": 0.8697, + "num_input_tokens_seen": 81781975, + "step": 3798 + }, + { + "epoch": 0.45680274153790656, + "flos": 15537766724160.0, + "grad_norm": 2.181796239973768, + "learning_rate": 2.3747932360826803e-06, + "loss": 0.7457, + "num_input_tokens_seen": 81798905, + "step": 3799 + }, + { + "epoch": 0.4569229844285457, + "flos": 19764707873280.0, + "grad_norm": 3.3106237892060806, + "learning_rate": 2.3740280360304205e-06, + "loss": 0.8235, + "num_input_tokens_seen": 81816630, + "step": 3800 + }, + { + "epoch": 0.45704322731918473, + "flos": 24059346622560.0, + "grad_norm": 1.8263783384082044, + "learning_rate": 2.3732627792384038e-06, + "loss": 0.6767, + "num_input_tokens_seen": 81837700, + "step": 3801 + }, + { + "epoch": 0.45716347020982384, + "flos": 31272679087200.0, + "grad_norm": 2.3443038286080617, + "learning_rate": 2.3724974658227207e-06, + "loss": 0.7551, + "num_input_tokens_seen": 81858490, + "step": 3802 + }, + { + "epoch": 0.45728371310046295, + "flos": 26463270400800.0, + "grad_norm": 2.027738499660188, + "learning_rate": 2.3717320958994687e-06, + "loss": 0.7124, + "num_input_tokens_seen": 81876050, + "step": 3803 + }, + { + "epoch": 0.457403955991102, + "flos": 17903845953120.0, + "grad_norm": 2.1985200412332158, + "learning_rate": 2.3709666695847534e-06, + "loss": 0.7049, + "num_input_tokens_seen": 81894230, + "step": 3804 + }, + { + "epoch": 0.4575241988817411, + "flos": 42169779333600.0, + "grad_norm": 3.053616777559741, + "learning_rate": 2.370201186994689e-06, + "loss": 0.7009, + "num_input_tokens_seen": 81917915, + "step": 3805 + }, + { + "epoch": 0.45764444177238023, + "flos": 30074062476480.0, + "grad_norm": 3.950722910725831, + "learning_rate": 2.369435648245399e-06, + "loss": 0.7006, + "num_input_tokens_seen": 81938130, + "step": 3806 + }, + { + "epoch": 0.4577646846630193, + "flos": 24025850935200.0, + "grad_norm": 28.488254201361766, + "learning_rate": 2.368670053453015e-06, + "loss": 0.8499, + "num_input_tokens_seen": 81959820, + "step": 3807 + }, + { + "epoch": 0.4578849275536584, + "flos": 17393759405760.0, + "grad_norm": 2.3634352956961058, + "learning_rate": 2.3679044027336757e-06, + "loss": 0.7424, + "num_input_tokens_seen": 81975710, + "step": 3808 + }, + { + "epoch": 0.4580051704442975, + "flos": 13490878656960.0, + "grad_norm": 2.7550432604043116, + "learning_rate": 2.3671386962035326e-06, + "loss": 0.693, + "num_input_tokens_seen": 81993180, + "step": 3809 + }, + { + "epoch": 0.45812541333493656, + "flos": 18011841306720.0, + "grad_norm": 2.7068492833749933, + "learning_rate": 2.3663729339787405e-06, + "loss": 0.6911, + "num_input_tokens_seen": 82010115, + "step": 3810 + }, + { + "epoch": 0.45824565622557567, + "flos": 20193083820960.0, + "grad_norm": 3.583236973967656, + "learning_rate": 2.365607116175466e-06, + "loss": 0.7382, + "num_input_tokens_seen": 82025540, + "step": 3811 + }, + { + "epoch": 0.4583658991162148, + "flos": 19836568486560.0, + "grad_norm": 2.829203026097353, + "learning_rate": 2.3648412429098825e-06, + "loss": 0.6684, + "num_input_tokens_seen": 82043890, + "step": 3812 + }, + { + "epoch": 0.45848614200685384, + "flos": 20999830921920.0, + "grad_norm": 2.348684989999931, + "learning_rate": 2.364075314298172e-06, + "loss": 0.82, + "num_input_tokens_seen": 82061345, + "step": 3813 + }, + { + "epoch": 0.45860638489749295, + "flos": 21039236601120.0, + "grad_norm": 2.487698329514295, + "learning_rate": 2.3633093304565267e-06, + "loss": 0.6989, + "num_input_tokens_seen": 82080400, + "step": 3814 + }, + { + "epoch": 0.458726627788132, + "flos": 26795433808800.0, + "grad_norm": 2.930895258864195, + "learning_rate": 2.3625432915011443e-06, + "loss": 0.6316, + "num_input_tokens_seen": 82102310, + "step": 3815 + }, + { + "epoch": 0.4588468706787711, + "flos": 24063472465920.0, + "grad_norm": 1.8055329427362736, + "learning_rate": 2.3617771975482334e-06, + "loss": 0.6562, + "num_input_tokens_seen": 82121140, + "step": 3816 + }, + { + "epoch": 0.4589671135694102, + "flos": 17863696878720.0, + "grad_norm": 2.3443029132968274, + "learning_rate": 2.3610110487140083e-06, + "loss": 0.7426, + "num_input_tokens_seen": 82139575, + "step": 3817 + }, + { + "epoch": 0.4590873564600493, + "flos": 25590163811040.0, + "grad_norm": 2.2471159358824537, + "learning_rate": 2.360244845114695e-06, + "loss": 0.8098, + "num_input_tokens_seen": 82159195, + "step": 3818 + }, + { + "epoch": 0.4592075993506884, + "flos": 18488432166720.0, + "grad_norm": 2.4037619273762156, + "learning_rate": 2.3594785868665245e-06, + "loss": 0.6819, + "num_input_tokens_seen": 82175500, + "step": 3819 + }, + { + "epoch": 0.4593278422413275, + "flos": 20601679722720.0, + "grad_norm": 6.067209204144478, + "learning_rate": 2.3587122740857386e-06, + "loss": 0.81, + "num_input_tokens_seen": 82192600, + "step": 3820 + }, + { + "epoch": 0.45944808513196655, + "flos": 21327942826080.0, + "grad_norm": 1.9029419132855172, + "learning_rate": 2.357945906888586e-06, + "loss": 0.7833, + "num_input_tokens_seen": 82212040, + "step": 3821 + }, + { + "epoch": 0.45956832802260567, + "flos": 21396978537600.0, + "grad_norm": 2.886367879767497, + "learning_rate": 2.357179485391324e-06, + "loss": 0.8001, + "num_input_tokens_seen": 82229895, + "step": 3822 + }, + { + "epoch": 0.4596885709132448, + "flos": 22350680634240.0, + "grad_norm": 2.2191392798840206, + "learning_rate": 2.3564130097102173e-06, + "loss": 0.858, + "num_input_tokens_seen": 82249550, + "step": 3823 + }, + { + "epoch": 0.45980881380388383, + "flos": 28940207054880.0, + "grad_norm": 2.413099275781378, + "learning_rate": 2.355646479961541e-06, + "loss": 0.7521, + "num_input_tokens_seen": 82268860, + "step": 3824 + }, + { + "epoch": 0.45992905669452294, + "flos": 33348416621760.0, + "grad_norm": 2.0581269576465644, + "learning_rate": 2.354879896261576e-06, + "loss": 0.7168, + "num_input_tokens_seen": 82289070, + "step": 3825 + }, + { + "epoch": 0.46004929958516205, + "flos": 36265846565280.0, + "grad_norm": 2.4464222885067066, + "learning_rate": 2.3541132587266133e-06, + "loss": 0.5683, + "num_input_tokens_seen": 82311790, + "step": 3826 + }, + { + "epoch": 0.4601695424758011, + "flos": 17213977778400.0, + "grad_norm": 3.4010337996963873, + "learning_rate": 2.3533465674729515e-06, + "loss": 0.6957, + "num_input_tokens_seen": 82329020, + "step": 3827 + }, + { + "epoch": 0.4602897853664402, + "flos": 15865915798080.0, + "grad_norm": 2.9067046107285757, + "learning_rate": 2.352579822616895e-06, + "loss": 0.7306, + "num_input_tokens_seen": 82346455, + "step": 3828 + }, + { + "epoch": 0.4604100282570793, + "flos": 25376291780160.0, + "grad_norm": 1.9687119737616074, + "learning_rate": 2.351813024274761e-06, + "loss": 0.7795, + "num_input_tokens_seen": 82367810, + "step": 3829 + }, + { + "epoch": 0.4605302711477184, + "flos": 27590881302720.0, + "grad_norm": 2.1916458080541723, + "learning_rate": 2.3510461725628693e-06, + "loss": 0.7388, + "num_input_tokens_seen": 82388275, + "step": 3830 + }, + { + "epoch": 0.4606505140383575, + "flos": 23805474196320.0, + "grad_norm": 2.3235143283232276, + "learning_rate": 2.350279267597554e-06, + "loss": 0.7107, + "num_input_tokens_seen": 82408915, + "step": 3831 + }, + { + "epoch": 0.46077075692899655, + "flos": 16084173860640.0, + "grad_norm": 2.96177077835096, + "learning_rate": 2.3495123094951515e-06, + "loss": 0.8309, + "num_input_tokens_seen": 82427260, + "step": 3832 + }, + { + "epoch": 0.46089099981963566, + "flos": 48727705724640.0, + "grad_norm": 2.560238188898217, + "learning_rate": 2.34874529837201e-06, + "loss": 0.7628, + "num_input_tokens_seen": 82450805, + "step": 3833 + }, + { + "epoch": 0.46101124271027477, + "flos": 19072051966560.0, + "grad_norm": 3.339952397949535, + "learning_rate": 2.347978234344483e-06, + "loss": 0.7941, + "num_input_tokens_seen": 82467010, + "step": 3834 + }, + { + "epoch": 0.4611314856009138, + "flos": 39291234692640.0, + "grad_norm": 2.9738266754848364, + "learning_rate": 2.347211117528935e-06, + "loss": 0.6879, + "num_input_tokens_seen": 82489310, + "step": 3835 + }, + { + "epoch": 0.46125172849155294, + "flos": 20780978143200.0, + "grad_norm": 1.853838202139527, + "learning_rate": 2.3464439480417374e-06, + "loss": 0.7175, + "num_input_tokens_seen": 82508830, + "step": 3836 + }, + { + "epoch": 0.46137197138219205, + "flos": 17905295573760.0, + "grad_norm": 2.9957002980737086, + "learning_rate": 2.3456767259992676e-06, + "loss": 0.7808, + "num_input_tokens_seen": 82526475, + "step": 3837 + }, + { + "epoch": 0.4614922142728311, + "flos": 16812704319360.0, + "grad_norm": 4.618543706475877, + "learning_rate": 2.3449094515179135e-06, + "loss": 0.8882, + "num_input_tokens_seen": 82543935, + "step": 3838 + }, + { + "epoch": 0.4616124571634702, + "flos": 26576766878880.0, + "grad_norm": 1.8676348034720318, + "learning_rate": 2.34414212471407e-06, + "loss": 0.8214, + "num_input_tokens_seen": 82563730, + "step": 3839 + }, + { + "epoch": 0.4617327000541093, + "flos": 20311152179520.0, + "grad_norm": 2.9978989555325257, + "learning_rate": 2.3433747457041394e-06, + "loss": 0.7301, + "num_input_tokens_seen": 82582435, + "step": 3840 + }, + { + "epoch": 0.4618529429447484, + "flos": 29528324395680.0, + "grad_norm": 1.8055902510349817, + "learning_rate": 2.342607314604533e-06, + "loss": 0.848, + "num_input_tokens_seen": 82602185, + "step": 3841 + }, + { + "epoch": 0.4619731858353875, + "flos": 19758649202400.0, + "grad_norm": 2.1659644407204786, + "learning_rate": 2.3418398315316694e-06, + "loss": 0.842, + "num_input_tokens_seen": 82620005, + "step": 3842 + }, + { + "epoch": 0.4620934287260266, + "flos": 18924093387360.0, + "grad_norm": 3.259381618474749, + "learning_rate": 2.3410722966019755e-06, + "loss": 0.7836, + "num_input_tokens_seen": 82635120, + "step": 3843 + }, + { + "epoch": 0.46221367161666566, + "flos": 37287915317760.0, + "grad_norm": 1.9323792005044549, + "learning_rate": 2.3403047099318848e-06, + "loss": 0.6556, + "num_input_tokens_seen": 82659190, + "step": 3844 + }, + { + "epoch": 0.46233391450730477, + "flos": 14408148655200.0, + "grad_norm": 2.668973719490702, + "learning_rate": 2.3395370716378405e-06, + "loss": 0.7472, + "num_input_tokens_seen": 82676635, + "step": 3845 + }, + { + "epoch": 0.4624541573979438, + "flos": 22460943343200.0, + "grad_norm": 3.038680260947421, + "learning_rate": 2.338769381836292e-06, + "loss": 0.7293, + "num_input_tokens_seen": 82696245, + "step": 3846 + }, + { + "epoch": 0.46257440028858293, + "flos": 14444617923360.0, + "grad_norm": 2.66326856668887, + "learning_rate": 2.3380016406436984e-06, + "loss": 0.7346, + "num_input_tokens_seen": 82713725, + "step": 3847 + }, + { + "epoch": 0.46269464317922204, + "flos": 23298881606400.0, + "grad_norm": 2.284586100080634, + "learning_rate": 2.337233848176524e-06, + "loss": 0.8144, + "num_input_tokens_seen": 82731495, + "step": 3848 + }, + { + "epoch": 0.4628148860698611, + "flos": 18525644830080.0, + "grad_norm": 3.242738681499658, + "learning_rate": 2.3364660045512435e-06, + "loss": 0.8359, + "num_input_tokens_seen": 82750255, + "step": 3849 + }, + { + "epoch": 0.4629351289605002, + "flos": 70567141815360.0, + "grad_norm": 0.7480576716109526, + "learning_rate": 2.335698109884337e-06, + "loss": 0.5822, + "num_input_tokens_seen": 82815460, + "step": 3850 + }, + { + "epoch": 0.4630553718511393, + "flos": 59602827175680.0, + "grad_norm": 0.7988879640733332, + "learning_rate": 2.334930164292294e-06, + "loss": 0.5983, + "num_input_tokens_seen": 82878010, + "step": 3851 + }, + { + "epoch": 0.4631756147417784, + "flos": 15937962260160.0, + "grad_norm": 7.211796834376456, + "learning_rate": 2.334162167891612e-06, + "loss": 0.804, + "num_input_tokens_seen": 82895275, + "step": 3852 + }, + { + "epoch": 0.4632958576324175, + "flos": 16449572767680.0, + "grad_norm": 4.691236710482567, + "learning_rate": 2.333394120798795e-06, + "loss": 0.7477, + "num_input_tokens_seen": 82914010, + "step": 3853 + }, + { + "epoch": 0.4634161005230566, + "flos": 22314583063680.0, + "grad_norm": 2.8182705490485254, + "learning_rate": 2.3326260231303545e-06, + "loss": 0.7195, + "num_input_tokens_seen": 82932610, + "step": 3854 + }, + { + "epoch": 0.46353634341369565, + "flos": 15719927216160.0, + "grad_norm": 2.3380468534632186, + "learning_rate": 2.331857875002811e-06, + "loss": 0.8664, + "num_input_tokens_seen": 82951210, + "step": 3855 + }, + { + "epoch": 0.46365658630433476, + "flos": 28289187012960.0, + "grad_norm": 2.228042806470971, + "learning_rate": 2.3310896765326916e-06, + "loss": 0.7604, + "num_input_tokens_seen": 82972210, + "step": 3856 + }, + { + "epoch": 0.46377682919497387, + "flos": 24572852787840.0, + "grad_norm": 1.8351176277731278, + "learning_rate": 2.330321427836531e-06, + "loss": 0.8442, + "num_input_tokens_seen": 82994080, + "step": 3857 + }, + { + "epoch": 0.4638970720856129, + "flos": 19033761380160.0, + "grad_norm": 2.2117177711950875, + "learning_rate": 2.3295531290308733e-06, + "loss": 0.8288, + "num_input_tokens_seen": 83012230, + "step": 3858 + }, + { + "epoch": 0.46401731497625204, + "flos": 18444454607040.0, + "grad_norm": 3.017759842323836, + "learning_rate": 2.3287847802322678e-06, + "loss": 0.7602, + "num_input_tokens_seen": 83027800, + "step": 3859 + }, + { + "epoch": 0.4641375578668911, + "flos": 26029690686720.0, + "grad_norm": 2.0173459062497554, + "learning_rate": 2.3280163815572723e-06, + "loss": 0.8372, + "num_input_tokens_seen": 83048395, + "step": 3860 + }, + { + "epoch": 0.4642578007575302, + "flos": 19542398306880.0, + "grad_norm": 2.5973846523017996, + "learning_rate": 2.3272479331224522e-06, + "loss": 0.7703, + "num_input_tokens_seen": 83065825, + "step": 3861 + }, + { + "epoch": 0.4643780436481693, + "flos": 28145911823520.0, + "grad_norm": 2.6361537299704487, + "learning_rate": 2.3264794350443817e-06, + "loss": 0.7769, + "num_input_tokens_seen": 83087920, + "step": 3862 + }, + { + "epoch": 0.46449828653880837, + "flos": 25338967607520.0, + "grad_norm": 2.835891765299266, + "learning_rate": 2.3257108874396396e-06, + "loss": 0.7876, + "num_input_tokens_seen": 83109015, + "step": 3863 + }, + { + "epoch": 0.4646185294294475, + "flos": 16011607021920.0, + "grad_norm": 2.3332014614260874, + "learning_rate": 2.3249422904248152e-06, + "loss": 0.7375, + "num_input_tokens_seen": 83127450, + "step": 3864 + }, + { + "epoch": 0.4647387723200866, + "flos": 26325459166080.0, + "grad_norm": 3.9629531938136013, + "learning_rate": 2.324173644116504e-06, + "loss": 0.8726, + "num_input_tokens_seen": 83150300, + "step": 3865 + }, + { + "epoch": 0.46485901521072565, + "flos": 27121204018080.0, + "grad_norm": 2.580910951824483, + "learning_rate": 2.3234049486313087e-06, + "loss": 0.8153, + "num_input_tokens_seen": 83171750, + "step": 3866 + }, + { + "epoch": 0.46497925810136476, + "flos": 24245112581280.0, + "grad_norm": 2.196993147167827, + "learning_rate": 2.322636204085839e-06, + "loss": 0.754, + "num_input_tokens_seen": 83191820, + "step": 3867 + }, + { + "epoch": 0.46509950099200387, + "flos": 16230088103040.0, + "grad_norm": 2.4267126190459827, + "learning_rate": 2.3218674105967143e-06, + "loss": 0.7895, + "num_input_tokens_seen": 83210085, + "step": 3868 + }, + { + "epoch": 0.4652197438826429, + "flos": 23408809787520.0, + "grad_norm": 2.533079639916243, + "learning_rate": 2.3210985682805593e-06, + "loss": 0.8365, + "num_input_tokens_seen": 83231865, + "step": 3869 + }, + { + "epoch": 0.46533998677328203, + "flos": 16193395816320.0, + "grad_norm": 2.9731887560991703, + "learning_rate": 2.320329677254007e-06, + "loss": 0.6843, + "num_input_tokens_seen": 83249195, + "step": 3870 + }, + { + "epoch": 0.46546022966392114, + "flos": 21111691930560.0, + "grad_norm": 3.601802925475635, + "learning_rate": 2.319560737633697e-06, + "loss": 0.7359, + "num_input_tokens_seen": 83267915, + "step": 3871 + }, + { + "epoch": 0.4655804725545602, + "flos": 41112579424320.0, + "grad_norm": 1.735395595182953, + "learning_rate": 2.3187917495362775e-06, + "loss": 0.6811, + "num_input_tokens_seen": 83292325, + "step": 3872 + }, + { + "epoch": 0.4657007154451993, + "flos": 19542993023040.0, + "grad_norm": 2.9414117618017745, + "learning_rate": 2.318022713078403e-06, + "loss": 0.7651, + "num_input_tokens_seen": 83306905, + "step": 3873 + }, + { + "epoch": 0.4658209583358384, + "flos": 15495647652480.0, + "grad_norm": 6.103540312450072, + "learning_rate": 2.3172536283767354e-06, + "loss": 0.8501, + "num_input_tokens_seen": 83320665, + "step": 3874 + }, + { + "epoch": 0.4659412012264775, + "flos": 14882100462240.0, + "grad_norm": 2.802212869327892, + "learning_rate": 2.3164844955479447e-06, + "loss": 0.8135, + "num_input_tokens_seen": 83336475, + "step": 3875 + }, + { + "epoch": 0.4660614441171166, + "flos": 24390655126080.0, + "grad_norm": 1.9781261184954382, + "learning_rate": 2.3157153147087082e-06, + "loss": 0.7059, + "num_input_tokens_seen": 83358365, + "step": 3876 + }, + { + "epoch": 0.46618168700775564, + "flos": 22059558374880.0, + "grad_norm": 2.065392286895419, + "learning_rate": 2.314946085975709e-06, + "loss": 0.831, + "num_input_tokens_seen": 83377345, + "step": 3877 + }, + { + "epoch": 0.46630192989839475, + "flos": 26139470188800.0, + "grad_norm": 1.8331516638246297, + "learning_rate": 2.3141768094656393e-06, + "loss": 0.8266, + "num_input_tokens_seen": 83395920, + "step": 3878 + }, + { + "epoch": 0.46642217278903386, + "flos": 11493469273920.0, + "grad_norm": 2.5651899036208183, + "learning_rate": 2.3134074852951966e-06, + "loss": 0.8307, + "num_input_tokens_seen": 83412510, + "step": 3879 + }, + { + "epoch": 0.4665424156796729, + "flos": 32260397247840.0, + "grad_norm": 2.0123247437549967, + "learning_rate": 2.312638113581088e-06, + "loss": 0.7809, + "num_input_tokens_seen": 83432995, + "step": 3880 + }, + { + "epoch": 0.46666265857031203, + "flos": 18409732317600.0, + "grad_norm": 3.0060849467627024, + "learning_rate": 2.311868694440027e-06, + "loss": 0.7871, + "num_input_tokens_seen": 83447415, + "step": 3881 + }, + { + "epoch": 0.46678290146095114, + "flos": 68341698722880.0, + "grad_norm": 0.7412207078855186, + "learning_rate": 2.3110992279887323e-06, + "loss": 0.6242, + "num_input_tokens_seen": 83519340, + "step": 3882 + }, + { + "epoch": 0.4669031443515902, + "flos": 17687334869280.0, + "grad_norm": 2.7882657662835793, + "learning_rate": 2.310329714343932e-06, + "loss": 0.8522, + "num_input_tokens_seen": 83535490, + "step": 3883 + }, + { + "epoch": 0.4670233872422293, + "flos": 23913358040640.0, + "grad_norm": 2.022914156010252, + "learning_rate": 2.309560153622361e-06, + "loss": 0.8196, + "num_input_tokens_seen": 83552400, + "step": 3884 + }, + { + "epoch": 0.4671436301328684, + "flos": 28070966120160.0, + "grad_norm": 3.114733655144959, + "learning_rate": 2.3087905459407602e-06, + "loss": 0.746, + "num_input_tokens_seen": 83571340, + "step": 3885 + }, + { + "epoch": 0.46726387302350747, + "flos": 69270385571040.0, + "grad_norm": 0.8676964580796401, + "learning_rate": 2.3080208914158795e-06, + "loss": 0.6284, + "num_input_tokens_seen": 83634340, + "step": 3886 + }, + { + "epoch": 0.4673841159141466, + "flos": 25483097701440.0, + "grad_norm": 2.8123810637015465, + "learning_rate": 2.3072511901644753e-06, + "loss": 0.7286, + "num_input_tokens_seen": 83653410, + "step": 3887 + }, + { + "epoch": 0.4675043588047857, + "flos": 24464225548320.0, + "grad_norm": 2.353071267374463, + "learning_rate": 2.306481442303309e-06, + "loss": 0.8104, + "num_input_tokens_seen": 83672985, + "step": 3888 + }, + { + "epoch": 0.46762460169542475, + "flos": 20930906719680.0, + "grad_norm": 2.4566131866858942, + "learning_rate": 2.3057116479491515e-06, + "loss": 0.7326, + "num_input_tokens_seen": 83692395, + "step": 3889 + }, + { + "epoch": 0.46774484458606386, + "flos": 19144358616960.0, + "grad_norm": 5.496999095320562, + "learning_rate": 2.30494180721878e-06, + "loss": 0.7602, + "num_input_tokens_seen": 83709735, + "step": 3890 + }, + { + "epoch": 0.4678650874767029, + "flos": 17942136539520.0, + "grad_norm": 2.521534677678914, + "learning_rate": 2.3041719202289794e-06, + "loss": 0.8982, + "num_input_tokens_seen": 83725910, + "step": 3891 + }, + { + "epoch": 0.467985330367342, + "flos": 21330321690720.0, + "grad_norm": 2.042481694413779, + "learning_rate": 2.30340198709654e-06, + "loss": 0.8071, + "num_input_tokens_seen": 83745745, + "step": 3892 + }, + { + "epoch": 0.46810557325798113, + "flos": 20492569276320.0, + "grad_norm": 2.1115993498459926, + "learning_rate": 2.3026320079382605e-06, + "loss": 0.7484, + "num_input_tokens_seen": 83762680, + "step": 3893 + }, + { + "epoch": 0.4682258161486202, + "flos": 30074620022880.0, + "grad_norm": 3.1846393389387457, + "learning_rate": 2.3018619828709454e-06, + "loss": 0.7614, + "num_input_tokens_seen": 83784220, + "step": 3894 + }, + { + "epoch": 0.4683460590392593, + "flos": 25256848140480.0, + "grad_norm": 2.6691443337339775, + "learning_rate": 2.3010919120114084e-06, + "loss": 0.8211, + "num_input_tokens_seen": 83800750, + "step": 3895 + }, + { + "epoch": 0.4684663019298984, + "flos": 15346239452640.0, + "grad_norm": 2.6974629102813905, + "learning_rate": 2.3003217954764672e-06, + "loss": 0.6616, + "num_input_tokens_seen": 83815455, + "step": 3896 + }, + { + "epoch": 0.46858654482053747, + "flos": 27738877051680.0, + "grad_norm": 2.0345536582772703, + "learning_rate": 2.299551633382949e-06, + "loss": 0.7912, + "num_input_tokens_seen": 83835765, + "step": 3897 + }, + { + "epoch": 0.4687067877111766, + "flos": 18014963566560.0, + "grad_norm": 3.1073135366681512, + "learning_rate": 2.2987814258476854e-06, + "loss": 0.8556, + "num_input_tokens_seen": 83853565, + "step": 3898 + }, + { + "epoch": 0.4688270306018157, + "flos": 16953080267520.0, + "grad_norm": 3.6204645130708397, + "learning_rate": 2.2980111729875177e-06, + "loss": 0.6853, + "num_input_tokens_seen": 83869815, + "step": 3899 + }, + { + "epoch": 0.46894727349245474, + "flos": 17796296636640.0, + "grad_norm": 2.9603966694603203, + "learning_rate": 2.2972408749192917e-06, + "loss": 0.8232, + "num_input_tokens_seen": 83887545, + "step": 3900 + }, + { + "epoch": 0.46906751638309385, + "flos": 21440733078720.0, + "grad_norm": 3.055769550651995, + "learning_rate": 2.296470531759861e-06, + "loss": 0.6728, + "num_input_tokens_seen": 83905645, + "step": 3901 + }, + { + "epoch": 0.46918775927373296, + "flos": 20308773314880.0, + "grad_norm": 2.3143686528624907, + "learning_rate": 2.2957001436260866e-06, + "loss": 0.796, + "num_input_tokens_seen": 83922705, + "step": 3902 + }, + { + "epoch": 0.469308002164372, + "flos": 18376980025440.0, + "grad_norm": 2.0430820645185004, + "learning_rate": 2.294929710634836e-06, + "loss": 0.7344, + "num_input_tokens_seen": 83940795, + "step": 3903 + }, + { + "epoch": 0.46942824505501113, + "flos": 37907409669600.0, + "grad_norm": 2.4269260033516944, + "learning_rate": 2.2941592329029823e-06, + "loss": 0.6131, + "num_input_tokens_seen": 83961900, + "step": 3904 + }, + { + "epoch": 0.46954848794565024, + "flos": 21841300312320.0, + "grad_norm": 1.9933574158458736, + "learning_rate": 2.2933887105474067e-06, + "loss": 0.791, + "num_input_tokens_seen": 83980075, + "step": 3905 + }, + { + "epoch": 0.4696687308362893, + "flos": 21985058708640.0, + "grad_norm": 2.2060761942496923, + "learning_rate": 2.2926181436849974e-06, + "loss": 0.8174, + "num_input_tokens_seen": 83999430, + "step": 3906 + }, + { + "epoch": 0.4697889737269284, + "flos": 21582707326560.0, + "grad_norm": 1.8332873107451035, + "learning_rate": 2.2918475324326478e-06, + "loss": 0.727, + "num_input_tokens_seen": 84019150, + "step": 3907 + }, + { + "epoch": 0.46990921661756746, + "flos": 25192198460640.0, + "grad_norm": 3.2055749967159266, + "learning_rate": 2.2910768769072603e-06, + "loss": 0.9117, + "num_input_tokens_seen": 84037930, + "step": 3908 + }, + { + "epoch": 0.47002945950820657, + "flos": 13824454515840.0, + "grad_norm": 4.617261237315706, + "learning_rate": 2.2903061772257417e-06, + "loss": 0.7635, + "num_input_tokens_seen": 84055915, + "step": 3909 + }, + { + "epoch": 0.4701497023988457, + "flos": 26209286465280.0, + "grad_norm": 2.0447498966976783, + "learning_rate": 2.289535433505007e-06, + "loss": 0.7864, + "num_input_tokens_seen": 84077270, + "step": 3910 + }, + { + "epoch": 0.47026994528948474, + "flos": 25592802864000.0, + "grad_norm": 2.4082051050529167, + "learning_rate": 2.2887646458619767e-06, + "loss": 0.6357, + "num_input_tokens_seen": 84098635, + "step": 3911 + }, + { + "epoch": 0.47039018818012385, + "flos": 20524949870880.0, + "grad_norm": 2.4646787270803, + "learning_rate": 2.2879938144135797e-06, + "loss": 0.7655, + "num_input_tokens_seen": 84114740, + "step": 3912 + }, + { + "epoch": 0.47051043107076296, + "flos": 21546349567680.0, + "grad_norm": 2.0393180833146363, + "learning_rate": 2.2872229392767496e-06, + "loss": 0.7507, + "num_input_tokens_seen": 84134845, + "step": 3913 + }, + { + "epoch": 0.470630673961402, + "flos": 18926212063680.0, + "grad_norm": 1.835157252961425, + "learning_rate": 2.286452020568428e-06, + "loss": 0.7468, + "num_input_tokens_seen": 84152920, + "step": 3914 + }, + { + "epoch": 0.4707509168520411, + "flos": 19910584945920.0, + "grad_norm": 2.1863025112399925, + "learning_rate": 2.2856810584055637e-06, + "loss": 0.7332, + "num_input_tokens_seen": 84170455, + "step": 3915 + }, + { + "epoch": 0.47087115974268023, + "flos": 40061103658080.0, + "grad_norm": 2.4157868680228036, + "learning_rate": 2.2849100529051085e-06, + "loss": 0.6799, + "num_input_tokens_seen": 84197390, + "step": 3916 + }, + { + "epoch": 0.4709914026333193, + "flos": 13533592444800.0, + "grad_norm": 3.802963586922314, + "learning_rate": 2.284139004184026e-06, + "loss": 0.8066, + "num_input_tokens_seen": 84214620, + "step": 3917 + }, + { + "epoch": 0.4711116455239584, + "flos": 19946162139840.0, + "grad_norm": 3.679472250828466, + "learning_rate": 2.2833679123592814e-06, + "loss": 0.7447, + "num_input_tokens_seen": 84231880, + "step": 3918 + }, + { + "epoch": 0.4712318884145975, + "flos": 32078682792960.0, + "grad_norm": 2.1067522485869734, + "learning_rate": 2.2825967775478508e-06, + "loss": 0.6353, + "num_input_tokens_seen": 84252980, + "step": 3919 + }, + { + "epoch": 0.47135213130523657, + "flos": 20019286524960.0, + "grad_norm": 2.3810318093448095, + "learning_rate": 2.2818255998667135e-06, + "loss": 0.8364, + "num_input_tokens_seen": 84271490, + "step": 3920 + }, + { + "epoch": 0.4714723741958757, + "flos": 19399308966240.0, + "grad_norm": 2.6786962516401567, + "learning_rate": 2.2810543794328566e-06, + "loss": 0.7939, + "num_input_tokens_seen": 84290525, + "step": 3921 + }, + { + "epoch": 0.4715926170865148, + "flos": 20344759376160.0, + "grad_norm": 4.715457803582096, + "learning_rate": 2.2802831163632735e-06, + "loss": 0.8252, + "num_input_tokens_seen": 84309245, + "step": 3922 + }, + { + "epoch": 0.47171285997715384, + "flos": 22640427612480.0, + "grad_norm": 2.823016700989809, + "learning_rate": 2.279511810774965e-06, + "loss": 0.7447, + "num_input_tokens_seen": 84330775, + "step": 3923 + }, + { + "epoch": 0.47183310286779295, + "flos": 21075297001920.0, + "grad_norm": 2.573268315510141, + "learning_rate": 2.2787404627849364e-06, + "loss": 0.7155, + "num_input_tokens_seen": 84349985, + "step": 3924 + }, + { + "epoch": 0.471953345758432, + "flos": 21695571918720.0, + "grad_norm": 1.8444720648544841, + "learning_rate": 2.277969072510202e-06, + "loss": 0.7886, + "num_input_tokens_seen": 84368965, + "step": 3925 + }, + { + "epoch": 0.4720735886490711, + "flos": 19832628492000.0, + "grad_norm": 2.482130848462498, + "learning_rate": 2.2771976400677803e-06, + "loss": 0.8148, + "num_input_tokens_seen": 84387795, + "step": 3926 + }, + { + "epoch": 0.47219383153971023, + "flos": 19144507296000.0, + "grad_norm": 3.285841862351986, + "learning_rate": 2.2764261655746965e-06, + "loss": 0.7954, + "num_input_tokens_seen": 84405290, + "step": 3927 + }, + { + "epoch": 0.4723140744303493, + "flos": 23190886252800.0, + "grad_norm": 1.8303745481613884, + "learning_rate": 2.2756546491479832e-06, + "loss": 0.7601, + "num_input_tokens_seen": 84426205, + "step": 3928 + }, + { + "epoch": 0.4724343173209884, + "flos": 18197347077120.0, + "grad_norm": 7.781339958927156, + "learning_rate": 2.274883090904679e-06, + "loss": 0.8018, + "num_input_tokens_seen": 84443970, + "step": 3929 + }, + { + "epoch": 0.4725545602116275, + "flos": 21220690867680.0, + "grad_norm": 2.642468232367198, + "learning_rate": 2.2741114909618283e-06, + "loss": 0.6798, + "num_input_tokens_seen": 84459865, + "step": 3930 + }, + { + "epoch": 0.47267480310226656, + "flos": 21403408906080.0, + "grad_norm": 2.8017294753653688, + "learning_rate": 2.2733398494364828e-06, + "loss": 0.7205, + "num_input_tokens_seen": 84479110, + "step": 3931 + }, + { + "epoch": 0.47279504599290567, + "flos": 18743791383360.0, + "grad_norm": 2.4169253068398233, + "learning_rate": 2.272568166445699e-06, + "loss": 0.8471, + "num_input_tokens_seen": 84497675, + "step": 3932 + }, + { + "epoch": 0.4729152888835448, + "flos": 21075631529760.0, + "grad_norm": 2.8591188409849284, + "learning_rate": 2.271796442106541e-06, + "loss": 0.6439, + "num_input_tokens_seen": 84517030, + "step": 3933 + }, + { + "epoch": 0.47303553177418384, + "flos": 70101112900800.0, + "grad_norm": 0.8032931125895043, + "learning_rate": 2.271024676536079e-06, + "loss": 0.5646, + "num_input_tokens_seen": 84577290, + "step": 3934 + }, + { + "epoch": 0.47315577466482295, + "flos": 22423024454400.0, + "grad_norm": 2.5143268149905587, + "learning_rate": 2.2702528698513894e-06, + "loss": 0.7321, + "num_input_tokens_seen": 84598415, + "step": 3935 + }, + { + "epoch": 0.47327601755546206, + "flos": 24318013947840.0, + "grad_norm": 2.0509737632855147, + "learning_rate": 2.269481022169554e-06, + "loss": 0.7855, + "num_input_tokens_seen": 84617010, + "step": 3936 + }, + { + "epoch": 0.4733962604461011, + "flos": 22893891171360.0, + "grad_norm": 3.683916193096856, + "learning_rate": 2.2687091336076614e-06, + "loss": 0.8003, + "num_input_tokens_seen": 84636350, + "step": 3937 + }, + { + "epoch": 0.4735165033367402, + "flos": 18301105078080.0, + "grad_norm": 3.071697670056887, + "learning_rate": 2.267937204282807e-06, + "loss": 0.8049, + "num_input_tokens_seen": 84653490, + "step": 3938 + }, + { + "epoch": 0.4736367462273793, + "flos": 23004005201280.0, + "grad_norm": 2.6812870156460717, + "learning_rate": 2.2671652343120926e-06, + "loss": 0.7892, + "num_input_tokens_seen": 84673965, + "step": 3939 + }, + { + "epoch": 0.4737569891180184, + "flos": 25338001193760.0, + "grad_norm": 1.835118926973196, + "learning_rate": 2.2663932238126236e-06, + "loss": 0.8066, + "num_input_tokens_seen": 84692525, + "step": 3940 + }, + { + "epoch": 0.4738772320086575, + "flos": 25812324698400.0, + "grad_norm": 1.9010630218839244, + "learning_rate": 2.265621172901515e-06, + "loss": 0.8014, + "num_input_tokens_seen": 84715640, + "step": 3941 + }, + { + "epoch": 0.47399747489929656, + "flos": 27524930681280.0, + "grad_norm": 3.0114334307740225, + "learning_rate": 2.2648490816958854e-06, + "loss": 0.7134, + "num_input_tokens_seen": 84736635, + "step": 3942 + }, + { + "epoch": 0.47411771778993567, + "flos": 24828063325440.0, + "grad_norm": 2.6229362549680335, + "learning_rate": 2.264076950312861e-06, + "loss": 0.7314, + "num_input_tokens_seen": 84755445, + "step": 3943 + }, + { + "epoch": 0.4742379606805748, + "flos": 22715521994880.0, + "grad_norm": 2.2973696636634244, + "learning_rate": 2.2633047788695727e-06, + "loss": 0.8252, + "num_input_tokens_seen": 84775750, + "step": 3944 + }, + { + "epoch": 0.47435820357121383, + "flos": 19653032713440.0, + "grad_norm": 2.3109000573726237, + "learning_rate": 2.262532567483159e-06, + "loss": 0.644, + "num_input_tokens_seen": 84794310, + "step": 3945 + }, + { + "epoch": 0.47447844646185294, + "flos": 25192607328000.0, + "grad_norm": 2.810275236341042, + "learning_rate": 2.2617603162707635e-06, + "loss": 0.8016, + "num_input_tokens_seen": 84814720, + "step": 3946 + }, + { + "epoch": 0.47459868935249205, + "flos": 24535417105920.0, + "grad_norm": 3.4228805388395465, + "learning_rate": 2.2609880253495363e-06, + "loss": 0.8256, + "num_input_tokens_seen": 84834355, + "step": 3947 + }, + { + "epoch": 0.4747189322431311, + "flos": 20529187223520.0, + "grad_norm": 1.9455840752326352, + "learning_rate": 2.260215694836633e-06, + "loss": 0.8683, + "num_input_tokens_seen": 84853530, + "step": 3948 + }, + { + "epoch": 0.4748391751337702, + "flos": 25958015922240.0, + "grad_norm": 2.601148967696475, + "learning_rate": 2.2594433248492157e-06, + "loss": 0.6471, + "num_input_tokens_seen": 84872505, + "step": 3949 + }, + { + "epoch": 0.47495941802440933, + "flos": 22788460531200.0, + "grad_norm": 2.164849093598207, + "learning_rate": 2.2586709155044527e-06, + "loss": 0.8041, + "num_input_tokens_seen": 84893105, + "step": 3950 + }, + { + "epoch": 0.4750796609150484, + "flos": 27851778813600.0, + "grad_norm": 1.7866051468055388, + "learning_rate": 2.2578984669195167e-06, + "loss": 0.7627, + "num_input_tokens_seen": 84914070, + "step": 3951 + }, + { + "epoch": 0.4751999038056875, + "flos": 35608656343200.0, + "grad_norm": 2.689354367683446, + "learning_rate": 2.2571259792115887e-06, + "loss": 0.6817, + "num_input_tokens_seen": 84935290, + "step": 3952 + }, + { + "epoch": 0.4753201466963266, + "flos": 22059335356320.0, + "grad_norm": 2.089406076394724, + "learning_rate": 2.2563534524978544e-06, + "loss": 0.7919, + "num_input_tokens_seen": 84952760, + "step": 3953 + }, + { + "epoch": 0.47544038958696566, + "flos": 30147521389440.0, + "grad_norm": 1.935866066859417, + "learning_rate": 2.2555808868955052e-06, + "loss": 0.7096, + "num_input_tokens_seen": 84974890, + "step": 3954 + }, + { + "epoch": 0.47556063247760477, + "flos": 23438476989600.0, + "grad_norm": 2.822124036781426, + "learning_rate": 2.254808282521738e-06, + "loss": 0.7445, + "num_input_tokens_seen": 84993640, + "step": 3955 + }, + { + "epoch": 0.4756808753682438, + "flos": 25119185584800.0, + "grad_norm": 3.6928124388460475, + "learning_rate": 2.2540356394937573e-06, + "loss": 0.8088, + "num_input_tokens_seen": 85012340, + "step": 3956 + }, + { + "epoch": 0.47580111825888294, + "flos": 15647546226240.0, + "grad_norm": 2.6882991933182234, + "learning_rate": 2.253262957928772e-06, + "loss": 0.8368, + "num_input_tokens_seen": 85029225, + "step": 3957 + }, + { + "epoch": 0.47592136114952205, + "flos": 17611385582400.0, + "grad_norm": 3.174228640594208, + "learning_rate": 2.2524902379439976e-06, + "loss": 0.7207, + "num_input_tokens_seen": 85047895, + "step": 3958 + }, + { + "epoch": 0.4760416040401611, + "flos": 61329489061440.0, + "grad_norm": 0.7546612215699378, + "learning_rate": 2.251717479656655e-06, + "loss": 0.6366, + "num_input_tokens_seen": 85112690, + "step": 3959 + }, + { + "epoch": 0.4761618469308002, + "flos": 18379693417920.0, + "grad_norm": 2.6935133084353797, + "learning_rate": 2.2509446831839704e-06, + "loss": 0.7634, + "num_input_tokens_seen": 85132130, + "step": 3960 + }, + { + "epoch": 0.4762820898214393, + "flos": 18015260924640.0, + "grad_norm": 3.449685157010001, + "learning_rate": 2.250171848643177e-06, + "loss": 0.818, + "num_input_tokens_seen": 85149420, + "step": 3961 + }, + { + "epoch": 0.4764023327120784, + "flos": 19290756066240.0, + "grad_norm": 4.71460223270707, + "learning_rate": 2.249398976151513e-06, + "loss": 0.8633, + "num_input_tokens_seen": 85166970, + "step": 3962 + }, + { + "epoch": 0.4765225756027175, + "flos": 22714741429920.0, + "grad_norm": 3.392487607636725, + "learning_rate": 2.248626065826223e-06, + "loss": 0.7852, + "num_input_tokens_seen": 85185175, + "step": 3963 + }, + { + "epoch": 0.4766428184933566, + "flos": 65838854746080.0, + "grad_norm": 0.7878924844972364, + "learning_rate": 2.2478531177845564e-06, + "loss": 0.6248, + "num_input_tokens_seen": 85246170, + "step": 3964 + }, + { + "epoch": 0.47676306138399566, + "flos": 24102060410400.0, + "grad_norm": 2.67572324369722, + "learning_rate": 2.247080132143769e-06, + "loss": 0.8507, + "num_input_tokens_seen": 85268525, + "step": 3965 + }, + { + "epoch": 0.47688330427463477, + "flos": 12586580904960.0, + "grad_norm": 2.7188454972957334, + "learning_rate": 2.246307109021121e-06, + "loss": 0.6935, + "num_input_tokens_seen": 85285930, + "step": 3966 + }, + { + "epoch": 0.4770035471652739, + "flos": 21359803044000.0, + "grad_norm": 1.8778724284773707, + "learning_rate": 2.2455340485338817e-06, + "loss": 0.8238, + "num_input_tokens_seen": 85303565, + "step": 3967 + }, + { + "epoch": 0.47712379005591293, + "flos": 25120263507840.0, + "grad_norm": 2.6867758196609457, + "learning_rate": 2.244760950799322e-06, + "loss": 0.68, + "num_input_tokens_seen": 85321835, + "step": 3968 + }, + { + "epoch": 0.47724403294655204, + "flos": 22022828918400.0, + "grad_norm": 2.07315639726234, + "learning_rate": 2.2439878159347203e-06, + "loss": 0.7206, + "num_input_tokens_seen": 85341260, + "step": 3969 + }, + { + "epoch": 0.4773642758371911, + "flos": 70129584936960.0, + "grad_norm": 0.8997236913813178, + "learning_rate": 2.2432146440573616e-06, + "loss": 0.5526, + "num_input_tokens_seen": 85407220, + "step": 3970 + }, + { + "epoch": 0.4774845187278302, + "flos": 23514872313600.0, + "grad_norm": 2.2636613425930556, + "learning_rate": 2.242441435284534e-06, + "loss": 0.6686, + "num_input_tokens_seen": 85426095, + "step": 3971 + }, + { + "epoch": 0.4776047616184693, + "flos": 23042370127200.0, + "grad_norm": 2.5303211156136127, + "learning_rate": 2.2416681897335337e-06, + "loss": 0.8569, + "num_input_tokens_seen": 85444245, + "step": 3972 + }, + { + "epoch": 0.4777250045091084, + "flos": 31852024364640.0, + "grad_norm": 3.0467366505048146, + "learning_rate": 2.240894907521661e-06, + "loss": 0.6706, + "num_input_tokens_seen": 85463325, + "step": 3973 + }, + { + "epoch": 0.4778452473997475, + "flos": 24243365602560.0, + "grad_norm": 3.6346659208517598, + "learning_rate": 2.240121588766223e-06, + "loss": 0.6384, + "num_input_tokens_seen": 85483375, + "step": 3974 + }, + { + "epoch": 0.4779654902903866, + "flos": 31529599433760.0, + "grad_norm": 2.666833594183818, + "learning_rate": 2.239348233584531e-06, + "loss": 0.7201, + "num_input_tokens_seen": 85504230, + "step": 3975 + }, + { + "epoch": 0.47808573318102565, + "flos": 19472656369920.0, + "grad_norm": 2.0249759908124094, + "learning_rate": 2.2385748420939013e-06, + "loss": 0.8115, + "num_input_tokens_seen": 85523425, + "step": 3976 + }, + { + "epoch": 0.47820597607166476, + "flos": 22568827187520.0, + "grad_norm": 2.470233616302334, + "learning_rate": 2.2378014144116583e-06, + "loss": 0.73, + "num_input_tokens_seen": 85542235, + "step": 3977 + }, + { + "epoch": 0.4783262189623039, + "flos": 22970137816320.0, + "grad_norm": 2.9538920380241125, + "learning_rate": 2.23702795065513e-06, + "loss": 0.7968, + "num_input_tokens_seen": 85561815, + "step": 3978 + }, + { + "epoch": 0.47844646185294293, + "flos": 49701888359040.0, + "grad_norm": 0.9916992488017752, + "learning_rate": 2.2362544509416493e-06, + "loss": 0.6736, + "num_input_tokens_seen": 85613930, + "step": 3979 + }, + { + "epoch": 0.47856670474358204, + "flos": 20200703621760.0, + "grad_norm": 2.5868622661976697, + "learning_rate": 2.2354809153885572e-06, + "loss": 0.8255, + "num_input_tokens_seen": 85631000, + "step": 3980 + }, + { + "epoch": 0.47868694763422115, + "flos": 20960796940320.0, + "grad_norm": 2.134798236398896, + "learning_rate": 2.234707344113197e-06, + "loss": 0.8259, + "num_input_tokens_seen": 85649095, + "step": 3981 + }, + { + "epoch": 0.4788071905248602, + "flos": 18999485127840.0, + "grad_norm": 1.879476734245711, + "learning_rate": 2.233933737232919e-06, + "loss": 0.7789, + "num_input_tokens_seen": 85666875, + "step": 3982 + }, + { + "epoch": 0.4789274334154993, + "flos": 22969320081600.0, + "grad_norm": 1.774865201747727, + "learning_rate": 2.2331600948650793e-06, + "loss": 0.7824, + "num_input_tokens_seen": 85687020, + "step": 3983 + }, + { + "epoch": 0.4790476763061384, + "flos": 23042593145760.0, + "grad_norm": 3.5422507907646588, + "learning_rate": 2.2323864171270386e-06, + "loss": 0.8021, + "num_input_tokens_seen": 85708290, + "step": 3984 + }, + { + "epoch": 0.4791679191967775, + "flos": 21149090442720.0, + "grad_norm": 2.4481978775719506, + "learning_rate": 2.231612704136164e-06, + "loss": 0.7271, + "num_input_tokens_seen": 85728660, + "step": 3985 + }, + { + "epoch": 0.4792881620874166, + "flos": 22269341732160.0, + "grad_norm": 8.564558909338148, + "learning_rate": 2.2308389560098253e-06, + "loss": 0.7497, + "num_input_tokens_seen": 85745035, + "step": 3986 + }, + { + "epoch": 0.47940840497805565, + "flos": 17396138270400.0, + "grad_norm": 2.3960115670512443, + "learning_rate": 2.2300651728654008e-06, + "loss": 0.7762, + "num_input_tokens_seen": 85760295, + "step": 3987 + }, + { + "epoch": 0.47952864786869476, + "flos": 65264787574560.0, + "grad_norm": 0.740070979219657, + "learning_rate": 2.229291354820272e-06, + "loss": 0.6009, + "num_input_tokens_seen": 85821305, + "step": 3988 + }, + { + "epoch": 0.47964889075933387, + "flos": 16775937693120.0, + "grad_norm": 2.3096265725507594, + "learning_rate": 2.228517501991828e-06, + "loss": 0.7592, + "num_input_tokens_seen": 85840105, + "step": 3989 + }, + { + "epoch": 0.4797691336499729, + "flos": 69979507681440.0, + "grad_norm": 0.822601938220912, + "learning_rate": 2.22774361449746e-06, + "loss": 0.6096, + "num_input_tokens_seen": 85896585, + "step": 3990 + }, + { + "epoch": 0.47988937654061203, + "flos": 18926360742720.0, + "grad_norm": 2.6559318800773726, + "learning_rate": 2.2269696924545668e-06, + "loss": 0.7113, + "num_input_tokens_seen": 85914415, + "step": 3991 + }, + { + "epoch": 0.48000961943125114, + "flos": 14441421324000.0, + "grad_norm": 3.16427362938535, + "learning_rate": 2.2261957359805523e-06, + "loss": 0.7807, + "num_input_tokens_seen": 85931650, + "step": 3992 + }, + { + "epoch": 0.4801298623218902, + "flos": 27012316590240.0, + "grad_norm": 3.1145801012581646, + "learning_rate": 2.225421745192823e-06, + "loss": 0.74, + "num_input_tokens_seen": 85951805, + "step": 3993 + }, + { + "epoch": 0.4802501052125293, + "flos": 26317318988640.0, + "grad_norm": 2.2035621588110783, + "learning_rate": 2.2246477202087955e-06, + "loss": 0.7833, + "num_input_tokens_seen": 85972485, + "step": 3994 + }, + { + "epoch": 0.4803703481031684, + "flos": 20963882030400.0, + "grad_norm": 1.7214205022621638, + "learning_rate": 2.223873661145887e-06, + "loss": 0.8304, + "num_input_tokens_seen": 85992540, + "step": 3995 + }, + { + "epoch": 0.4804905909938075, + "flos": 20674172221920.0, + "grad_norm": 1.793568230434209, + "learning_rate": 2.2230995681215226e-06, + "loss": 0.717, + "num_input_tokens_seen": 86012065, + "step": 3996 + }, + { + "epoch": 0.4806108338844466, + "flos": 16631212883040.0, + "grad_norm": 2.86651287405602, + "learning_rate": 2.2223254412531305e-06, + "loss": 0.7808, + "num_input_tokens_seen": 86029435, + "step": 3997 + }, + { + "epoch": 0.4807310767750857, + "flos": 19982817256800.0, + "grad_norm": 3.735151266075947, + "learning_rate": 2.221551280658146e-06, + "loss": 0.8262, + "num_input_tokens_seen": 86048495, + "step": 3998 + }, + { + "epoch": 0.48085131966572475, + "flos": 23152112459520.0, + "grad_norm": 2.0097181724232778, + "learning_rate": 2.2207770864540085e-06, + "loss": 0.7401, + "num_input_tokens_seen": 86067470, + "step": 3999 + }, + { + "epoch": 0.48097156255636386, + "flos": 20529335902560.0, + "grad_norm": 2.8587398970224847, + "learning_rate": 2.220002858758162e-06, + "loss": 0.7272, + "num_input_tokens_seen": 86085460, + "step": 4000 + }, + { + "epoch": 0.481091805447003, + "flos": 70411043058720.0, + "grad_norm": 0.8986367709547949, + "learning_rate": 2.2192285976880573e-06, + "loss": 0.6085, + "num_input_tokens_seen": 86149715, + "step": 4001 + }, + { + "epoch": 0.48121204833764203, + "flos": 36375737576640.0, + "grad_norm": 2.191613933765268, + "learning_rate": 2.2184543033611485e-06, + "loss": 0.811, + "num_input_tokens_seen": 86170795, + "step": 4002 + }, + { + "epoch": 0.48133229122828114, + "flos": 27450839882400.0, + "grad_norm": 2.832325550441128, + "learning_rate": 2.2176799758948957e-06, + "loss": 0.8221, + "num_input_tokens_seen": 86190955, + "step": 4003 + }, + { + "epoch": 0.4814525341189202, + "flos": 43010802686880.0, + "grad_norm": 2.25864804129069, + "learning_rate": 2.2169056154067635e-06, + "loss": 0.7336, + "num_input_tokens_seen": 86214875, + "step": 4004 + }, + { + "epoch": 0.4815727770095593, + "flos": 24201989926080.0, + "grad_norm": 2.026172336264286, + "learning_rate": 2.216131222014222e-06, + "loss": 0.8273, + "num_input_tokens_seen": 86232950, + "step": 4005 + }, + { + "epoch": 0.4816930199001984, + "flos": 18087864933120.0, + "grad_norm": 2.4857074355358324, + "learning_rate": 2.2153567958347455e-06, + "loss": 0.8053, + "num_input_tokens_seen": 86249160, + "step": 4006 + }, + { + "epoch": 0.48181326279083747, + "flos": 17250186858240.0, + "grad_norm": 2.3342184930958676, + "learning_rate": 2.214582336985815e-06, + "loss": 0.8016, + "num_input_tokens_seen": 86267060, + "step": 4007 + }, + { + "epoch": 0.4819335056814766, + "flos": 14882174801760.0, + "grad_norm": 3.9531616691728138, + "learning_rate": 2.2138078455849142e-06, + "loss": 0.6667, + "num_input_tokens_seen": 86285055, + "step": 4008 + }, + { + "epoch": 0.4820537485721157, + "flos": 19217483002080.0, + "grad_norm": 2.0290622373598683, + "learning_rate": 2.2130333217495334e-06, + "loss": 0.7888, + "num_input_tokens_seen": 86304225, + "step": 4009 + }, + { + "epoch": 0.48217399146275475, + "flos": 16011049475520.0, + "grad_norm": 3.4357264711975684, + "learning_rate": 2.2122587655971665e-06, + "loss": 0.6822, + "num_input_tokens_seen": 86319170, + "step": 4010 + }, + { + "epoch": 0.48229423435339386, + "flos": 24099681545760.0, + "grad_norm": 10.895916347053706, + "learning_rate": 2.211484177245314e-06, + "loss": 0.6426, + "num_input_tokens_seen": 86338760, + "step": 4011 + }, + { + "epoch": 0.48241447724403297, + "flos": 23771383792800.0, + "grad_norm": 2.6491683806679522, + "learning_rate": 2.21070955681148e-06, + "loss": 0.7222, + "num_input_tokens_seen": 86357865, + "step": 4012 + }, + { + "epoch": 0.482534720134672, + "flos": 23077203925920.0, + "grad_norm": 2.3997751289559743, + "learning_rate": 2.209934904413174e-06, + "loss": 0.782, + "num_input_tokens_seen": 86379470, + "step": 4013 + }, + { + "epoch": 0.48265496302531113, + "flos": 20893991414400.0, + "grad_norm": 2.254792090598637, + "learning_rate": 2.2091602201679095e-06, + "loss": 0.719, + "num_input_tokens_seen": 86399080, + "step": 4014 + }, + { + "epoch": 0.48277520591595025, + "flos": 15209580480480.0, + "grad_norm": 2.90428948674035, + "learning_rate": 2.208385504193206e-06, + "loss": 0.8407, + "num_input_tokens_seen": 86416580, + "step": 4015 + }, + { + "epoch": 0.4828954488065893, + "flos": 17833286281440.0, + "grad_norm": 4.7606867985559935, + "learning_rate": 2.2076107566065873e-06, + "loss": 0.8118, + "num_input_tokens_seen": 86434365, + "step": 4016 + }, + { + "epoch": 0.4830156916972284, + "flos": 32041284280800.0, + "grad_norm": 3.457424887105371, + "learning_rate": 2.2068359775255816e-06, + "loss": 0.7563, + "num_input_tokens_seen": 86452675, + "step": 4017 + }, + { + "epoch": 0.48313593458786747, + "flos": 21840482577600.0, + "grad_norm": 6.1238561068123865, + "learning_rate": 2.206061167067723e-06, + "loss": 0.7861, + "num_input_tokens_seen": 86470785, + "step": 4018 + }, + { + "epoch": 0.4832561774785066, + "flos": 22568901527040.0, + "grad_norm": 3.6839111450281874, + "learning_rate": 2.205286325350549e-06, + "loss": 0.7987, + "num_input_tokens_seen": 86489850, + "step": 4019 + }, + { + "epoch": 0.4833764203691457, + "flos": 13418163139200.0, + "grad_norm": 2.809341776718224, + "learning_rate": 2.204511452491603e-06, + "loss": 0.7272, + "num_input_tokens_seen": 86506475, + "step": 4020 + }, + { + "epoch": 0.48349666325978474, + "flos": 44064397129440.0, + "grad_norm": 1.7455264541776176, + "learning_rate": 2.2037365486084316e-06, + "loss": 0.7496, + "num_input_tokens_seen": 86529715, + "step": 4021 + }, + { + "epoch": 0.48361690615042385, + "flos": 25990991232960.0, + "grad_norm": 3.4671833521374187, + "learning_rate": 2.2029616138185886e-06, + "loss": 0.7781, + "num_input_tokens_seen": 86548715, + "step": 4022 + }, + { + "epoch": 0.48373714904106296, + "flos": 22241495848320.0, + "grad_norm": 1.9202120029110477, + "learning_rate": 2.202186648239629e-06, + "loss": 0.8336, + "num_input_tokens_seen": 86568650, + "step": 4023 + }, + { + "epoch": 0.483857391931702, + "flos": 28251676991520.0, + "grad_norm": 1.8513760372223607, + "learning_rate": 2.201411651989117e-06, + "loss": 0.7137, + "num_input_tokens_seen": 86590630, + "step": 4024 + }, + { + "epoch": 0.48397763482234113, + "flos": 27378756250560.0, + "grad_norm": 3.1792636748212764, + "learning_rate": 2.2006366251846167e-06, + "loss": 0.7825, + "num_input_tokens_seen": 86611270, + "step": 4025 + }, + { + "epoch": 0.48409787771298024, + "flos": 16773298640160.0, + "grad_norm": 4.2788834539714795, + "learning_rate": 2.1998615679436997e-06, + "loss": 0.757, + "num_input_tokens_seen": 86628810, + "step": 4026 + }, + { + "epoch": 0.4842181206036193, + "flos": 25047622329600.0, + "grad_norm": 3.2212705650161633, + "learning_rate": 2.199086480383942e-06, + "loss": 0.778, + "num_input_tokens_seen": 86648185, + "step": 4027 + }, + { + "epoch": 0.4843383634942584, + "flos": 30328678297920.0, + "grad_norm": 3.3105653115268345, + "learning_rate": 2.1983113626229234e-06, + "loss": 0.6772, + "num_input_tokens_seen": 86667435, + "step": 4028 + }, + { + "epoch": 0.4844586063848975, + "flos": 20384499583200.0, + "grad_norm": 2.344675822187605, + "learning_rate": 2.1975362147782293e-06, + "loss": 0.7878, + "num_input_tokens_seen": 86686630, + "step": 4029 + }, + { + "epoch": 0.48457884927553657, + "flos": 70203861584640.0, + "grad_norm": 0.6939640256300301, + "learning_rate": 2.196761036967448e-06, + "loss": 0.5405, + "num_input_tokens_seen": 86754595, + "step": 4030 + }, + { + "epoch": 0.4846990921661757, + "flos": 19905975895680.0, + "grad_norm": 3.2758608203615553, + "learning_rate": 2.1959858293081743e-06, + "loss": 0.7759, + "num_input_tokens_seen": 86773730, + "step": 4031 + }, + { + "epoch": 0.4848193350568148, + "flos": 23042964843360.0, + "grad_norm": 1.8965300722440832, + "learning_rate": 2.1952105919180056e-06, + "loss": 0.762, + "num_input_tokens_seen": 86792985, + "step": 4032 + }, + { + "epoch": 0.48493957794745385, + "flos": 22423396152000.0, + "grad_norm": 6.9263650698656285, + "learning_rate": 2.1944353249145456e-06, + "loss": 0.6798, + "num_input_tokens_seen": 86812095, + "step": 4033 + }, + { + "epoch": 0.48505982083809296, + "flos": 25010260987200.0, + "grad_norm": 1.8569592081169557, + "learning_rate": 2.193660028415401e-06, + "loss": 0.746, + "num_input_tokens_seen": 86832875, + "step": 4034 + }, + { + "epoch": 0.485180063728732, + "flos": 26723052818880.0, + "grad_norm": 2.2773434359254177, + "learning_rate": 2.1928847025381852e-06, + "loss": 0.8205, + "num_input_tokens_seen": 86853480, + "step": 4035 + }, + { + "epoch": 0.4853003066193711, + "flos": 24024661502880.0, + "grad_norm": 1.9403597738018405, + "learning_rate": 2.192109347400512e-06, + "loss": 0.8382, + "num_input_tokens_seen": 86873695, + "step": 4036 + }, + { + "epoch": 0.48542054951001024, + "flos": 23043410880480.0, + "grad_norm": 3.492533212229397, + "learning_rate": 2.191333963120004e-06, + "loss": 0.7892, + "num_input_tokens_seen": 86892675, + "step": 4037 + }, + { + "epoch": 0.4855407924006493, + "flos": 25628082699840.0, + "grad_norm": 2.554434107541576, + "learning_rate": 2.190558549814286e-06, + "loss": 0.6982, + "num_input_tokens_seen": 86912835, + "step": 4038 + }, + { + "epoch": 0.4856610352912884, + "flos": 23953618624320.0, + "grad_norm": 2.4478244737078785, + "learning_rate": 2.1897831076009872e-06, + "loss": 0.799, + "num_input_tokens_seen": 86932475, + "step": 4039 + }, + { + "epoch": 0.4857812781819275, + "flos": 24062320203360.0, + "grad_norm": 2.265380635730705, + "learning_rate": 2.1890076365977426e-06, + "loss": 0.7974, + "num_input_tokens_seen": 86952300, + "step": 4040 + }, + { + "epoch": 0.48590152107256657, + "flos": 56186244327840.0, + "grad_norm": 0.8578742013777759, + "learning_rate": 2.188232136922189e-06, + "loss": 0.5277, + "num_input_tokens_seen": 87010420, + "step": 4041 + }, + { + "epoch": 0.4860217639632057, + "flos": 20018394450720.0, + "grad_norm": 2.61678800807737, + "learning_rate": 2.187456608691971e-06, + "loss": 0.7591, + "num_input_tokens_seen": 87029295, + "step": 4042 + }, + { + "epoch": 0.4861420068538448, + "flos": 17796817013280.0, + "grad_norm": 4.482565272079931, + "learning_rate": 2.1866810520247334e-06, + "loss": 0.8778, + "num_input_tokens_seen": 87048165, + "step": 4043 + }, + { + "epoch": 0.48626224974448384, + "flos": 26212520234400.0, + "grad_norm": 2.142962439538535, + "learning_rate": 2.185905467038129e-06, + "loss": 0.652, + "num_input_tokens_seen": 87067070, + "step": 4044 + }, + { + "epoch": 0.48638249263512295, + "flos": 22023014767200.0, + "grad_norm": 1.939494613710788, + "learning_rate": 2.1851298538498127e-06, + "loss": 0.7743, + "num_input_tokens_seen": 87086785, + "step": 4045 + }, + { + "epoch": 0.48650273552576206, + "flos": 25084017258240.0, + "grad_norm": 2.853909464661223, + "learning_rate": 2.184354212577446e-06, + "loss": 0.802, + "num_input_tokens_seen": 87107090, + "step": 4046 + }, + { + "epoch": 0.4866229784164011, + "flos": 17432124331680.0, + "grad_norm": 4.6029570818882375, + "learning_rate": 2.1835785433386907e-06, + "loss": 0.6257, + "num_input_tokens_seen": 87125905, + "step": 4047 + }, + { + "epoch": 0.48674322130704023, + "flos": 23297766513600.0, + "grad_norm": 2.1946434005919317, + "learning_rate": 2.182802846251216e-06, + "loss": 0.65, + "num_input_tokens_seen": 87146175, + "step": 4048 + }, + { + "epoch": 0.4868634641976793, + "flos": 28763027310720.0, + "grad_norm": 3.3961982452975716, + "learning_rate": 2.182027121432696e-06, + "loss": 0.7241, + "num_input_tokens_seen": 87166800, + "step": 4049 + }, + { + "epoch": 0.4869837070883184, + "flos": 18998890411680.0, + "grad_norm": 2.131450287924872, + "learning_rate": 2.1812513690008054e-06, + "loss": 0.8246, + "num_input_tokens_seen": 87185920, + "step": 4050 + }, + { + "epoch": 0.4871039499789575, + "flos": 15099986827200.0, + "grad_norm": 2.8690638418291097, + "learning_rate": 2.180475589073227e-06, + "loss": 0.7982, + "num_input_tokens_seen": 87203620, + "step": 4051 + }, + { + "epoch": 0.48722419286959656, + "flos": 26136459438240.0, + "grad_norm": 1.8955097621021337, + "learning_rate": 2.1796997817676456e-06, + "loss": 0.7336, + "num_input_tokens_seen": 87224630, + "step": 4052 + }, + { + "epoch": 0.4873444357602357, + "flos": 23991834871200.0, + "grad_norm": 1.92727064992194, + "learning_rate": 2.1789239472017494e-06, + "loss": 0.6737, + "num_input_tokens_seen": 87246280, + "step": 4053 + }, + { + "epoch": 0.4874646786508748, + "flos": 22787828645280.0, + "grad_norm": 2.577219208630635, + "learning_rate": 2.1781480854932326e-06, + "loss": 0.7278, + "num_input_tokens_seen": 87266045, + "step": 4054 + }, + { + "epoch": 0.48758492154151384, + "flos": 21257197305600.0, + "grad_norm": 2.4479263521909176, + "learning_rate": 2.1773721967597933e-06, + "loss": 0.7914, + "num_input_tokens_seen": 87284495, + "step": 4055 + }, + { + "epoch": 0.48770516443215295, + "flos": 62154789606240.0, + "grad_norm": 0.8521846633265208, + "learning_rate": 2.1765962811191322e-06, + "loss": 0.5739, + "num_input_tokens_seen": 87338960, + "step": 4056 + }, + { + "epoch": 0.48782540732279206, + "flos": 66039265590240.0, + "grad_norm": 0.8268235254160324, + "learning_rate": 2.1758203386889566e-06, + "loss": 0.6191, + "num_input_tokens_seen": 87401730, + "step": 4057 + }, + { + "epoch": 0.4879456502134311, + "flos": 14587149717600.0, + "grad_norm": 9.653377167336998, + "learning_rate": 2.1750443695869746e-06, + "loss": 0.8453, + "num_input_tokens_seen": 87417300, + "step": 4058 + }, + { + "epoch": 0.4880658931040702, + "flos": 19472693539680.0, + "grad_norm": 5.859618484524371, + "learning_rate": 2.174268373930901e-06, + "loss": 0.8574, + "num_input_tokens_seen": 87434815, + "step": 4059 + }, + { + "epoch": 0.48818613599470934, + "flos": 16699170671520.0, + "grad_norm": 2.866056213881781, + "learning_rate": 2.1734923518384537e-06, + "loss": 0.8013, + "num_input_tokens_seen": 87451420, + "step": 4060 + }, + { + "epoch": 0.4883063788853484, + "flos": 26718443768640.0, + "grad_norm": 2.352312443682309, + "learning_rate": 2.1727163034273547e-06, + "loss": 0.8221, + "num_input_tokens_seen": 87469585, + "step": 4061 + }, + { + "epoch": 0.4884266217759875, + "flos": 16739579934240.0, + "grad_norm": 2.575074211458583, + "learning_rate": 2.17194022881533e-06, + "loss": 0.7891, + "num_input_tokens_seen": 87485675, + "step": 4062 + }, + { + "epoch": 0.4885468646666266, + "flos": 24172508572800.0, + "grad_norm": 5.874596439867068, + "learning_rate": 2.1711641281201092e-06, + "loss": 0.6771, + "num_input_tokens_seen": 87505605, + "step": 4063 + }, + { + "epoch": 0.48866710755726567, + "flos": 14590123298400.0, + "grad_norm": 3.4378079363138396, + "learning_rate": 2.1703880014594264e-06, + "loss": 0.7921, + "num_input_tokens_seen": 87523195, + "step": 4064 + }, + { + "epoch": 0.4887873504479048, + "flos": 28763138820000.0, + "grad_norm": 1.9175543249956941, + "learning_rate": 2.1696118489510182e-06, + "loss": 0.7423, + "num_input_tokens_seen": 87544125, + "step": 4065 + }, + { + "epoch": 0.48890759333854383, + "flos": 22751842584000.0, + "grad_norm": 2.708597982143092, + "learning_rate": 2.1688356707126286e-06, + "loss": 0.7234, + "num_input_tokens_seen": 87563745, + "step": 4066 + }, + { + "epoch": 0.48902783622918294, + "flos": 17760830952000.0, + "grad_norm": 3.2655337380246503, + "learning_rate": 2.168059466862001e-06, + "loss": 0.7027, + "num_input_tokens_seen": 87581895, + "step": 4067 + }, + { + "epoch": 0.48914807911982205, + "flos": 22278448323360.0, + "grad_norm": 2.6083792425468384, + "learning_rate": 2.167283237516887e-06, + "loss": 0.8186, + "num_input_tokens_seen": 87600170, + "step": 4068 + }, + { + "epoch": 0.4892683220104611, + "flos": 16339830435360.0, + "grad_norm": 2.6200614067336154, + "learning_rate": 2.1665069827950383e-06, + "loss": 0.7501, + "num_input_tokens_seen": 87617455, + "step": 4069 + }, + { + "epoch": 0.4893885649011002, + "flos": 15716730616800.0, + "grad_norm": 2.2828411667085895, + "learning_rate": 2.1657307028142126e-06, + "loss": 0.8658, + "num_input_tokens_seen": 87634995, + "step": 4070 + }, + { + "epoch": 0.48950880779173933, + "flos": 28540866423360.0, + "grad_norm": 2.9775669275508205, + "learning_rate": 2.164954397692171e-06, + "loss": 0.6761, + "num_input_tokens_seen": 87654420, + "step": 4071 + }, + { + "epoch": 0.4896290506823784, + "flos": 66091314721440.0, + "grad_norm": 1.0937466212629174, + "learning_rate": 2.164178067546678e-06, + "loss": 0.7729, + "num_input_tokens_seen": 87713585, + "step": 4072 + }, + { + "epoch": 0.4897492935730175, + "flos": 12513419350080.0, + "grad_norm": 66.34826144576058, + "learning_rate": 2.163401712495504e-06, + "loss": 0.909, + "num_input_tokens_seen": 87731280, + "step": 4073 + }, + { + "epoch": 0.4898695364636566, + "flos": 23443234718880.0, + "grad_norm": 2.3282435854210357, + "learning_rate": 2.1626253326564194e-06, + "loss": 0.7946, + "num_input_tokens_seen": 87750230, + "step": 4074 + }, + { + "epoch": 0.48998977935429566, + "flos": 27121204018080.0, + "grad_norm": 1.9265739248761504, + "learning_rate": 2.161848928147201e-06, + "loss": 0.7677, + "num_input_tokens_seen": 87770500, + "step": 4075 + }, + { + "epoch": 0.4901100222449348, + "flos": 20310594633120.0, + "grad_norm": 2.5652830316498614, + "learning_rate": 2.161072499085629e-06, + "loss": 0.8097, + "num_input_tokens_seen": 87789495, + "step": 4076 + }, + { + "epoch": 0.4902302651355739, + "flos": 30402657587520.0, + "grad_norm": 1.8813000694531596, + "learning_rate": 2.160296045589487e-06, + "loss": 0.8314, + "num_input_tokens_seen": 87812955, + "step": 4077 + }, + { + "epoch": 0.49035050802621294, + "flos": 19146960500160.0, + "grad_norm": 3.98901384371031, + "learning_rate": 2.159519567776562e-06, + "loss": 0.6973, + "num_input_tokens_seen": 87832605, + "step": 4078 + }, + { + "epoch": 0.49047075091685205, + "flos": 22196440365600.0, + "grad_norm": 3.6274714759425546, + "learning_rate": 2.1587430657646463e-06, + "loss": 0.7035, + "num_input_tokens_seen": 87846040, + "step": 4079 + }, + { + "epoch": 0.4905909938074911, + "flos": 20127765085440.0, + "grad_norm": 1.9683697045583795, + "learning_rate": 2.157966539671533e-06, + "loss": 0.781, + "num_input_tokens_seen": 87865680, + "step": 4080 + }, + { + "epoch": 0.4907112366981302, + "flos": 17177805868320.0, + "grad_norm": 2.4301466921888975, + "learning_rate": 2.157189989615021e-06, + "loss": 0.674, + "num_input_tokens_seen": 87884270, + "step": 4081 + }, + { + "epoch": 0.4908314795887693, + "flos": 21657987557760.0, + "grad_norm": 2.226024361880678, + "learning_rate": 2.156413415712913e-06, + "loss": 0.7532, + "num_input_tokens_seen": 87906320, + "step": 4082 + }, + { + "epoch": 0.4909517224794084, + "flos": 26176125305760.0, + "grad_norm": 13.87725756725336, + "learning_rate": 2.155636818083014e-06, + "loss": 0.7835, + "num_input_tokens_seen": 87927690, + "step": 4083 + }, + { + "epoch": 0.4910719653700475, + "flos": 23115271493760.0, + "grad_norm": 3.017290041113003, + "learning_rate": 2.154860196843134e-06, + "loss": 0.8436, + "num_input_tokens_seen": 87946885, + "step": 4084 + }, + { + "epoch": 0.4911922082606866, + "flos": 23298509908800.0, + "grad_norm": 2.3134659476658657, + "learning_rate": 2.154083552111085e-06, + "loss": 0.7719, + "num_input_tokens_seen": 87966055, + "step": 4085 + }, + { + "epoch": 0.49131245115132566, + "flos": 29161252849440.0, + "grad_norm": 2.0985470377495035, + "learning_rate": 2.1533068840046834e-06, + "loss": 0.8186, + "num_input_tokens_seen": 87986320, + "step": 4086 + }, + { + "epoch": 0.49143269404196477, + "flos": 20118398305920.0, + "grad_norm": 3.7427129515950206, + "learning_rate": 2.152530192641749e-06, + "loss": 0.6173, + "num_input_tokens_seen": 88001230, + "step": 4087 + }, + { + "epoch": 0.4915529369326039, + "flos": 24354929253120.0, + "grad_norm": 2.2391898822420915, + "learning_rate": 2.1517534781401068e-06, + "loss": 0.7251, + "num_input_tokens_seen": 88019110, + "step": 4088 + }, + { + "epoch": 0.49167317982324293, + "flos": 10509579598560.0, + "grad_norm": 2.6844441983382143, + "learning_rate": 2.150976740617581e-06, + "loss": 0.6962, + "num_input_tokens_seen": 88035670, + "step": 4089 + }, + { + "epoch": 0.49179342271388204, + "flos": 25556779632960.0, + "grad_norm": 1.977905661428642, + "learning_rate": 2.150199980192006e-06, + "loss": 0.7127, + "num_input_tokens_seen": 88054790, + "step": 4090 + }, + { + "epoch": 0.49191366560452116, + "flos": 21071803044480.0, + "grad_norm": 1.9376486444366299, + "learning_rate": 2.1494231969812114e-06, + "loss": 0.8136, + "num_input_tokens_seen": 88073780, + "step": 4091 + }, + { + "epoch": 0.4920339084951602, + "flos": 26030173893600.0, + "grad_norm": 2.2588232609278855, + "learning_rate": 2.1486463911030372e-06, + "loss": 0.8081, + "num_input_tokens_seen": 88094430, + "step": 4092 + }, + { + "epoch": 0.4921541513857993, + "flos": 25045317804480.0, + "grad_norm": 3.800884876622405, + "learning_rate": 2.147869562675324e-06, + "loss": 0.745, + "num_input_tokens_seen": 88113400, + "step": 4093 + }, + { + "epoch": 0.49227439427643843, + "flos": 24354929253120.0, + "grad_norm": 2.2581007618346316, + "learning_rate": 2.147092711815915e-06, + "loss": 0.7259, + "num_input_tokens_seen": 88132750, + "step": 4094 + }, + { + "epoch": 0.4923946371670775, + "flos": 11347815219840.0, + "grad_norm": 3.01037995383152, + "learning_rate": 2.1463158386426593e-06, + "loss": 0.8609, + "num_input_tokens_seen": 88150560, + "step": 4095 + }, + { + "epoch": 0.4925148800577166, + "flos": 30402248720160.0, + "grad_norm": 7.87528966946167, + "learning_rate": 2.145538943273407e-06, + "loss": 0.7749, + "num_input_tokens_seen": 88170835, + "step": 4096 + }, + { + "epoch": 0.49263512294835565, + "flos": 20820718350240.0, + "grad_norm": 3.174329287355133, + "learning_rate": 2.144762025826013e-06, + "loss": 0.7211, + "num_input_tokens_seen": 88189925, + "step": 4097 + }, + { + "epoch": 0.49275536583899476, + "flos": 23733427734240.0, + "grad_norm": 3.1929887375697192, + "learning_rate": 2.143985086418334e-06, + "loss": 0.8693, + "num_input_tokens_seen": 88205105, + "step": 4098 + }, + { + "epoch": 0.4928756087296339, + "flos": 22241718866880.0, + "grad_norm": 1.4650012122399845, + "learning_rate": 2.1432081251682324e-06, + "loss": 0.769, + "num_input_tokens_seen": 88225475, + "step": 4099 + }, + { + "epoch": 0.49299585162027293, + "flos": 19617455519520.0, + "grad_norm": 2.1179828346428584, + "learning_rate": 2.142431142193572e-06, + "loss": 0.8733, + "num_input_tokens_seen": 88243255, + "step": 4100 + }, + { + "epoch": 0.49311609451091204, + "flos": 38781185315040.0, + "grad_norm": 2.6762890855909642, + "learning_rate": 2.1416541376122207e-06, + "loss": 0.7187, + "num_input_tokens_seen": 88263435, + "step": 4101 + }, + { + "epoch": 0.49323633740155115, + "flos": 28288443617760.0, + "grad_norm": 2.296864412244411, + "learning_rate": 2.1408771115420496e-06, + "loss": 0.7299, + "num_input_tokens_seen": 88283295, + "step": 4102 + }, + { + "epoch": 0.4933565802921902, + "flos": 21105187222560.0, + "grad_norm": 1.923967168595275, + "learning_rate": 2.140100064100932e-06, + "loss": 0.6462, + "num_input_tokens_seen": 88299270, + "step": 4103 + }, + { + "epoch": 0.4934768231828293, + "flos": 18013290927360.0, + "grad_norm": 3.172610794188882, + "learning_rate": 2.139322995406746e-06, + "loss": 0.7601, + "num_input_tokens_seen": 88316905, + "step": 4104 + }, + { + "epoch": 0.4935970660734684, + "flos": 23436135294720.0, + "grad_norm": 2.594896086300943, + "learning_rate": 2.1385459055773727e-06, + "loss": 0.7973, + "num_input_tokens_seen": 88335095, + "step": 4105 + }, + { + "epoch": 0.4937173089641075, + "flos": 64385702387040.0, + "grad_norm": 2.5952590780713995, + "learning_rate": 2.137768794730696e-06, + "loss": 0.7366, + "num_input_tokens_seen": 88358545, + "step": 4106 + }, + { + "epoch": 0.4938375518547466, + "flos": 22314099856800.0, + "grad_norm": 2.2240210233216127, + "learning_rate": 2.1369916629846026e-06, + "loss": 0.8058, + "num_input_tokens_seen": 88378295, + "step": 4107 + }, + { + "epoch": 0.4939577947453857, + "flos": 17833323451200.0, + "grad_norm": 5.022389826233164, + "learning_rate": 2.136214510456983e-06, + "loss": 0.7519, + "num_input_tokens_seen": 88396750, + "step": 4108 + }, + { + "epoch": 0.49407803763602476, + "flos": 69966572604960.0, + "grad_norm": 0.9035478624924959, + "learning_rate": 2.1354373372657296e-06, + "loss": 0.6313, + "num_input_tokens_seen": 88455190, + "step": 4109 + }, + { + "epoch": 0.49419828052666387, + "flos": 24282734112000.0, + "grad_norm": 1.6964740202129016, + "learning_rate": 2.1346601435287404e-06, + "loss": 0.7073, + "num_input_tokens_seen": 88477460, + "step": 4110 + }, + { + "epoch": 0.494318523417303, + "flos": 29343822208800.0, + "grad_norm": 5.6075382041224975, + "learning_rate": 2.1338829293639144e-06, + "loss": 0.8041, + "num_input_tokens_seen": 88497820, + "step": 4111 + }, + { + "epoch": 0.49443876630794203, + "flos": 15246681634560.0, + "grad_norm": 2.759531495785636, + "learning_rate": 2.1331056948891547e-06, + "loss": 0.8302, + "num_input_tokens_seen": 88514920, + "step": 4112 + }, + { + "epoch": 0.49455900919858115, + "flos": 12330032256000.0, + "grad_norm": 2.6870233867211004, + "learning_rate": 2.1323284402223666e-06, + "loss": 0.7649, + "num_input_tokens_seen": 88530910, + "step": 4113 + }, + { + "epoch": 0.4946792520892202, + "flos": 22746899005920.0, + "grad_norm": 2.676124272321942, + "learning_rate": 2.1315511654814597e-06, + "loss": 0.8821, + "num_input_tokens_seen": 88549320, + "step": 4114 + }, + { + "epoch": 0.4947994949798593, + "flos": 23115011305440.0, + "grad_norm": 4.0184546568276085, + "learning_rate": 2.1307738707843456e-06, + "loss": 0.7815, + "num_input_tokens_seen": 88568985, + "step": 4115 + }, + { + "epoch": 0.4949197378704984, + "flos": 23626175775840.0, + "grad_norm": 2.252157766104336, + "learning_rate": 2.1299965562489385e-06, + "loss": 0.693, + "num_input_tokens_seen": 88588790, + "step": 4116 + }, + { + "epoch": 0.4950399807611375, + "flos": 26873092904640.0, + "grad_norm": 1.6305855344102855, + "learning_rate": 2.129219221993158e-06, + "loss": 0.7919, + "num_input_tokens_seen": 88613575, + "step": 4117 + }, + { + "epoch": 0.4951602236517766, + "flos": 67218553925760.0, + "grad_norm": 0.7946946477774823, + "learning_rate": 2.128441868134924e-06, + "loss": 0.5988, + "num_input_tokens_seen": 88675510, + "step": 4118 + }, + { + "epoch": 0.4952804665424157, + "flos": 19873074924480.0, + "grad_norm": 5.779550993483776, + "learning_rate": 2.1276644947921606e-06, + "loss": 0.8302, + "num_input_tokens_seen": 88694425, + "step": 4119 + }, + { + "epoch": 0.49540070943305475, + "flos": 18780037632960.0, + "grad_norm": 2.4507613731212996, + "learning_rate": 2.126887102082795e-06, + "loss": 0.8286, + "num_input_tokens_seen": 88714450, + "step": 4120 + }, + { + "epoch": 0.49552095232369386, + "flos": 24898399978560.0, + "grad_norm": 1.8338099062546893, + "learning_rate": 2.126109690124757e-06, + "loss": 0.7059, + "num_input_tokens_seen": 88735265, + "step": 4121 + }, + { + "epoch": 0.495641195214333, + "flos": 22824595271520.0, + "grad_norm": 1.8465738170339607, + "learning_rate": 2.1253322590359786e-06, + "loss": 0.7125, + "num_input_tokens_seen": 88754600, + "step": 4122 + }, + { + "epoch": 0.49576143810497203, + "flos": 25732621265760.0, + "grad_norm": 2.831617359303131, + "learning_rate": 2.124554808934397e-06, + "loss": 0.7397, + "num_input_tokens_seen": 88775180, + "step": 4123 + }, + { + "epoch": 0.49588168099561114, + "flos": 22097291414880.0, + "grad_norm": 2.3521815907644963, + "learning_rate": 2.1237773399379496e-06, + "loss": 0.73, + "num_input_tokens_seen": 88796145, + "step": 4124 + }, + { + "epoch": 0.49600192388625025, + "flos": 24352364539680.0, + "grad_norm": 2.4361955135353317, + "learning_rate": 2.122999852164578e-06, + "loss": 0.8689, + "num_input_tokens_seen": 88816765, + "step": 4125 + }, + { + "epoch": 0.4961221667768893, + "flos": 22824781120320.0, + "grad_norm": 9.79036475650574, + "learning_rate": 2.122222345732227e-06, + "loss": 0.5868, + "num_input_tokens_seen": 88836680, + "step": 4126 + }, + { + "epoch": 0.4962424096675284, + "flos": 17832691565280.0, + "grad_norm": 1.9806008078583406, + "learning_rate": 2.121444820758843e-06, + "loss": 0.8342, + "num_input_tokens_seen": 88855320, + "step": 4127 + }, + { + "epoch": 0.49636265255816747, + "flos": 21762303105120.0, + "grad_norm": 2.177851077660565, + "learning_rate": 2.120667277362376e-06, + "loss": 0.7892, + "num_input_tokens_seen": 88874035, + "step": 4128 + }, + { + "epoch": 0.4964828954488066, + "flos": 16335258554880.0, + "grad_norm": 2.3595735580059523, + "learning_rate": 2.1198897156607796e-06, + "loss": 0.8483, + "num_input_tokens_seen": 88891390, + "step": 4129 + }, + { + "epoch": 0.4966031383394457, + "flos": 24675495696000.0, + "grad_norm": 2.9456320672746683, + "learning_rate": 2.1191121357720085e-06, + "loss": 0.7382, + "num_input_tokens_seen": 88909450, + "step": 4130 + }, + { + "epoch": 0.49672338123008475, + "flos": 22897868335680.0, + "grad_norm": 2.4410091473924895, + "learning_rate": 2.1183345378140206e-06, + "loss": 0.7491, + "num_input_tokens_seen": 88929195, + "step": 4131 + }, + { + "epoch": 0.49684362412072386, + "flos": 65882386268640.0, + "grad_norm": 0.8804252891340961, + "learning_rate": 2.1175569219047783e-06, + "loss": 0.6197, + "num_input_tokens_seen": 88990635, + "step": 4132 + }, + { + "epoch": 0.49696386701136297, + "flos": 19942705352160.0, + "grad_norm": 4.715142898740397, + "learning_rate": 2.1167792881622437e-06, + "loss": 0.7347, + "num_input_tokens_seen": 89009655, + "step": 4133 + }, + { + "epoch": 0.497084109902002, + "flos": 24715273072800.0, + "grad_norm": 1.7641750222093688, + "learning_rate": 2.116001636704384e-06, + "loss": 0.8077, + "num_input_tokens_seen": 89030040, + "step": 4134 + }, + { + "epoch": 0.49720435279264114, + "flos": 21840594086880.0, + "grad_norm": 2.025194502543478, + "learning_rate": 2.1152239676491685e-06, + "loss": 0.8034, + "num_input_tokens_seen": 89048145, + "step": 4135 + }, + { + "epoch": 0.49732459568328025, + "flos": 23771755490400.0, + "grad_norm": 2.2118255664899005, + "learning_rate": 2.114446281114569e-06, + "loss": 0.7346, + "num_input_tokens_seen": 89067165, + "step": 4136 + }, + { + "epoch": 0.4974448385739193, + "flos": 20019175015680.0, + "grad_norm": 2.063671010826113, + "learning_rate": 2.1136685772185587e-06, + "loss": 0.7642, + "num_input_tokens_seen": 89086190, + "step": 4137 + }, + { + "epoch": 0.4975650814645584, + "flos": 24785609725920.0, + "grad_norm": 2.1458462711672572, + "learning_rate": 2.1128908560791163e-06, + "loss": 0.7773, + "num_input_tokens_seen": 89106020, + "step": 4138 + }, + { + "epoch": 0.4976853243551975, + "flos": 19801251480960.0, + "grad_norm": 2.4070553113210824, + "learning_rate": 2.1121131178142203e-06, + "loss": 0.7818, + "num_input_tokens_seen": 89125385, + "step": 4139 + }, + { + "epoch": 0.4978055672458366, + "flos": 23109696029760.0, + "grad_norm": 1.9887470531621616, + "learning_rate": 2.1113353625418544e-06, + "loss": 0.8252, + "num_input_tokens_seen": 89143935, + "step": 4140 + }, + { + "epoch": 0.4979258101364757, + "flos": 15537878233440.0, + "grad_norm": 1.9299069175608492, + "learning_rate": 2.1105575903800017e-06, + "loss": 0.7929, + "num_input_tokens_seen": 89162655, + "step": 4141 + }, + { + "epoch": 0.4980460530271148, + "flos": 26318285402400.0, + "grad_norm": 2.0649089039436324, + "learning_rate": 2.1097798014466502e-06, + "loss": 0.8469, + "num_input_tokens_seen": 89182255, + "step": 4142 + }, + { + "epoch": 0.49816629591775385, + "flos": 17250001009440.0, + "grad_norm": 2.927806525076206, + "learning_rate": 2.109001995859791e-06, + "loss": 0.5871, + "num_input_tokens_seen": 89201155, + "step": 4143 + }, + { + "epoch": 0.49828653880839296, + "flos": 64838567472960.0, + "grad_norm": 0.812474763918483, + "learning_rate": 2.108224173737415e-06, + "loss": 0.6001, + "num_input_tokens_seen": 89264455, + "step": 4144 + }, + { + "epoch": 0.498406781699032, + "flos": 27445190078880.0, + "grad_norm": 2.337935699818713, + "learning_rate": 2.1074463351975183e-06, + "loss": 0.7631, + "num_input_tokens_seen": 89283340, + "step": 4145 + }, + { + "epoch": 0.49852702458967113, + "flos": 31454467881600.0, + "grad_norm": 3.110151184430533, + "learning_rate": 2.106668480358098e-06, + "loss": 0.7152, + "num_input_tokens_seen": 89303565, + "step": 4146 + }, + { + "epoch": 0.49864726748031024, + "flos": 22820060560800.0, + "grad_norm": 1.9145598096374412, + "learning_rate": 2.105890609337154e-06, + "loss": 0.71, + "num_input_tokens_seen": 89322080, + "step": 4147 + }, + { + "epoch": 0.4987675103709493, + "flos": 70305426569760.0, + "grad_norm": 0.7094763948009938, + "learning_rate": 2.1051127222526883e-06, + "loss": 0.6374, + "num_input_tokens_seen": 89394195, + "step": 4148 + }, + { + "epoch": 0.4988877532615884, + "flos": 28726297854240.0, + "grad_norm": 2.470735934194214, + "learning_rate": 2.1043348192227067e-06, + "loss": 0.8138, + "num_input_tokens_seen": 89414565, + "step": 4149 + }, + { + "epoch": 0.4990079961522275, + "flos": 16848727550400.0, + "grad_norm": 3.778340883610645, + "learning_rate": 2.1035569003652156e-06, + "loss": 0.6205, + "num_input_tokens_seen": 89433195, + "step": 4150 + }, + { + "epoch": 0.4991282390428666, + "flos": 13272063048000.0, + "grad_norm": 2.793739223246089, + "learning_rate": 2.1027789657982255e-06, + "loss": 0.819, + "num_input_tokens_seen": 89447410, + "step": 4151 + }, + { + "epoch": 0.4992484819335057, + "flos": 21506535021120.0, + "grad_norm": 2.2982611593081144, + "learning_rate": 2.1020010156397482e-06, + "loss": 0.7767, + "num_input_tokens_seen": 89464630, + "step": 4152 + }, + { + "epoch": 0.4993687248241448, + "flos": 24824383519200.0, + "grad_norm": 1.837685000909334, + "learning_rate": 2.101223050007797e-06, + "loss": 0.7764, + "num_input_tokens_seen": 89483180, + "step": 4153 + }, + { + "epoch": 0.49948896771478385, + "flos": 62851422677280.0, + "grad_norm": 0.8328391437670085, + "learning_rate": 2.1004450690203904e-06, + "loss": 0.5368, + "num_input_tokens_seen": 89539260, + "step": 4154 + }, + { + "epoch": 0.49960921060542296, + "flos": 68186826568320.0, + "grad_norm": 0.8638911099693255, + "learning_rate": 2.099667072795546e-06, + "loss": 0.6333, + "num_input_tokens_seen": 89601380, + "step": 4155 + }, + { + "epoch": 0.49972945349606207, + "flos": 23625506720160.0, + "grad_norm": 2.3440151827519027, + "learning_rate": 2.0988890614512864e-06, + "loss": 0.7939, + "num_input_tokens_seen": 89621270, + "step": 4156 + }, + { + "epoch": 0.4998496963867011, + "flos": 19727681058720.0, + "grad_norm": 7.455421461834601, + "learning_rate": 2.098111035105635e-06, + "loss": 0.8421, + "num_input_tokens_seen": 89639695, + "step": 4157 + }, + { + "epoch": 0.49996993927734024, + "flos": 22233653028960.0, + "grad_norm": 2.598084653936156, + "learning_rate": 2.0973329938766176e-06, + "loss": 0.7339, + "num_input_tokens_seen": 89657125, + "step": 4158 + }, + { + "epoch": 0.5000901821679793, + "flos": 23293715009760.0, + "grad_norm": 2.869759136553285, + "learning_rate": 2.0965549378822618e-06, + "loss": 0.7909, + "num_input_tokens_seen": 89678415, + "step": 4159 + }, + { + "epoch": 0.5002104250586185, + "flos": 20309999916960.0, + "grad_norm": 2.8307917546170387, + "learning_rate": 2.095776867240599e-06, + "loss": 0.8386, + "num_input_tokens_seen": 89695405, + "step": 4160 + }, + { + "epoch": 0.5003306679492575, + "flos": 13971632530080.0, + "grad_norm": 2.965284875111321, + "learning_rate": 2.094998782069661e-06, + "loss": 0.8286, + "num_input_tokens_seen": 89713110, + "step": 4161 + }, + { + "epoch": 0.5004509108398966, + "flos": 27633074713920.0, + "grad_norm": 3.061390578506755, + "learning_rate": 2.0942206824874845e-06, + "loss": 0.7585, + "num_input_tokens_seen": 89735885, + "step": 4162 + }, + { + "epoch": 0.5005711537305357, + "flos": 14955001828800.0, + "grad_norm": 5.220117994999143, + "learning_rate": 2.093442568612105e-06, + "loss": 0.7932, + "num_input_tokens_seen": 89753690, + "step": 4163 + }, + { + "epoch": 0.5006913966211748, + "flos": 26465203228320.0, + "grad_norm": 3.059228356249522, + "learning_rate": 2.0926644405615613e-06, + "loss": 0.85, + "num_input_tokens_seen": 89774590, + "step": 4164 + }, + { + "epoch": 0.5008116395118138, + "flos": 20420151116640.0, + "grad_norm": 4.006583518013372, + "learning_rate": 2.091886298453897e-06, + "loss": 0.8142, + "num_input_tokens_seen": 89792775, + "step": 4165 + }, + { + "epoch": 0.500931882402453, + "flos": 21549062960160.0, + "grad_norm": 3.689186615655521, + "learning_rate": 2.091108142407153e-06, + "loss": 0.7288, + "num_input_tokens_seen": 89811515, + "step": 4166 + }, + { + "epoch": 0.5010521252930921, + "flos": 57703228632000.0, + "grad_norm": 0.8421512220846159, + "learning_rate": 2.090329972539377e-06, + "loss": 0.6232, + "num_input_tokens_seen": 89870080, + "step": 4167 + }, + { + "epoch": 0.5011723681837311, + "flos": 18599029403520.0, + "grad_norm": 3.7681498946485714, + "learning_rate": 2.089551788968616e-06, + "loss": 0.6843, + "num_input_tokens_seen": 89888045, + "step": 4168 + }, + { + "epoch": 0.5012926110743702, + "flos": 55803964616160.0, + "grad_norm": 0.8480633579868022, + "learning_rate": 2.08877359181292e-06, + "loss": 0.6055, + "num_input_tokens_seen": 89944610, + "step": 4169 + }, + { + "epoch": 0.5014128539650093, + "flos": 24203402376960.0, + "grad_norm": 5.250609771194233, + "learning_rate": 2.0879953811903396e-06, + "loss": 0.8556, + "num_input_tokens_seen": 89959495, + "step": 4170 + }, + { + "epoch": 0.5015330968556484, + "flos": 27488015376000.0, + "grad_norm": 2.4969691023681393, + "learning_rate": 2.08721715721893e-06, + "loss": 0.7865, + "num_input_tokens_seen": 89978820, + "step": 4171 + }, + { + "epoch": 0.5016533397462875, + "flos": 23766588893760.0, + "grad_norm": 2.431204368339598, + "learning_rate": 2.0864389200167477e-06, + "loss": 0.7695, + "num_input_tokens_seen": 89997075, + "step": 4172 + }, + { + "epoch": 0.5017735826369266, + "flos": 25258557949440.0, + "grad_norm": 2.5295705993294963, + "learning_rate": 2.0856606697018504e-06, + "loss": 0.792, + "num_input_tokens_seen": 90015085, + "step": 4173 + }, + { + "epoch": 0.5018938255275657, + "flos": 16849247927040.0, + "grad_norm": 2.6585944207947563, + "learning_rate": 2.084882406392297e-06, + "loss": 0.7311, + "num_input_tokens_seen": 90033045, + "step": 4174 + }, + { + "epoch": 0.5020140684182047, + "flos": 25478711669760.0, + "grad_norm": 4.247173042309244, + "learning_rate": 2.0841041302061496e-06, + "loss": 0.7113, + "num_input_tokens_seen": 90052505, + "step": 4175 + }, + { + "epoch": 0.5021343113088439, + "flos": 23625432380640.0, + "grad_norm": 3.1060644574081553, + "learning_rate": 2.083325841261473e-06, + "loss": 0.7562, + "num_input_tokens_seen": 90071565, + "step": 4176 + }, + { + "epoch": 0.502254554199483, + "flos": 24499542553920.0, + "grad_norm": 34.605824803803806, + "learning_rate": 2.0825475396763322e-06, + "loss": 0.6614, + "num_input_tokens_seen": 90092215, + "step": 4177 + }, + { + "epoch": 0.502374797090122, + "flos": 34191038274720.0, + "grad_norm": 2.555599038357687, + "learning_rate": 2.081769225568796e-06, + "loss": 0.6584, + "num_input_tokens_seen": 90114665, + "step": 4178 + }, + { + "epoch": 0.5024950399807612, + "flos": 25993816134720.0, + "grad_norm": 2.076960416604889, + "learning_rate": 2.0809908990569327e-06, + "loss": 0.7592, + "num_input_tokens_seen": 90136445, + "step": 4179 + }, + { + "epoch": 0.5026152828714002, + "flos": 21222140488320.0, + "grad_norm": 2.6844410010007675, + "learning_rate": 2.0802125602588146e-06, + "loss": 0.788, + "num_input_tokens_seen": 90155710, + "step": 4180 + }, + { + "epoch": 0.5027355257620393, + "flos": 30911629042080.0, + "grad_norm": 3.6474081216892293, + "learning_rate": 2.0794342092925146e-06, + "loss": 0.6634, + "num_input_tokens_seen": 90176845, + "step": 4181 + }, + { + "epoch": 0.5028557686526784, + "flos": 24755682335520.0, + "grad_norm": 3.2311488476974506, + "learning_rate": 2.078655846276108e-06, + "loss": 0.6774, + "num_input_tokens_seen": 90197175, + "step": 4182 + }, + { + "epoch": 0.5029760115433175, + "flos": 22934188924800.0, + "grad_norm": 4.183443296489317, + "learning_rate": 2.0778774713276727e-06, + "loss": 0.6839, + "num_input_tokens_seen": 90216650, + "step": 4183 + }, + { + "epoch": 0.5030962544339566, + "flos": 15283522600320.0, + "grad_norm": 4.284534391836992, + "learning_rate": 2.077099084565287e-06, + "loss": 0.6748, + "num_input_tokens_seen": 90234055, + "step": 4184 + }, + { + "epoch": 0.5032164973245957, + "flos": 24459653667840.0, + "grad_norm": 3.972495646873475, + "learning_rate": 2.0763206861070313e-06, + "loss": 0.6496, + "num_input_tokens_seen": 90253115, + "step": 4185 + }, + { + "epoch": 0.5033367402152348, + "flos": 16190347896000.0, + "grad_norm": 11.90280205570704, + "learning_rate": 2.0755422760709876e-06, + "loss": 0.7559, + "num_input_tokens_seen": 90270515, + "step": 4186 + }, + { + "epoch": 0.5034569831058738, + "flos": 21360620778720.0, + "grad_norm": 2.6472043567742665, + "learning_rate": 2.0747638545752417e-06, + "loss": 0.7702, + "num_input_tokens_seen": 90289075, + "step": 4187 + }, + { + "epoch": 0.503577225996513, + "flos": 20529447411840.0, + "grad_norm": 26.702097561056064, + "learning_rate": 2.073985421737878e-06, + "loss": 0.8339, + "num_input_tokens_seen": 90306385, + "step": 4188 + }, + { + "epoch": 0.5036974688871521, + "flos": 27189905201760.0, + "grad_norm": 3.3332151074106373, + "learning_rate": 2.0732069776769844e-06, + "loss": 0.7357, + "num_input_tokens_seen": 90323795, + "step": 4189 + }, + { + "epoch": 0.5038177117777911, + "flos": 20383049962560.0, + "grad_norm": 3.088698762154536, + "learning_rate": 2.072428522510651e-06, + "loss": 0.7326, + "num_input_tokens_seen": 90340360, + "step": 4190 + }, + { + "epoch": 0.5039379546684303, + "flos": 21876691657440.0, + "grad_norm": 4.460580039859411, + "learning_rate": 2.071650056356968e-06, + "loss": 0.7633, + "num_input_tokens_seen": 90357900, + "step": 4191 + }, + { + "epoch": 0.5040581975590693, + "flos": 19982185370880.0, + "grad_norm": 2.535425580358287, + "learning_rate": 2.070871579334028e-06, + "loss": 0.8018, + "num_input_tokens_seen": 90377010, + "step": 4192 + }, + { + "epoch": 0.5041784404497084, + "flos": 20929940305920.0, + "grad_norm": 3.767570278066608, + "learning_rate": 2.0700930915599264e-06, + "loss": 0.7156, + "num_input_tokens_seen": 90396740, + "step": 4193 + }, + { + "epoch": 0.5042986833403476, + "flos": 12477507628320.0, + "grad_norm": 3.0680776135252423, + "learning_rate": 2.0693145931527583e-06, + "loss": 0.783, + "num_input_tokens_seen": 90413220, + "step": 4194 + }, + { + "epoch": 0.5044189262309866, + "flos": 29160286435680.0, + "grad_norm": 2.109556628284302, + "learning_rate": 2.068536084230622e-06, + "loss": 0.7799, + "num_input_tokens_seen": 90435085, + "step": 4195 + }, + { + "epoch": 0.5045391691216257, + "flos": 23838858374400.0, + "grad_norm": 4.914737430672281, + "learning_rate": 2.067757564911616e-06, + "loss": 0.8855, + "num_input_tokens_seen": 90453660, + "step": 4196 + }, + { + "epoch": 0.5046594120122648, + "flos": 24609730923360.0, + "grad_norm": 13.334570700132643, + "learning_rate": 2.0669790353138407e-06, + "loss": 0.9315, + "num_input_tokens_seen": 90472625, + "step": 4197 + }, + { + "epoch": 0.5047796549029039, + "flos": 23329255033920.0, + "grad_norm": 8.894243037161823, + "learning_rate": 2.0662004955553995e-06, + "loss": 0.7283, + "num_input_tokens_seen": 90492325, + "step": 4198 + }, + { + "epoch": 0.5048998977935429, + "flos": 17279259344160.0, + "grad_norm": 8.978946644741859, + "learning_rate": 2.065421945754395e-06, + "loss": 0.769, + "num_input_tokens_seen": 90510055, + "step": 4199 + }, + { + "epoch": 0.505020140684182, + "flos": 34806704141280.0, + "grad_norm": 3.3271111748537088, + "learning_rate": 2.0646433860289344e-06, + "loss": 0.7831, + "num_input_tokens_seen": 90528980, + "step": 4200 + }, + { + "epoch": 0.5051403835748212, + "flos": 24828397853280.0, + "grad_norm": 2.8339406278928774, + "learning_rate": 2.0638648164971233e-06, + "loss": 0.8257, + "num_input_tokens_seen": 90547445, + "step": 4201 + }, + { + "epoch": 0.5052606264654602, + "flos": 20929791626880.0, + "grad_norm": 3.0841573868407233, + "learning_rate": 2.06308623727707e-06, + "loss": 0.8862, + "num_input_tokens_seen": 90567545, + "step": 4202 + }, + { + "epoch": 0.5053808693560993, + "flos": 19614258920160.0, + "grad_norm": 3.9849245178655437, + "learning_rate": 2.0623076484868846e-06, + "loss": 0.7695, + "num_input_tokens_seen": 90585000, + "step": 4203 + }, + { + "epoch": 0.5055011122467384, + "flos": 67408185539520.0, + "grad_norm": 0.8622727177570072, + "learning_rate": 2.061529050244679e-06, + "loss": 0.6069, + "num_input_tokens_seen": 90644745, + "step": 4204 + }, + { + "epoch": 0.5056213551373775, + "flos": 16848987738720.0, + "grad_norm": 3.100400615770724, + "learning_rate": 2.060750442668565e-06, + "loss": 0.7366, + "num_input_tokens_seen": 90662500, + "step": 4205 + }, + { + "epoch": 0.5057415980280165, + "flos": 15312223388640.0, + "grad_norm": 4.131370711186189, + "learning_rate": 2.059971825876657e-06, + "loss": 0.6465, + "num_input_tokens_seen": 90677010, + "step": 4206 + }, + { + "epoch": 0.5058618409186557, + "flos": 18998890411680.0, + "grad_norm": 2.7799696814990105, + "learning_rate": 2.0591931999870713e-06, + "loss": 0.7647, + "num_input_tokens_seen": 90695935, + "step": 4207 + }, + { + "epoch": 0.5059820838092948, + "flos": 63362215450080.0, + "grad_norm": 0.8552322409134416, + "learning_rate": 2.0584145651179234e-06, + "loss": 0.5753, + "num_input_tokens_seen": 90751645, + "step": 4208 + }, + { + "epoch": 0.5061023266999338, + "flos": 15419624026080.0, + "grad_norm": 4.455017972453949, + "learning_rate": 2.0576359213873327e-06, + "loss": 0.7968, + "num_input_tokens_seen": 90766795, + "step": 4209 + }, + { + "epoch": 0.506222569590573, + "flos": 22418824271520.0, + "grad_norm": 6.727658633286558, + "learning_rate": 2.056857268913419e-06, + "loss": 0.6972, + "num_input_tokens_seen": 90786845, + "step": 4210 + }, + { + "epoch": 0.506342812481212, + "flos": 17533429128480.0, + "grad_norm": 4.132833960320581, + "learning_rate": 2.056078607814303e-06, + "loss": 0.8423, + "num_input_tokens_seen": 90802630, + "step": 4211 + }, + { + "epoch": 0.5064630553718511, + "flos": 23368512034080.0, + "grad_norm": 3.381416720018967, + "learning_rate": 2.055299938208106e-06, + "loss": 0.7845, + "num_input_tokens_seen": 90823620, + "step": 4212 + }, + { + "epoch": 0.5065832982624903, + "flos": 23952689380320.0, + "grad_norm": 1.9697742522579582, + "learning_rate": 2.0545212602129526e-06, + "loss": 0.8604, + "num_input_tokens_seen": 90843475, + "step": 4213 + }, + { + "epoch": 0.5067035411531293, + "flos": 21472221599040.0, + "grad_norm": 3.301559807866123, + "learning_rate": 2.0537425739469673e-06, + "loss": 0.6588, + "num_input_tokens_seen": 90862410, + "step": 4214 + }, + { + "epoch": 0.5068237840437684, + "flos": 65839895499360.0, + "grad_norm": 0.8440633121445614, + "learning_rate": 2.052963879528276e-06, + "loss": 0.5931, + "num_input_tokens_seen": 90922280, + "step": 4215 + }, + { + "epoch": 0.5069440269344075, + "flos": 27224850509760.0, + "grad_norm": 2.652285141220098, + "learning_rate": 2.052185177075007e-06, + "loss": 0.7656, + "num_input_tokens_seen": 90941850, + "step": 4216 + }, + { + "epoch": 0.5070642698250466, + "flos": 23332860500640.0, + "grad_norm": 2.4515049924603174, + "learning_rate": 2.051406466705288e-06, + "loss": 0.831, + "num_input_tokens_seen": 90961390, + "step": 4217 + }, + { + "epoch": 0.5071845127156857, + "flos": 20311152179520.0, + "grad_norm": 2.965666605232475, + "learning_rate": 2.0506277485372486e-06, + "loss": 0.81, + "num_input_tokens_seen": 90980215, + "step": 4218 + }, + { + "epoch": 0.5073047556063248, + "flos": 12075936811200.0, + "grad_norm": 3.3061135103432133, + "learning_rate": 2.04984902268902e-06, + "loss": 0.6676, + "num_input_tokens_seen": 90997415, + "step": 4219 + }, + { + "epoch": 0.5074249984969639, + "flos": 19654816861920.0, + "grad_norm": 3.0966906202996065, + "learning_rate": 2.0490702892787345e-06, + "loss": 0.757, + "num_input_tokens_seen": 91016475, + "step": 4220 + }, + { + "epoch": 0.5075452413876029, + "flos": 28724365026720.0, + "grad_norm": 2.409622405296813, + "learning_rate": 2.0482915484245246e-06, + "loss": 0.6226, + "num_input_tokens_seen": 91038095, + "step": 4221 + }, + { + "epoch": 0.5076654842782421, + "flos": 20310445954080.0, + "grad_norm": 3.083218615348139, + "learning_rate": 2.047512800244526e-06, + "loss": 0.8399, + "num_input_tokens_seen": 91053360, + "step": 4222 + }, + { + "epoch": 0.5077857271688812, + "flos": 26321890869120.0, + "grad_norm": 2.966410628710638, + "learning_rate": 2.046734044856873e-06, + "loss": 0.7895, + "num_input_tokens_seen": 91072770, + "step": 4223 + }, + { + "epoch": 0.5079059700595202, + "flos": 21767841399360.0, + "grad_norm": 3.271156116926949, + "learning_rate": 2.045955282379702e-06, + "loss": 0.8121, + "num_input_tokens_seen": 91091430, + "step": 4224 + }, + { + "epoch": 0.5080262129501594, + "flos": 13169345800320.0, + "grad_norm": 4.347517177554432, + "learning_rate": 2.045176512931152e-06, + "loss": 0.7526, + "num_input_tokens_seen": 91106060, + "step": 4225 + }, + { + "epoch": 0.5081464558407984, + "flos": 25265248506240.0, + "grad_norm": 3.3304359400415158, + "learning_rate": 2.0443977366293604e-06, + "loss": 0.7586, + "num_input_tokens_seen": 91124100, + "step": 4226 + }, + { + "epoch": 0.5082666987314375, + "flos": 30906573954720.0, + "grad_norm": 2.97600656358959, + "learning_rate": 2.043618953592468e-06, + "loss": 0.7686, + "num_input_tokens_seen": 91146555, + "step": 4227 + }, + { + "epoch": 0.5083869416220766, + "flos": 19654445164320.0, + "grad_norm": 5.3626608159776135, + "learning_rate": 2.0428401639386144e-06, + "loss": 0.8116, + "num_input_tokens_seen": 91167120, + "step": 4228 + }, + { + "epoch": 0.5085071845127157, + "flos": 71714792951520.0, + "grad_norm": 0.831120632975421, + "learning_rate": 2.042061367785943e-06, + "loss": 0.5796, + "num_input_tokens_seen": 91220260, + "step": 4229 + }, + { + "epoch": 0.5086274274033548, + "flos": 35899741432800.0, + "grad_norm": 3.562941942074383, + "learning_rate": 2.041282565252594e-06, + "loss": 0.7475, + "num_input_tokens_seen": 91240425, + "step": 4230 + }, + { + "epoch": 0.5087476702939938, + "flos": 23480112854400.0, + "grad_norm": 2.5478720556321774, + "learning_rate": 2.040503756456714e-06, + "loss": 0.7688, + "num_input_tokens_seen": 91259635, + "step": 4231 + }, + { + "epoch": 0.508867913184633, + "flos": 15100693052640.0, + "grad_norm": 2.444943921278818, + "learning_rate": 2.0397249415164456e-06, + "loss": 0.7843, + "num_input_tokens_seen": 91276990, + "step": 4232 + }, + { + "epoch": 0.508988156075272, + "flos": 25848236420160.0, + "grad_norm": 2.462235656337071, + "learning_rate": 2.0389461205499354e-06, + "loss": 0.803, + "num_input_tokens_seen": 91296125, + "step": 4233 + }, + { + "epoch": 0.5091083989659111, + "flos": 13825309420320.0, + "grad_norm": 2.4069369437845647, + "learning_rate": 2.03816729367533e-06, + "loss": 0.7346, + "num_input_tokens_seen": 91314280, + "step": 4234 + }, + { + "epoch": 0.5092286418565503, + "flos": 21074739455520.0, + "grad_norm": 2.635210071854374, + "learning_rate": 2.0373884610107765e-06, + "loss": 0.7131, + "num_input_tokens_seen": 91334595, + "step": 4235 + }, + { + "epoch": 0.5093488847471893, + "flos": 18594680541600.0, + "grad_norm": 4.164994527348361, + "learning_rate": 2.0366096226744225e-06, + "loss": 0.6943, + "num_input_tokens_seen": 91349790, + "step": 4236 + }, + { + "epoch": 0.5094691276378284, + "flos": 23769673983840.0, + "grad_norm": 3.5332618316277387, + "learning_rate": 2.035830778784418e-06, + "loss": 0.7721, + "num_input_tokens_seen": 91370465, + "step": 4237 + }, + { + "epoch": 0.5095893705284675, + "flos": 17395655063520.0, + "grad_norm": 2.753211227261753, + "learning_rate": 2.0350519294589134e-06, + "loss": 0.8008, + "num_input_tokens_seen": 91388505, + "step": 4238 + }, + { + "epoch": 0.5097096134191066, + "flos": 25812361868160.0, + "grad_norm": 2.7941884185402466, + "learning_rate": 2.0342730748160588e-06, + "loss": 0.8279, + "num_input_tokens_seen": 91408970, + "step": 4239 + }, + { + "epoch": 0.5098298563097456, + "flos": 27705307024800.0, + "grad_norm": 2.7078951163216303, + "learning_rate": 2.033494214974006e-06, + "loss": 0.7036, + "num_input_tokens_seen": 91430115, + "step": 4240 + }, + { + "epoch": 0.5099500992003848, + "flos": 21328351693440.0, + "grad_norm": 3.3119094610140087, + "learning_rate": 2.0327153500509067e-06, + "loss": 0.8379, + "num_input_tokens_seen": 91449140, + "step": 4241 + }, + { + "epoch": 0.5100703420910239, + "flos": 19837720749120.0, + "grad_norm": 2.8193704400166273, + "learning_rate": 2.031936480164916e-06, + "loss": 0.845, + "num_input_tokens_seen": 91466880, + "step": 4242 + }, + { + "epoch": 0.5101905849816629, + "flos": 24613670917920.0, + "grad_norm": 2.6938170776512385, + "learning_rate": 2.0311576054341857e-06, + "loss": 0.8032, + "num_input_tokens_seen": 91487495, + "step": 4243 + }, + { + "epoch": 0.5103108278723021, + "flos": 22897310789280.0, + "grad_norm": 2.366964499667874, + "learning_rate": 2.0303787259768715e-06, + "loss": 0.627, + "num_input_tokens_seen": 91508395, + "step": 4244 + }, + { + "epoch": 0.5104310707629411, + "flos": 21475901405280.0, + "grad_norm": 3.4043030380394588, + "learning_rate": 2.0295998419111294e-06, + "loss": 0.687, + "num_input_tokens_seen": 91525685, + "step": 4245 + }, + { + "epoch": 0.5105513136535802, + "flos": 14882100462240.0, + "grad_norm": 3.046604242276574, + "learning_rate": 2.028820953355115e-06, + "loss": 0.7341, + "num_input_tokens_seen": 91543785, + "step": 4246 + }, + { + "epoch": 0.5106715565442194, + "flos": 22569942280320.0, + "grad_norm": 4.041784671489623, + "learning_rate": 2.0280420604269834e-06, + "loss": 0.7837, + "num_input_tokens_seen": 91563325, + "step": 4247 + }, + { + "epoch": 0.5107917994348584, + "flos": 70926519221280.0, + "grad_norm": 0.7124714494305453, + "learning_rate": 2.027263163244895e-06, + "loss": 0.5887, + "num_input_tokens_seen": 91632450, + "step": 4248 + }, + { + "epoch": 0.5109120423254975, + "flos": 24791891415360.0, + "grad_norm": 2.817886873371929, + "learning_rate": 2.026484261927005e-06, + "loss": 0.7437, + "num_input_tokens_seen": 91651945, + "step": 4249 + }, + { + "epoch": 0.5110322852161366, + "flos": 21217196910240.0, + "grad_norm": 3.693422636703716, + "learning_rate": 2.025705356591475e-06, + "loss": 0.7431, + "num_input_tokens_seen": 91669635, + "step": 4250 + }, + { + "epoch": 0.5111525281067757, + "flos": 66362285237280.0, + "grad_norm": 0.7689894497126594, + "learning_rate": 2.024926447356462e-06, + "loss": 0.5785, + "num_input_tokens_seen": 91731920, + "step": 4251 + }, + { + "epoch": 0.5112727709974147, + "flos": 14845705533600.0, + "grad_norm": 3.938949318227817, + "learning_rate": 2.024147534340127e-06, + "loss": 0.7847, + "num_input_tokens_seen": 91749780, + "step": 4252 + }, + { + "epoch": 0.5113930138880539, + "flos": 21147343464000.0, + "grad_norm": 2.248897070239485, + "learning_rate": 2.02336861766063e-06, + "loss": 0.8004, + "num_input_tokens_seen": 91768840, + "step": 4253 + }, + { + "epoch": 0.511513256778693, + "flos": 20379890532960.0, + "grad_norm": 2.574122425644715, + "learning_rate": 2.0225896974361327e-06, + "loss": 0.7885, + "num_input_tokens_seen": 91788945, + "step": 4254 + }, + { + "epoch": 0.511633499669332, + "flos": 69779914572000.0, + "grad_norm": 0.873666598824192, + "learning_rate": 2.0218107737847962e-06, + "loss": 0.599, + "num_input_tokens_seen": 91850990, + "step": 4255 + }, + { + "epoch": 0.5117537425599712, + "flos": 24063398126400.0, + "grad_norm": 3.166299062439873, + "learning_rate": 2.0210318468247826e-06, + "loss": 0.7473, + "num_input_tokens_seen": 91869960, + "step": 4256 + }, + { + "epoch": 0.5118739854506102, + "flos": 20929642947840.0, + "grad_norm": 2.505905424484324, + "learning_rate": 2.020252916674255e-06, + "loss": 0.8174, + "num_input_tokens_seen": 91889075, + "step": 4257 + }, + { + "epoch": 0.5119942283412493, + "flos": 17432533199040.0, + "grad_norm": 4.183191388724899, + "learning_rate": 2.019473983451375e-06, + "loss": 0.8095, + "num_input_tokens_seen": 91907495, + "step": 4258 + }, + { + "epoch": 0.5121144712318885, + "flos": 21036597548160.0, + "grad_norm": 3.020566854013973, + "learning_rate": 2.0186950472743076e-06, + "loss": 0.7176, + "num_input_tokens_seen": 91927400, + "step": 4259 + }, + { + "epoch": 0.5122347141225275, + "flos": 19832331133920.0, + "grad_norm": 3.5069005243721696, + "learning_rate": 2.0179161082612162e-06, + "loss": 0.7395, + "num_input_tokens_seen": 91946790, + "step": 4260 + }, + { + "epoch": 0.5123549570131666, + "flos": 22496297518560.0, + "grad_norm": 5.829470563505853, + "learning_rate": 2.017137166530266e-06, + "loss": 0.7247, + "num_input_tokens_seen": 91966325, + "step": 4261 + }, + { + "epoch": 0.5124751999038056, + "flos": 20304164264640.0, + "grad_norm": 3.0476955776028936, + "learning_rate": 2.0163582221996213e-06, + "loss": 0.7982, + "num_input_tokens_seen": 91984700, + "step": 4262 + }, + { + "epoch": 0.5125954427944448, + "flos": 39728642892000.0, + "grad_norm": 3.110843790770628, + "learning_rate": 2.015579275387446e-06, + "loss": 0.682, + "num_input_tokens_seen": 92010020, + "step": 4263 + }, + { + "epoch": 0.5127156856850839, + "flos": 29163483035040.0, + "grad_norm": 18.722716151714106, + "learning_rate": 2.0148003262119085e-06, + "loss": 0.6841, + "num_input_tokens_seen": 92030990, + "step": 4264 + }, + { + "epoch": 0.5128359285757229, + "flos": 13534596028320.0, + "grad_norm": 2.5256199330649824, + "learning_rate": 2.0140213747911728e-06, + "loss": 0.7666, + "num_input_tokens_seen": 92049525, + "step": 4265 + }, + { + "epoch": 0.5129561714663621, + "flos": 25156026550560.0, + "grad_norm": 2.718241501533066, + "learning_rate": 2.013242421243406e-06, + "loss": 0.8072, + "num_input_tokens_seen": 92068430, + "step": 4266 + }, + { + "epoch": 0.5130764143570011, + "flos": 18124408540800.0, + "grad_norm": 2.9278044078306693, + "learning_rate": 2.012463465686774e-06, + "loss": 0.7928, + "num_input_tokens_seen": 92088180, + "step": 4267 + }, + { + "epoch": 0.5131966572476402, + "flos": 59710302152640.0, + "grad_norm": 0.7704890917303527, + "learning_rate": 2.0116845082394446e-06, + "loss": 0.547, + "num_input_tokens_seen": 92153015, + "step": 4268 + }, + { + "epoch": 0.5133169001382794, + "flos": 18488878203840.0, + "grad_norm": 2.626330174888625, + "learning_rate": 2.0109055490195836e-06, + "loss": 0.7886, + "num_input_tokens_seen": 92171470, + "step": 4269 + }, + { + "epoch": 0.5134371430289184, + "flos": 15501557644320.0, + "grad_norm": 2.670244221179896, + "learning_rate": 2.0101265881453605e-06, + "loss": 0.6384, + "num_input_tokens_seen": 92189945, + "step": 4270 + }, + { + "epoch": 0.5135573859195575, + "flos": 21403185887520.0, + "grad_norm": 3.682006882478248, + "learning_rate": 2.009347625734941e-06, + "loss": 0.7769, + "num_input_tokens_seen": 92209855, + "step": 4271 + }, + { + "epoch": 0.5136776288101966, + "flos": 17687483548320.0, + "grad_norm": 3.8877337026724845, + "learning_rate": 2.0085686619064954e-06, + "loss": 0.7552, + "num_input_tokens_seen": 92226295, + "step": 4272 + }, + { + "epoch": 0.5137978717008357, + "flos": 16558162837440.0, + "grad_norm": 3.4715051105455608, + "learning_rate": 2.00778969677819e-06, + "loss": 0.8275, + "num_input_tokens_seen": 92243330, + "step": 4273 + }, + { + "epoch": 0.5139181145914747, + "flos": 20638780876800.0, + "grad_norm": 2.3985637420333794, + "learning_rate": 2.0070107304681934e-06, + "loss": 0.6382, + "num_input_tokens_seen": 92262285, + "step": 4274 + }, + { + "epoch": 0.5140383574821139, + "flos": 32879668581120.0, + "grad_norm": 2.213854695046359, + "learning_rate": 2.006231763094675e-06, + "loss": 0.7825, + "num_input_tokens_seen": 92283305, + "step": 4275 + }, + { + "epoch": 0.514158600372753, + "flos": 19509274317120.0, + "grad_norm": 2.93491217070768, + "learning_rate": 2.0054527947758027e-06, + "loss": 0.8709, + "num_input_tokens_seen": 92302105, + "step": 4276 + }, + { + "epoch": 0.514278843263392, + "flos": 62435015392320.0, + "grad_norm": 0.7350869076139589, + "learning_rate": 2.004673825629746e-06, + "loss": 0.558, + "num_input_tokens_seen": 92362885, + "step": 4277 + }, + { + "epoch": 0.5143990861540312, + "flos": 25845151330080.0, + "grad_norm": 1.9981959073894284, + "learning_rate": 2.0038948557746744e-06, + "loss": 0.7223, + "num_input_tokens_seen": 92384740, + "step": 4278 + }, + { + "epoch": 0.5145193290446702, + "flos": 23297283306720.0, + "grad_norm": 3.2685093165640238, + "learning_rate": 2.0031158853287558e-06, + "loss": 0.7498, + "num_input_tokens_seen": 92405175, + "step": 4279 + }, + { + "epoch": 0.5146395719353093, + "flos": 22823926215840.0, + "grad_norm": 8.43697643774906, + "learning_rate": 2.0023369144101593e-06, + "loss": 0.7025, + "num_input_tokens_seen": 92423345, + "step": 4280 + }, + { + "epoch": 0.5147598148259485, + "flos": 26353602408000.0, + "grad_norm": 5.0789213993895155, + "learning_rate": 2.0015579431370555e-06, + "loss": 0.7707, + "num_input_tokens_seen": 92445025, + "step": 4281 + }, + { + "epoch": 0.5148800577165875, + "flos": 29922461260800.0, + "grad_norm": 9.965565331209538, + "learning_rate": 2.000778971627612e-06, + "loss": 0.699, + "num_input_tokens_seen": 92464490, + "step": 4282 + }, + { + "epoch": 0.5150003006072266, + "flos": 17905890289920.0, + "grad_norm": 3.1655273060914118, + "learning_rate": 2e-06, + "loss": 0.9092, + "num_input_tokens_seen": 92482085, + "step": 4283 + }, + { + "epoch": 0.5151205434978657, + "flos": 18452446105440.0, + "grad_norm": 2.4752260085672075, + "learning_rate": 1.9992210283723878e-06, + "loss": 0.8559, + "num_input_tokens_seen": 92499325, + "step": 4284 + }, + { + "epoch": 0.5152407863885048, + "flos": 25305100222560.0, + "grad_norm": 2.3460240839301654, + "learning_rate": 1.9984420568629448e-06, + "loss": 0.7925, + "num_input_tokens_seen": 92522090, + "step": 4285 + }, + { + "epoch": 0.5153610292791438, + "flos": 18303558282240.0, + "grad_norm": 5.60935476825434, + "learning_rate": 1.9976630855898405e-06, + "loss": 0.7873, + "num_input_tokens_seen": 92539845, + "step": 4286 + }, + { + "epoch": 0.515481272169783, + "flos": 30366299828640.0, + "grad_norm": 3.2241734118147436, + "learning_rate": 1.9968841146712445e-06, + "loss": 0.747, + "num_input_tokens_seen": 92559460, + "step": 4287 + }, + { + "epoch": 0.5156015150604221, + "flos": 23004302559360.0, + "grad_norm": 3.2874605531827923, + "learning_rate": 1.996105144225326e-06, + "loss": 0.7132, + "num_input_tokens_seen": 92579695, + "step": 4288 + }, + { + "epoch": 0.5157217579510611, + "flos": 17833174772160.0, + "grad_norm": 3.8425782441293057, + "learning_rate": 1.995326174370254e-06, + "loss": 0.7907, + "num_input_tokens_seen": 92598795, + "step": 4289 + }, + { + "epoch": 0.5158420008417003, + "flos": 19144172768160.0, + "grad_norm": 2.873912771559059, + "learning_rate": 1.994547205224197e-06, + "loss": 0.7291, + "num_input_tokens_seen": 92617700, + "step": 4290 + }, + { + "epoch": 0.5159622437323393, + "flos": 22387521600000.0, + "grad_norm": 5.734583214246113, + "learning_rate": 1.993768236905325e-06, + "loss": 0.6731, + "num_input_tokens_seen": 92638370, + "step": 4291 + }, + { + "epoch": 0.5160824866229784, + "flos": 24567611851680.0, + "grad_norm": 3.9552369036415786, + "learning_rate": 1.992989269531807e-06, + "loss": 0.6633, + "num_input_tokens_seen": 92657180, + "step": 4292 + }, + { + "epoch": 0.5162027295136175, + "flos": 17977081847520.0, + "grad_norm": 7.094933768158836, + "learning_rate": 1.99221030322181e-06, + "loss": 0.6807, + "num_input_tokens_seen": 92673985, + "step": 4293 + }, + { + "epoch": 0.5163229724042566, + "flos": 27305408846880.0, + "grad_norm": 2.1388367471766045, + "learning_rate": 1.991431338093505e-06, + "loss": 0.8092, + "num_input_tokens_seen": 92696340, + "step": 4294 + }, + { + "epoch": 0.5164432152948957, + "flos": 21731706659040.0, + "grad_norm": 2.4142372043440785, + "learning_rate": 1.9906523742650587e-06, + "loss": 0.7913, + "num_input_tokens_seen": 92715200, + "step": 4295 + }, + { + "epoch": 0.5165634581855347, + "flos": 25513508298720.0, + "grad_norm": 2.8114685710444034, + "learning_rate": 1.9898734118546397e-06, + "loss": 0.7741, + "num_input_tokens_seen": 92735150, + "step": 4296 + }, + { + "epoch": 0.5166837010761739, + "flos": 19872814736160.0, + "grad_norm": 2.5891254055966004, + "learning_rate": 1.989094450980416e-06, + "loss": 0.8027, + "num_input_tokens_seen": 92755250, + "step": 4297 + }, + { + "epoch": 0.516803943966813, + "flos": 26607697852800.0, + "grad_norm": 55.0837607932229, + "learning_rate": 1.9883154917605556e-06, + "loss": 0.7618, + "num_input_tokens_seen": 92774460, + "step": 4298 + }, + { + "epoch": 0.516924186857452, + "flos": 19654816861920.0, + "grad_norm": 13.197837213138074, + "learning_rate": 1.9875365343132262e-06, + "loss": 0.8339, + "num_input_tokens_seen": 92791580, + "step": 4299 + }, + { + "epoch": 0.5170444297480912, + "flos": 15933799247040.0, + "grad_norm": 3.3181044473333094, + "learning_rate": 1.9867575787565946e-06, + "loss": 0.8479, + "num_input_tokens_seen": 92806240, + "step": 4300 + }, + { + "epoch": 0.5171646726387302, + "flos": 14154164719680.0, + "grad_norm": 3.323196128326497, + "learning_rate": 1.9859786252088275e-06, + "loss": 0.8638, + "num_input_tokens_seen": 92823420, + "step": 4301 + }, + { + "epoch": 0.5172849155293693, + "flos": 23544539515680.0, + "grad_norm": 3.2982877624149607, + "learning_rate": 1.9851996737880914e-06, + "loss": 0.676, + "num_input_tokens_seen": 92838605, + "step": 4302 + }, + { + "epoch": 0.5174051584200084, + "flos": 14263386675360.0, + "grad_norm": 2.75784094258195, + "learning_rate": 1.9844207246125537e-06, + "loss": 0.7411, + "num_input_tokens_seen": 92855185, + "step": 4303 + }, + { + "epoch": 0.5175254013106475, + "flos": 37834842830880.0, + "grad_norm": 2.4338396416580435, + "learning_rate": 1.983641777800379e-06, + "loss": 0.6831, + "num_input_tokens_seen": 92877830, + "step": 4304 + }, + { + "epoch": 0.5176456442012866, + "flos": 68452333129440.0, + "grad_norm": 0.7525605361769656, + "learning_rate": 1.9828628334697343e-06, + "loss": 0.5871, + "num_input_tokens_seen": 92945040, + "step": 4305 + }, + { + "epoch": 0.5177658870919257, + "flos": 63992712050880.0, + "grad_norm": 0.772438560190653, + "learning_rate": 1.982083891738784e-06, + "loss": 0.5463, + "num_input_tokens_seen": 93005910, + "step": 4306 + }, + { + "epoch": 0.5178861299825648, + "flos": 26613050298240.0, + "grad_norm": 1.9162063736997814, + "learning_rate": 1.9813049527256923e-06, + "loss": 0.8282, + "num_input_tokens_seen": 93026305, + "step": 4307 + }, + { + "epoch": 0.5180063728732038, + "flos": 17906410666560.0, + "grad_norm": 3.0772127455296294, + "learning_rate": 1.9805260165486252e-06, + "loss": 0.8195, + "num_input_tokens_seen": 93045675, + "step": 4308 + }, + { + "epoch": 0.518126615763843, + "flos": 19473139576800.0, + "grad_norm": 3.9390928891442303, + "learning_rate": 1.9797470833257457e-06, + "loss": 0.8607, + "num_input_tokens_seen": 93065375, + "step": 4309 + }, + { + "epoch": 0.5182468586544821, + "flos": 20674543919520.0, + "grad_norm": 3.0870394899976863, + "learning_rate": 1.9789681531752177e-06, + "loss": 0.7734, + "num_input_tokens_seen": 93085595, + "step": 4310 + }, + { + "epoch": 0.5183671015451211, + "flos": 23079062413920.0, + "grad_norm": 2.083059776148025, + "learning_rate": 1.978189226215204e-06, + "loss": 0.7256, + "num_input_tokens_seen": 93106140, + "step": 4311 + }, + { + "epoch": 0.5184873444357603, + "flos": 17571868393920.0, + "grad_norm": 2.5874693790411483, + "learning_rate": 1.9774103025638675e-06, + "loss": 0.7711, + "num_input_tokens_seen": 93124265, + "step": 4312 + }, + { + "epoch": 0.5186075873263993, + "flos": 24901113371040.0, + "grad_norm": 2.159893551075501, + "learning_rate": 1.9766313823393696e-06, + "loss": 0.7669, + "num_input_tokens_seen": 93145130, + "step": 4313 + }, + { + "epoch": 0.5187278302170384, + "flos": 15168427822560.0, + "grad_norm": 3.0550808470372095, + "learning_rate": 1.975852465659873e-06, + "loss": 0.688, + "num_input_tokens_seen": 93161110, + "step": 4314 + }, + { + "epoch": 0.5188480731076776, + "flos": 24973605870240.0, + "grad_norm": 5.629720106197207, + "learning_rate": 1.9750735526435377e-06, + "loss": 0.7029, + "num_input_tokens_seen": 93178055, + "step": 4315 + }, + { + "epoch": 0.5189683159983166, + "flos": 24754641582240.0, + "grad_norm": 5.917800830435338, + "learning_rate": 1.974294643408525e-06, + "loss": 0.7936, + "num_input_tokens_seen": 93195850, + "step": 4316 + }, + { + "epoch": 0.5190885588889557, + "flos": 24718841369760.0, + "grad_norm": 5.436819801317858, + "learning_rate": 1.9735157380729947e-06, + "loss": 0.6695, + "num_input_tokens_seen": 93216260, + "step": 4317 + }, + { + "epoch": 0.5192088017795948, + "flos": 24677093995680.0, + "grad_norm": 2.7290556693376544, + "learning_rate": 1.9727368367551053e-06, + "loss": 0.8431, + "num_input_tokens_seen": 93234810, + "step": 4318 + }, + { + "epoch": 0.5193290446702339, + "flos": 27190351238880.0, + "grad_norm": 2.560358408634241, + "learning_rate": 1.9719579395730164e-06, + "loss": 0.6825, + "num_input_tokens_seen": 93254185, + "step": 4319 + }, + { + "epoch": 0.5194492875608729, + "flos": 11457260194080.0, + "grad_norm": 3.030372885997041, + "learning_rate": 1.9711790466448854e-06, + "loss": 0.9366, + "num_input_tokens_seen": 93268205, + "step": 4320 + }, + { + "epoch": 0.5195695304515121, + "flos": 20309888407680.0, + "grad_norm": 3.621768695103949, + "learning_rate": 1.9704001580888704e-06, + "loss": 0.7111, + "num_input_tokens_seen": 93286945, + "step": 4321 + }, + { + "epoch": 0.5196897733421512, + "flos": 20019769731840.0, + "grad_norm": 3.406161105894594, + "learning_rate": 1.9696212740231283e-06, + "loss": 0.8685, + "num_input_tokens_seen": 93305595, + "step": 4322 + }, + { + "epoch": 0.5198100162327902, + "flos": 23771458132320.0, + "grad_norm": 5.930792945480242, + "learning_rate": 1.9688423945658146e-06, + "loss": 0.8183, + "num_input_tokens_seen": 93326460, + "step": 4323 + }, + { + "epoch": 0.5199302591234293, + "flos": 23989567515840.0, + "grad_norm": 2.8943793582574546, + "learning_rate": 1.9680635198350845e-06, + "loss": 0.7185, + "num_input_tokens_seen": 93346170, + "step": 4324 + }, + { + "epoch": 0.5200505020140684, + "flos": 26321593511040.0, + "grad_norm": 2.791515137464834, + "learning_rate": 1.967284649949093e-06, + "loss": 0.7282, + "num_input_tokens_seen": 93365380, + "step": 4325 + }, + { + "epoch": 0.5201707449047075, + "flos": 39547225795200.0, + "grad_norm": 6.156246282121339, + "learning_rate": 1.966505785025994e-06, + "loss": 0.725, + "num_input_tokens_seen": 93387285, + "step": 4326 + }, + { + "epoch": 0.5202909877953465, + "flos": 53605072199520.0, + "grad_norm": 5.560629883907986, + "learning_rate": 1.965726925183941e-06, + "loss": 0.7635, + "num_input_tokens_seen": 93410390, + "step": 4327 + }, + { + "epoch": 0.5204112306859857, + "flos": 19509199977600.0, + "grad_norm": 2.575839237373799, + "learning_rate": 1.964948070541087e-06, + "loss": 0.8477, + "num_input_tokens_seen": 93429050, + "step": 4328 + }, + { + "epoch": 0.5205314735766248, + "flos": 15283039393440.0, + "grad_norm": 2.8506978435447046, + "learning_rate": 1.9641692212155816e-06, + "loss": 0.6954, + "num_input_tokens_seen": 93446815, + "step": 4329 + }, + { + "epoch": 0.5206517164672638, + "flos": 59177807973600.0, + "grad_norm": 2.961799113743763, + "learning_rate": 1.9633903773255777e-06, + "loss": 0.7288, + "num_input_tokens_seen": 93469130, + "step": 4330 + }, + { + "epoch": 0.520771959357903, + "flos": 26832646472160.0, + "grad_norm": 2.7641344094505067, + "learning_rate": 1.9626115389892237e-06, + "loss": 0.7463, + "num_input_tokens_seen": 93489920, + "step": 4331 + }, + { + "epoch": 0.520892202248542, + "flos": 26868521024160.0, + "grad_norm": 2.889153183152921, + "learning_rate": 1.96183270632467e-06, + "loss": 0.8544, + "num_input_tokens_seen": 93509770, + "step": 4332 + }, + { + "epoch": 0.5210124451391811, + "flos": 25812436207680.0, + "grad_norm": 2.177830277456723, + "learning_rate": 1.9610538794500644e-06, + "loss": 0.7938, + "num_input_tokens_seen": 93529320, + "step": 4333 + }, + { + "epoch": 0.5211326880298203, + "flos": 70453385148960.0, + "grad_norm": 0.8095430707777158, + "learning_rate": 1.9602750584835542e-06, + "loss": 0.5935, + "num_input_tokens_seen": 93595255, + "step": 4334 + }, + { + "epoch": 0.5212529309204593, + "flos": 15606616586880.0, + "grad_norm": 3.0709526035398977, + "learning_rate": 1.959496243543286e-06, + "loss": 0.8313, + "num_input_tokens_seen": 93613135, + "step": 4335 + }, + { + "epoch": 0.5213731738110984, + "flos": 26204788924320.0, + "grad_norm": 2.5941805194991776, + "learning_rate": 1.9587174347474057e-06, + "loss": 0.7893, + "num_input_tokens_seen": 93629645, + "step": 4336 + }, + { + "epoch": 0.5214934167017375, + "flos": 19390425393600.0, + "grad_norm": 4.832307750589891, + "learning_rate": 1.9579386322140574e-06, + "loss": 0.8235, + "num_input_tokens_seen": 93645325, + "step": 4337 + }, + { + "epoch": 0.5216136595923766, + "flos": 30911294514240.0, + "grad_norm": 2.086262424036588, + "learning_rate": 1.9571598360613854e-06, + "loss": 0.8082, + "num_input_tokens_seen": 93668200, + "step": 4338 + }, + { + "epoch": 0.5217339024830157, + "flos": 21913718472000.0, + "grad_norm": 3.4756641384501044, + "learning_rate": 1.956381046407532e-06, + "loss": 0.6977, + "num_input_tokens_seen": 93688495, + "step": 4339 + }, + { + "epoch": 0.5218541453736548, + "flos": 20893359528480.0, + "grad_norm": 18.567101932448207, + "learning_rate": 1.9556022633706394e-06, + "loss": 0.8582, + "num_input_tokens_seen": 93707285, + "step": 4340 + }, + { + "epoch": 0.5219743882642939, + "flos": 23917409544480.0, + "grad_norm": 5.5564604128290105, + "learning_rate": 1.954823487068848e-06, + "loss": 0.7975, + "num_input_tokens_seen": 93727495, + "step": 4341 + }, + { + "epoch": 0.5220946311549329, + "flos": 28757786374560.0, + "grad_norm": 1.9314628187526672, + "learning_rate": 1.9540447176202976e-06, + "loss": 0.8136, + "num_input_tokens_seen": 93748740, + "step": 4342 + }, + { + "epoch": 0.5222148740455721, + "flos": 67093566258240.0, + "grad_norm": 0.8810377595865522, + "learning_rate": 1.9532659551431272e-06, + "loss": 0.6065, + "num_input_tokens_seen": 93815770, + "step": 4343 + }, + { + "epoch": 0.5223351169362112, + "flos": 61766568466560.0, + "grad_norm": 1.910057517778437, + "learning_rate": 1.9524871997554744e-06, + "loss": 0.6749, + "num_input_tokens_seen": 93843530, + "step": 4344 + }, + { + "epoch": 0.5224553598268502, + "flos": 14626666906080.0, + "grad_norm": 3.032182530376754, + "learning_rate": 1.951708451575475e-06, + "loss": 0.8048, + "num_input_tokens_seen": 93860595, + "step": 4345 + }, + { + "epoch": 0.5225756027174894, + "flos": 14804924573280.0, + "grad_norm": 3.0698933011875624, + "learning_rate": 1.9509297107212657e-06, + "loss": 0.8224, + "num_input_tokens_seen": 93877365, + "step": 4346 + }, + { + "epoch": 0.5226958456081284, + "flos": 23478328705920.0, + "grad_norm": 1.9885001970712604, + "learning_rate": 1.95015097731098e-06, + "loss": 0.7903, + "num_input_tokens_seen": 93896855, + "step": 4347 + }, + { + "epoch": 0.5228160884987675, + "flos": 19035805716960.0, + "grad_norm": 3.4370033598975818, + "learning_rate": 1.949372251462751e-06, + "loss": 0.8157, + "num_input_tokens_seen": 93914270, + "step": 4348 + }, + { + "epoch": 0.5229363313894067, + "flos": 21032954911680.0, + "grad_norm": 10.141796355931985, + "learning_rate": 1.9485935332947124e-06, + "loss": 0.8254, + "num_input_tokens_seen": 93932180, + "step": 4349 + }, + { + "epoch": 0.5230565742800457, + "flos": 14809422114240.0, + "grad_norm": 3.0663962977545505, + "learning_rate": 1.947814822924993e-06, + "loss": 0.8369, + "num_input_tokens_seen": 93949690, + "step": 4350 + }, + { + "epoch": 0.5231768171706848, + "flos": 25776078448800.0, + "grad_norm": 2.4802296921281526, + "learning_rate": 1.9470361204717236e-06, + "loss": 0.8354, + "num_input_tokens_seen": 93968830, + "step": 4351 + }, + { + "epoch": 0.5232970600613239, + "flos": 22711173132960.0, + "grad_norm": 1.9883858104260137, + "learning_rate": 1.9462574260530326e-06, + "loss": 0.8089, + "num_input_tokens_seen": 93989110, + "step": 4352 + }, + { + "epoch": 0.523417302951963, + "flos": 17286098580000.0, + "grad_norm": 3.4654445535860896, + "learning_rate": 1.9454787397870472e-06, + "loss": 0.8118, + "num_input_tokens_seen": 94006430, + "step": 4353 + }, + { + "epoch": 0.523537545842602, + "flos": 18525347472000.0, + "grad_norm": 3.2648635779675783, + "learning_rate": 1.944700061791894e-06, + "loss": 0.7156, + "num_input_tokens_seen": 94024825, + "step": 4354 + }, + { + "epoch": 0.5236577887332411, + "flos": 19691174620800.0, + "grad_norm": 3.050088606000745, + "learning_rate": 1.943921392185698e-06, + "loss": 0.6534, + "num_input_tokens_seen": 94043450, + "step": 4355 + }, + { + "epoch": 0.5237780316238803, + "flos": 23516544952800.0, + "grad_norm": 2.406994394864005, + "learning_rate": 1.9431427310865814e-06, + "loss": 0.7664, + "num_input_tokens_seen": 94063055, + "step": 4356 + }, + { + "epoch": 0.5238982745145193, + "flos": 22459753910880.0, + "grad_norm": 3.039185823032041, + "learning_rate": 1.942364078612667e-06, + "loss": 0.7857, + "num_input_tokens_seen": 94081785, + "step": 4357 + }, + { + "epoch": 0.5240185174051584, + "flos": 27050086800000.0, + "grad_norm": 2.598573084062034, + "learning_rate": 1.9415854348820765e-06, + "loss": 0.7593, + "num_input_tokens_seen": 94101635, + "step": 4358 + }, + { + "epoch": 0.5241387602957975, + "flos": 22642286100480.0, + "grad_norm": 3.3299522577477587, + "learning_rate": 1.940806800012929e-06, + "loss": 0.6806, + "num_input_tokens_seen": 94121705, + "step": 4359 + }, + { + "epoch": 0.5242590031864366, + "flos": 40495055069760.0, + "grad_norm": 2.23607288219979, + "learning_rate": 1.9400281741233432e-06, + "loss": 0.6357, + "num_input_tokens_seen": 94146595, + "step": 4360 + }, + { + "epoch": 0.5243792460770756, + "flos": 66581175185760.0, + "grad_norm": 0.6658689275815225, + "learning_rate": 1.939249557331435e-06, + "loss": 0.5248, + "num_input_tokens_seen": 94212410, + "step": 4361 + }, + { + "epoch": 0.5244994889677148, + "flos": 28143235600800.0, + "grad_norm": 3.4304181756548693, + "learning_rate": 1.938470949755321e-06, + "loss": 0.7261, + "num_input_tokens_seen": 94232290, + "step": 4362 + }, + { + "epoch": 0.5246197318583539, + "flos": 65856436042560.0, + "grad_norm": 0.8210827072664093, + "learning_rate": 1.937692351513115e-06, + "loss": 0.5552, + "num_input_tokens_seen": 94291285, + "step": 4363 + }, + { + "epoch": 0.5247399747489929, + "flos": 21003510728160.0, + "grad_norm": 3.9706704606993894, + "learning_rate": 1.9369137627229297e-06, + "loss": 0.8076, + "num_input_tokens_seen": 94309685, + "step": 4364 + }, + { + "epoch": 0.5248602176396321, + "flos": 18998853241920.0, + "grad_norm": 2.682741076896715, + "learning_rate": 1.936135183502877e-06, + "loss": 0.8841, + "num_input_tokens_seen": 94326985, + "step": 4365 + }, + { + "epoch": 0.5249804605302711, + "flos": 22168371463200.0, + "grad_norm": 4.218265759969198, + "learning_rate": 1.935356613971066e-06, + "loss": 0.8028, + "num_input_tokens_seen": 94347645, + "step": 4366 + }, + { + "epoch": 0.5251007034209102, + "flos": 23771643981120.0, + "grad_norm": 2.581311932206435, + "learning_rate": 1.9345780542456047e-06, + "loss": 0.7678, + "num_input_tokens_seen": 94367020, + "step": 4367 + }, + { + "epoch": 0.5252209463115494, + "flos": 23261297245440.0, + "grad_norm": 3.177766486757696, + "learning_rate": 1.9337995044446007e-06, + "loss": 0.7216, + "num_input_tokens_seen": 94385860, + "step": 4368 + }, + { + "epoch": 0.5253411892021884, + "flos": 19800433746240.0, + "grad_norm": 2.7310278760729805, + "learning_rate": 1.9330209646861596e-06, + "loss": 0.7966, + "num_input_tokens_seen": 94405010, + "step": 4369 + }, + { + "epoch": 0.5254614320928275, + "flos": 24099198338880.0, + "grad_norm": 2.1317096695527966, + "learning_rate": 1.9322424350883843e-06, + "loss": 0.7749, + "num_input_tokens_seen": 94426845, + "step": 4370 + }, + { + "epoch": 0.5255816749834666, + "flos": 24609693753600.0, + "grad_norm": 3.064513187050824, + "learning_rate": 1.931463915769379e-06, + "loss": 0.7856, + "num_input_tokens_seen": 94446960, + "step": 4371 + }, + { + "epoch": 0.5257019178741057, + "flos": 14116617528480.0, + "grad_norm": 3.927717135793222, + "learning_rate": 1.930685406847242e-06, + "loss": 0.74, + "num_input_tokens_seen": 94461535, + "step": 4372 + }, + { + "epoch": 0.5258221607647448, + "flos": 23515244011200.0, + "grad_norm": 2.63064304699987, + "learning_rate": 1.9299069084400734e-06, + "loss": 0.8184, + "num_input_tokens_seen": 94482395, + "step": 4373 + }, + { + "epoch": 0.5259424036553839, + "flos": 24930111517440.0, + "grad_norm": 2.5436690109069495, + "learning_rate": 1.9291284206659717e-06, + "loss": 0.7007, + "num_input_tokens_seen": 94500580, + "step": 4374 + }, + { + "epoch": 0.526062646546023, + "flos": 28722506538720.0, + "grad_norm": 5.055967095267988, + "learning_rate": 1.928349943643032e-06, + "loss": 0.7171, + "num_input_tokens_seen": 94519680, + "step": 4375 + }, + { + "epoch": 0.526182889436662, + "flos": 22788795059040.0, + "grad_norm": 3.639204280150205, + "learning_rate": 1.9275714774893493e-06, + "loss": 0.8196, + "num_input_tokens_seen": 94539135, + "step": 4376 + }, + { + "epoch": 0.5263031323273012, + "flos": 22896678903360.0, + "grad_norm": 4.083126348322715, + "learning_rate": 1.9267930223230154e-06, + "loss": 0.7305, + "num_input_tokens_seen": 94557610, + "step": 4377 + }, + { + "epoch": 0.5264233752179402, + "flos": 17723395270080.0, + "grad_norm": 3.057849856905255, + "learning_rate": 1.9260145782621224e-06, + "loss": 0.7829, + "num_input_tokens_seen": 94575390, + "step": 4378 + }, + { + "epoch": 0.5265436181085793, + "flos": 24386529282720.0, + "grad_norm": 2.154771934000193, + "learning_rate": 1.925236145424758e-06, + "loss": 0.8843, + "num_input_tokens_seen": 94594210, + "step": 4379 + }, + { + "epoch": 0.5266638609992185, + "flos": 69108965805120.0, + "grad_norm": 0.7013097005988311, + "learning_rate": 1.924457723929012e-06, + "loss": 0.5751, + "num_input_tokens_seen": 94665020, + "step": 4380 + }, + { + "epoch": 0.5267841038898575, + "flos": 20709043190400.0, + "grad_norm": 3.095659995756345, + "learning_rate": 1.9236793138929685e-06, + "loss": 0.8288, + "num_input_tokens_seen": 94684290, + "step": 4381 + }, + { + "epoch": 0.5269043467804966, + "flos": 17209777595520.0, + "grad_norm": 2.73553477078573, + "learning_rate": 1.9229009154347133e-06, + "loss": 0.8101, + "num_input_tokens_seen": 94701880, + "step": 4382 + }, + { + "epoch": 0.5270245896711357, + "flos": 18197235567840.0, + "grad_norm": 3.54170852337207, + "learning_rate": 1.922122528672327e-06, + "loss": 0.8074, + "num_input_tokens_seen": 94720545, + "step": 4383 + }, + { + "epoch": 0.5271448325617748, + "flos": 21257160135840.0, + "grad_norm": 3.4447396076714774, + "learning_rate": 1.9213441537238914e-06, + "loss": 0.7833, + "num_input_tokens_seen": 94737420, + "step": 4384 + }, + { + "epoch": 0.5272650754524139, + "flos": 65401595225760.0, + "grad_norm": 0.8513726539025013, + "learning_rate": 1.920565790707485e-06, + "loss": 0.5731, + "num_input_tokens_seen": 94802810, + "step": 4385 + }, + { + "epoch": 0.527385318343053, + "flos": 19648163474880.0, + "grad_norm": 4.3509209808910585, + "learning_rate": 1.9197874397411853e-06, + "loss": 0.6592, + "num_input_tokens_seen": 94819440, + "step": 4386 + }, + { + "epoch": 0.5275055612336921, + "flos": 12694055881920.0, + "grad_norm": 5.148078685986112, + "learning_rate": 1.919009100943067e-06, + "loss": 0.6628, + "num_input_tokens_seen": 94835130, + "step": 4387 + }, + { + "epoch": 0.5276258041243311, + "flos": 17723804137440.0, + "grad_norm": 2.584764842805345, + "learning_rate": 1.9182307744312043e-06, + "loss": 0.6571, + "num_input_tokens_seen": 94852235, + "step": 4388 + }, + { + "epoch": 0.5277460470149702, + "flos": 22677974803680.0, + "grad_norm": 4.815566634035494, + "learning_rate": 1.9174524603236676e-06, + "loss": 0.7633, + "num_input_tokens_seen": 94871185, + "step": 4389 + }, + { + "epoch": 0.5278662899056094, + "flos": 19874375866080.0, + "grad_norm": 3.2570965389553175, + "learning_rate": 1.916674158738527e-06, + "loss": 0.7602, + "num_input_tokens_seen": 94890090, + "step": 4390 + }, + { + "epoch": 0.5279865327962484, + "flos": 17979646560960.0, + "grad_norm": 2.577554813101882, + "learning_rate": 1.9158958697938506e-06, + "loss": 0.6019, + "num_input_tokens_seen": 94908470, + "step": 4391 + }, + { + "epoch": 0.5281067756868875, + "flos": 15901864689600.0, + "grad_norm": 3.84382319186857, + "learning_rate": 1.9151175936077032e-06, + "loss": 0.8572, + "num_input_tokens_seen": 94923440, + "step": 4392 + }, + { + "epoch": 0.5282270185775266, + "flos": 19398788589600.0, + "grad_norm": 5.789613355202087, + "learning_rate": 1.9143393302981507e-06, + "loss": 0.7943, + "num_input_tokens_seen": 94941125, + "step": 4393 + }, + { + "epoch": 0.5283472614681657, + "flos": 16376374043040.0, + "grad_norm": 4.442181548907265, + "learning_rate": 1.913561079983252e-06, + "loss": 0.8315, + "num_input_tokens_seen": 94959950, + "step": 4394 + }, + { + "epoch": 0.5284675043588047, + "flos": 26722160744640.0, + "grad_norm": 2.786101321347618, + "learning_rate": 1.9127828427810693e-06, + "loss": 0.7504, + "num_input_tokens_seen": 94980515, + "step": 4395 + }, + { + "epoch": 0.5285877472494439, + "flos": 19871476624800.0, + "grad_norm": 2.8086932380892855, + "learning_rate": 1.9120046188096607e-06, + "loss": 0.808, + "num_input_tokens_seen": 94998560, + "step": 4396 + }, + { + "epoch": 0.528707990140083, + "flos": 19982742917280.0, + "grad_norm": 2.7167300031646895, + "learning_rate": 1.9112264081870804e-06, + "loss": 0.7427, + "num_input_tokens_seen": 95016950, + "step": 4397 + }, + { + "epoch": 0.528828233030722, + "flos": 20638483518720.0, + "grad_norm": 2.62885564384102, + "learning_rate": 1.9104482110313843e-06, + "loss": 0.761, + "num_input_tokens_seen": 95036540, + "step": 4398 + }, + { + "epoch": 0.5289484759213612, + "flos": 25156398248160.0, + "grad_norm": 3.467893065030385, + "learning_rate": 1.909670027460623e-06, + "loss": 0.7406, + "num_input_tokens_seen": 95053840, + "step": 4399 + }, + { + "epoch": 0.5290687188120002, + "flos": 31094532929280.0, + "grad_norm": 4.734857213607603, + "learning_rate": 1.908891857592847e-06, + "loss": 0.7151, + "num_input_tokens_seen": 95074910, + "step": 4400 + }, + { + "epoch": 0.5291889617026393, + "flos": 20091147138240.0, + "grad_norm": 3.53470786607413, + "learning_rate": 1.9081137015461034e-06, + "loss": 0.9024, + "num_input_tokens_seen": 95090985, + "step": 4401 + }, + { + "epoch": 0.5293092045932785, + "flos": 19615671371040.0, + "grad_norm": 2.755150133947259, + "learning_rate": 1.9073355594384383e-06, + "loss": 0.9051, + "num_input_tokens_seen": 95109225, + "step": 4402 + }, + { + "epoch": 0.5294294474839175, + "flos": 24282994300320.0, + "grad_norm": 2.6452468061767656, + "learning_rate": 1.906557431387895e-06, + "loss": 0.8055, + "num_input_tokens_seen": 95128215, + "step": 4403 + }, + { + "epoch": 0.5295496903745566, + "flos": 18848924665440.0, + "grad_norm": 2.7839788700979056, + "learning_rate": 1.905779317512516e-06, + "loss": 0.7896, + "num_input_tokens_seen": 95145760, + "step": 4404 + }, + { + "epoch": 0.5296699332651957, + "flos": 20893248019200.0, + "grad_norm": 2.533299593533486, + "learning_rate": 1.9050012179303385e-06, + "loss": 0.807, + "num_input_tokens_seen": 95164270, + "step": 4405 + }, + { + "epoch": 0.5297901761558348, + "flos": 22014986099040.0, + "grad_norm": 3.9385324971232616, + "learning_rate": 1.904223132759401e-06, + "loss": 0.6929, + "num_input_tokens_seen": 95182870, + "step": 4406 + }, + { + "epoch": 0.5299104190464738, + "flos": 21767358192480.0, + "grad_norm": 3.079573701858857, + "learning_rate": 1.9034450621177383e-06, + "loss": 0.6901, + "num_input_tokens_seen": 95201190, + "step": 4407 + }, + { + "epoch": 0.530030661937113, + "flos": 14699977140000.0, + "grad_norm": 2.802493881913108, + "learning_rate": 1.9026670061233824e-06, + "loss": 0.6991, + "num_input_tokens_seen": 95218625, + "step": 4408 + }, + { + "epoch": 0.5301509048277521, + "flos": 21221174074560.0, + "grad_norm": 2.3501406201434505, + "learning_rate": 1.901888964894365e-06, + "loss": 0.8016, + "num_input_tokens_seen": 95237665, + "step": 4409 + }, + { + "epoch": 0.5302711477183911, + "flos": 25919836845120.0, + "grad_norm": 2.0923024197982363, + "learning_rate": 1.9011109385487134e-06, + "loss": 0.6732, + "num_input_tokens_seen": 95258915, + "step": 4410 + }, + { + "epoch": 0.5303913906090303, + "flos": 22241309999520.0, + "grad_norm": 4.484129789340555, + "learning_rate": 1.900332927204454e-06, + "loss": 0.6607, + "num_input_tokens_seen": 95277365, + "step": 4411 + }, + { + "epoch": 0.5305116334996693, + "flos": 24900295636320.0, + "grad_norm": 2.5866292621752747, + "learning_rate": 1.8995549309796097e-06, + "loss": 0.7646, + "num_input_tokens_seen": 95296670, + "step": 4412 + }, + { + "epoch": 0.5306318763903084, + "flos": 20160405868320.0, + "grad_norm": 3.9257440821468155, + "learning_rate": 1.8987769499922028e-06, + "loss": 0.7647, + "num_input_tokens_seen": 95315640, + "step": 4413 + }, + { + "epoch": 0.5307521192809476, + "flos": 20237470248000.0, + "grad_norm": 2.8658038134513646, + "learning_rate": 1.897998984360252e-06, + "loss": 0.7085, + "num_input_tokens_seen": 95334610, + "step": 4414 + }, + { + "epoch": 0.5308723621715866, + "flos": 28803027706080.0, + "grad_norm": 1.7737960789013367, + "learning_rate": 1.897221034201775e-06, + "loss": 0.7863, + "num_input_tokens_seen": 95358350, + "step": 4415 + }, + { + "epoch": 0.5309926050622257, + "flos": 27418533627360.0, + "grad_norm": 3.7442454371659175, + "learning_rate": 1.8964430996347842e-06, + "loss": 0.6692, + "num_input_tokens_seen": 95379900, + "step": 4416 + }, + { + "epoch": 0.5311128479528648, + "flos": 20485023815040.0, + "grad_norm": 2.6347646296154994, + "learning_rate": 1.8956651807772931e-06, + "loss": 0.8215, + "num_input_tokens_seen": 95397935, + "step": 4417 + }, + { + "epoch": 0.5312330908435039, + "flos": 21367125486720.0, + "grad_norm": 2.130383124482561, + "learning_rate": 1.8948872777473115e-06, + "loss": 0.841, + "num_input_tokens_seen": 95415890, + "step": 4418 + }, + { + "epoch": 0.531353333734143, + "flos": 24681963234240.0, + "grad_norm": 3.1018767305792445, + "learning_rate": 1.8941093906628458e-06, + "loss": 0.6344, + "num_input_tokens_seen": 95433675, + "step": 4419 + }, + { + "epoch": 0.531473576624782, + "flos": 30436859500320.0, + "grad_norm": 2.3104663876777924, + "learning_rate": 1.893331519641902e-06, + "loss": 0.71, + "num_input_tokens_seen": 95455820, + "step": 4420 + }, + { + "epoch": 0.5315938195154212, + "flos": 22970100646560.0, + "grad_norm": 3.0408690116491868, + "learning_rate": 1.8925536648024815e-06, + "loss": 0.7339, + "num_input_tokens_seen": 95473240, + "step": 4421 + }, + { + "epoch": 0.5317140624060602, + "flos": 22716116711040.0, + "grad_norm": 2.5731611148578235, + "learning_rate": 1.8917758262625849e-06, + "loss": 0.7572, + "num_input_tokens_seen": 95493480, + "step": 4422 + }, + { + "epoch": 0.5318343052966993, + "flos": 22788497700960.0, + "grad_norm": 12.009028535123974, + "learning_rate": 1.8909980041402089e-06, + "loss": 0.8138, + "num_input_tokens_seen": 95512075, + "step": 4423 + }, + { + "epoch": 0.5319545481873384, + "flos": 13606865508960.0, + "grad_norm": 3.952911733068877, + "learning_rate": 1.8902201985533494e-06, + "loss": 0.6521, + "num_input_tokens_seen": 95529340, + "step": 4424 + }, + { + "epoch": 0.5320747910779775, + "flos": 22130824272000.0, + "grad_norm": 2.133964478988113, + "learning_rate": 1.8894424096199983e-06, + "loss": 0.7506, + "num_input_tokens_seen": 95548545, + "step": 4425 + }, + { + "epoch": 0.5321950339686166, + "flos": 18561630891360.0, + "grad_norm": 2.734843709828315, + "learning_rate": 1.8886646374581463e-06, + "loss": 0.861, + "num_input_tokens_seen": 95567770, + "step": 4426 + }, + { + "epoch": 0.5323152768592557, + "flos": 22824149234400.0, + "grad_norm": 2.283124910945144, + "learning_rate": 1.8878868821857795e-06, + "loss": 0.7138, + "num_input_tokens_seen": 95587420, + "step": 4427 + }, + { + "epoch": 0.5324355197498948, + "flos": 33899878845600.0, + "grad_norm": 5.170224390233109, + "learning_rate": 1.8871091439208838e-06, + "loss": 0.7486, + "num_input_tokens_seen": 95609030, + "step": 4428 + }, + { + "epoch": 0.5325557626405338, + "flos": 23223415526400.0, + "grad_norm": 5.103968194748848, + "learning_rate": 1.8863314227814414e-06, + "loss": 0.775, + "num_input_tokens_seen": 95627340, + "step": 4429 + }, + { + "epoch": 0.532676005531173, + "flos": 26680190352000.0, + "grad_norm": 4.749798820652294, + "learning_rate": 1.8855537188854313e-06, + "loss": 0.4853, + "num_input_tokens_seen": 95646315, + "step": 4430 + }, + { + "epoch": 0.5327962484218121, + "flos": 17869160833440.0, + "grad_norm": 2.7180309002363994, + "learning_rate": 1.8847760323508315e-06, + "loss": 0.7832, + "num_input_tokens_seen": 95665220, + "step": 4431 + }, + { + "epoch": 0.5329164913124511, + "flos": 17899682940000.0, + "grad_norm": 2.234496640182158, + "learning_rate": 1.883998363295616e-06, + "loss": 0.7555, + "num_input_tokens_seen": 95682700, + "step": 4432 + }, + { + "epoch": 0.5330367342030903, + "flos": 57172438528320.0, + "grad_norm": 0.8812355540829806, + "learning_rate": 1.8832207118377565e-06, + "loss": 0.6259, + "num_input_tokens_seen": 95738070, + "step": 4433 + }, + { + "epoch": 0.5331569770937293, + "flos": 17395766572800.0, + "grad_norm": 2.677079068389543, + "learning_rate": 1.882443078095222e-06, + "loss": 0.7007, + "num_input_tokens_seen": 95756950, + "step": 4434 + }, + { + "epoch": 0.5332772199843684, + "flos": 56670046121280.0, + "grad_norm": 0.8674111636295057, + "learning_rate": 1.8816654621859794e-06, + "loss": 0.667, + "num_input_tokens_seen": 95816615, + "step": 4435 + }, + { + "epoch": 0.5333974628750076, + "flos": 18671150205120.0, + "grad_norm": 3.5412510048285504, + "learning_rate": 1.8808878642279915e-06, + "loss": 0.725, + "num_input_tokens_seen": 95832020, + "step": 4436 + }, + { + "epoch": 0.5335177057656466, + "flos": 23771458132320.0, + "grad_norm": 2.757634403000215, + "learning_rate": 1.8801102843392209e-06, + "loss": 0.659, + "num_input_tokens_seen": 95851425, + "step": 4437 + }, + { + "epoch": 0.5336379486562857, + "flos": 25042641581760.0, + "grad_norm": 4.45270116027122, + "learning_rate": 1.8793327226376238e-06, + "loss": 0.8506, + "num_input_tokens_seen": 95870745, + "step": 4438 + }, + { + "epoch": 0.5337581915469248, + "flos": 21366121903200.0, + "grad_norm": 2.640761332406601, + "learning_rate": 1.8785551792411569e-06, + "loss": 0.8008, + "num_input_tokens_seen": 95889755, + "step": 4439 + }, + { + "epoch": 0.5338784344375639, + "flos": 14844850629120.0, + "grad_norm": 3.7077161439640207, + "learning_rate": 1.8777776542677733e-06, + "loss": 0.8264, + "num_input_tokens_seen": 95907805, + "step": 4440 + }, + { + "epoch": 0.5339986773282029, + "flos": 20783803044960.0, + "grad_norm": 3.2901258765724157, + "learning_rate": 1.8770001478354216e-06, + "loss": 0.7313, + "num_input_tokens_seen": 95927480, + "step": 4441 + }, + { + "epoch": 0.5341189202188421, + "flos": 17943586160160.0, + "grad_norm": 9.503063711754645, + "learning_rate": 1.8762226600620504e-06, + "loss": 0.8363, + "num_input_tokens_seen": 95946095, + "step": 4442 + }, + { + "epoch": 0.5342391631094812, + "flos": 11016283697760.0, + "grad_norm": 3.6949018770437525, + "learning_rate": 1.8754451910656031e-06, + "loss": 0.5923, + "num_input_tokens_seen": 95959995, + "step": 4443 + }, + { + "epoch": 0.5343594060001202, + "flos": 15317352815520.0, + "grad_norm": 2.663298197989169, + "learning_rate": 1.8746677409640212e-06, + "loss": 0.8267, + "num_input_tokens_seen": 95977095, + "step": 4444 + }, + { + "epoch": 0.5344796488907594, + "flos": 26862127825440.0, + "grad_norm": 3.2566047460577554, + "learning_rate": 1.8738903098752432e-06, + "loss": 0.8471, + "num_input_tokens_seen": 95996225, + "step": 4445 + }, + { + "epoch": 0.5345998917813984, + "flos": 25374842159520.0, + "grad_norm": 2.7922283765194824, + "learning_rate": 1.8731128979172052e-06, + "loss": 0.7297, + "num_input_tokens_seen": 96015425, + "step": 4446 + }, + { + "epoch": 0.5347201346720375, + "flos": 32806135328640.0, + "grad_norm": 3.538621348660285, + "learning_rate": 1.8723355052078394e-06, + "loss": 0.6706, + "num_input_tokens_seen": 96035460, + "step": 4447 + }, + { + "epoch": 0.5348403775626767, + "flos": 17942396727840.0, + "grad_norm": 2.8233051009344785, + "learning_rate": 1.8715581318650765e-06, + "loss": 0.7739, + "num_input_tokens_seen": 96054515, + "step": 4448 + }, + { + "epoch": 0.5349606204533157, + "flos": 17578335932160.0, + "grad_norm": 5.883941987768887, + "learning_rate": 1.8707807780068422e-06, + "loss": 0.818, + "num_input_tokens_seen": 96072330, + "step": 4449 + }, + { + "epoch": 0.5350808633439548, + "flos": 29125935843840.0, + "grad_norm": 3.772278159402068, + "learning_rate": 1.8700034437510611e-06, + "loss": 0.6622, + "num_input_tokens_seen": 96092425, + "step": 4450 + }, + { + "epoch": 0.5352011062345938, + "flos": 19472210332800.0, + "grad_norm": 2.791673992394092, + "learning_rate": 1.8692261292156549e-06, + "loss": 0.8147, + "num_input_tokens_seen": 96111660, + "step": 4451 + }, + { + "epoch": 0.535321349125233, + "flos": 23444089623360.0, + "grad_norm": 2.3644999238706124, + "learning_rate": 1.8684488345185401e-06, + "loss": 0.8139, + "num_input_tokens_seen": 96131835, + "step": 4452 + }, + { + "epoch": 0.535441592015872, + "flos": 20448591716640.0, + "grad_norm": 3.056114390354385, + "learning_rate": 1.8676715597776332e-06, + "loss": 0.7892, + "num_input_tokens_seen": 96150180, + "step": 4453 + }, + { + "epoch": 0.5355618349065111, + "flos": 19545260378400.0, + "grad_norm": 2.4146176783105164, + "learning_rate": 1.8668943051108455e-06, + "loss": 0.7604, + "num_input_tokens_seen": 96167485, + "step": 4454 + }, + { + "epoch": 0.5356820777971503, + "flos": 23990273741280.0, + "grad_norm": 2.255866667500388, + "learning_rate": 1.8661170706360856e-06, + "loss": 0.7625, + "num_input_tokens_seen": 96186650, + "step": 4455 + }, + { + "epoch": 0.5358023206877893, + "flos": 20854808753760.0, + "grad_norm": 1.971803688658137, + "learning_rate": 1.8653398564712594e-06, + "loss": 0.8148, + "num_input_tokens_seen": 96205595, + "step": 4456 + }, + { + "epoch": 0.5359225635784284, + "flos": 22386964053600.0, + "grad_norm": 2.5693082847832525, + "learning_rate": 1.8645626627342704e-06, + "loss": 0.8227, + "num_input_tokens_seen": 96226360, + "step": 4457 + }, + { + "epoch": 0.5360428064690675, + "flos": 24063323786880.0, + "grad_norm": 3.162813820930524, + "learning_rate": 1.8637854895430172e-06, + "loss": 0.8126, + "num_input_tokens_seen": 96245625, + "step": 4458 + }, + { + "epoch": 0.5361630493597066, + "flos": 21403929282720.0, + "grad_norm": 2.582907939054518, + "learning_rate": 1.8630083370153978e-06, + "loss": 0.6979, + "num_input_tokens_seen": 96265265, + "step": 4459 + }, + { + "epoch": 0.5362832922503457, + "flos": 68789854716480.0, + "grad_norm": 0.7537114183157964, + "learning_rate": 1.8622312052693041e-06, + "loss": 0.5532, + "num_input_tokens_seen": 96326445, + "step": 4460 + }, + { + "epoch": 0.5364035351409848, + "flos": 9780008386560.0, + "grad_norm": 3.576642658073664, + "learning_rate": 1.8614540944226267e-06, + "loss": 0.7194, + "num_input_tokens_seen": 96343115, + "step": 4461 + }, + { + "epoch": 0.5365237780316239, + "flos": 23256613855680.0, + "grad_norm": 2.6118426504365932, + "learning_rate": 1.8606770045932537e-06, + "loss": 0.6776, + "num_input_tokens_seen": 96362230, + "step": 4462 + }, + { + "epoch": 0.5366440209222629, + "flos": 26540037422400.0, + "grad_norm": 2.5527808864474495, + "learning_rate": 1.859899935899068e-06, + "loss": 0.8171, + "num_input_tokens_seen": 96382085, + "step": 4463 + }, + { + "epoch": 0.5367642638129021, + "flos": 19580057007360.0, + "grad_norm": 4.3115826059431, + "learning_rate": 1.8591228884579506e-06, + "loss": 0.7893, + "num_input_tokens_seen": 96400580, + "step": 4464 + }, + { + "epoch": 0.5368845067035412, + "flos": 23881014615840.0, + "grad_norm": 2.9533865676571325, + "learning_rate": 1.8583458623877795e-06, + "loss": 0.8209, + "num_input_tokens_seen": 96419680, + "step": 4465 + }, + { + "epoch": 0.5370047495941802, + "flos": 16849173587520.0, + "grad_norm": 2.5396884557441273, + "learning_rate": 1.8575688578064281e-06, + "loss": 0.7433, + "num_input_tokens_seen": 96437805, + "step": 4466 + }, + { + "epoch": 0.5371249924848194, + "flos": 20711347715520.0, + "grad_norm": 2.913013422821475, + "learning_rate": 1.8567918748317674e-06, + "loss": 0.7666, + "num_input_tokens_seen": 96457155, + "step": 4467 + }, + { + "epoch": 0.5372452353754584, + "flos": 17942582576640.0, + "grad_norm": 2.902094669872158, + "learning_rate": 1.8560149135816659e-06, + "loss": 0.8277, + "num_input_tokens_seen": 96473990, + "step": 4468 + }, + { + "epoch": 0.5373654782660975, + "flos": 15355234534560.0, + "grad_norm": 3.7269491638859833, + "learning_rate": 1.8552379741739873e-06, + "loss": 0.8421, + "num_input_tokens_seen": 96491335, + "step": 4469 + }, + { + "epoch": 0.5374857211567367, + "flos": 68901158178720.0, + "grad_norm": 0.9107582491727026, + "learning_rate": 1.8544610567265935e-06, + "loss": 0.5558, + "num_input_tokens_seen": 96545275, + "step": 4470 + }, + { + "epoch": 0.5376059640473757, + "flos": 15064000765920.0, + "grad_norm": 2.7571047125887693, + "learning_rate": 1.853684161357341e-06, + "loss": 0.8326, + "num_input_tokens_seen": 96562935, + "step": 4471 + }, + { + "epoch": 0.5377262069380148, + "flos": 19764410515200.0, + "grad_norm": 26.762295473004087, + "learning_rate": 1.852907288184085e-06, + "loss": 0.7647, + "num_input_tokens_seen": 96581820, + "step": 4472 + }, + { + "epoch": 0.5378464498286539, + "flos": 29959822603200.0, + "grad_norm": 2.7007648550280234, + "learning_rate": 1.8521304373246762e-06, + "loss": 0.7, + "num_input_tokens_seen": 96602865, + "step": 4473 + }, + { + "epoch": 0.537966692719293, + "flos": 21220876716480.0, + "grad_norm": 3.7921826916237027, + "learning_rate": 1.8513536088969626e-06, + "loss": 0.8901, + "num_input_tokens_seen": 96620530, + "step": 4474 + }, + { + "epoch": 0.538086935609932, + "flos": 21512407843200.0, + "grad_norm": 4.152619116795424, + "learning_rate": 1.8505768030187884e-06, + "loss": 0.805, + "num_input_tokens_seen": 96640660, + "step": 4475 + }, + { + "epoch": 0.5382071785005712, + "flos": 22714704260160.0, + "grad_norm": 1.7721096226601591, + "learning_rate": 1.849800019807995e-06, + "loss": 0.8007, + "num_input_tokens_seen": 96661885, + "step": 4476 + }, + { + "epoch": 0.5383274213912103, + "flos": 24898808845920.0, + "grad_norm": 3.147485631738762, + "learning_rate": 1.8490232593824186e-06, + "loss": 0.7109, + "num_input_tokens_seen": 96678340, + "step": 4477 + }, + { + "epoch": 0.5384476642818493, + "flos": 22278336814080.0, + "grad_norm": 2.110874236632111, + "learning_rate": 1.8482465218598935e-06, + "loss": 0.8474, + "num_input_tokens_seen": 96698285, + "step": 4478 + }, + { + "epoch": 0.5385679071724885, + "flos": 22678569519840.0, + "grad_norm": 2.5017027778060816, + "learning_rate": 1.8474698073582508e-06, + "loss": 0.8334, + "num_input_tokens_seen": 96719655, + "step": 4479 + }, + { + "epoch": 0.5386881500631275, + "flos": 15930862836000.0, + "grad_norm": 3.0390681484925013, + "learning_rate": 1.8466931159953166e-06, + "loss": 0.8756, + "num_input_tokens_seen": 96736290, + "step": 4480 + }, + { + "epoch": 0.5388083929537666, + "flos": 24025590746880.0, + "grad_norm": 2.7186026697688317, + "learning_rate": 1.8459164478889158e-06, + "loss": 0.8427, + "num_input_tokens_seen": 96757685, + "step": 4481 + }, + { + "epoch": 0.5389286358444056, + "flos": 22205138089440.0, + "grad_norm": 4.49811980916771, + "learning_rate": 1.8451398031568663e-06, + "loss": 0.7577, + "num_input_tokens_seen": 96777310, + "step": 4482 + }, + { + "epoch": 0.5390488787350448, + "flos": 24247045408800.0, + "grad_norm": 2.323461482957457, + "learning_rate": 1.844363181916986e-06, + "loss": 0.7481, + "num_input_tokens_seen": 96798035, + "step": 4483 + }, + { + "epoch": 0.5391691216256839, + "flos": 16557865479360.0, + "grad_norm": 2.6990264514928626, + "learning_rate": 1.8435865842870868e-06, + "loss": 0.829, + "num_input_tokens_seen": 96815975, + "step": 4484 + }, + { + "epoch": 0.5392893645163229, + "flos": 23296651420800.0, + "grad_norm": 2.6685326623974426, + "learning_rate": 1.8428100103849787e-06, + "loss": 0.7207, + "num_input_tokens_seen": 96835770, + "step": 4485 + }, + { + "epoch": 0.5394096074069621, + "flos": 15647620565760.0, + "grad_norm": 3.473170582429449, + "learning_rate": 1.842033460328467e-06, + "loss": 0.7304, + "num_input_tokens_seen": 96854490, + "step": 4486 + }, + { + "epoch": 0.5395298502976011, + "flos": 22860730011840.0, + "grad_norm": 2.244778508430264, + "learning_rate": 1.8412569342353541e-06, + "loss": 0.7513, + "num_input_tokens_seen": 96874645, + "step": 4487 + }, + { + "epoch": 0.5396500931882402, + "flos": 23808038909760.0, + "grad_norm": 2.779507466190396, + "learning_rate": 1.840480432223438e-06, + "loss": 0.8439, + "num_input_tokens_seen": 96893650, + "step": 4488 + }, + { + "epoch": 0.5397703360788794, + "flos": 26285272921920.0, + "grad_norm": 2.49391062130798, + "learning_rate": 1.8397039544105131e-06, + "loss": 0.7802, + "num_input_tokens_seen": 96912735, + "step": 4489 + }, + { + "epoch": 0.5398905789695184, + "flos": 21184295939040.0, + "grad_norm": 2.2515695183966886, + "learning_rate": 1.8389275009143711e-06, + "loss": 0.6976, + "num_input_tokens_seen": 96932475, + "step": 4490 + }, + { + "epoch": 0.5400108218601575, + "flos": 25044351390720.0, + "grad_norm": 2.3835255040764225, + "learning_rate": 1.8381510718527988e-06, + "loss": 0.7358, + "num_input_tokens_seen": 96952085, + "step": 4491 + }, + { + "epoch": 0.5401310647507966, + "flos": 26319660683520.0, + "grad_norm": 2.7310812158175297, + "learning_rate": 1.8373746673435812e-06, + "loss": 0.6309, + "num_input_tokens_seen": 96973110, + "step": 4492 + }, + { + "epoch": 0.5402513076414357, + "flos": 27815383884960.0, + "grad_norm": 2.3749125615136246, + "learning_rate": 1.8365982875044964e-06, + "loss": 0.7899, + "num_input_tokens_seen": 96993420, + "step": 4493 + }, + { + "epoch": 0.5403715505320748, + "flos": 22861064539680.0, + "grad_norm": 8.177547550185622, + "learning_rate": 1.8358219324533217e-06, + "loss": 0.7602, + "num_input_tokens_seen": 97013685, + "step": 4494 + }, + { + "epoch": 0.5404917934227139, + "flos": 30180645379200.0, + "grad_norm": 2.4317048458826194, + "learning_rate": 1.8350456023078292e-06, + "loss": 0.7018, + "num_input_tokens_seen": 97036060, + "step": 4495 + }, + { + "epoch": 0.540612036313353, + "flos": 19910101739040.0, + "grad_norm": 3.119401418046286, + "learning_rate": 1.8342692971857874e-06, + "loss": 0.7801, + "num_input_tokens_seen": 97055260, + "step": 4496 + }, + { + "epoch": 0.540732279203992, + "flos": 24244852392960.0, + "grad_norm": 4.799670669066175, + "learning_rate": 1.833493017204962e-06, + "loss": 0.7136, + "num_input_tokens_seen": 97075365, + "step": 4497 + }, + { + "epoch": 0.5408525220946312, + "flos": 20164420202400.0, + "grad_norm": 7.560183840337947, + "learning_rate": 1.8327167624831134e-06, + "loss": 0.7775, + "num_input_tokens_seen": 97093095, + "step": 4498 + }, + { + "epoch": 0.5409727649852702, + "flos": 24099941734080.0, + "grad_norm": 2.7623544093643324, + "learning_rate": 1.831940533137999e-06, + "loss": 0.7106, + "num_input_tokens_seen": 97112000, + "step": 4499 + }, + { + "epoch": 0.5410930078759093, + "flos": 23878412732640.0, + "grad_norm": 2.3419594600892255, + "learning_rate": 1.8311643292873718e-06, + "loss": 0.7237, + "num_input_tokens_seen": 97130820, + "step": 4500 + }, + { + "epoch": 0.5412132507665485, + "flos": 21074999643840.0, + "grad_norm": 3.9759702110249164, + "learning_rate": 1.8303881510489818e-06, + "loss": 0.8778, + "num_input_tokens_seen": 97149210, + "step": 4501 + }, + { + "epoch": 0.5413334936571875, + "flos": 30183953487840.0, + "grad_norm": 2.717968707894689, + "learning_rate": 1.829611998540574e-06, + "loss": 0.6941, + "num_input_tokens_seen": 97170415, + "step": 4502 + }, + { + "epoch": 0.5414537365478266, + "flos": 24244889562720.0, + "grad_norm": 4.342617628297544, + "learning_rate": 1.8288358718798914e-06, + "loss": 0.8007, + "num_input_tokens_seen": 97189605, + "step": 4503 + }, + { + "epoch": 0.5415739794384657, + "flos": 16630915524960.0, + "grad_norm": 2.199456806748197, + "learning_rate": 1.8280597711846703e-06, + "loss": 0.723, + "num_input_tokens_seen": 97207845, + "step": 4504 + }, + { + "epoch": 0.5416942223291048, + "flos": 23152595666400.0, + "grad_norm": 2.2008434412804885, + "learning_rate": 1.8272836965726455e-06, + "loss": 0.8328, + "num_input_tokens_seen": 97226780, + "step": 4505 + }, + { + "epoch": 0.5418144652197439, + "flos": 20274311213760.0, + "grad_norm": 2.9146629717083004, + "learning_rate": 1.8265076481615461e-06, + "loss": 0.7829, + "num_input_tokens_seen": 97246695, + "step": 4506 + }, + { + "epoch": 0.541934708110383, + "flos": 12441112699680.0, + "grad_norm": 3.367005239275852, + "learning_rate": 1.8257316260690987e-06, + "loss": 0.8704, + "num_input_tokens_seen": 97264480, + "step": 4507 + }, + { + "epoch": 0.5420549510010221, + "flos": 21223552939200.0, + "grad_norm": 1.8697489004814096, + "learning_rate": 1.8249556304130254e-06, + "loss": 0.7592, + "num_input_tokens_seen": 97285760, + "step": 4508 + }, + { + "epoch": 0.5421751938916611, + "flos": 29447951907360.0, + "grad_norm": 2.3724543330115635, + "learning_rate": 1.824179661311044e-06, + "loss": 0.6827, + "num_input_tokens_seen": 97304510, + "step": 4509 + }, + { + "epoch": 0.5422954367823003, + "flos": 18707582303520.0, + "grad_norm": 2.5381422047617854, + "learning_rate": 1.823403718880868e-06, + "loss": 0.7979, + "num_input_tokens_seen": 97323270, + "step": 4510 + }, + { + "epoch": 0.5424156796729394, + "flos": 39911435269920.0, + "grad_norm": 2.8816466416353923, + "learning_rate": 1.822627803240207e-06, + "loss": 0.6639, + "num_input_tokens_seen": 97345600, + "step": 4511 + }, + { + "epoch": 0.5425359225635784, + "flos": 11530347409440.0, + "grad_norm": 2.5993026678574505, + "learning_rate": 1.8218519145067675e-06, + "loss": 0.8498, + "num_input_tokens_seen": 97361220, + "step": 4512 + }, + { + "epoch": 0.5426561654542175, + "flos": 20200852300800.0, + "grad_norm": 3.1269519050591, + "learning_rate": 1.8210760527982508e-06, + "loss": 0.8931, + "num_input_tokens_seen": 97381505, + "step": 4513 + }, + { + "epoch": 0.5427764083448566, + "flos": 21840445407840.0, + "grad_norm": 2.7590165144069037, + "learning_rate": 1.8203002182323552e-06, + "loss": 0.751, + "num_input_tokens_seen": 97399060, + "step": 4514 + }, + { + "epoch": 0.5428966512354957, + "flos": 19612772129760.0, + "grad_norm": 2.2515680359369292, + "learning_rate": 1.819524410926773e-06, + "loss": 0.7583, + "num_input_tokens_seen": 97416740, + "step": 4515 + }, + { + "epoch": 0.5430168941261347, + "flos": 22350197427360.0, + "grad_norm": 2.4083563762825184, + "learning_rate": 1.8187486309991944e-06, + "loss": 0.7679, + "num_input_tokens_seen": 97437620, + "step": 4516 + }, + { + "epoch": 0.5431371370167739, + "flos": 18737249505600.0, + "grad_norm": 6.223921116248124, + "learning_rate": 1.817972878567304e-06, + "loss": 0.7755, + "num_input_tokens_seen": 97456275, + "step": 4517 + }, + { + "epoch": 0.543257379907413, + "flos": 18779219898240.0, + "grad_norm": 2.8375557310034094, + "learning_rate": 1.8171971537487834e-06, + "loss": 0.7655, + "num_input_tokens_seen": 97474925, + "step": 4518 + }, + { + "epoch": 0.543377622798052, + "flos": 17468519260320.0, + "grad_norm": 4.783670049816125, + "learning_rate": 1.8164214566613093e-06, + "loss": 0.8087, + "num_input_tokens_seen": 97493550, + "step": 4519 + }, + { + "epoch": 0.5434978656886912, + "flos": 18962272464480.0, + "grad_norm": 4.905390767466234, + "learning_rate": 1.8156457874225547e-06, + "loss": 0.6641, + "num_input_tokens_seen": 97512010, + "step": 4520 + }, + { + "epoch": 0.5436181085793302, + "flos": 17250409876800.0, + "grad_norm": 2.3740269524796584, + "learning_rate": 1.814870146150187e-06, + "loss": 0.8046, + "num_input_tokens_seen": 97530120, + "step": 4521 + }, + { + "epoch": 0.5437383514699693, + "flos": 19071382910880.0, + "grad_norm": 3.2170903640022686, + "learning_rate": 1.814094532961871e-06, + "loss": 0.7874, + "num_input_tokens_seen": 97548695, + "step": 4522 + }, + { + "epoch": 0.5438585943606085, + "flos": 22570462656960.0, + "grad_norm": 2.966722940775599, + "learning_rate": 1.8133189479752666e-06, + "loss": 0.8403, + "num_input_tokens_seen": 97567625, + "step": 4523 + }, + { + "epoch": 0.5439788372512475, + "flos": 21767358192480.0, + "grad_norm": 2.561506962841369, + "learning_rate": 1.8125433913080292e-06, + "loss": 0.8191, + "num_input_tokens_seen": 97585325, + "step": 4524 + }, + { + "epoch": 0.5440990801418866, + "flos": 16376113854720.0, + "grad_norm": 3.7178176865225288, + "learning_rate": 1.811767863077811e-06, + "loss": 0.8232, + "num_input_tokens_seen": 97604275, + "step": 4525 + }, + { + "epoch": 0.5442193230325257, + "flos": 21584751663360.0, + "grad_norm": 3.2197292282746015, + "learning_rate": 1.8109923634022577e-06, + "loss": 0.7811, + "num_input_tokens_seen": 97624300, + "step": 4526 + }, + { + "epoch": 0.5443395659231648, + "flos": 15459066875040.0, + "grad_norm": 2.727090866354341, + "learning_rate": 1.8102168923990128e-06, + "loss": 0.8619, + "num_input_tokens_seen": 97637845, + "step": 4527 + }, + { + "epoch": 0.5444598088138038, + "flos": 18743717043840.0, + "grad_norm": 2.2917169276419, + "learning_rate": 1.809441450185714e-06, + "loss": 0.7995, + "num_input_tokens_seen": 97656330, + "step": 4528 + }, + { + "epoch": 0.544580051704443, + "flos": 21112175137440.0, + "grad_norm": 2.6073090375280206, + "learning_rate": 1.8086660368799958e-06, + "loss": 0.7325, + "num_input_tokens_seen": 97674380, + "step": 4529 + }, + { + "epoch": 0.5447002945950821, + "flos": 32444899434720.0, + "grad_norm": 3.339494939067745, + "learning_rate": 1.807890652599488e-06, + "loss": 0.7724, + "num_input_tokens_seen": 97698765, + "step": 4530 + }, + { + "epoch": 0.5448205374857211, + "flos": 11781617952480.0, + "grad_norm": 2.6259143690048132, + "learning_rate": 1.8071152974618156e-06, + "loss": 0.8239, + "num_input_tokens_seen": 97716565, + "step": 4531 + }, + { + "epoch": 0.5449407803763603, + "flos": 24099235508640.0, + "grad_norm": 2.6015686401661307, + "learning_rate": 1.806339971584599e-06, + "loss": 0.7789, + "num_input_tokens_seen": 97733225, + "step": 4532 + }, + { + "epoch": 0.5450610232669993, + "flos": 23225311184160.0, + "grad_norm": 6.1592087195338046, + "learning_rate": 1.8055646750854546e-06, + "loss": 0.8508, + "num_input_tokens_seen": 97752735, + "step": 4533 + }, + { + "epoch": 0.5451812661576384, + "flos": 17759790198720.0, + "grad_norm": 4.33526573132278, + "learning_rate": 1.8047894080819945e-06, + "loss": 0.819, + "num_input_tokens_seen": 97769985, + "step": 4534 + }, + { + "epoch": 0.5453015090482776, + "flos": 70961353020000.0, + "grad_norm": 0.7289974470420897, + "learning_rate": 1.8040141706918258e-06, + "loss": 0.6315, + "num_input_tokens_seen": 97831225, + "step": 4535 + }, + { + "epoch": 0.5454217519389166, + "flos": 25515441126240.0, + "grad_norm": 2.4471514432411468, + "learning_rate": 1.8032389630325525e-06, + "loss": 0.7716, + "num_input_tokens_seen": 97849975, + "step": 4536 + }, + { + "epoch": 0.5455419948295557, + "flos": 23624465966880.0, + "grad_norm": 4.438394805524519, + "learning_rate": 1.8024637852217707e-06, + "loss": 0.7615, + "num_input_tokens_seen": 97869700, + "step": 4537 + }, + { + "epoch": 0.5456622377201948, + "flos": 23370519201120.0, + "grad_norm": 2.304516905597, + "learning_rate": 1.8016886373770766e-06, + "loss": 0.8493, + "num_input_tokens_seen": 97888215, + "step": 4538 + }, + { + "epoch": 0.5457824806108339, + "flos": 23953730133600.0, + "grad_norm": 1.9203556537047242, + "learning_rate": 1.8009135196160579e-06, + "loss": 0.7914, + "num_input_tokens_seen": 97907090, + "step": 4539 + }, + { + "epoch": 0.545902723501473, + "flos": 22533658860960.0, + "grad_norm": 2.1199328651234723, + "learning_rate": 1.8001384320563e-06, + "loss": 0.8402, + "num_input_tokens_seen": 97927345, + "step": 4540 + }, + { + "epoch": 0.5460229663921121, + "flos": 55119194432160.0, + "grad_norm": 0.7917014457357933, + "learning_rate": 1.7993633748153833e-06, + "loss": 0.5769, + "num_input_tokens_seen": 97981950, + "step": 4541 + }, + { + "epoch": 0.5461432092827512, + "flos": 15392038330560.0, + "grad_norm": 2.4671755248655507, + "learning_rate": 1.7985883480108834e-06, + "loss": 0.7297, + "num_input_tokens_seen": 97999585, + "step": 4542 + }, + { + "epoch": 0.5462634521733902, + "flos": 23990571099360.0, + "grad_norm": 2.0509309841170493, + "learning_rate": 1.797813351760371e-06, + "loss": 0.7198, + "num_input_tokens_seen": 98021290, + "step": 4543 + }, + { + "epoch": 0.5463836950640293, + "flos": 22788349021920.0, + "grad_norm": 2.701059599754471, + "learning_rate": 1.7970383861814116e-06, + "loss": 0.7801, + "num_input_tokens_seen": 98041060, + "step": 4544 + }, + { + "epoch": 0.5465039379546685, + "flos": 20419110363360.0, + "grad_norm": 3.2486064564196457, + "learning_rate": 1.7962634513915684e-06, + "loss": 0.7418, + "num_input_tokens_seen": 98058410, + "step": 4545 + }, + { + "epoch": 0.5466241808453075, + "flos": 17318516344320.0, + "grad_norm": 2.16205831879222, + "learning_rate": 1.7954885475083969e-06, + "loss": 0.7942, + "num_input_tokens_seen": 98076080, + "step": 4546 + }, + { + "epoch": 0.5467444237359466, + "flos": 21586201284000.0, + "grad_norm": 2.941157721852402, + "learning_rate": 1.7947136746494513e-06, + "loss": 0.7306, + "num_input_tokens_seen": 98096995, + "step": 4547 + }, + { + "epoch": 0.5468646666265857, + "flos": 24135890625600.0, + "grad_norm": 3.5635979282309327, + "learning_rate": 1.793938832932277e-06, + "loss": 0.8794, + "num_input_tokens_seen": 98115105, + "step": 4548 + }, + { + "epoch": 0.5469849095172248, + "flos": 27487532169120.0, + "grad_norm": 3.3667852953667325, + "learning_rate": 1.7931640224744185e-06, + "loss": 0.7067, + "num_input_tokens_seen": 98135655, + "step": 4549 + }, + { + "epoch": 0.5471051524078638, + "flos": 27924568670880.0, + "grad_norm": 2.3984206009558364, + "learning_rate": 1.7923892433934127e-06, + "loss": 0.7348, + "num_input_tokens_seen": 98156730, + "step": 4550 + }, + { + "epoch": 0.547225395298503, + "flos": 18124705898880.0, + "grad_norm": 4.673247170916963, + "learning_rate": 1.7916144958067939e-06, + "loss": 0.7917, + "num_input_tokens_seen": 98174590, + "step": 4551 + }, + { + "epoch": 0.5473456381891421, + "flos": 21331027916160.0, + "grad_norm": 2.67197226185592, + "learning_rate": 1.7908397798320905e-06, + "loss": 0.7912, + "num_input_tokens_seen": 98194325, + "step": 4552 + }, + { + "epoch": 0.5474658810797811, + "flos": 19903522691520.0, + "grad_norm": 3.693413470279081, + "learning_rate": 1.7900650955868265e-06, + "loss": 0.7485, + "num_input_tokens_seen": 98213970, + "step": 4553 + }, + { + "epoch": 0.5475861239704203, + "flos": 50403322062720.0, + "grad_norm": 1.65660062713298, + "learning_rate": 1.7892904431885202e-06, + "loss": 0.7649, + "num_input_tokens_seen": 98241145, + "step": 4554 + }, + { + "epoch": 0.5477063668610593, + "flos": 20675621842560.0, + "grad_norm": 3.5372972949287087, + "learning_rate": 1.788515822754686e-06, + "loss": 0.7552, + "num_input_tokens_seen": 98261565, + "step": 4555 + }, + { + "epoch": 0.5478266097516984, + "flos": 19581803986080.0, + "grad_norm": 2.7818184603839993, + "learning_rate": 1.7877412344028335e-06, + "loss": 0.7836, + "num_input_tokens_seen": 98281370, + "step": 4556 + }, + { + "epoch": 0.5479468526423376, + "flos": 12878446559520.0, + "grad_norm": 3.927972923861413, + "learning_rate": 1.7869666782504668e-06, + "loss": 0.7738, + "num_input_tokens_seen": 98297950, + "step": 4557 + }, + { + "epoch": 0.5480670955329766, + "flos": 18562374286560.0, + "grad_norm": 2.944634721313965, + "learning_rate": 1.7861921544150867e-06, + "loss": 0.6921, + "num_input_tokens_seen": 98316800, + "step": 4558 + }, + { + "epoch": 0.5481873384236157, + "flos": 15931754910240.0, + "grad_norm": 6.236179482647603, + "learning_rate": 1.7854176630141856e-06, + "loss": 0.77, + "num_input_tokens_seen": 98333935, + "step": 4559 + }, + { + "epoch": 0.5483075813142548, + "flos": 22751247867840.0, + "grad_norm": 3.900899411686076, + "learning_rate": 1.784643204165255e-06, + "loss": 0.8441, + "num_input_tokens_seen": 98353255, + "step": 4560 + }, + { + "epoch": 0.5484278242048939, + "flos": 19289678143200.0, + "grad_norm": 2.274513226088038, + "learning_rate": 1.7838687779857783e-06, + "loss": 0.7723, + "num_input_tokens_seen": 98371130, + "step": 4561 + }, + { + "epoch": 0.5485480670955329, + "flos": 22783256764800.0, + "grad_norm": 2.8067532638214154, + "learning_rate": 1.7830943845932366e-06, + "loss": 0.6438, + "num_input_tokens_seen": 98390455, + "step": 4562 + }, + { + "epoch": 0.5486683099861721, + "flos": 22638792143040.0, + "grad_norm": 2.559980689214352, + "learning_rate": 1.7823200241051044e-06, + "loss": 0.7524, + "num_input_tokens_seen": 98409370, + "step": 4563 + }, + { + "epoch": 0.5487885528768112, + "flos": 23116312247040.0, + "grad_norm": 2.5060689694289966, + "learning_rate": 1.7815456966388513e-06, + "loss": 0.8075, + "num_input_tokens_seen": 98428465, + "step": 4564 + }, + { + "epoch": 0.5489087957674502, + "flos": 22022420051040.0, + "grad_norm": 2.6266369030164953, + "learning_rate": 1.780771402311943e-06, + "loss": 0.8125, + "num_input_tokens_seen": 98447300, + "step": 4565 + }, + { + "epoch": 0.5490290386580894, + "flos": 24281172982080.0, + "grad_norm": 2.6805127710893117, + "learning_rate": 1.7799971412418374e-06, + "loss": 0.7878, + "num_input_tokens_seen": 98468695, + "step": 4566 + }, + { + "epoch": 0.5491492815487284, + "flos": 18268873162560.0, + "grad_norm": 2.453126652225011, + "learning_rate": 1.7792229135459918e-06, + "loss": 0.741, + "num_input_tokens_seen": 98485345, + "step": 4567 + }, + { + "epoch": 0.5492695244393675, + "flos": 64457854624800.0, + "grad_norm": 0.7460590773909245, + "learning_rate": 1.7784487193418538e-06, + "loss": 0.6148, + "num_input_tokens_seen": 98543195, + "step": 4568 + }, + { + "epoch": 0.5493897673300067, + "flos": 17354725424160.0, + "grad_norm": 2.7391034903889246, + "learning_rate": 1.7776745587468698e-06, + "loss": 0.6092, + "num_input_tokens_seen": 98560620, + "step": 4569 + }, + { + "epoch": 0.5495100102206457, + "flos": 19873409452320.0, + "grad_norm": 3.0950391761383336, + "learning_rate": 1.7769004318784776e-06, + "loss": 0.8201, + "num_input_tokens_seen": 98578265, + "step": 4570 + }, + { + "epoch": 0.5496302531112848, + "flos": 16703445193920.0, + "grad_norm": 2.166989216881977, + "learning_rate": 1.776126338854113e-06, + "loss": 0.8082, + "num_input_tokens_seen": 98596055, + "step": 4571 + }, + { + "epoch": 0.5497504960019239, + "flos": 24537201254400.0, + "grad_norm": 1.9958561645773467, + "learning_rate": 1.7753522797912044e-06, + "loss": 0.8461, + "num_input_tokens_seen": 98616140, + "step": 4572 + }, + { + "epoch": 0.549870738892563, + "flos": 15428619108000.0, + "grad_norm": 2.8184021796493917, + "learning_rate": 1.7745782548071765e-06, + "loss": 0.7029, + "num_input_tokens_seen": 98633035, + "step": 4573 + }, + { + "epoch": 0.549990981783202, + "flos": 21039050752320.0, + "grad_norm": 1.9197762667913993, + "learning_rate": 1.7738042640194482e-06, + "loss": 0.7409, + "num_input_tokens_seen": 98652700, + "step": 4574 + }, + { + "epoch": 0.5501112246738411, + "flos": 21365304168480.0, + "grad_norm": 2.4784671426995772, + "learning_rate": 1.7730303075454335e-06, + "loss": 0.705, + "num_input_tokens_seen": 98671590, + "step": 4575 + }, + { + "epoch": 0.5502314675644803, + "flos": 17432124331680.0, + "grad_norm": 2.3660656376276843, + "learning_rate": 1.7722563855025402e-06, + "loss": 0.8471, + "num_input_tokens_seen": 98689620, + "step": 4576 + }, + { + "epoch": 0.5503517104551193, + "flos": 24275411669280.0, + "grad_norm": 9.492416567452098, + "learning_rate": 1.7714824980081721e-06, + "loss": 0.7109, + "num_input_tokens_seen": 98708915, + "step": 4577 + }, + { + "epoch": 0.5504719533457584, + "flos": 22387410090720.0, + "grad_norm": 2.7148275910040685, + "learning_rate": 1.7707086451797276e-06, + "loss": 0.7395, + "num_input_tokens_seen": 98729790, + "step": 4578 + }, + { + "epoch": 0.5505921962363975, + "flos": 67198030484640.0, + "grad_norm": 0.7146238395250448, + "learning_rate": 1.7699348271345993e-06, + "loss": 0.5227, + "num_input_tokens_seen": 98792415, + "step": 4579 + }, + { + "epoch": 0.5507124391270366, + "flos": 45620118057120.0, + "grad_norm": 0.718685478963315, + "learning_rate": 1.7691610439901753e-06, + "loss": 0.5438, + "num_input_tokens_seen": 98855350, + "step": 4580 + }, + { + "epoch": 0.5508326820176757, + "flos": 22242053394720.0, + "grad_norm": 2.2914383051984535, + "learning_rate": 1.7683872958638367e-06, + "loss": 0.7547, + "num_input_tokens_seen": 98874230, + "step": 4581 + }, + { + "epoch": 0.5509529249083148, + "flos": 19983969519360.0, + "grad_norm": 3.362931041671839, + "learning_rate": 1.7676135828729614e-06, + "loss": 0.8426, + "num_input_tokens_seen": 98893940, + "step": 4582 + }, + { + "epoch": 0.5510731677989539, + "flos": 21803604442080.0, + "grad_norm": 2.7003020612093085, + "learning_rate": 1.7668399051349205e-06, + "loss": 0.8303, + "num_input_tokens_seen": 98913620, + "step": 4583 + }, + { + "epoch": 0.5511934106895929, + "flos": 21437164781760.0, + "grad_norm": 2.3129836942672743, + "learning_rate": 1.766066262767081e-06, + "loss": 0.8292, + "num_input_tokens_seen": 98931460, + "step": 4584 + }, + { + "epoch": 0.5513136535802321, + "flos": 21038864903520.0, + "grad_norm": 3.1421728163214375, + "learning_rate": 1.765292655886803e-06, + "loss": 0.7758, + "num_input_tokens_seen": 98950105, + "step": 4585 + }, + { + "epoch": 0.5514338964708712, + "flos": 27774974622240.0, + "grad_norm": 3.466494540073921, + "learning_rate": 1.764519084611443e-06, + "loss": 0.7048, + "num_input_tokens_seen": 98970560, + "step": 4586 + }, + { + "epoch": 0.5515541393615102, + "flos": 21877249203840.0, + "grad_norm": 2.1289970728548138, + "learning_rate": 1.7637455490583505e-06, + "loss": 0.7786, + "num_input_tokens_seen": 98989885, + "step": 4587 + }, + { + "epoch": 0.5516743822521494, + "flos": 20456211517440.0, + "grad_norm": 2.631182157316918, + "learning_rate": 1.7629720493448701e-06, + "loss": 0.7728, + "num_input_tokens_seen": 99007180, + "step": 4588 + }, + { + "epoch": 0.5517946251427884, + "flos": 14919275955840.0, + "grad_norm": 2.338852610373415, + "learning_rate": 1.7621985855883418e-06, + "loss": 0.8522, + "num_input_tokens_seen": 99023880, + "step": 4589 + }, + { + "epoch": 0.5519148680334275, + "flos": 18378652664640.0, + "grad_norm": 3.198720294654958, + "learning_rate": 1.7614251579060983e-06, + "loss": 0.726, + "num_input_tokens_seen": 99042475, + "step": 4590 + }, + { + "epoch": 0.5520351109240667, + "flos": 25077178022400.0, + "grad_norm": 2.205920887030826, + "learning_rate": 1.76065176641547e-06, + "loss": 0.8508, + "num_input_tokens_seen": 99065185, + "step": 4591 + }, + { + "epoch": 0.5521553538147057, + "flos": 21039496789440.0, + "grad_norm": 2.4013675846380313, + "learning_rate": 1.759878411233777e-06, + "loss": 0.7819, + "num_input_tokens_seen": 99083645, + "step": 4592 + }, + { + "epoch": 0.5522755967053448, + "flos": 18853124848320.0, + "grad_norm": 2.6168666358232815, + "learning_rate": 1.7591050924783388e-06, + "loss": 0.7601, + "num_input_tokens_seen": 99100835, + "step": 4593 + }, + { + "epoch": 0.5523958395959839, + "flos": 64583617123680.0, + "grad_norm": 0.8394717079970753, + "learning_rate": 1.7583318102664661e-06, + "loss": 0.5783, + "num_input_tokens_seen": 99168400, + "step": 4594 + }, + { + "epoch": 0.552516082486623, + "flos": 10874495298720.0, + "grad_norm": 2.344988991359744, + "learning_rate": 1.757558564715466e-06, + "loss": 0.7905, + "num_input_tokens_seen": 99186475, + "step": 4595 + }, + { + "epoch": 0.552636325377262, + "flos": 22168074105120.0, + "grad_norm": 5.7251114613706156, + "learning_rate": 1.7567853559426386e-06, + "loss": 0.7435, + "num_input_tokens_seen": 99203525, + "step": 4596 + }, + { + "epoch": 0.5527565682679012, + "flos": 23953878812640.0, + "grad_norm": 3.3515795727274953, + "learning_rate": 1.7560121840652797e-06, + "loss": 0.7504, + "num_input_tokens_seen": 99222910, + "step": 4597 + }, + { + "epoch": 0.5528768111585403, + "flos": 19691174620800.0, + "grad_norm": 2.4039533479705364, + "learning_rate": 1.7552390492006782e-06, + "loss": 0.6891, + "num_input_tokens_seen": 99241230, + "step": 4598 + }, + { + "epoch": 0.5529970540491793, + "flos": 26178504170400.0, + "grad_norm": 3.126794986192915, + "learning_rate": 1.7544659514661184e-06, + "loss": 0.6456, + "num_input_tokens_seen": 99264320, + "step": 4599 + }, + { + "epoch": 0.5531172969398185, + "flos": 24391026823680.0, + "grad_norm": 2.2221041833469353, + "learning_rate": 1.7536928909788786e-06, + "loss": 0.7974, + "num_input_tokens_seen": 99283625, + "step": 4600 + }, + { + "epoch": 0.5532375398304575, + "flos": 64814109770880.0, + "grad_norm": 0.890588893493113, + "learning_rate": 1.752919867856231e-06, + "loss": 0.6194, + "num_input_tokens_seen": 99335270, + "step": 4601 + }, + { + "epoch": 0.5533577827210966, + "flos": 19655077050240.0, + "grad_norm": 30.806856021504167, + "learning_rate": 1.7521468822154436e-06, + "loss": 0.7911, + "num_input_tokens_seen": 99354185, + "step": 4602 + }, + { + "epoch": 0.5534780256117358, + "flos": 32260062720000.0, + "grad_norm": 2.1958262429554187, + "learning_rate": 1.751373934173777e-06, + "loss": 0.7519, + "num_input_tokens_seen": 99376930, + "step": 4603 + }, + { + "epoch": 0.5535982685023748, + "flos": 23189250783360.0, + "grad_norm": 1.9887788221187659, + "learning_rate": 1.750601023848487e-06, + "loss": 0.7319, + "num_input_tokens_seen": 99397655, + "step": 4604 + }, + { + "epoch": 0.5537185113930139, + "flos": 24317679420000.0, + "grad_norm": 2.330033205147743, + "learning_rate": 1.749828151356823e-06, + "loss": 0.7379, + "num_input_tokens_seen": 99417850, + "step": 4605 + }, + { + "epoch": 0.553838754283653, + "flos": 23515355520480.0, + "grad_norm": 2.1503371839266596, + "learning_rate": 1.7490553168160297e-06, + "loss": 0.7563, + "num_input_tokens_seen": 99438920, + "step": 4606 + }, + { + "epoch": 0.5539589971742921, + "flos": 17250521386080.0, + "grad_norm": 5.120688881426143, + "learning_rate": 1.748282520343345e-06, + "loss": 0.7661, + "num_input_tokens_seen": 99457950, + "step": 4607 + }, + { + "epoch": 0.5540792400649311, + "flos": 27524670492960.0, + "grad_norm": 2.3375925178374146, + "learning_rate": 1.7475097620560023e-06, + "loss": 0.7883, + "num_input_tokens_seen": 99478015, + "step": 4608 + }, + { + "epoch": 0.5541994829555702, + "flos": 23844508177920.0, + "grad_norm": 3.204856144240087, + "learning_rate": 1.746737042071228e-06, + "loss": 0.7104, + "num_input_tokens_seen": 99496520, + "step": 4609 + }, + { + "epoch": 0.5543197258462094, + "flos": 20086872615840.0, + "grad_norm": 2.3825519356765947, + "learning_rate": 1.7459643605062424e-06, + "loss": 0.7927, + "num_input_tokens_seen": 99513780, + "step": 4610 + }, + { + "epoch": 0.5544399687368484, + "flos": 20886483122880.0, + "grad_norm": 2.328052084376198, + "learning_rate": 1.745191717478262e-06, + "loss": 0.8066, + "num_input_tokens_seen": 99532360, + "step": 4611 + }, + { + "epoch": 0.5545602116274875, + "flos": 25482056948160.0, + "grad_norm": 2.8663728060659683, + "learning_rate": 1.7444191131044948e-06, + "loss": 0.7969, + "num_input_tokens_seen": 99552635, + "step": 4612 + }, + { + "epoch": 0.5546804545181266, + "flos": 20965591839360.0, + "grad_norm": 2.823687364914507, + "learning_rate": 1.7436465475021456e-06, + "loss": 0.7313, + "num_input_tokens_seen": 99572685, + "step": 4613 + }, + { + "epoch": 0.5548006974087657, + "flos": 26795173620480.0, + "grad_norm": 7.244117718267234, + "learning_rate": 1.7428740207884111e-06, + "loss": 0.7137, + "num_input_tokens_seen": 99593565, + "step": 4614 + }, + { + "epoch": 0.5549209402994048, + "flos": 33608013191040.0, + "grad_norm": 3.377879009765486, + "learning_rate": 1.7421015330804833e-06, + "loss": 0.6134, + "num_input_tokens_seen": 99613485, + "step": 4615 + }, + { + "epoch": 0.5550411831900439, + "flos": 23735546410560.0, + "grad_norm": 4.64186649862579, + "learning_rate": 1.7413290844955475e-06, + "loss": 0.7265, + "num_input_tokens_seen": 99633515, + "step": 4616 + }, + { + "epoch": 0.555161426080683, + "flos": 21620217348000.0, + "grad_norm": 2.3132272556217046, + "learning_rate": 1.7405566751507843e-06, + "loss": 0.7832, + "num_input_tokens_seen": 99651560, + "step": 4617 + }, + { + "epoch": 0.555281668971322, + "flos": 49491516019200.0, + "grad_norm": 2.728813925119711, + "learning_rate": 1.7397843051633668e-06, + "loss": 0.6754, + "num_input_tokens_seen": 99674255, + "step": 4618 + }, + { + "epoch": 0.5554019118619612, + "flos": 20712053940960.0, + "grad_norm": 2.403321502945662, + "learning_rate": 1.739011974650464e-06, + "loss": 0.7131, + "num_input_tokens_seen": 99693585, + "step": 4619 + }, + { + "epoch": 0.5555221547526003, + "flos": 25446702772800.0, + "grad_norm": 2.5312019626156497, + "learning_rate": 1.7382396837292365e-06, + "loss": 0.7715, + "num_input_tokens_seen": 99714045, + "step": 4620 + }, + { + "epoch": 0.5556423976432393, + "flos": 21731446470720.0, + "grad_norm": 2.0503630293422446, + "learning_rate": 1.737467432516841e-06, + "loss": 0.7356, + "num_input_tokens_seen": 99734145, + "step": 4621 + }, + { + "epoch": 0.5557626405338785, + "flos": 24864532593600.0, + "grad_norm": 6.874617687345818, + "learning_rate": 1.7366952211304274e-06, + "loss": 0.7354, + "num_input_tokens_seen": 99751990, + "step": 4622 + }, + { + "epoch": 0.5558828834245175, + "flos": 18670890016800.0, + "grad_norm": 2.454881173111375, + "learning_rate": 1.735923049687139e-06, + "loss": 0.8305, + "num_input_tokens_seen": 99768565, + "step": 4623 + }, + { + "epoch": 0.5560031263151566, + "flos": 27233102196480.0, + "grad_norm": 2.125236498070154, + "learning_rate": 1.7351509183041144e-06, + "loss": 0.7403, + "num_input_tokens_seen": 99788895, + "step": 4624 + }, + { + "epoch": 0.5561233692057957, + "flos": 23370259012800.0, + "grad_norm": 2.1161699277867654, + "learning_rate": 1.7343788270984852e-06, + "loss": 0.7211, + "num_input_tokens_seen": 99808035, + "step": 4625 + }, + { + "epoch": 0.5562436120964348, + "flos": 37616659107840.0, + "grad_norm": 2.051629288945206, + "learning_rate": 1.7336067761873764e-06, + "loss": 0.7484, + "num_input_tokens_seen": 99830460, + "step": 4626 + }, + { + "epoch": 0.5563638549870739, + "flos": 25119222754560.0, + "grad_norm": 2.4770746995044477, + "learning_rate": 1.7328347656879076e-06, + "loss": 0.7619, + "num_input_tokens_seen": 99850320, + "step": 4627 + }, + { + "epoch": 0.556484097877713, + "flos": 13562367572640.0, + "grad_norm": 3.3408644159523098, + "learning_rate": 1.7320627957171927e-06, + "loss": 0.6824, + "num_input_tokens_seen": 99864980, + "step": 4628 + }, + { + "epoch": 0.5566043407683521, + "flos": 24646051512480.0, + "grad_norm": 2.260398571686166, + "learning_rate": 1.7312908663923382e-06, + "loss": 0.8139, + "num_input_tokens_seen": 99881155, + "step": 4629 + }, + { + "epoch": 0.5567245836589911, + "flos": 20558557067520.0, + "grad_norm": 2.6844406457408163, + "learning_rate": 1.7305189778304463e-06, + "loss": 0.6748, + "num_input_tokens_seen": 99899330, + "step": 4630 + }, + { + "epoch": 0.5568448265496303, + "flos": 20674655428800.0, + "grad_norm": 2.3675873170116275, + "learning_rate": 1.729747130148611e-06, + "loss": 0.8014, + "num_input_tokens_seen": 99918525, + "step": 4631 + }, + { + "epoch": 0.5569650694402694, + "flos": 25266735296640.0, + "grad_norm": 3.0220758097220246, + "learning_rate": 1.7289753234639208e-06, + "loss": 0.7744, + "num_input_tokens_seen": 99937500, + "step": 4632 + }, + { + "epoch": 0.5570853123309084, + "flos": 19684298215200.0, + "grad_norm": 2.5599115836253024, + "learning_rate": 1.7282035578934592e-06, + "loss": 0.7661, + "num_input_tokens_seen": 99955665, + "step": 4633 + }, + { + "epoch": 0.5572055552215476, + "flos": 16085288953440.0, + "grad_norm": 3.499765933248167, + "learning_rate": 1.727431833554301e-06, + "loss": 0.7879, + "num_input_tokens_seen": 99974140, + "step": 4634 + }, + { + "epoch": 0.5573257981121866, + "flos": 17104495634400.0, + "grad_norm": 2.7426521798738617, + "learning_rate": 1.7266601505635175e-06, + "loss": 0.7769, + "num_input_tokens_seen": 99991480, + "step": 4635 + }, + { + "epoch": 0.5574460410028257, + "flos": 18780260651520.0, + "grad_norm": 3.248933910699217, + "learning_rate": 1.7258885090381717e-06, + "loss": 0.7538, + "num_input_tokens_seen": 100009520, + "step": 4636 + }, + { + "epoch": 0.5575662838934649, + "flos": 29600110669440.0, + "grad_norm": 3.216434275347755, + "learning_rate": 1.7251169090953213e-06, + "loss": 0.7872, + "num_input_tokens_seen": 100029995, + "step": 4637 + }, + { + "epoch": 0.5576865267841039, + "flos": 22022977597440.0, + "grad_norm": 3.821701452599071, + "learning_rate": 1.7243453508520168e-06, + "loss": 0.7594, + "num_input_tokens_seen": 100046980, + "step": 4638 + }, + { + "epoch": 0.557806769674743, + "flos": 17171412669600.0, + "grad_norm": 2.9113501640702775, + "learning_rate": 1.7235738344253038e-06, + "loss": 0.8502, + "num_input_tokens_seen": 100060610, + "step": 4639 + }, + { + "epoch": 0.557927012565382, + "flos": 24677168335200.0, + "grad_norm": 2.564688143743446, + "learning_rate": 1.72280235993222e-06, + "loss": 0.828, + "num_input_tokens_seen": 100078915, + "step": 4640 + }, + { + "epoch": 0.5580472554560212, + "flos": 16958990259360.0, + "grad_norm": 2.770076810963056, + "learning_rate": 1.722030927489798e-06, + "loss": 0.6959, + "num_input_tokens_seen": 100096195, + "step": 4641 + }, + { + "epoch": 0.5581674983466602, + "flos": 23475317955360.0, + "grad_norm": 2.1098510205072185, + "learning_rate": 1.7212595372150634e-06, + "loss": 0.7402, + "num_input_tokens_seen": 100116175, + "step": 4642 + }, + { + "epoch": 0.5582877412372993, + "flos": 13461285794400.0, + "grad_norm": 4.264267254725783, + "learning_rate": 1.720488189225035e-06, + "loss": 0.7267, + "num_input_tokens_seen": 100133875, + "step": 4643 + }, + { + "epoch": 0.5584079841279385, + "flos": 21872156946720.0, + "grad_norm": 3.2170857691787518, + "learning_rate": 1.7197168836367265e-06, + "loss": 0.7896, + "num_input_tokens_seen": 100150685, + "step": 4644 + }, + { + "epoch": 0.5585282270185775, + "flos": 18816729919680.0, + "grad_norm": 2.8809576916486184, + "learning_rate": 1.7189456205671433e-06, + "loss": 0.8186, + "num_input_tokens_seen": 100169530, + "step": 4645 + }, + { + "epoch": 0.5586484699092166, + "flos": 21837100129440.0, + "grad_norm": 2.430333287808802, + "learning_rate": 1.7181744001332866e-06, + "loss": 0.8248, + "num_input_tokens_seen": 100188140, + "step": 4646 + }, + { + "epoch": 0.5587687127998557, + "flos": 22860506993280.0, + "grad_norm": 2.252413832427284, + "learning_rate": 1.7174032224521493e-06, + "loss": 0.6335, + "num_input_tokens_seen": 100206850, + "step": 4647 + }, + { + "epoch": 0.5588889556904948, + "flos": 20274125364960.0, + "grad_norm": 7.340192996613854, + "learning_rate": 1.7166320876407184e-06, + "loss": 0.6929, + "num_input_tokens_seen": 100226270, + "step": 4648 + }, + { + "epoch": 0.5590091985811338, + "flos": 16448792202720.0, + "grad_norm": 2.404725322463121, + "learning_rate": 1.7158609958159742e-06, + "loss": 0.6788, + "num_input_tokens_seen": 100243675, + "step": 4649 + }, + { + "epoch": 0.559129441471773, + "flos": 14511423449280.0, + "grad_norm": 2.812109008738169, + "learning_rate": 1.7150899470948911e-06, + "loss": 0.7835, + "num_input_tokens_seen": 100258940, + "step": 4650 + }, + { + "epoch": 0.5592496843624121, + "flos": 60435227144640.0, + "grad_norm": 0.8143352908145937, + "learning_rate": 1.7143189415944365e-06, + "loss": 0.5661, + "num_input_tokens_seen": 100323400, + "step": 4651 + }, + { + "epoch": 0.5593699272530511, + "flos": 20856444223200.0, + "grad_norm": 1.9817088080252072, + "learning_rate": 1.7135479794315714e-06, + "loss": 0.7623, + "num_input_tokens_seen": 100340355, + "step": 4652 + }, + { + "epoch": 0.5594901701436903, + "flos": 12878595238560.0, + "grad_norm": 2.539799791540589, + "learning_rate": 1.7127770607232502e-06, + "loss": 0.7913, + "num_input_tokens_seen": 100358900, + "step": 4653 + }, + { + "epoch": 0.5596104130343293, + "flos": 23079062413920.0, + "grad_norm": 2.662821281999715, + "learning_rate": 1.7120061855864204e-06, + "loss": 0.7944, + "num_input_tokens_seen": 100378910, + "step": 4654 + }, + { + "epoch": 0.5597306559249684, + "flos": 25920617410080.0, + "grad_norm": 2.1366995578719887, + "learning_rate": 1.7112353541380233e-06, + "loss": 0.7105, + "num_input_tokens_seen": 100398770, + "step": 4655 + }, + { + "epoch": 0.5598508988156076, + "flos": 22460162778240.0, + "grad_norm": 2.8159059881805786, + "learning_rate": 1.7104645664949931e-06, + "loss": 0.7223, + "num_input_tokens_seen": 100421240, + "step": 4656 + }, + { + "epoch": 0.5599711417062466, + "flos": 23079508451040.0, + "grad_norm": 2.501938449837426, + "learning_rate": 1.7096938227742584e-06, + "loss": 0.7162, + "num_input_tokens_seen": 100442370, + "step": 4657 + }, + { + "epoch": 0.5600913845968857, + "flos": 22307186281440.0, + "grad_norm": 2.5868398699664232, + "learning_rate": 1.70892312309274e-06, + "loss": 0.8452, + "num_input_tokens_seen": 100460260, + "step": 4658 + }, + { + "epoch": 0.5602116274875248, + "flos": 17608523510880.0, + "grad_norm": 3.282401763770875, + "learning_rate": 1.7081524675673523e-06, + "loss": 0.6823, + "num_input_tokens_seen": 100475750, + "step": 4659 + }, + { + "epoch": 0.5603318703781639, + "flos": 70016162798400.0, + "grad_norm": 0.777263963141406, + "learning_rate": 1.7073818563150026e-06, + "loss": 0.5954, + "num_input_tokens_seen": 100537830, + "step": 4660 + }, + { + "epoch": 0.560452113268803, + "flos": 18519251631360.0, + "grad_norm": 2.7561348196862774, + "learning_rate": 1.7066112894525935e-06, + "loss": 0.864, + "num_input_tokens_seen": 100555910, + "step": 4661 + }, + { + "epoch": 0.5605723561594421, + "flos": 25228481880000.0, + "grad_norm": 2.797383011868037, + "learning_rate": 1.7058407670970177e-06, + "loss": 0.727, + "num_input_tokens_seen": 100577385, + "step": 4662 + }, + { + "epoch": 0.5606925990500812, + "flos": 20565582152160.0, + "grad_norm": 2.4348145388055547, + "learning_rate": 1.7050702893651643e-06, + "loss": 0.6081, + "num_input_tokens_seen": 100596965, + "step": 4663 + }, + { + "epoch": 0.5608128419407202, + "flos": 34955666304000.0, + "grad_norm": 2.819409762222958, + "learning_rate": 1.7042998563739134e-06, + "loss": 0.7557, + "num_input_tokens_seen": 100615315, + "step": 4664 + }, + { + "epoch": 0.5609330848313594, + "flos": 24604378477920.0, + "grad_norm": 3.4751854387382215, + "learning_rate": 1.703529468240139e-06, + "loss": 0.7149, + "num_input_tokens_seen": 100634185, + "step": 4665 + }, + { + "epoch": 0.5610533277219985, + "flos": 18735911394240.0, + "grad_norm": 3.3698558457329457, + "learning_rate": 1.7027591250807088e-06, + "loss": 0.7375, + "num_input_tokens_seen": 100651360, + "step": 4666 + }, + { + "epoch": 0.5611735706126375, + "flos": 14991136569120.0, + "grad_norm": 3.2482293514117133, + "learning_rate": 1.7019888270124825e-06, + "loss": 0.8514, + "num_input_tokens_seen": 100668525, + "step": 4667 + }, + { + "epoch": 0.5612938135032767, + "flos": 16444666359360.0, + "grad_norm": 2.4891130862510527, + "learning_rate": 1.7012185741523147e-06, + "loss": 0.825, + "num_input_tokens_seen": 100684850, + "step": 4668 + }, + { + "epoch": 0.5614140563939157, + "flos": 25629680999520.0, + "grad_norm": 3.351169803261686, + "learning_rate": 1.7004483666170514e-06, + "loss": 0.6283, + "num_input_tokens_seen": 100705345, + "step": 4669 + }, + { + "epoch": 0.5615342992845548, + "flos": 24682186252800.0, + "grad_norm": 3.029698554929225, + "learning_rate": 1.699678204523533e-06, + "loss": 0.8066, + "num_input_tokens_seen": 100725210, + "step": 4670 + }, + { + "epoch": 0.5616545421751938, + "flos": 21984426822720.0, + "grad_norm": 8.264751308828194, + "learning_rate": 1.6989080879885918e-06, + "loss": 0.6961, + "num_input_tokens_seen": 100743070, + "step": 4671 + }, + { + "epoch": 0.561774785065833, + "flos": 53684026503360.0, + "grad_norm": 0.9279276114544762, + "learning_rate": 1.6981380171290544e-06, + "loss": 0.6092, + "num_input_tokens_seen": 100804095, + "step": 4672 + }, + { + "epoch": 0.5618950279564721, + "flos": 19721659557600.0, + "grad_norm": 2.7755993651159967, + "learning_rate": 1.6973679920617396e-06, + "loss": 0.7412, + "num_input_tokens_seen": 100821630, + "step": 4673 + }, + { + "epoch": 0.5620152708471111, + "flos": 16776755427840.0, + "grad_norm": 5.4936458990065855, + "learning_rate": 1.6965980129034603e-06, + "loss": 0.8564, + "num_input_tokens_seen": 100839015, + "step": 4674 + }, + { + "epoch": 0.5621355137377503, + "flos": 26759410577760.0, + "grad_norm": 3.6181614267028186, + "learning_rate": 1.6958280797710209e-06, + "loss": 0.7667, + "num_input_tokens_seen": 100860940, + "step": 4675 + }, + { + "epoch": 0.5622557566283893, + "flos": 61120003062240.0, + "grad_norm": 0.7368874284959614, + "learning_rate": 1.6950581927812198e-06, + "loss": 0.5465, + "num_input_tokens_seen": 100920510, + "step": 4676 + }, + { + "epoch": 0.5623759995190284, + "flos": 26430592448160.0, + "grad_norm": 3.099505071201105, + "learning_rate": 1.6942883520508486e-06, + "loss": 0.7899, + "num_input_tokens_seen": 100939245, + "step": 4677 + }, + { + "epoch": 0.5624962424096676, + "flos": 19363285735200.0, + "grad_norm": 136.2880035875315, + "learning_rate": 1.693518557696691e-06, + "loss": 0.7723, + "num_input_tokens_seen": 100958385, + "step": 4678 + }, + { + "epoch": 0.5626164853003066, + "flos": 20637851632800.0, + "grad_norm": 3.2127246941384735, + "learning_rate": 1.6927488098355252e-06, + "loss": 0.8927, + "num_input_tokens_seen": 100976015, + "step": 4679 + }, + { + "epoch": 0.5627367281909457, + "flos": 62676398779200.0, + "grad_norm": 0.8992349817052202, + "learning_rate": 1.6919791085841201e-06, + "loss": 0.6311, + "num_input_tokens_seen": 101033425, + "step": 4680 + }, + { + "epoch": 0.5628569710815848, + "flos": 12769150264320.0, + "grad_norm": 4.0508683580617255, + "learning_rate": 1.6912094540592396e-06, + "loss": 0.7848, + "num_input_tokens_seen": 101048945, + "step": 4681 + }, + { + "epoch": 0.5629772139722239, + "flos": 13742297879040.0, + "grad_norm": 3.910600604590431, + "learning_rate": 1.6904398463776393e-06, + "loss": 0.8131, + "num_input_tokens_seen": 101062745, + "step": 4682 + }, + { + "epoch": 0.5630974568628629, + "flos": 21437090442240.0, + "grad_norm": 2.371687536659374, + "learning_rate": 1.6896702856560683e-06, + "loss": 0.7263, + "num_input_tokens_seen": 101081840, + "step": 4683 + }, + { + "epoch": 0.5632176997535021, + "flos": 14225765144640.0, + "grad_norm": 5.41836619002347, + "learning_rate": 1.6889007720112677e-06, + "loss": 0.6942, + "num_input_tokens_seen": 101099585, + "step": 4684 + }, + { + "epoch": 0.5633379426441412, + "flos": 20782502103360.0, + "grad_norm": 1.8875957843653415, + "learning_rate": 1.6881313055599734e-06, + "loss": 0.7752, + "num_input_tokens_seen": 101119515, + "step": 4685 + }, + { + "epoch": 0.5634581855347802, + "flos": 22568566999200.0, + "grad_norm": 4.165280963630358, + "learning_rate": 1.6873618864189117e-06, + "loss": 0.8262, + "num_input_tokens_seen": 101135285, + "step": 4686 + }, + { + "epoch": 0.5635784284254194, + "flos": 21476273102880.0, + "grad_norm": 4.448647051244665, + "learning_rate": 1.686592514704803e-06, + "loss": 0.7809, + "num_input_tokens_seen": 101152680, + "step": 4687 + }, + { + "epoch": 0.5636986713160584, + "flos": 19799169974400.0, + "grad_norm": 2.9469711530740814, + "learning_rate": 1.685823190534361e-06, + "loss": 0.7073, + "num_input_tokens_seen": 101170920, + "step": 4688 + }, + { + "epoch": 0.5638189142066975, + "flos": 19764261836160.0, + "grad_norm": 8.322518871924242, + "learning_rate": 1.6850539140242907e-06, + "loss": 0.8358, + "num_input_tokens_seen": 101190180, + "step": 4689 + }, + { + "epoch": 0.5639391570973367, + "flos": 22861064539680.0, + "grad_norm": 2.3723622782763236, + "learning_rate": 1.684284685291292e-06, + "loss": 0.8254, + "num_input_tokens_seen": 101211825, + "step": 4690 + }, + { + "epoch": 0.5640593999879757, + "flos": 23692944132000.0, + "grad_norm": 26.232368287977618, + "learning_rate": 1.683515504452055e-06, + "loss": 0.8142, + "num_input_tokens_seen": 101229200, + "step": 4691 + }, + { + "epoch": 0.5641796428786148, + "flos": 22677640275840.0, + "grad_norm": 4.881955979564069, + "learning_rate": 1.6827463716232648e-06, + "loss": 0.664, + "num_input_tokens_seen": 101249860, + "step": 4692 + }, + { + "epoch": 0.5642998857692539, + "flos": 19763778629280.0, + "grad_norm": 4.666698432995898, + "learning_rate": 1.6819772869215972e-06, + "loss": 0.7597, + "num_input_tokens_seen": 101268940, + "step": 4693 + }, + { + "epoch": 0.564420128659893, + "flos": 23152781515200.0, + "grad_norm": 2.5417020722304793, + "learning_rate": 1.6812082504637228e-06, + "loss": 0.8233, + "num_input_tokens_seen": 101290260, + "step": 4694 + }, + { + "epoch": 0.564540371550532, + "flos": 23225013826080.0, + "grad_norm": 2.2230010879726723, + "learning_rate": 1.6804392623663025e-06, + "loss": 0.7427, + "num_input_tokens_seen": 101311465, + "step": 4695 + }, + { + "epoch": 0.5646606144411712, + "flos": 24974274925920.0, + "grad_norm": 4.2818885939536395, + "learning_rate": 1.6796703227459935e-06, + "loss": 0.7843, + "num_input_tokens_seen": 101329420, + "step": 4696 + }, + { + "epoch": 0.5647808573318103, + "flos": 36487078208640.0, + "grad_norm": 2.6030116355363355, + "learning_rate": 1.6789014317194407e-06, + "loss": 0.7604, + "num_input_tokens_seen": 101350900, + "step": 4697 + }, + { + "epoch": 0.5649011002224493, + "flos": 22496260348800.0, + "grad_norm": 5.656506948799445, + "learning_rate": 1.6781325894032853e-06, + "loss": 0.7312, + "num_input_tokens_seen": 101369860, + "step": 4698 + }, + { + "epoch": 0.5650213431130885, + "flos": 18488803864320.0, + "grad_norm": 2.32751856336556, + "learning_rate": 1.6773637959141608e-06, + "loss": 0.9182, + "num_input_tokens_seen": 101386835, + "step": 4699 + }, + { + "epoch": 0.5651415860037275, + "flos": 17501271552480.0, + "grad_norm": 8.187315494700666, + "learning_rate": 1.6765950513686915e-06, + "loss": 0.6609, + "num_input_tokens_seen": 101404945, + "step": 4700 + }, + { + "epoch": 0.5652618288943666, + "flos": 25483469399040.0, + "grad_norm": 2.6029048354888205, + "learning_rate": 1.675826355883496e-06, + "loss": 0.7639, + "num_input_tokens_seen": 101424160, + "step": 4701 + }, + { + "epoch": 0.5653820717850057, + "flos": 19655077050240.0, + "grad_norm": 2.2675322824554316, + "learning_rate": 1.6750577095751848e-06, + "loss": 0.7938, + "num_input_tokens_seen": 101443270, + "step": 4702 + }, + { + "epoch": 0.5655023146756448, + "flos": 26941050693120.0, + "grad_norm": 2.0314082157432707, + "learning_rate": 1.6742891125603605e-06, + "loss": 0.7277, + "num_input_tokens_seen": 101464175, + "step": 4703 + }, + { + "epoch": 0.5656225575662839, + "flos": 27629543586720.0, + "grad_norm": 4.456869271199406, + "learning_rate": 1.6735205649556185e-06, + "loss": 0.7189, + "num_input_tokens_seen": 101484045, + "step": 4704 + }, + { + "epoch": 0.5657428004569229, + "flos": 24314780178720.0, + "grad_norm": 2.3062058441979953, + "learning_rate": 1.6727520668775476e-06, + "loss": 0.8487, + "num_input_tokens_seen": 101504330, + "step": 4705 + }, + { + "epoch": 0.5658630433475621, + "flos": 21912677718720.0, + "grad_norm": 2.588564181213916, + "learning_rate": 1.6719836184427275e-06, + "loss": 0.7537, + "num_input_tokens_seen": 101524635, + "step": 4706 + }, + { + "epoch": 0.5659832862382012, + "flos": 30365333414880.0, + "grad_norm": 2.266023061576047, + "learning_rate": 1.671215219767733e-06, + "loss": 0.6435, + "num_input_tokens_seen": 101544095, + "step": 4707 + }, + { + "epoch": 0.5661035291288402, + "flos": 13169605988640.0, + "grad_norm": 2.650198382023657, + "learning_rate": 1.670446870969127e-06, + "loss": 0.7587, + "num_input_tokens_seen": 101561760, + "step": 4708 + }, + { + "epoch": 0.5662237720194794, + "flos": 16120122752160.0, + "grad_norm": 3.3799984969186543, + "learning_rate": 1.6696785721634685e-06, + "loss": 0.8019, + "num_input_tokens_seen": 101576245, + "step": 4709 + }, + { + "epoch": 0.5663440149101184, + "flos": 17651757675360.0, + "grad_norm": 5.417367430131251, + "learning_rate": 1.6689103234673086e-06, + "loss": 0.7381, + "num_input_tokens_seen": 101594565, + "step": 4710 + }, + { + "epoch": 0.5664642578007575, + "flos": 23334942007200.0, + "grad_norm": 3.4897788529161473, + "learning_rate": 1.668142124997189e-06, + "loss": 0.7682, + "num_input_tokens_seen": 101614225, + "step": 4711 + }, + { + "epoch": 0.5665845006913967, + "flos": 65423302365600.0, + "grad_norm": 0.7469760012467593, + "learning_rate": 1.6673739768696453e-06, + "loss": 0.5967, + "num_input_tokens_seen": 101680795, + "step": 4712 + }, + { + "epoch": 0.5667047435820357, + "flos": 26103446957760.0, + "grad_norm": 2.192931706761526, + "learning_rate": 1.6666058792012052e-06, + "loss": 0.7755, + "num_input_tokens_seen": 101701680, + "step": 4713 + }, + { + "epoch": 0.5668249864726748, + "flos": 71764606163520.0, + "grad_norm": 0.8921981436767626, + "learning_rate": 1.6658378321083878e-06, + "loss": 0.6865, + "num_input_tokens_seen": 101766010, + "step": 4714 + }, + { + "epoch": 0.5669452293633139, + "flos": 22163799582720.0, + "grad_norm": 2.453419661592503, + "learning_rate": 1.6650698357077055e-06, + "loss": 0.8281, + "num_input_tokens_seen": 101783055, + "step": 4715 + }, + { + "epoch": 0.567065472253953, + "flos": 18197272737600.0, + "grad_norm": 3.3477405728878358, + "learning_rate": 1.6643018901156632e-06, + "loss": 0.8106, + "num_input_tokens_seen": 101802150, + "step": 4716 + }, + { + "epoch": 0.567185715144592, + "flos": 20342343341760.0, + "grad_norm": 3.8271987728732513, + "learning_rate": 1.6635339954487566e-06, + "loss": 0.7997, + "num_input_tokens_seen": 101818025, + "step": 4717 + }, + { + "epoch": 0.5673059580352312, + "flos": 23188135690560.0, + "grad_norm": 2.9542390851253932, + "learning_rate": 1.6627661518234765e-06, + "loss": 0.8215, + "num_input_tokens_seen": 101837280, + "step": 4718 + }, + { + "epoch": 0.5674262009258703, + "flos": 21688360985280.0, + "grad_norm": 1.9989773997049716, + "learning_rate": 1.661998359356302e-06, + "loss": 0.8558, + "num_input_tokens_seen": 101856310, + "step": 4719 + }, + { + "epoch": 0.5675464438165093, + "flos": 67373760608160.0, + "grad_norm": 0.7535630072314905, + "learning_rate": 1.6612306181637077e-06, + "loss": 0.556, + "num_input_tokens_seen": 101915635, + "step": 4720 + }, + { + "epoch": 0.5676666867071485, + "flos": 18852269943840.0, + "grad_norm": 2.877959511106923, + "learning_rate": 1.6604629283621598e-06, + "loss": 0.6643, + "num_input_tokens_seen": 101933565, + "step": 4721 + }, + { + "epoch": 0.5677869295977875, + "flos": 33498531047040.0, + "grad_norm": 2.0776127922572054, + "learning_rate": 1.6596952900681152e-06, + "loss": 0.744, + "num_input_tokens_seen": 101954325, + "step": 4722 + }, + { + "epoch": 0.5679071724884266, + "flos": 27997024000320.0, + "grad_norm": 9.14672189160164, + "learning_rate": 1.658927703398025e-06, + "loss": 0.8186, + "num_input_tokens_seen": 101974390, + "step": 4723 + }, + { + "epoch": 0.5680274153790658, + "flos": 23516433443520.0, + "grad_norm": 3.0389109281972386, + "learning_rate": 1.6581601684683309e-06, + "loss": 0.7765, + "num_input_tokens_seen": 101994815, + "step": 4724 + }, + { + "epoch": 0.5681476582697048, + "flos": 22423061624160.0, + "grad_norm": 3.1843013796096753, + "learning_rate": 1.6573926853954674e-06, + "loss": 0.6892, + "num_input_tokens_seen": 102012400, + "step": 4725 + }, + { + "epoch": 0.5682679011603439, + "flos": 19509088468320.0, + "grad_norm": 2.7535153375312396, + "learning_rate": 1.6566252542958608e-06, + "loss": 0.836, + "num_input_tokens_seen": 102031900, + "step": 4726 + }, + { + "epoch": 0.568388144050983, + "flos": 28724142008160.0, + "grad_norm": 2.1875540590418487, + "learning_rate": 1.6558578752859305e-06, + "loss": 0.7849, + "num_input_tokens_seen": 102050860, + "step": 4727 + }, + { + "epoch": 0.5685083869416221, + "flos": 21178980663360.0, + "grad_norm": 3.5425031833459393, + "learning_rate": 1.6550905484820865e-06, + "loss": 0.7922, + "num_input_tokens_seen": 102069320, + "step": 4728 + }, + { + "epoch": 0.5686286298322611, + "flos": 24791482548000.0, + "grad_norm": 5.892648364575572, + "learning_rate": 1.6543232740007328e-06, + "loss": 0.7901, + "num_input_tokens_seen": 102086435, + "step": 4729 + }, + { + "epoch": 0.5687488727229003, + "flos": 26576989897440.0, + "grad_norm": 7.234233945456, + "learning_rate": 1.653556051958263e-06, + "loss": 0.6672, + "num_input_tokens_seen": 102106115, + "step": 4730 + }, + { + "epoch": 0.5688691156135394, + "flos": 20778859466880.0, + "grad_norm": 2.6380514210279893, + "learning_rate": 1.6527888824710642e-06, + "loss": 0.7383, + "num_input_tokens_seen": 102125715, + "step": 4731 + }, + { + "epoch": 0.5689893585041784, + "flos": 25843738879200.0, + "grad_norm": 3.5008527534331835, + "learning_rate": 1.6520217656555166e-06, + "loss": 0.768, + "num_input_tokens_seen": 102145005, + "step": 4732 + }, + { + "epoch": 0.5691096013948175, + "flos": 23443903774560.0, + "grad_norm": 1.9273393065761253, + "learning_rate": 1.65125470162799e-06, + "loss": 0.7121, + "num_input_tokens_seen": 102165360, + "step": 4733 + }, + { + "epoch": 0.5692298442854566, + "flos": 18049834535040.0, + "grad_norm": 3.6031605780797813, + "learning_rate": 1.6504876905048485e-06, + "loss": 0.6979, + "num_input_tokens_seen": 102182320, + "step": 4734 + }, + { + "epoch": 0.5693500871760957, + "flos": 23006384065920.0, + "grad_norm": 2.6910753641905263, + "learning_rate": 1.6497207324024464e-06, + "loss": 0.7185, + "num_input_tokens_seen": 102201220, + "step": 4735 + }, + { + "epoch": 0.5694703300667348, + "flos": 18962644162080.0, + "grad_norm": 4.952057251027703, + "learning_rate": 1.6489538274371305e-06, + "loss": 0.8249, + "num_input_tokens_seen": 102219825, + "step": 4736 + }, + { + "epoch": 0.5695905729573739, + "flos": 21877435052640.0, + "grad_norm": 2.906084445626924, + "learning_rate": 1.6481869757252396e-06, + "loss": 0.8307, + "num_input_tokens_seen": 102238160, + "step": 4737 + }, + { + "epoch": 0.569710815848013, + "flos": 28435844650560.0, + "grad_norm": 2.311233173513359, + "learning_rate": 1.647420177383105e-06, + "loss": 0.719, + "num_input_tokens_seen": 102261030, + "step": 4738 + }, + { + "epoch": 0.569831058738652, + "flos": 28325470432320.0, + "grad_norm": 2.630848409492506, + "learning_rate": 1.646653432527049e-06, + "loss": 0.7261, + "num_input_tokens_seen": 102281670, + "step": 4739 + }, + { + "epoch": 0.5699513016292912, + "flos": 25812993754080.0, + "grad_norm": 16.420988543510216, + "learning_rate": 1.645886741273387e-06, + "loss": 0.7439, + "num_input_tokens_seen": 102303485, + "step": 4740 + }, + { + "epoch": 0.5700715445199303, + "flos": 18012101495040.0, + "grad_norm": 3.747995731070417, + "learning_rate": 1.645120103738424e-06, + "loss": 0.7402, + "num_input_tokens_seen": 102320995, + "step": 4741 + }, + { + "epoch": 0.5701917874105693, + "flos": 11457260194080.0, + "grad_norm": 4.0826437750668125, + "learning_rate": 1.6443535200384591e-06, + "loss": 0.8342, + "num_input_tokens_seen": 102339170, + "step": 4742 + }, + { + "epoch": 0.5703120303012085, + "flos": 21730665905760.0, + "grad_norm": 2.0187767286863707, + "learning_rate": 1.6435869902897827e-06, + "loss": 0.7077, + "num_input_tokens_seen": 102357750, + "step": 4743 + }, + { + "epoch": 0.5704322731918475, + "flos": 56665362731520.0, + "grad_norm": 0.8085520268239866, + "learning_rate": 1.6428205146086764e-06, + "loss": 0.6186, + "num_input_tokens_seen": 102419730, + "step": 4744 + }, + { + "epoch": 0.5705525160824866, + "flos": 20711570734080.0, + "grad_norm": 2.1604770746372104, + "learning_rate": 1.6420540931114142e-06, + "loss": 0.7071, + "num_input_tokens_seen": 102440320, + "step": 4745 + }, + { + "epoch": 0.5706727589731257, + "flos": 18744571948320.0, + "grad_norm": 1.7926707300575389, + "learning_rate": 1.6412877259142616e-06, + "loss": 0.79, + "num_input_tokens_seen": 102460240, + "step": 4746 + }, + { + "epoch": 0.5707930018637648, + "flos": 27595044315840.0, + "grad_norm": 4.977884018328145, + "learning_rate": 1.6405214131334757e-06, + "loss": 0.7384, + "num_input_tokens_seen": 102478765, + "step": 4747 + }, + { + "epoch": 0.5709132447544039, + "flos": 27558017501280.0, + "grad_norm": 2.7546320831211917, + "learning_rate": 1.6397551548853052e-06, + "loss": 0.794, + "num_input_tokens_seen": 102496930, + "step": 4748 + }, + { + "epoch": 0.571033487645043, + "flos": 21655013976960.0, + "grad_norm": 2.617216559860822, + "learning_rate": 1.6389889512859917e-06, + "loss": 0.7071, + "num_input_tokens_seen": 102516590, + "step": 4749 + }, + { + "epoch": 0.5711537305356821, + "flos": 70081295685120.0, + "grad_norm": 0.8418263299661153, + "learning_rate": 1.638222802451767e-06, + "loss": 0.6029, + "num_input_tokens_seen": 102578070, + "step": 4750 + }, + { + "epoch": 0.5712739734263211, + "flos": 24682111913280.0, + "grad_norm": 2.1945560550340297, + "learning_rate": 1.6374567084988561e-06, + "loss": 0.7506, + "num_input_tokens_seen": 102599255, + "step": 4751 + }, + { + "epoch": 0.5713942163169603, + "flos": 26540074592160.0, + "grad_norm": 2.9577814760510623, + "learning_rate": 1.6366906695434738e-06, + "loss": 0.7644, + "num_input_tokens_seen": 102621775, + "step": 4752 + }, + { + "epoch": 0.5715144592075994, + "flos": 21112286646720.0, + "grad_norm": 3.2437543277307355, + "learning_rate": 1.6359246857018275e-06, + "loss": 0.8578, + "num_input_tokens_seen": 102639305, + "step": 4753 + }, + { + "epoch": 0.5716347020982384, + "flos": 23296874439360.0, + "grad_norm": 2.4802264237868954, + "learning_rate": 1.6351587570901178e-06, + "loss": 0.7803, + "num_input_tokens_seen": 102659430, + "step": 4754 + }, + { + "epoch": 0.5717549449888776, + "flos": 16985051994720.0, + "grad_norm": 3.5026356446318263, + "learning_rate": 1.634392883824534e-06, + "loss": 0.7585, + "num_input_tokens_seen": 102671430, + "step": 4755 + }, + { + "epoch": 0.5718751878795166, + "flos": 35467053792960.0, + "grad_norm": 2.1536174681585667, + "learning_rate": 1.6336270660212595e-06, + "loss": 0.6791, + "num_input_tokens_seen": 102693025, + "step": 4756 + }, + { + "epoch": 0.5719954307701557, + "flos": 38558206692960.0, + "grad_norm": 2.6937925614183134, + "learning_rate": 1.6328613037964676e-06, + "loss": 0.661, + "num_input_tokens_seen": 102716090, + "step": 4757 + }, + { + "epoch": 0.5721156736607949, + "flos": 20601679722720.0, + "grad_norm": 2.4399745802258757, + "learning_rate": 1.6320955972663241e-06, + "loss": 0.6805, + "num_input_tokens_seen": 102735685, + "step": 4758 + }, + { + "epoch": 0.5722359165514339, + "flos": 37361225551680.0, + "grad_norm": 3.0529333673347248, + "learning_rate": 1.6313299465469857e-06, + "loss": 0.6523, + "num_input_tokens_seen": 102758415, + "step": 4759 + }, + { + "epoch": 0.572356159442073, + "flos": 21941192658240.0, + "grad_norm": 2.8169770951915787, + "learning_rate": 1.6305643517546014e-06, + "loss": 0.7985, + "num_input_tokens_seen": 102773795, + "step": 4760 + }, + { + "epoch": 0.5724764023327121, + "flos": 19108298216160.0, + "grad_norm": 2.321892300075396, + "learning_rate": 1.629798813005311e-06, + "loss": 0.8456, + "num_input_tokens_seen": 102793515, + "step": 4761 + }, + { + "epoch": 0.5725966452233512, + "flos": 22787122419840.0, + "grad_norm": 2.373639469958213, + "learning_rate": 1.6290333304152473e-06, + "loss": 0.7087, + "num_input_tokens_seen": 102813390, + "step": 4762 + }, + { + "epoch": 0.5727168881139902, + "flos": 41436639824640.0, + "grad_norm": 2.8875711341451895, + "learning_rate": 1.6282679041005314e-06, + "loss": 0.5675, + "num_input_tokens_seen": 102838980, + "step": 4763 + }, + { + "epoch": 0.5728371310046293, + "flos": 14626964264160.0, + "grad_norm": 2.801871753820938, + "learning_rate": 1.6275025341772789e-06, + "loss": 0.8699, + "num_input_tokens_seen": 102855285, + "step": 4764 + }, + { + "epoch": 0.5729573738952685, + "flos": 21475901405280.0, + "grad_norm": 8.251660266622816, + "learning_rate": 1.626737220761596e-06, + "loss": 0.8196, + "num_input_tokens_seen": 102872750, + "step": 4765 + }, + { + "epoch": 0.5730776167859075, + "flos": 23587848019680.0, + "grad_norm": 3.640102889315054, + "learning_rate": 1.62597196396958e-06, + "loss": 0.7913, + "num_input_tokens_seen": 102890920, + "step": 4766 + }, + { + "epoch": 0.5731978596765466, + "flos": 25702359347520.0, + "grad_norm": 2.618800972599509, + "learning_rate": 1.6252067639173197e-06, + "loss": 0.8523, + "num_input_tokens_seen": 102912830, + "step": 4767 + }, + { + "epoch": 0.5733181025671857, + "flos": 26321742190080.0, + "grad_norm": 3.5075107096306675, + "learning_rate": 1.6244416207208956e-06, + "loss": 0.6951, + "num_input_tokens_seen": 102933765, + "step": 4768 + }, + { + "epoch": 0.5734383454578248, + "flos": 29381629588320.0, + "grad_norm": 2.2448790873227935, + "learning_rate": 1.6236765344963787e-06, + "loss": 0.7328, + "num_input_tokens_seen": 102955025, + "step": 4769 + }, + { + "epoch": 0.5735585883484638, + "flos": 34919643072960.0, + "grad_norm": 2.784068704491785, + "learning_rate": 1.6229115053598322e-06, + "loss": 0.6976, + "num_input_tokens_seen": 102976780, + "step": 4770 + }, + { + "epoch": 0.573678831239103, + "flos": 18744051571680.0, + "grad_norm": 3.4796433417298003, + "learning_rate": 1.6221465334273108e-06, + "loss": 0.7215, + "num_input_tokens_seen": 102995990, + "step": 4771 + }, + { + "epoch": 0.5737990741297421, + "flos": 25666075928160.0, + "grad_norm": 5.497721633641589, + "learning_rate": 1.6213816188148593e-06, + "loss": 0.6154, + "num_input_tokens_seen": 103014695, + "step": 4772 + }, + { + "epoch": 0.5739193170203811, + "flos": 27230054276160.0, + "grad_norm": 2.1373267834174934, + "learning_rate": 1.6206167616385162e-06, + "loss": 0.7704, + "num_input_tokens_seen": 103035760, + "step": 4773 + }, + { + "epoch": 0.5740395599110203, + "flos": 12222520109280.0, + "grad_norm": 2.66528748570191, + "learning_rate": 1.6198519620143078e-06, + "loss": 0.7405, + "num_input_tokens_seen": 103052915, + "step": 4774 + }, + { + "epoch": 0.5741598028016593, + "flos": 25884259651200.0, + "grad_norm": 1.9297598388950834, + "learning_rate": 1.6190872200582546e-06, + "loss": 0.7822, + "num_input_tokens_seen": 103074655, + "step": 4775 + }, + { + "epoch": 0.5742800456922984, + "flos": 19216702437120.0, + "grad_norm": 2.7639593076015334, + "learning_rate": 1.6183225358863676e-06, + "loss": 0.7779, + "num_input_tokens_seen": 103091550, + "step": 4776 + }, + { + "epoch": 0.5744002885829376, + "flos": 30876089017920.0, + "grad_norm": 3.3636657282416835, + "learning_rate": 1.617557909614648e-06, + "loss": 0.7178, + "num_input_tokens_seen": 103110985, + "step": 4777 + }, + { + "epoch": 0.5745205314735766, + "flos": 23805845893920.0, + "grad_norm": 2.1932794789057577, + "learning_rate": 1.6167933413590899e-06, + "loss": 0.8611, + "num_input_tokens_seen": 103129890, + "step": 4778 + }, + { + "epoch": 0.5746407743642157, + "flos": 12294566571360.0, + "grad_norm": 4.856221736791347, + "learning_rate": 1.6160288312356773e-06, + "loss": 0.9128, + "num_input_tokens_seen": 103147020, + "step": 4779 + }, + { + "epoch": 0.5747610172548548, + "flos": 24099123999360.0, + "grad_norm": 3.714165177328216, + "learning_rate": 1.6152643793603857e-06, + "loss": 0.8178, + "num_input_tokens_seen": 103167370, + "step": 4780 + }, + { + "epoch": 0.5748812601454939, + "flos": 25371980088000.0, + "grad_norm": 2.3945531548498495, + "learning_rate": 1.6144999858491815e-06, + "loss": 0.8774, + "num_input_tokens_seen": 103186000, + "step": 4781 + }, + { + "epoch": 0.575001503036133, + "flos": 30580840915200.0, + "grad_norm": 2.359357416959006, + "learning_rate": 1.6137356508180232e-06, + "loss": 0.8587, + "num_input_tokens_seen": 103207710, + "step": 4782 + }, + { + "epoch": 0.5751217459267721, + "flos": 21695571918720.0, + "grad_norm": 2.1239463213371543, + "learning_rate": 1.6129713743828593e-06, + "loss": 0.8163, + "num_input_tokens_seen": 103226720, + "step": 4783 + }, + { + "epoch": 0.5752419888174112, + "flos": 21620663385120.0, + "grad_norm": 2.0224057663889656, + "learning_rate": 1.6122071566596306e-06, + "loss": 0.7511, + "num_input_tokens_seen": 103246995, + "step": 4784 + }, + { + "epoch": 0.5753622317080502, + "flos": 17750386249440.0, + "grad_norm": 3.119764902109387, + "learning_rate": 1.6114429977642674e-06, + "loss": 0.8343, + "num_input_tokens_seen": 103261500, + "step": 4785 + }, + { + "epoch": 0.5754824745986894, + "flos": 19761325425120.0, + "grad_norm": 2.673755987516443, + "learning_rate": 1.6106788978126926e-06, + "loss": 0.7367, + "num_input_tokens_seen": 103280475, + "step": 4786 + }, + { + "epoch": 0.5756027174893285, + "flos": 30940887376800.0, + "grad_norm": 10.420283022970565, + "learning_rate": 1.6099148569208196e-06, + "loss": 0.792, + "num_input_tokens_seen": 103297695, + "step": 4787 + }, + { + "epoch": 0.5757229603799675, + "flos": 28505846775840.0, + "grad_norm": 2.6391928135265554, + "learning_rate": 1.6091508752045523e-06, + "loss": 0.627, + "num_input_tokens_seen": 103318575, + "step": 4788 + }, + { + "epoch": 0.5758432032706067, + "flos": 22966011972960.0, + "grad_norm": 1.8887099348018515, + "learning_rate": 1.608386952779787e-06, + "loss": 0.8647, + "num_input_tokens_seen": 103337945, + "step": 4789 + }, + { + "epoch": 0.5759634461612457, + "flos": 25702879724160.0, + "grad_norm": 3.6257865315558924, + "learning_rate": 1.6076230897624098e-06, + "loss": 0.7475, + "num_input_tokens_seen": 103360150, + "step": 4790 + }, + { + "epoch": 0.5760836890518848, + "flos": 30547828434720.0, + "grad_norm": 4.534381074066616, + "learning_rate": 1.6068592862682974e-06, + "loss": 0.7734, + "num_input_tokens_seen": 103378305, + "step": 4791 + }, + { + "epoch": 0.576203931942524, + "flos": 36486371983200.0, + "grad_norm": 2.4086125653051993, + "learning_rate": 1.6060955424133187e-06, + "loss": 0.7392, + "num_input_tokens_seen": 103399230, + "step": 4792 + }, + { + "epoch": 0.576324174833163, + "flos": 25480272799680.0, + "grad_norm": 2.1570517113809045, + "learning_rate": 1.6053318583133332e-06, + "loss": 0.8965, + "num_input_tokens_seen": 103420100, + "step": 4793 + }, + { + "epoch": 0.5764444177238021, + "flos": 25083311032800.0, + "grad_norm": 2.604377026009006, + "learning_rate": 1.6045682340841907e-06, + "loss": 0.7537, + "num_input_tokens_seen": 103439025, + "step": 4794 + }, + { + "epoch": 0.5765646606144411, + "flos": 62123598444000.0, + "grad_norm": 0.7951067960130844, + "learning_rate": 1.6038046698417336e-06, + "loss": 0.5782, + "num_input_tokens_seen": 103499920, + "step": 4795 + }, + { + "epoch": 0.5766849035050803, + "flos": 25082790656160.0, + "grad_norm": 3.486053473075058, + "learning_rate": 1.6030411657017919e-06, + "loss": 0.69, + "num_input_tokens_seen": 103519575, + "step": 4796 + }, + { + "epoch": 0.5768051463957193, + "flos": 15968187008640.0, + "grad_norm": 2.2087687177140234, + "learning_rate": 1.6022777217801903e-06, + "loss": 0.8431, + "num_input_tokens_seen": 103536530, + "step": 4797 + }, + { + "epoch": 0.5769253892863584, + "flos": 22131976534560.0, + "grad_norm": 2.3862482034747403, + "learning_rate": 1.601514338192742e-06, + "loss": 0.7403, + "num_input_tokens_seen": 103556460, + "step": 4798 + }, + { + "epoch": 0.5770456321769976, + "flos": 22823703197280.0, + "grad_norm": 3.196552810020278, + "learning_rate": 1.6007510150552514e-06, + "loss": 0.7126, + "num_input_tokens_seen": 103574520, + "step": 4799 + }, + { + "epoch": 0.5771658750676366, + "flos": 46286792301600.0, + "grad_norm": 2.1536916399646704, + "learning_rate": 1.599987752483515e-06, + "loss": 0.6246, + "num_input_tokens_seen": 103598965, + "step": 4800 + }, + { + "epoch": 0.5772861179582757, + "flos": 22128185219040.0, + "grad_norm": 2.7354523198597476, + "learning_rate": 1.5992245505933184e-06, + "loss": 0.68, + "num_input_tokens_seen": 103618665, + "step": 4801 + }, + { + "epoch": 0.5774063608489148, + "flos": 31204052243040.0, + "grad_norm": 9.86497548877051, + "learning_rate": 1.5984614095004388e-06, + "loss": 0.7135, + "num_input_tokens_seen": 103639800, + "step": 4802 + }, + { + "epoch": 0.5775266037395539, + "flos": 22494885067680.0, + "grad_norm": 3.8052749963506662, + "learning_rate": 1.5976983293206438e-06, + "loss": 0.8102, + "num_input_tokens_seen": 103655605, + "step": 4803 + }, + { + "epoch": 0.577646846630193, + "flos": 21038716224480.0, + "grad_norm": 2.4417357199563674, + "learning_rate": 1.5969353101696928e-06, + "loss": 0.7154, + "num_input_tokens_seen": 103674045, + "step": 4804 + }, + { + "epoch": 0.5777670895208321, + "flos": 29671562415360.0, + "grad_norm": 3.0223459249623823, + "learning_rate": 1.5961723521633341e-06, + "loss": 0.8, + "num_input_tokens_seen": 103695285, + "step": 4805 + }, + { + "epoch": 0.5778873324114712, + "flos": 19472656369920.0, + "grad_norm": 2.6991503826928787, + "learning_rate": 1.5954094554173097e-06, + "loss": 0.9088, + "num_input_tokens_seen": 103713275, + "step": 4806 + }, + { + "epoch": 0.5780075753021102, + "flos": 14116877716800.0, + "grad_norm": 3.006776309188719, + "learning_rate": 1.5946466200473482e-06, + "loss": 0.7851, + "num_input_tokens_seen": 103731185, + "step": 4807 + }, + { + "epoch": 0.5781278181927494, + "flos": 15240362775360.0, + "grad_norm": 2.2167770855394227, + "learning_rate": 1.5938838461691723e-06, + "loss": 0.8331, + "num_input_tokens_seen": 103748095, + "step": 4808 + }, + { + "epoch": 0.5782480610833884, + "flos": 16703073496320.0, + "grad_norm": 8.670380554647087, + "learning_rate": 1.593121133898494e-06, + "loss": 0.8325, + "num_input_tokens_seen": 103765300, + "step": 4809 + }, + { + "epoch": 0.5783683039740275, + "flos": 25445587680000.0, + "grad_norm": 4.624399507047509, + "learning_rate": 1.592358483351016e-06, + "loss": 0.7899, + "num_input_tokens_seen": 103785475, + "step": 4810 + }, + { + "epoch": 0.5784885468646667, + "flos": 18379656248160.0, + "grad_norm": 3.0201716959063614, + "learning_rate": 1.5915958946424326e-06, + "loss": 0.7227, + "num_input_tokens_seen": 103804160, + "step": 4811 + }, + { + "epoch": 0.5786087897553057, + "flos": 46034295156480.0, + "grad_norm": 2.1563175577830975, + "learning_rate": 1.5908333678884271e-06, + "loss": 0.7468, + "num_input_tokens_seen": 103830580, + "step": 4812 + }, + { + "epoch": 0.5787290326459448, + "flos": 12368545860960.0, + "grad_norm": 2.3188054492006605, + "learning_rate": 1.5900709032046743e-06, + "loss": 0.7347, + "num_input_tokens_seen": 103848655, + "step": 4813 + }, + { + "epoch": 0.5788492755365839, + "flos": 23256874044000.0, + "grad_norm": 3.514910409146465, + "learning_rate": 1.5893085007068391e-06, + "loss": 0.7846, + "num_input_tokens_seen": 103866215, + "step": 4814 + }, + { + "epoch": 0.578969518427223, + "flos": 24026631500160.0, + "grad_norm": 2.4107582169016633, + "learning_rate": 1.5885461605105786e-06, + "loss": 0.7066, + "num_input_tokens_seen": 103887815, + "step": 4815 + }, + { + "epoch": 0.579089761317862, + "flos": 21840594086880.0, + "grad_norm": 2.7916891941660933, + "learning_rate": 1.5877838827315375e-06, + "loss": 0.7691, + "num_input_tokens_seen": 103906360, + "step": 4816 + }, + { + "epoch": 0.5792100042085012, + "flos": 22896901921920.0, + "grad_norm": 2.252776763713006, + "learning_rate": 1.587021667485355e-06, + "loss": 0.703, + "num_input_tokens_seen": 103925730, + "step": 4817 + }, + { + "epoch": 0.5793302470991403, + "flos": 21439394967360.0, + "grad_norm": 2.4660996322305486, + "learning_rate": 1.5862595148876559e-06, + "loss": 0.7852, + "num_input_tokens_seen": 103945835, + "step": 4818 + }, + { + "epoch": 0.5794504899897793, + "flos": 12693015128640.0, + "grad_norm": 3.7173842799345604, + "learning_rate": 1.58549742505406e-06, + "loss": 0.7631, + "num_input_tokens_seen": 103960580, + "step": 4819 + }, + { + "epoch": 0.5795707328804185, + "flos": 14846040061440.0, + "grad_norm": 3.477313631330543, + "learning_rate": 1.5847353981001747e-06, + "loss": 0.7593, + "num_input_tokens_seen": 103977760, + "step": 4820 + }, + { + "epoch": 0.5796909757710575, + "flos": 36377633234400.0, + "grad_norm": 2.7332385508068, + "learning_rate": 1.5839734341415993e-06, + "loss": 0.6955, + "num_input_tokens_seen": 103999960, + "step": 4821 + }, + { + "epoch": 0.5798112186616966, + "flos": 23006978782080.0, + "grad_norm": 2.309973702879668, + "learning_rate": 1.5832115332939238e-06, + "loss": 0.7671, + "num_input_tokens_seen": 104018400, + "step": 4822 + }, + { + "epoch": 0.5799314615523358, + "flos": 16628908357920.0, + "grad_norm": 2.4503253097987483, + "learning_rate": 1.5824496956727272e-06, + "loss": 0.7458, + "num_input_tokens_seen": 104035605, + "step": 4823 + }, + { + "epoch": 0.5800517044429748, + "flos": 20456471705760.0, + "grad_norm": 2.113843014665331, + "learning_rate": 1.5816879213935797e-06, + "loss": 0.7333, + "num_input_tokens_seen": 104054415, + "step": 4824 + }, + { + "epoch": 0.5801719473336139, + "flos": 31492869977280.0, + "grad_norm": 2.006226027871348, + "learning_rate": 1.5809262105720416e-06, + "loss": 0.797, + "num_input_tokens_seen": 104075455, + "step": 4825 + }, + { + "epoch": 0.580292190224253, + "flos": 20346915222240.0, + "grad_norm": 2.0151778323927125, + "learning_rate": 1.5801645633236644e-06, + "loss": 0.7925, + "num_input_tokens_seen": 104096440, + "step": 4826 + }, + { + "epoch": 0.5804124331148921, + "flos": 26577250085760.0, + "grad_norm": 5.144312501281376, + "learning_rate": 1.579402979763989e-06, + "loss": 0.7736, + "num_input_tokens_seen": 104116775, + "step": 4827 + }, + { + "epoch": 0.5805326760055312, + "flos": 13459427306400.0, + "grad_norm": 4.51630626146129, + "learning_rate": 1.578641460008548e-06, + "loss": 0.8126, + "num_input_tokens_seen": 104134510, + "step": 4828 + }, + { + "epoch": 0.5806529188961702, + "flos": 12074301341760.0, + "grad_norm": 3.115382233462098, + "learning_rate": 1.5778800041728613e-06, + "loss": 0.6808, + "num_input_tokens_seen": 104150800, + "step": 4829 + }, + { + "epoch": 0.5807731617868094, + "flos": 26176571342880.0, + "grad_norm": 2.0734672262537424, + "learning_rate": 1.577118612372443e-06, + "loss": 0.6617, + "num_input_tokens_seen": 104172640, + "step": 4830 + }, + { + "epoch": 0.5808934046774484, + "flos": 37907484009120.0, + "grad_norm": 3.425582479802192, + "learning_rate": 1.5763572847227943e-06, + "loss": 0.7078, + "num_input_tokens_seen": 104193525, + "step": 4831 + }, + { + "epoch": 0.5810136475680875, + "flos": 20456657554560.0, + "grad_norm": 2.660501376490962, + "learning_rate": 1.5755960213394091e-06, + "loss": 0.8104, + "num_input_tokens_seen": 104210625, + "step": 4832 + }, + { + "epoch": 0.5811338904587267, + "flos": 17505248716800.0, + "grad_norm": 2.5951382701985217, + "learning_rate": 1.5748348223377703e-06, + "loss": 0.7816, + "num_input_tokens_seen": 104228180, + "step": 4833 + }, + { + "epoch": 0.5812541333493657, + "flos": 19430091261120.0, + "grad_norm": 2.1977449563977123, + "learning_rate": 1.5740736878333507e-06, + "loss": 0.7819, + "num_input_tokens_seen": 104246020, + "step": 4834 + }, + { + "epoch": 0.5813743762400048, + "flos": 20565582152160.0, + "grad_norm": 4.466499241969669, + "learning_rate": 1.5733126179416143e-06, + "loss": 0.7825, + "num_input_tokens_seen": 104260505, + "step": 4835 + }, + { + "epoch": 0.5814946191306439, + "flos": 33130381577760.0, + "grad_norm": 3.0074346924347757, + "learning_rate": 1.5725516127780137e-06, + "loss": 0.7279, + "num_input_tokens_seen": 104282640, + "step": 4836 + }, + { + "epoch": 0.581614862021283, + "flos": 16120382940480.0, + "grad_norm": 3.437784911399475, + "learning_rate": 1.5717906724579943e-06, + "loss": 0.8871, + "num_input_tokens_seen": 104299700, + "step": 4837 + }, + { + "epoch": 0.581735104911922, + "flos": 33754038942720.0, + "grad_norm": 2.779341482205278, + "learning_rate": 1.571029797096989e-06, + "loss": 0.6822, + "num_input_tokens_seen": 104320805, + "step": 4838 + }, + { + "epoch": 0.5818553478025612, + "flos": 23297803683360.0, + "grad_norm": 2.7547916801602845, + "learning_rate": 1.570268986810423e-06, + "loss": 0.7899, + "num_input_tokens_seen": 104340700, + "step": 4839 + }, + { + "epoch": 0.5819755906932003, + "flos": 20966186555520.0, + "grad_norm": 3.180033297094526, + "learning_rate": 1.5695082417137096e-06, + "loss": 0.752, + "num_input_tokens_seen": 104358575, + "step": 4840 + }, + { + "epoch": 0.5820958335838393, + "flos": 21400992871680.0, + "grad_norm": 2.358315135045374, + "learning_rate": 1.5687475619222539e-06, + "loss": 0.7551, + "num_input_tokens_seen": 104378530, + "step": 4841 + }, + { + "epoch": 0.5822160764744785, + "flos": 17941913520960.0, + "grad_norm": 3.069958848459991, + "learning_rate": 1.5679869475514496e-06, + "loss": 0.7336, + "num_input_tokens_seen": 104395465, + "step": 4842 + }, + { + "epoch": 0.5823363193651175, + "flos": 23000734262400.0, + "grad_norm": 3.052485538232836, + "learning_rate": 1.567226398716682e-06, + "loss": 0.8178, + "num_input_tokens_seen": 104415380, + "step": 4843 + }, + { + "epoch": 0.5824565622557566, + "flos": 32844611763840.0, + "grad_norm": 2.6397197006716477, + "learning_rate": 1.566465915533326e-06, + "loss": 0.6192, + "num_input_tokens_seen": 104437410, + "step": 4844 + }, + { + "epoch": 0.5825768051463958, + "flos": 22197666967680.0, + "grad_norm": 3.4753408964616193, + "learning_rate": 1.5657054981167458e-06, + "loss": 0.8855, + "num_input_tokens_seen": 104456305, + "step": 4845 + }, + { + "epoch": 0.5826970480370348, + "flos": 27961409636640.0, + "grad_norm": 3.1944581920678097, + "learning_rate": 1.5649451465822965e-06, + "loss": 0.6793, + "num_input_tokens_seen": 104477695, + "step": 4846 + }, + { + "epoch": 0.5828172909276739, + "flos": 17832654395520.0, + "grad_norm": 3.3559658050977443, + "learning_rate": 1.5641848610453218e-06, + "loss": 0.8376, + "num_input_tokens_seen": 104496355, + "step": 4847 + }, + { + "epoch": 0.582937533818313, + "flos": 19836940184160.0, + "grad_norm": 2.5047453665994603, + "learning_rate": 1.563424641621158e-06, + "loss": 0.8608, + "num_input_tokens_seen": 104515535, + "step": 4848 + }, + { + "epoch": 0.5830577767089521, + "flos": 26832237604800.0, + "grad_norm": 3.4147899559739265, + "learning_rate": 1.5626644884251282e-06, + "loss": 0.6979, + "num_input_tokens_seen": 104535055, + "step": 4849 + }, + { + "epoch": 0.5831780195995911, + "flos": 25261866058080.0, + "grad_norm": 2.242480428348729, + "learning_rate": 1.5619044015725488e-06, + "loss": 0.8792, + "num_input_tokens_seen": 104554745, + "step": 4850 + }, + { + "epoch": 0.5832982624902303, + "flos": 14736111880320.0, + "grad_norm": 2.8696803759661247, + "learning_rate": 1.5611443811787224e-06, + "loss": 0.8647, + "num_input_tokens_seen": 104568870, + "step": 4851 + }, + { + "epoch": 0.5834185053808694, + "flos": 20415244708320.0, + "grad_norm": 2.936511583624645, + "learning_rate": 1.560384427358945e-06, + "loss": 0.6908, + "num_input_tokens_seen": 104588415, + "step": 4852 + }, + { + "epoch": 0.5835387482715084, + "flos": 27161464601760.0, + "grad_norm": 2.2997204030136693, + "learning_rate": 1.5596245402284998e-06, + "loss": 0.7308, + "num_input_tokens_seen": 104611940, + "step": 4853 + }, + { + "epoch": 0.5836589911621476, + "flos": 16630989864480.0, + "grad_norm": 2.1134176439421126, + "learning_rate": 1.5588647199026619e-06, + "loss": 0.8158, + "num_input_tokens_seen": 104629675, + "step": 4854 + }, + { + "epoch": 0.5837792340527866, + "flos": 20417586403200.0, + "grad_norm": 3.3495679491334127, + "learning_rate": 1.5581049664966956e-06, + "loss": 0.8767, + "num_input_tokens_seen": 104647435, + "step": 4855 + }, + { + "epoch": 0.5838994769434257, + "flos": 65900456505600.0, + "grad_norm": 0.9933630819847792, + "learning_rate": 1.5573452801258545e-06, + "loss": 0.6495, + "num_input_tokens_seen": 104694605, + "step": 4856 + }, + { + "epoch": 0.5840197198340649, + "flos": 21439878174240.0, + "grad_norm": 2.4429089125626695, + "learning_rate": 1.5565856609053824e-06, + "loss": 0.6315, + "num_input_tokens_seen": 104713400, + "step": 4857 + }, + { + "epoch": 0.5841399627247039, + "flos": 19107889348800.0, + "grad_norm": 2.65728640813144, + "learning_rate": 1.5558261089505127e-06, + "loss": 0.7979, + "num_input_tokens_seen": 104732130, + "step": 4858 + }, + { + "epoch": 0.584260205615343, + "flos": 26387581302240.0, + "grad_norm": 2.4025067033783345, + "learning_rate": 1.5550666243764697e-06, + "loss": 0.7971, + "num_input_tokens_seen": 104750290, + "step": 4859 + }, + { + "epoch": 0.584380448505982, + "flos": 13862224725600.0, + "grad_norm": 3.089041469504316, + "learning_rate": 1.554307207298465e-06, + "loss": 0.7709, + "num_input_tokens_seen": 104768550, + "step": 4860 + }, + { + "epoch": 0.5845006913966212, + "flos": 21512965389600.0, + "grad_norm": 3.542525527635988, + "learning_rate": 1.553547857831704e-06, + "loss": 0.7886, + "num_input_tokens_seen": 104785060, + "step": 4861 + }, + { + "epoch": 0.5846209342872603, + "flos": 58291723404000.0, + "grad_norm": 0.8850559996005304, + "learning_rate": 1.5527885760913771e-06, + "loss": 0.6411, + "num_input_tokens_seen": 104834950, + "step": 4862 + }, + { + "epoch": 0.5847411771778993, + "flos": 18561593721600.0, + "grad_norm": 1.8632678785184191, + "learning_rate": 1.552029362192668e-06, + "loss": 0.7653, + "num_input_tokens_seen": 104854210, + "step": 4863 + }, + { + "epoch": 0.5848614200685385, + "flos": 24206004260160.0, + "grad_norm": 2.43731767021394, + "learning_rate": 1.5512702162507478e-06, + "loss": 0.726, + "num_input_tokens_seen": 104874525, + "step": 4864 + }, + { + "epoch": 0.5849816629591775, + "flos": 71557318913760.0, + "grad_norm": 1.0994705464892414, + "learning_rate": 1.5505111383807792e-06, + "loss": 0.5574, + "num_input_tokens_seen": 104937195, + "step": 4865 + }, + { + "epoch": 0.5851019058498166, + "flos": 23767927005120.0, + "grad_norm": 2.682181129523557, + "learning_rate": 1.5497521286979138e-06, + "loss": 0.8082, + "num_input_tokens_seen": 104957435, + "step": 4866 + }, + { + "epoch": 0.5852221487404557, + "flos": 24353591141760.0, + "grad_norm": 2.4176647383410756, + "learning_rate": 1.5489931873172927e-06, + "loss": 0.7435, + "num_input_tokens_seen": 104978755, + "step": 4867 + }, + { + "epoch": 0.5853423916310948, + "flos": 27232321631520.0, + "grad_norm": 2.256360388932236, + "learning_rate": 1.5482343143540467e-06, + "loss": 0.7898, + "num_input_tokens_seen": 105000020, + "step": 4868 + }, + { + "epoch": 0.5854626345217339, + "flos": 11966937874080.0, + "grad_norm": 2.3535944375419104, + "learning_rate": 1.547475509923295e-06, + "loss": 0.8288, + "num_input_tokens_seen": 105017060, + "step": 4869 + }, + { + "epoch": 0.585582877412373, + "flos": 64250635980960.0, + "grad_norm": 0.7428133684730187, + "learning_rate": 1.5467167741401495e-06, + "loss": 0.5596, + "num_input_tokens_seen": 105078540, + "step": 4870 + }, + { + "epoch": 0.5857031203030121, + "flos": 16987505198880.0, + "grad_norm": 2.571758570764618, + "learning_rate": 1.5459581071197083e-06, + "loss": 0.716, + "num_input_tokens_seen": 105094200, + "step": 4871 + }, + { + "epoch": 0.5858233631936511, + "flos": 20855849507040.0, + "grad_norm": 4.276674596274188, + "learning_rate": 1.5451995089770624e-06, + "loss": 0.8345, + "num_input_tokens_seen": 105113985, + "step": 4872 + }, + { + "epoch": 0.5859436060842903, + "flos": 23158988865120.0, + "grad_norm": 1.7339474950186835, + "learning_rate": 1.5444409798272885e-06, + "loss": 0.72, + "num_input_tokens_seen": 105138075, + "step": 4873 + }, + { + "epoch": 0.5860638489749294, + "flos": 22460534475840.0, + "grad_norm": 2.444282485432342, + "learning_rate": 1.543682519785456e-06, + "loss": 0.8074, + "num_input_tokens_seen": 105156555, + "step": 4874 + }, + { + "epoch": 0.5861840918655684, + "flos": 17541271947840.0, + "grad_norm": 3.7007031468530314, + "learning_rate": 1.5429241289666219e-06, + "loss": 0.7939, + "num_input_tokens_seen": 105174055, + "step": 4875 + }, + { + "epoch": 0.5863043347562076, + "flos": 25519529799840.0, + "grad_norm": 3.5606408873182342, + "learning_rate": 1.5421658074858342e-06, + "loss": 0.6985, + "num_input_tokens_seen": 105192160, + "step": 4876 + }, + { + "epoch": 0.5864245776468466, + "flos": 20638037481600.0, + "grad_norm": 3.558910636317996, + "learning_rate": 1.5414075554581298e-06, + "loss": 0.6658, + "num_input_tokens_seen": 105210205, + "step": 4877 + }, + { + "epoch": 0.5865448205374857, + "flos": 28872435115200.0, + "grad_norm": 3.2611103883908923, + "learning_rate": 1.5406493729985348e-06, + "loss": 0.7814, + "num_input_tokens_seen": 105229400, + "step": 4878 + }, + { + "epoch": 0.5866650634281249, + "flos": 25805336783520.0, + "grad_norm": 3.998033039943087, + "learning_rate": 1.5398912602220644e-06, + "loss": 0.721, + "num_input_tokens_seen": 105248675, + "step": 4879 + }, + { + "epoch": 0.5867853063187639, + "flos": 17027988801120.0, + "grad_norm": 2.8887124822830414, + "learning_rate": 1.539133217243724e-06, + "loss": 0.7921, + "num_input_tokens_seen": 105265695, + "step": 4880 + }, + { + "epoch": 0.586905549209403, + "flos": 24609768093120.0, + "grad_norm": 4.659999080543264, + "learning_rate": 1.5383752441785081e-06, + "loss": 0.7621, + "num_input_tokens_seen": 105284920, + "step": 4881 + }, + { + "epoch": 0.5870257921000421, + "flos": 14699716951680.0, + "grad_norm": 2.6177482345905894, + "learning_rate": 1.5376173411414003e-06, + "loss": 0.8537, + "num_input_tokens_seen": 105302960, + "step": 4882 + }, + { + "epoch": 0.5871460349906812, + "flos": 23880791597280.0, + "grad_norm": 2.7168428594014533, + "learning_rate": 1.5368595082473753e-06, + "loss": 0.7886, + "num_input_tokens_seen": 105321260, + "step": 4883 + }, + { + "epoch": 0.5872662778813202, + "flos": 22132534080960.0, + "grad_norm": 2.6528601164852135, + "learning_rate": 1.5361017456113935e-06, + "loss": 0.7821, + "num_input_tokens_seen": 105341610, + "step": 4884 + }, + { + "epoch": 0.5873865207719594, + "flos": 18415679479200.0, + "grad_norm": 2.535177033843204, + "learning_rate": 1.5353440533484085e-06, + "loss": 0.857, + "num_input_tokens_seen": 105360465, + "step": 4885 + }, + { + "epoch": 0.5875067636625985, + "flos": 53937272777280.0, + "grad_norm": 2.8612046355071423, + "learning_rate": 1.534586431573361e-06, + "loss": 0.6584, + "num_input_tokens_seen": 105385405, + "step": 4886 + }, + { + "epoch": 0.5876270065532375, + "flos": 27955350965760.0, + "grad_norm": 2.870050565161184, + "learning_rate": 1.5338288804011817e-06, + "loss": 0.7926, + "num_input_tokens_seen": 105404905, + "step": 4887 + }, + { + "epoch": 0.5877472494438767, + "flos": 21330990746400.0, + "grad_norm": 2.8953314330238022, + "learning_rate": 1.533071399946791e-06, + "loss": 0.7116, + "num_input_tokens_seen": 105423650, + "step": 4888 + }, + { + "epoch": 0.5878674923345157, + "flos": 22351052331840.0, + "grad_norm": 6.672651402058888, + "learning_rate": 1.5323139903250977e-06, + "loss": 0.5748, + "num_input_tokens_seen": 105444255, + "step": 4889 + }, + { + "epoch": 0.5879877352251548, + "flos": 21837323148000.0, + "grad_norm": 1.9568185900120938, + "learning_rate": 1.5315566516510002e-06, + "loss": 0.7683, + "num_input_tokens_seen": 105462425, + "step": 4890 + }, + { + "epoch": 0.5881079781157939, + "flos": 17468965297440.0, + "grad_norm": 3.1064887524960985, + "learning_rate": 1.5307993840393857e-06, + "loss": 0.6764, + "num_input_tokens_seen": 105480505, + "step": 4891 + }, + { + "epoch": 0.588228221006433, + "flos": 22569570582720.0, + "grad_norm": 2.300222435851483, + "learning_rate": 1.530042187605132e-06, + "loss": 0.8039, + "num_input_tokens_seen": 105499760, + "step": 4892 + }, + { + "epoch": 0.5883484638970721, + "flos": 26139433019040.0, + "grad_norm": 2.8526269885456124, + "learning_rate": 1.5292850624631044e-06, + "loss": 0.8426, + "num_input_tokens_seen": 105521955, + "step": 4893 + }, + { + "epoch": 0.5884687067877111, + "flos": 30399944195040.0, + "grad_norm": 2.348740922180466, + "learning_rate": 1.5285280087281593e-06, + "loss": 0.8092, + "num_input_tokens_seen": 105542685, + "step": 4894 + }, + { + "epoch": 0.5885889496783503, + "flos": 70406842875840.0, + "grad_norm": 0.6552363241632211, + "learning_rate": 1.5277710265151398e-06, + "loss": 0.5653, + "num_input_tokens_seen": 105612165, + "step": 4895 + }, + { + "epoch": 0.5887091925689893, + "flos": 19071568759680.0, + "grad_norm": 3.409494657016366, + "learning_rate": 1.5270141159388803e-06, + "loss": 0.7696, + "num_input_tokens_seen": 105629185, + "step": 4896 + }, + { + "epoch": 0.5888294354596284, + "flos": 23260962717600.0, + "grad_norm": 7.196036137776923, + "learning_rate": 1.526257277114203e-06, + "loss": 0.804, + "num_input_tokens_seen": 105650260, + "step": 4897 + }, + { + "epoch": 0.5889496783502676, + "flos": 21950150570400.0, + "grad_norm": 2.8310088739178583, + "learning_rate": 1.5255005101559201e-06, + "loss": 0.7993, + "num_input_tokens_seen": 105667870, + "step": 4898 + }, + { + "epoch": 0.5890699212409066, + "flos": 21654121902720.0, + "grad_norm": 2.7054452805323175, + "learning_rate": 1.524743815178833e-06, + "loss": 0.769, + "num_input_tokens_seen": 105685830, + "step": 4899 + }, + { + "epoch": 0.5891901641315457, + "flos": 19436856157440.0, + "grad_norm": 2.9786079967746186, + "learning_rate": 1.5239871922977315e-06, + "loss": 0.8093, + "num_input_tokens_seen": 105705745, + "step": 4900 + }, + { + "epoch": 0.5893104070221848, + "flos": 19581952665120.0, + "grad_norm": 2.163514766853423, + "learning_rate": 1.523230641627394e-06, + "loss": 0.8984, + "num_input_tokens_seen": 105724730, + "step": 4901 + }, + { + "epoch": 0.5894306499128239, + "flos": 29018386527360.0, + "grad_norm": 2.7560493517306264, + "learning_rate": 1.5224741632825888e-06, + "loss": 0.7314, + "num_input_tokens_seen": 105744920, + "step": 4902 + }, + { + "epoch": 0.589550892803463, + "flos": 42237737122080.0, + "grad_norm": 2.5729338198770844, + "learning_rate": 1.521717757378074e-06, + "loss": 0.6949, + "num_input_tokens_seen": 105765465, + "step": 4903 + }, + { + "epoch": 0.5896711356941021, + "flos": 14117100735360.0, + "grad_norm": 2.3277339734933156, + "learning_rate": 1.5209614240285943e-06, + "loss": 0.6887, + "num_input_tokens_seen": 105783120, + "step": 4904 + }, + { + "epoch": 0.5897913785847412, + "flos": 17176913794080.0, + "grad_norm": 3.2277424942979995, + "learning_rate": 1.520205163348887e-06, + "loss": 0.8502, + "num_input_tokens_seen": 105801055, + "step": 4905 + }, + { + "epoch": 0.5899116214753802, + "flos": 48724912259040.0, + "grad_norm": 0.7376586872651234, + "learning_rate": 1.519448975453674e-06, + "loss": 0.569, + "num_input_tokens_seen": 105856155, + "step": 4906 + }, + { + "epoch": 0.5900318643660194, + "flos": 21073475683680.0, + "grad_norm": 3.7060754024042284, + "learning_rate": 1.5186928604576696e-06, + "loss": 0.7617, + "num_input_tokens_seen": 105873350, + "step": 4907 + }, + { + "epoch": 0.5901521072566585, + "flos": 21148161198720.0, + "grad_norm": 2.8257252327721973, + "learning_rate": 1.5179368184755752e-06, + "loss": 0.774, + "num_input_tokens_seen": 105891435, + "step": 4908 + }, + { + "epoch": 0.5902723501472975, + "flos": 20196986645760.0, + "grad_norm": 3.0316020328565747, + "learning_rate": 1.5171808496220821e-06, + "loss": 0.8273, + "num_input_tokens_seen": 105910190, + "step": 4909 + }, + { + "epoch": 0.5903925930379367, + "flos": 22931512702080.0, + "grad_norm": 14.07122884474458, + "learning_rate": 1.5164249540118708e-06, + "loss": 0.8146, + "num_input_tokens_seen": 105929315, + "step": 4910 + }, + { + "epoch": 0.5905128359285757, + "flos": 23334310121280.0, + "grad_norm": 2.6878041827141126, + "learning_rate": 1.5156691317596093e-06, + "loss": 0.8354, + "num_input_tokens_seen": 105950740, + "step": 4911 + }, + { + "epoch": 0.5906330788192148, + "flos": 27991931743200.0, + "grad_norm": 3.059426770342107, + "learning_rate": 1.5149133829799556e-06, + "loss": 0.6694, + "num_input_tokens_seen": 105968735, + "step": 4912 + }, + { + "epoch": 0.590753321709854, + "flos": 18451628370720.0, + "grad_norm": 3.2405144108853707, + "learning_rate": 1.5141577077875556e-06, + "loss": 0.8047, + "num_input_tokens_seen": 105986060, + "step": 4913 + }, + { + "epoch": 0.590873564600493, + "flos": 16849768303680.0, + "grad_norm": 3.318963420780133, + "learning_rate": 1.5134021062970451e-06, + "loss": 0.7212, + "num_input_tokens_seen": 106004440, + "step": 4914 + }, + { + "epoch": 0.5909938074911321, + "flos": 13497086006880.0, + "grad_norm": 2.415407565300823, + "learning_rate": 1.5126465786230483e-06, + "loss": 0.8084, + "num_input_tokens_seen": 106021215, + "step": 4915 + }, + { + "epoch": 0.5911140503817712, + "flos": 25986530861760.0, + "grad_norm": 3.16293367571163, + "learning_rate": 1.5118911248801787e-06, + "loss": 0.8207, + "num_input_tokens_seen": 106039225, + "step": 4916 + }, + { + "epoch": 0.5912342932724103, + "flos": 23225497032960.0, + "grad_norm": 2.7177558045769614, + "learning_rate": 1.5111357451830364e-06, + "loss": 0.7904, + "num_input_tokens_seen": 106057920, + "step": 4917 + }, + { + "epoch": 0.5913545361630493, + "flos": 19545148869120.0, + "grad_norm": 2.5762357320672686, + "learning_rate": 1.5103804396462131e-06, + "loss": 0.7137, + "num_input_tokens_seen": 106077855, + "step": 4918 + }, + { + "epoch": 0.5914747790536885, + "flos": 26175976626720.0, + "grad_norm": 11.272175274755298, + "learning_rate": 1.5096252083842877e-06, + "loss": 0.8033, + "num_input_tokens_seen": 106097065, + "step": 4919 + }, + { + "epoch": 0.5915950219443276, + "flos": 27378235873920.0, + "grad_norm": 2.9566409862597216, + "learning_rate": 1.5088700515118285e-06, + "loss": 0.8565, + "num_input_tokens_seen": 106116385, + "step": 4920 + }, + { + "epoch": 0.5917152648349666, + "flos": 21876580148160.0, + "grad_norm": 2.9104501870222235, + "learning_rate": 1.508114969143392e-06, + "loss": 0.6691, + "num_input_tokens_seen": 106136370, + "step": 4921 + }, + { + "epoch": 0.5918355077256057, + "flos": 28068810274080.0, + "grad_norm": 2.7637318311462433, + "learning_rate": 1.5073599613935238e-06, + "loss": 0.7755, + "num_input_tokens_seen": 106158490, + "step": 4922 + }, + { + "epoch": 0.5919557506162448, + "flos": 28143384279840.0, + "grad_norm": 2.382888342478969, + "learning_rate": 1.5066050283767574e-06, + "loss": 0.5754, + "num_input_tokens_seen": 106179170, + "step": 4923 + }, + { + "epoch": 0.5920759935068839, + "flos": 12077386431840.0, + "grad_norm": 3.061310361903375, + "learning_rate": 1.505850170207616e-06, + "loss": 0.8283, + "num_input_tokens_seen": 106196035, + "step": 4924 + }, + { + "epoch": 0.592196236397523, + "flos": 29382558832320.0, + "grad_norm": 2.800741410869901, + "learning_rate": 1.505095387000611e-06, + "loss": 0.7823, + "num_input_tokens_seen": 106218335, + "step": 4925 + }, + { + "epoch": 0.5923164792881621, + "flos": 24349651147200.0, + "grad_norm": 2.3826873246129754, + "learning_rate": 1.504340678870242e-06, + "loss": 0.7408, + "num_input_tokens_seen": 106236550, + "step": 4926 + }, + { + "epoch": 0.5924367221788012, + "flos": 23989864873920.0, + "grad_norm": 2.461319763527795, + "learning_rate": 1.5035860459309989e-06, + "loss": 0.9011, + "num_input_tokens_seen": 106254265, + "step": 4927 + }, + { + "epoch": 0.5925569650694402, + "flos": 26832088925760.0, + "grad_norm": 2.7431452021590426, + "learning_rate": 1.5028314882973568e-06, + "loss": 0.6382, + "num_input_tokens_seen": 106274670, + "step": 4928 + }, + { + "epoch": 0.5926772079600794, + "flos": 22270679843520.0, + "grad_norm": 2.226588734254714, + "learning_rate": 1.502077006083783e-06, + "loss": 0.8443, + "num_input_tokens_seen": 106293330, + "step": 4929 + }, + { + "epoch": 0.5927974508507184, + "flos": 19836940184160.0, + "grad_norm": 2.5551978485186746, + "learning_rate": 1.5013225994047315e-06, + "loss": 0.7629, + "num_input_tokens_seen": 106310960, + "step": 4930 + }, + { + "epoch": 0.5929176937413575, + "flos": 15754314977760.0, + "grad_norm": 2.1571237756323667, + "learning_rate": 1.5005682683746452e-06, + "loss": 0.8073, + "num_input_tokens_seen": 106329475, + "step": 4931 + }, + { + "epoch": 0.5930379366319967, + "flos": 17576403104640.0, + "grad_norm": 2.504684446336547, + "learning_rate": 1.4998140131079553e-06, + "loss": 0.7283, + "num_input_tokens_seen": 106345565, + "step": 4932 + }, + { + "epoch": 0.5931581795226357, + "flos": 17678822994240.0, + "grad_norm": 2.0516070928324934, + "learning_rate": 1.4990598337190821e-06, + "loss": 0.7342, + "num_input_tokens_seen": 106361920, + "step": 4933 + }, + { + "epoch": 0.5932784224132748, + "flos": 24245372769600.0, + "grad_norm": 3.1408295066636773, + "learning_rate": 1.4983057303224338e-06, + "loss": 0.6816, + "num_input_tokens_seen": 106381735, + "step": 4934 + }, + { + "epoch": 0.5933986653039139, + "flos": 22893816831840.0, + "grad_norm": 2.308053458515314, + "learning_rate": 1.4975517030324072e-06, + "loss": 0.8809, + "num_input_tokens_seen": 106399745, + "step": 4935 + }, + { + "epoch": 0.593518908194553, + "flos": 71019906859200.0, + "grad_norm": 0.7905201431521391, + "learning_rate": 1.4967977519633882e-06, + "loss": 0.616, + "num_input_tokens_seen": 106457205, + "step": 4936 + }, + { + "epoch": 0.593639151085192, + "flos": 20419482060960.0, + "grad_norm": 2.7499132142677953, + "learning_rate": 1.4960438772297494e-06, + "loss": 0.7861, + "num_input_tokens_seen": 106474250, + "step": 4937 + }, + { + "epoch": 0.5937593939758312, + "flos": 30839062203360.0, + "grad_norm": 3.69995986298107, + "learning_rate": 1.495290078945855e-06, + "loss": 0.7396, + "num_input_tokens_seen": 106494015, + "step": 4938 + }, + { + "epoch": 0.5938796368664703, + "flos": 36844745805600.0, + "grad_norm": 2.5852610461061127, + "learning_rate": 1.4945363572260529e-06, + "loss": 0.749, + "num_input_tokens_seen": 106515010, + "step": 4939 + }, + { + "epoch": 0.5939998797571093, + "flos": 23808856644480.0, + "grad_norm": 3.038190309551998, + "learning_rate": 1.4937827121846845e-06, + "loss": 0.6831, + "num_input_tokens_seen": 106535265, + "step": 4940 + }, + { + "epoch": 0.5941201226477485, + "flos": 25155320325120.0, + "grad_norm": 2.630827565839856, + "learning_rate": 1.4930291439360755e-06, + "loss": 0.7345, + "num_input_tokens_seen": 106557830, + "step": 4941 + }, + { + "epoch": 0.5942403655383875, + "flos": 22386964053600.0, + "grad_norm": 2.488389041308561, + "learning_rate": 1.4922756525945427e-06, + "loss": 0.7924, + "num_input_tokens_seen": 106577140, + "step": 4942 + }, + { + "epoch": 0.5943606084290266, + "flos": 67532944454880.0, + "grad_norm": 0.7814116882855048, + "learning_rate": 1.4915222382743894e-06, + "loss": 0.594, + "num_input_tokens_seen": 106635975, + "step": 4943 + }, + { + "epoch": 0.5944808513196658, + "flos": 18197235567840.0, + "grad_norm": 3.2541083265114863, + "learning_rate": 1.4907689010899085e-06, + "loss": 0.7253, + "num_input_tokens_seen": 106653555, + "step": 4944 + }, + { + "epoch": 0.5946010942103048, + "flos": 24755347807680.0, + "grad_norm": 4.796430834357908, + "learning_rate": 1.4900156411553804e-06, + "loss": 0.6231, + "num_input_tokens_seen": 106673385, + "step": 4945 + }, + { + "epoch": 0.5947213371009439, + "flos": 15210063687360.0, + "grad_norm": 2.4507475588291285, + "learning_rate": 1.4892624585850739e-06, + "loss": 0.8544, + "num_input_tokens_seen": 106691740, + "step": 4946 + }, + { + "epoch": 0.594841579991583, + "flos": 25811023756800.0, + "grad_norm": 2.7213693913424803, + "learning_rate": 1.4885093534932465e-06, + "loss": 0.7991, + "num_input_tokens_seen": 106709705, + "step": 4947 + }, + { + "epoch": 0.5949618228822221, + "flos": 23946556369920.0, + "grad_norm": 2.4989887099482324, + "learning_rate": 1.4877563259941433e-06, + "loss": 0.717, + "num_input_tokens_seen": 106727560, + "step": 4948 + }, + { + "epoch": 0.5950820657728612, + "flos": 40488884889600.0, + "grad_norm": 3.1180889766934405, + "learning_rate": 1.4870033762019988e-06, + "loss": 0.6787, + "num_input_tokens_seen": 106747725, + "step": 4949 + }, + { + "epoch": 0.5952023086635003, + "flos": 23151294724800.0, + "grad_norm": 1.8955026554782373, + "learning_rate": 1.4862505042310334e-06, + "loss": 0.7341, + "num_input_tokens_seen": 106767010, + "step": 4950 + }, + { + "epoch": 0.5953225515541394, + "flos": 33604779421920.0, + "grad_norm": 3.1650880258380982, + "learning_rate": 1.4854977101954587e-06, + "loss": 0.6916, + "num_input_tokens_seen": 106789230, + "step": 4951 + }, + { + "epoch": 0.5954427944447784, + "flos": 24424522511040.0, + "grad_norm": 2.343863115759946, + "learning_rate": 1.4847449942094716e-06, + "loss": 0.8626, + "num_input_tokens_seen": 106808110, + "step": 4952 + }, + { + "epoch": 0.5955630373354175, + "flos": 18525458981280.0, + "grad_norm": 3.3892680322592708, + "learning_rate": 1.4839923563872598e-06, + "loss": 0.8655, + "num_input_tokens_seen": 106826650, + "step": 4953 + }, + { + "epoch": 0.5956832802260567, + "flos": 19763481271200.0, + "grad_norm": 3.0985334003514087, + "learning_rate": 1.483239796842997e-06, + "loss": 0.7608, + "num_input_tokens_seen": 106844680, + "step": 4954 + }, + { + "epoch": 0.5958035231166957, + "flos": 19722551631840.0, + "grad_norm": 3.0592105871955075, + "learning_rate": 1.4824873156908462e-06, + "loss": 0.8395, + "num_input_tokens_seen": 106862605, + "step": 4955 + }, + { + "epoch": 0.5959237660073348, + "flos": 21621889987200.0, + "grad_norm": 2.1727675106636486, + "learning_rate": 1.4817349130449584e-06, + "loss": 0.7562, + "num_input_tokens_seen": 106882325, + "step": 4956 + }, + { + "epoch": 0.5960440088979739, + "flos": 21140727246720.0, + "grad_norm": 2.8416127987526147, + "learning_rate": 1.4809825890194717e-06, + "loss": 0.8354, + "num_input_tokens_seen": 106900995, + "step": 4957 + }, + { + "epoch": 0.596164251788613, + "flos": 14736669426720.0, + "grad_norm": 2.434831283176097, + "learning_rate": 1.4802303437285139e-06, + "loss": 0.7757, + "num_input_tokens_seen": 106918265, + "step": 4958 + }, + { + "epoch": 0.596284494679252, + "flos": 20456731894080.0, + "grad_norm": 3.558232746593548, + "learning_rate": 1.4794781772861994e-06, + "loss": 0.8044, + "num_input_tokens_seen": 106934760, + "step": 4959 + }, + { + "epoch": 0.5964047375698912, + "flos": 31167434295840.0, + "grad_norm": 4.386878753760504, + "learning_rate": 1.4787260898066324e-06, + "loss": 0.6658, + "num_input_tokens_seen": 106954760, + "step": 4960 + }, + { + "epoch": 0.5965249804605303, + "flos": 27443963476800.0, + "grad_norm": 2.7581489509045696, + "learning_rate": 1.4779740814039023e-06, + "loss": 0.8574, + "num_input_tokens_seen": 106974845, + "step": 4961 + }, + { + "epoch": 0.5966452233511693, + "flos": 30729728738400.0, + "grad_norm": 2.242132950494705, + "learning_rate": 1.4772221521920894e-06, + "loss": 0.6864, + "num_input_tokens_seen": 106994870, + "step": 4962 + }, + { + "epoch": 0.5967654662418085, + "flos": 25444584096480.0, + "grad_norm": 2.466438466762783, + "learning_rate": 1.4764703022852598e-06, + "loss": 0.7415, + "num_input_tokens_seen": 107015390, + "step": 4963 + }, + { + "epoch": 0.5968857091324475, + "flos": 19071940457280.0, + "grad_norm": 3.3148193876336363, + "learning_rate": 1.4757185317974696e-06, + "loss": 0.7707, + "num_input_tokens_seen": 107033775, + "step": 4964 + }, + { + "epoch": 0.5970059520230866, + "flos": 23658705049440.0, + "grad_norm": 3.672176916817958, + "learning_rate": 1.474966840842761e-06, + "loss": 0.7094, + "num_input_tokens_seen": 107053355, + "step": 4965 + }, + { + "epoch": 0.5971261949137258, + "flos": 23152409817600.0, + "grad_norm": 2.2750653603355726, + "learning_rate": 1.4742152295351655e-06, + "loss": 0.8686, + "num_input_tokens_seen": 107072510, + "step": 4966 + }, + { + "epoch": 0.5972464378043648, + "flos": 20528369488800.0, + "grad_norm": 3.388076316541436, + "learning_rate": 1.4734636979887016e-06, + "loss": 0.6408, + "num_input_tokens_seen": 107089970, + "step": 4967 + }, + { + "epoch": 0.5973666806950039, + "flos": 29344825792320.0, + "grad_norm": 2.93180327247017, + "learning_rate": 1.4727122463173755e-06, + "loss": 0.9047, + "num_input_tokens_seen": 107108500, + "step": 4968 + }, + { + "epoch": 0.597486923585643, + "flos": 22241718866880.0, + "grad_norm": 2.275070495351645, + "learning_rate": 1.471960874635183e-06, + "loss": 0.6438, + "num_input_tokens_seen": 107128060, + "step": 4969 + }, + { + "epoch": 0.5976071664762821, + "flos": 13753039939680.0, + "grad_norm": 2.7232715234391804, + "learning_rate": 1.4712095830561055e-06, + "loss": 0.7159, + "num_input_tokens_seen": 107146550, + "step": 4970 + }, + { + "epoch": 0.5977274093669211, + "flos": 19071420080640.0, + "grad_norm": 2.922923017419254, + "learning_rate": 1.4704583716941147e-06, + "loss": 0.8119, + "num_input_tokens_seen": 107165415, + "step": 4971 + }, + { + "epoch": 0.5978476522575603, + "flos": 20346952392000.0, + "grad_norm": 46.66939282401797, + "learning_rate": 1.4697072406631672e-06, + "loss": 0.7218, + "num_input_tokens_seen": 107185320, + "step": 4972 + }, + { + "epoch": 0.5979678951481994, + "flos": 28981508391840.0, + "grad_norm": 2.373324053867644, + "learning_rate": 1.4689561900772097e-06, + "loss": 0.7284, + "num_input_tokens_seen": 107205380, + "step": 4973 + }, + { + "epoch": 0.5980881380388384, + "flos": 17942136539520.0, + "grad_norm": 3.1959529318846776, + "learning_rate": 1.4682052200501758e-06, + "loss": 0.7239, + "num_input_tokens_seen": 107222585, + "step": 4974 + }, + { + "epoch": 0.5982083809294776, + "flos": 22929877232640.0, + "grad_norm": 2.376266242327456, + "learning_rate": 1.4674543306959876e-06, + "loss": 0.8007, + "num_input_tokens_seen": 107242055, + "step": 4975 + }, + { + "epoch": 0.5983286238201166, + "flos": 20961837693600.0, + "grad_norm": 3.909654034384513, + "learning_rate": 1.4667035221285535e-06, + "loss": 0.8477, + "num_input_tokens_seen": 107259695, + "step": 4976 + }, + { + "epoch": 0.5984488667107557, + "flos": 28143012582240.0, + "grad_norm": 12.849297506522792, + "learning_rate": 1.4659527944617715e-06, + "loss": 0.7449, + "num_input_tokens_seen": 107279640, + "step": 4977 + }, + { + "epoch": 0.5985691096013949, + "flos": 16448420505120.0, + "grad_norm": 4.864399658046523, + "learning_rate": 1.465202147809526e-06, + "loss": 0.7579, + "num_input_tokens_seen": 107298135, + "step": 4978 + }, + { + "epoch": 0.5986893524920339, + "flos": 26680562049600.0, + "grad_norm": 2.7958739636382997, + "learning_rate": 1.4644515822856888e-06, + "loss": 0.7693, + "num_input_tokens_seen": 107316570, + "step": 4979 + }, + { + "epoch": 0.598809595382673, + "flos": 61520421616800.0, + "grad_norm": 0.7755334756169593, + "learning_rate": 1.4637010980041215e-06, + "loss": 0.5647, + "num_input_tokens_seen": 107378680, + "step": 4980 + }, + { + "epoch": 0.5989298382733121, + "flos": 11785149079680.0, + "grad_norm": 2.8675184344735953, + "learning_rate": 1.4629506950786707e-06, + "loss": 0.8929, + "num_input_tokens_seen": 107394860, + "step": 4981 + }, + { + "epoch": 0.5990500811639512, + "flos": 59939977065120.0, + "grad_norm": 0.817055439535324, + "learning_rate": 1.4622003736231733e-06, + "loss": 0.5598, + "num_input_tokens_seen": 107453850, + "step": 4982 + }, + { + "epoch": 0.5991703240545903, + "flos": 18196789530720.0, + "grad_norm": 2.78116067464308, + "learning_rate": 1.461450133751451e-06, + "loss": 0.8029, + "num_input_tokens_seen": 107471920, + "step": 4983 + }, + { + "epoch": 0.5992905669452293, + "flos": 27670175868000.0, + "grad_norm": 2.0481148961143627, + "learning_rate": 1.4606999755773153e-06, + "loss": 0.7578, + "num_input_tokens_seen": 107493125, + "step": 4984 + }, + { + "epoch": 0.5994108098358685, + "flos": 20419853758560.0, + "grad_norm": 3.9893075367923077, + "learning_rate": 1.4599498992145643e-06, + "loss": 0.8215, + "num_input_tokens_seen": 107513150, + "step": 4985 + }, + { + "epoch": 0.5995310527265075, + "flos": 22238039060640.0, + "grad_norm": 2.3583772077848826, + "learning_rate": 1.4591999047769846e-06, + "loss": 0.7071, + "num_input_tokens_seen": 107532005, + "step": 4986 + }, + { + "epoch": 0.5996512956171466, + "flos": 18890040153600.0, + "grad_norm": 2.2232870001399854, + "learning_rate": 1.4584499923783486e-06, + "loss": 0.7566, + "num_input_tokens_seen": 107550585, + "step": 4987 + }, + { + "epoch": 0.5997715385077858, + "flos": 15348915675360.0, + "grad_norm": 3.2725556489949934, + "learning_rate": 1.457700162132419e-06, + "loss": 0.7585, + "num_input_tokens_seen": 107567575, + "step": 4988 + }, + { + "epoch": 0.5998917813984248, + "flos": 25228667728800.0, + "grad_norm": 2.8758333076588327, + "learning_rate": 1.4569504141529433e-06, + "loss": 0.728, + "num_input_tokens_seen": 107585410, + "step": 4989 + }, + { + "epoch": 0.6000120242890639, + "flos": 22023014767200.0, + "grad_norm": 3.6255432248046304, + "learning_rate": 1.456200748553658e-06, + "loss": 0.7194, + "num_input_tokens_seen": 107603405, + "step": 4990 + }, + { + "epoch": 0.600132267179703, + "flos": 29820598917600.0, + "grad_norm": 2.01607029453976, + "learning_rate": 1.455451165448287e-06, + "loss": 0.7881, + "num_input_tokens_seen": 107626060, + "step": 4991 + }, + { + "epoch": 0.6002525100703421, + "flos": 25009480422240.0, + "grad_norm": 3.2410345394432714, + "learning_rate": 1.4547016649505407e-06, + "loss": 0.7374, + "num_input_tokens_seen": 107644070, + "step": 4992 + }, + { + "epoch": 0.6003727529609811, + "flos": 20819826276000.0, + "grad_norm": 2.5375414238916987, + "learning_rate": 1.4539522471741193e-06, + "loss": 0.8503, + "num_input_tokens_seen": 107661495, + "step": 4993 + }, + { + "epoch": 0.6004929958516203, + "flos": 15572451843840.0, + "grad_norm": 3.6422141867453246, + "learning_rate": 1.4532029122327067e-06, + "loss": 0.7138, + "num_input_tokens_seen": 107678280, + "step": 4994 + }, + { + "epoch": 0.6006132387422594, + "flos": 21732227035680.0, + "grad_norm": 2.6570844349393514, + "learning_rate": 1.4524536602399783e-06, + "loss": 0.7541, + "num_input_tokens_seen": 107697370, + "step": 4995 + }, + { + "epoch": 0.6007334816328984, + "flos": 22825933382880.0, + "grad_norm": 1.9430207997503086, + "learning_rate": 1.4517044913095938e-06, + "loss": 0.7751, + "num_input_tokens_seen": 107717790, + "step": 4996 + }, + { + "epoch": 0.6008537245235376, + "flos": 28283760228000.0, + "grad_norm": 2.583779901443148, + "learning_rate": 1.4509554055552022e-06, + "loss": 0.8134, + "num_input_tokens_seen": 107738020, + "step": 4997 + }, + { + "epoch": 0.6009739674141766, + "flos": 20856295544160.0, + "grad_norm": 2.921862433911398, + "learning_rate": 1.450206403090439e-06, + "loss": 0.8418, + "num_input_tokens_seen": 107756215, + "step": 4998 + }, + { + "epoch": 0.6010942103048157, + "flos": 20451639636960.0, + "grad_norm": 3.351714444217492, + "learning_rate": 1.4494574840289274e-06, + "loss": 0.8621, + "num_input_tokens_seen": 107773330, + "step": 4999 + }, + { + "epoch": 0.6012144531954549, + "flos": 23771941339200.0, + "grad_norm": 2.118685822091594, + "learning_rate": 1.4487086484842782e-06, + "loss": 0.7395, + "num_input_tokens_seen": 107791975, + "step": 5000 + }, + { + "epoch": 0.6013346960860939, + "flos": 18961306050720.0, + "grad_norm": 3.927489862556414, + "learning_rate": 1.4479598965700878e-06, + "loss": 0.6031, + "num_input_tokens_seen": 107809240, + "step": 5001 + }, + { + "epoch": 0.601454938976733, + "flos": 23990979966720.0, + "grad_norm": 3.6752532352728826, + "learning_rate": 1.4472112283999427e-06, + "loss": 0.6872, + "num_input_tokens_seen": 107827895, + "step": 5002 + }, + { + "epoch": 0.6015751818673721, + "flos": 26389068092640.0, + "grad_norm": 2.5627520367490826, + "learning_rate": 1.4464626440874143e-06, + "loss": 0.6918, + "num_input_tokens_seen": 107847010, + "step": 5003 + }, + { + "epoch": 0.6016954247580112, + "flos": 13096778961600.0, + "grad_norm": 3.666063750433319, + "learning_rate": 1.4457141437460636e-06, + "loss": 0.7377, + "num_input_tokens_seen": 107864235, + "step": 5004 + }, + { + "epoch": 0.6018156676486502, + "flos": 23734431317760.0, + "grad_norm": 3.377929122849765, + "learning_rate": 1.444965727489436e-06, + "loss": 0.7324, + "num_input_tokens_seen": 107883680, + "step": 5005 + }, + { + "epoch": 0.6019359105392894, + "flos": 26431447352640.0, + "grad_norm": 2.0537383349337532, + "learning_rate": 1.444217395431066e-06, + "loss": 0.6277, + "num_input_tokens_seen": 107906220, + "step": 5006 + }, + { + "epoch": 0.6020561534299285, + "flos": 69091942055040.0, + "grad_norm": 0.8097095622830034, + "learning_rate": 1.4434691476844755e-06, + "loss": 0.5575, + "num_input_tokens_seen": 107967195, + "step": 5007 + }, + { + "epoch": 0.6021763963205675, + "flos": 21804124818720.0, + "grad_norm": 2.7214643586891296, + "learning_rate": 1.4427209843631729e-06, + "loss": 0.6734, + "num_input_tokens_seen": 107984245, + "step": 5008 + }, + { + "epoch": 0.6022966392112067, + "flos": 26540557799040.0, + "grad_norm": 2.3451813459734216, + "learning_rate": 1.4419729055806534e-06, + "loss": 0.8142, + "num_input_tokens_seen": 108003680, + "step": 5009 + }, + { + "epoch": 0.6024168821018457, + "flos": 20674172221920.0, + "grad_norm": 2.0608557880852962, + "learning_rate": 1.441224911450401e-06, + "loss": 0.8248, + "num_input_tokens_seen": 108019635, + "step": 5010 + }, + { + "epoch": 0.6025371249924848, + "flos": 24645642645120.0, + "grad_norm": 2.2347057604560447, + "learning_rate": 1.4404770020858851e-06, + "loss": 0.8241, + "num_input_tokens_seen": 108039120, + "step": 5011 + }, + { + "epoch": 0.602657367883124, + "flos": 25918535903520.0, + "grad_norm": 2.2422898963647855, + "learning_rate": 1.439729177600563e-06, + "loss": 0.8607, + "num_input_tokens_seen": 108059615, + "step": 5012 + }, + { + "epoch": 0.602777610773763, + "flos": 16666938756000.0, + "grad_norm": 2.9630739259693266, + "learning_rate": 1.4389814381078793e-06, + "loss": 0.7284, + "num_input_tokens_seen": 108078855, + "step": 5013 + }, + { + "epoch": 0.6028978536644021, + "flos": 13315445891520.0, + "grad_norm": 3.6168655680679627, + "learning_rate": 1.438233783721265e-06, + "loss": 0.8047, + "num_input_tokens_seen": 108095020, + "step": 5014 + }, + { + "epoch": 0.6030180965550412, + "flos": 19616786463840.0, + "grad_norm": 2.9296207674691486, + "learning_rate": 1.43748621455414e-06, + "loss": 0.7803, + "num_input_tokens_seen": 108111455, + "step": 5015 + }, + { + "epoch": 0.6031383394456803, + "flos": 14437704348000.0, + "grad_norm": 5.39742975172486, + "learning_rate": 1.4367387307199082e-06, + "loss": 0.8053, + "num_input_tokens_seen": 108128305, + "step": 5016 + }, + { + "epoch": 0.6032585823363193, + "flos": 13898136447360.0, + "grad_norm": 2.3500213135097723, + "learning_rate": 1.4359913323319632e-06, + "loss": 0.8237, + "num_input_tokens_seen": 108145475, + "step": 5017 + }, + { + "epoch": 0.6033788252269584, + "flos": 24318422815200.0, + "grad_norm": 3.1790020928005496, + "learning_rate": 1.4352440195036847e-06, + "loss": 0.778, + "num_input_tokens_seen": 108165650, + "step": 5018 + }, + { + "epoch": 0.6034990681175976, + "flos": 25484733170880.0, + "grad_norm": 1.9383468622974813, + "learning_rate": 1.4344967923484395e-06, + "loss": 0.7992, + "num_input_tokens_seen": 108187620, + "step": 5019 + }, + { + "epoch": 0.6036193110082366, + "flos": 25921212126240.0, + "grad_norm": 2.9940003165000895, + "learning_rate": 1.433749650979581e-06, + "loss": 0.719, + "num_input_tokens_seen": 108207605, + "step": 5020 + }, + { + "epoch": 0.6037395538988757, + "flos": 25556779632960.0, + "grad_norm": 2.672987499968489, + "learning_rate": 1.433002595510451e-06, + "loss": 0.6814, + "num_input_tokens_seen": 108226385, + "step": 5021 + }, + { + "epoch": 0.6038597967895148, + "flos": 17790721172640.0, + "grad_norm": 2.250307380025215, + "learning_rate": 1.4322556260543757e-06, + "loss": 0.7199, + "num_input_tokens_seen": 108242960, + "step": 5022 + }, + { + "epoch": 0.6039800396801539, + "flos": 65075862186240.0, + "grad_norm": 0.9099595671883127, + "learning_rate": 1.4315087427246703e-06, + "loss": 0.6264, + "num_input_tokens_seen": 108296785, + "step": 5023 + }, + { + "epoch": 0.604100282570793, + "flos": 67289522464800.0, + "grad_norm": 0.8840403162794642, + "learning_rate": 1.4307619456346372e-06, + "loss": 0.5839, + "num_input_tokens_seen": 108341090, + "step": 5024 + }, + { + "epoch": 0.6042205254614321, + "flos": 35246677054080.0, + "grad_norm": 3.3214640696045215, + "learning_rate": 1.430015234897564e-06, + "loss": 0.7397, + "num_input_tokens_seen": 108364370, + "step": 5025 + }, + { + "epoch": 0.6043407683520712, + "flos": 45592315076640.0, + "grad_norm": 2.63609638922381, + "learning_rate": 1.4292686106267274e-06, + "loss": 0.6634, + "num_input_tokens_seen": 108387220, + "step": 5026 + }, + { + "epoch": 0.6044610112427102, + "flos": 16157075227200.0, + "grad_norm": 2.079184928717009, + "learning_rate": 1.4285220729353876e-06, + "loss": 0.773, + "num_input_tokens_seen": 108405760, + "step": 5027 + }, + { + "epoch": 0.6045812541333494, + "flos": 13784416950720.0, + "grad_norm": 3.4407226714944974, + "learning_rate": 1.4277756219367957e-06, + "loss": 0.7871, + "num_input_tokens_seen": 108422785, + "step": 5028 + }, + { + "epoch": 0.6047014970239885, + "flos": 19946608176960.0, + "grad_norm": 3.1973661431396776, + "learning_rate": 1.4270292577441864e-06, + "loss": 0.8021, + "num_input_tokens_seen": 108442290, + "step": 5029 + }, + { + "epoch": 0.6048217399146275, + "flos": 25300788530400.0, + "grad_norm": 2.354487124437215, + "learning_rate": 1.4262829804707836e-06, + "loss": 0.7188, + "num_input_tokens_seen": 108464280, + "step": 5030 + }, + { + "epoch": 0.6049419828052667, + "flos": 25993518776640.0, + "grad_norm": 2.01815400670702, + "learning_rate": 1.4255367902297958e-06, + "loss": 0.6972, + "num_input_tokens_seen": 108485610, + "step": 5031 + }, + { + "epoch": 0.6050622256959057, + "flos": 14627410301280.0, + "grad_norm": 3.698957456164129, + "learning_rate": 1.4247906871344215e-06, + "loss": 0.7875, + "num_input_tokens_seen": 108501005, + "step": 5032 + }, + { + "epoch": 0.6051824685865448, + "flos": 23298101041440.0, + "grad_norm": 4.801545593649052, + "learning_rate": 1.4240446712978415e-06, + "loss": 0.7539, + "num_input_tokens_seen": 108519990, + "step": 5033 + }, + { + "epoch": 0.605302711477184, + "flos": 27523778418720.0, + "grad_norm": 6.063921889473892, + "learning_rate": 1.423298742833227e-06, + "loss": 0.7447, + "num_input_tokens_seen": 108538650, + "step": 5034 + }, + { + "epoch": 0.605422954367823, + "flos": 15133110816960.0, + "grad_norm": 6.043905470385927, + "learning_rate": 1.4225529018537352e-06, + "loss": 0.7199, + "num_input_tokens_seen": 108554155, + "step": 5035 + }, + { + "epoch": 0.6055431972584621, + "flos": 27634264146240.0, + "grad_norm": 1.9638015075960635, + "learning_rate": 1.4218071484725082e-06, + "loss": 0.7752, + "num_input_tokens_seen": 108576275, + "step": 5036 + }, + { + "epoch": 0.6056634401491012, + "flos": 19359754608000.0, + "grad_norm": 51.09498039420678, + "learning_rate": 1.4210614828026786e-06, + "loss": 0.7626, + "num_input_tokens_seen": 108594125, + "step": 5037 + }, + { + "epoch": 0.6057836830397403, + "flos": 24753935356800.0, + "grad_norm": 1.983307077998061, + "learning_rate": 1.4203159049573605e-06, + "loss": 0.7446, + "num_input_tokens_seen": 108615360, + "step": 5038 + }, + { + "epoch": 0.6059039259303793, + "flos": 20528927035200.0, + "grad_norm": 6.040540427984021, + "learning_rate": 1.4195704150496593e-06, + "loss": 0.8659, + "num_input_tokens_seen": 108633075, + "step": 5039 + }, + { + "epoch": 0.6060241688210185, + "flos": 21039682638240.0, + "grad_norm": 2.2635335805941326, + "learning_rate": 1.4188250131926639e-06, + "loss": 0.7375, + "num_input_tokens_seen": 108651875, + "step": 5040 + }, + { + "epoch": 0.6061444117116576, + "flos": 16337823268320.0, + "grad_norm": 2.3574341123400795, + "learning_rate": 1.4180796994994525e-06, + "loss": 0.8068, + "num_input_tokens_seen": 108669290, + "step": 5041 + }, + { + "epoch": 0.6062646546022966, + "flos": 21476607630720.0, + "grad_norm": 2.420634290311455, + "learning_rate": 1.4173344740830877e-06, + "loss": 0.7209, + "num_input_tokens_seen": 108688420, + "step": 5042 + }, + { + "epoch": 0.6063848974929358, + "flos": 38926988048160.0, + "grad_norm": 2.0952283280524786, + "learning_rate": 1.4165893370566206e-06, + "loss": 0.7016, + "num_input_tokens_seen": 108712175, + "step": 5043 + }, + { + "epoch": 0.6065051403835748, + "flos": 19618607782080.0, + "grad_norm": 2.455397215677205, + "learning_rate": 1.4158442885330865e-06, + "loss": 0.7781, + "num_input_tokens_seen": 108730745, + "step": 5044 + }, + { + "epoch": 0.6066253832742139, + "flos": 23479443798720.0, + "grad_norm": 3.6004905843252724, + "learning_rate": 1.4150993286255094e-06, + "loss": 0.785, + "num_input_tokens_seen": 108749995, + "step": 5045 + }, + { + "epoch": 0.6067456261648531, + "flos": 19106402558400.0, + "grad_norm": 3.1020470536166216, + "learning_rate": 1.4143544574468993e-06, + "loss": 0.8009, + "num_input_tokens_seen": 108768355, + "step": 5046 + }, + { + "epoch": 0.6068658690554921, + "flos": 20491305504480.0, + "grad_norm": 2.1803186521461124, + "learning_rate": 1.4136096751102523e-06, + "loss": 0.827, + "num_input_tokens_seen": 108786685, + "step": 5047 + }, + { + "epoch": 0.6069861119461312, + "flos": 27335076048960.0, + "grad_norm": 2.2791800317831528, + "learning_rate": 1.4128649817285516e-06, + "loss": 0.831, + "num_input_tokens_seen": 108804820, + "step": 5048 + }, + { + "epoch": 0.6071063548367702, + "flos": 25589680604160.0, + "grad_norm": 3.182487885481318, + "learning_rate": 1.412120377414766e-06, + "loss": 0.633, + "num_input_tokens_seen": 108825010, + "step": 5049 + }, + { + "epoch": 0.6072265977274094, + "flos": 24425451755040.0, + "grad_norm": 2.6697391651204114, + "learning_rate": 1.4113758622818522e-06, + "loss": 0.7115, + "num_input_tokens_seen": 108845075, + "step": 5050 + }, + { + "epoch": 0.6073468406180484, + "flos": 18123404957280.0, + "grad_norm": 2.5707928518841388, + "learning_rate": 1.410631436442751e-06, + "loss": 0.8294, + "num_input_tokens_seen": 108863870, + "step": 5051 + }, + { + "epoch": 0.6074670835086875, + "flos": 20667853362720.0, + "grad_norm": 2.9872949505909476, + "learning_rate": 1.4098871000103936e-06, + "loss": 0.8651, + "num_input_tokens_seen": 108882470, + "step": 5052 + }, + { + "epoch": 0.6075873263993267, + "flos": 23735992447680.0, + "grad_norm": 2.6151991849644216, + "learning_rate": 1.409142853097693e-06, + "loss": 0.8285, + "num_input_tokens_seen": 108902710, + "step": 5053 + }, + { + "epoch": 0.6077075692899657, + "flos": 24419207235360.0, + "grad_norm": 13.365916938081844, + "learning_rate": 1.408398695817553e-06, + "loss": 0.7952, + "num_input_tokens_seen": 108919935, + "step": 5054 + }, + { + "epoch": 0.6078278121806048, + "flos": 27342881698560.0, + "grad_norm": 2.5891942839617306, + "learning_rate": 1.4076546282828593e-06, + "loss": 0.7001, + "num_input_tokens_seen": 108942790, + "step": 5055 + }, + { + "epoch": 0.6079480550712439, + "flos": 38381101288320.0, + "grad_norm": 8.450543303770845, + "learning_rate": 1.4069106506064874e-06, + "loss": 0.6587, + "num_input_tokens_seen": 108964215, + "step": 5056 + }, + { + "epoch": 0.608068297961883, + "flos": 25299933625920.0, + "grad_norm": 88.83168006298484, + "learning_rate": 1.4061667629012989e-06, + "loss": 0.783, + "num_input_tokens_seen": 108984885, + "step": 5057 + }, + { + "epoch": 0.608188540852522, + "flos": 24167676504000.0, + "grad_norm": 3.4863352372126335, + "learning_rate": 1.40542296528014e-06, + "loss": 0.834, + "num_input_tokens_seen": 109004545, + "step": 5058 + }, + { + "epoch": 0.6083087837431612, + "flos": 21253629008640.0, + "grad_norm": 2.6181673407756425, + "learning_rate": 1.4046792578558452e-06, + "loss": 0.7643, + "num_input_tokens_seen": 109022440, + "step": 5059 + }, + { + "epoch": 0.6084290266338003, + "flos": 16448048807520.0, + "grad_norm": 25.794941913489794, + "learning_rate": 1.4039356407412325e-06, + "loss": 0.7588, + "num_input_tokens_seen": 109035680, + "step": 5060 + }, + { + "epoch": 0.6085492695244393, + "flos": 66348272237760.0, + "grad_norm": 0.7895643319197998, + "learning_rate": 1.40319211404911e-06, + "loss": 0.5702, + "num_input_tokens_seen": 109090680, + "step": 5061 + }, + { + "epoch": 0.6086695124150785, + "flos": 23585060287680.0, + "grad_norm": 2.7462378690622704, + "learning_rate": 1.4024486778922691e-06, + "loss": 0.9077, + "num_input_tokens_seen": 109108995, + "step": 5062 + }, + { + "epoch": 0.6087897553057176, + "flos": 20128322631840.0, + "grad_norm": 2.8725205386019335, + "learning_rate": 1.4017053323834884e-06, + "loss": 0.7762, + "num_input_tokens_seen": 109128825, + "step": 5063 + }, + { + "epoch": 0.6089099981963566, + "flos": 25445996547360.0, + "grad_norm": 2.8452885416730544, + "learning_rate": 1.4009620776355333e-06, + "loss": 0.7642, + "num_input_tokens_seen": 109150425, + "step": 5064 + }, + { + "epoch": 0.6090302410869958, + "flos": 25296439668480.0, + "grad_norm": 2.137646238444253, + "learning_rate": 1.4002189137611553e-06, + "loss": 0.7937, + "num_input_tokens_seen": 109169765, + "step": 5065 + }, + { + "epoch": 0.6091504839776348, + "flos": 23953023908160.0, + "grad_norm": 3.004808069786635, + "learning_rate": 1.3994758408730901e-06, + "loss": 0.6994, + "num_input_tokens_seen": 109188950, + "step": 5066 + }, + { + "epoch": 0.6092707268682739, + "flos": 29600482367040.0, + "grad_norm": 3.093241640189866, + "learning_rate": 1.3987328590840629e-06, + "loss": 0.766, + "num_input_tokens_seen": 109209895, + "step": 5067 + }, + { + "epoch": 0.609390969758913, + "flos": 23990608269120.0, + "grad_norm": 2.5655317282594763, + "learning_rate": 1.397989968506783e-06, + "loss": 0.8642, + "num_input_tokens_seen": 109228905, + "step": 5068 + }, + { + "epoch": 0.6095112126495521, + "flos": 11086360162560.0, + "grad_norm": 3.204837992318307, + "learning_rate": 1.3972471692539458e-06, + "loss": 0.7297, + "num_input_tokens_seen": 109243620, + "step": 5069 + }, + { + "epoch": 0.6096314555401912, + "flos": 17240634229920.0, + "grad_norm": 2.5396098793985216, + "learning_rate": 1.3965044614382348e-06, + "loss": 0.7547, + "num_input_tokens_seen": 109259070, + "step": 5070 + }, + { + "epoch": 0.6097516984308303, + "flos": 21614753393280.0, + "grad_norm": 2.9163596582050473, + "learning_rate": 1.3957618451723162e-06, + "loss": 0.7557, + "num_input_tokens_seen": 109275100, + "step": 5071 + }, + { + "epoch": 0.6098719413214694, + "flos": 27160237999680.0, + "grad_norm": 3.118942493315856, + "learning_rate": 1.3950193205688457e-06, + "loss": 0.7186, + "num_input_tokens_seen": 109294825, + "step": 5072 + }, + { + "epoch": 0.6099921842121084, + "flos": 20383867697280.0, + "grad_norm": 2.1491629190795614, + "learning_rate": 1.3942768877404627e-06, + "loss": 0.839, + "num_input_tokens_seen": 109313790, + "step": 5073 + }, + { + "epoch": 0.6101124271027476, + "flos": 23332674651840.0, + "grad_norm": 2.3362202042689537, + "learning_rate": 1.393534546799795e-06, + "loss": 0.7369, + "num_input_tokens_seen": 109333490, + "step": 5074 + }, + { + "epoch": 0.6102326699933867, + "flos": 26649593905920.0, + "grad_norm": 4.261899781687212, + "learning_rate": 1.3927922978594536e-06, + "loss": 0.6775, + "num_input_tokens_seen": 109354575, + "step": 5075 + }, + { + "epoch": 0.6103529128840257, + "flos": 60557835947520.0, + "grad_norm": 0.7823261478577904, + "learning_rate": 1.3920501410320387e-06, + "loss": 0.5736, + "num_input_tokens_seen": 109410690, + "step": 5076 + }, + { + "epoch": 0.6104731557746649, + "flos": 18994058342880.0, + "grad_norm": 2.586286907273793, + "learning_rate": 1.3913080764301333e-06, + "loss": 0.7623, + "num_input_tokens_seen": 109427650, + "step": 5077 + }, + { + "epoch": 0.6105933986653039, + "flos": 23333455216800.0, + "grad_norm": 2.2118923969344566, + "learning_rate": 1.3905661041663085e-06, + "loss": 0.717, + "num_input_tokens_seen": 109448835, + "step": 5078 + }, + { + "epoch": 0.610713641555943, + "flos": 34587591174240.0, + "grad_norm": 3.0472136480472947, + "learning_rate": 1.389824224353122e-06, + "loss": 0.6539, + "num_input_tokens_seen": 109470425, + "step": 5079 + }, + { + "epoch": 0.610833884446582, + "flos": 26608106720160.0, + "grad_norm": 2.1375427331054815, + "learning_rate": 1.389082437103115e-06, + "loss": 0.7716, + "num_input_tokens_seen": 109489695, + "step": 5080 + }, + { + "epoch": 0.6109541273372212, + "flos": 21184853485440.0, + "grad_norm": 5.646562742667858, + "learning_rate": 1.3883407425288172e-06, + "loss": 0.7761, + "num_input_tokens_seen": 109510030, + "step": 5081 + }, + { + "epoch": 0.6110743702278603, + "flos": 20055458435040.0, + "grad_norm": 2.749189951107065, + "learning_rate": 1.3875991407427417e-06, + "loss": 0.7973, + "num_input_tokens_seen": 109528255, + "step": 5082 + }, + { + "epoch": 0.6111946131184993, + "flos": 68204896805280.0, + "grad_norm": 0.7749389562869254, + "learning_rate": 1.38685763185739e-06, + "loss": 0.581, + "num_input_tokens_seen": 109581915, + "step": 5083 + }, + { + "epoch": 0.6113148560091385, + "flos": 19909246834560.0, + "grad_norm": 3.6501174751084773, + "learning_rate": 1.3861162159852476e-06, + "loss": 0.6776, + "num_input_tokens_seen": 109600050, + "step": 5084 + }, + { + "epoch": 0.6114350988997775, + "flos": 23697887710080.0, + "grad_norm": 2.2577689935525003, + "learning_rate": 1.3853748932387875e-06, + "loss": 0.8019, + "num_input_tokens_seen": 109620690, + "step": 5085 + }, + { + "epoch": 0.6115553417904166, + "flos": 23989307327520.0, + "grad_norm": 3.346929446791754, + "learning_rate": 1.3846336637304671e-06, + "loss": 0.7523, + "num_input_tokens_seen": 109638960, + "step": 5086 + }, + { + "epoch": 0.6116755846810558, + "flos": 23699634688800.0, + "grad_norm": 2.7882523415206015, + "learning_rate": 1.3838925275727316e-06, + "loss": 0.8309, + "num_input_tokens_seen": 109659700, + "step": 5087 + }, + { + "epoch": 0.6117958275716948, + "flos": 18635089804320.0, + "grad_norm": 2.01614999957103, + "learning_rate": 1.3831514848780089e-06, + "loss": 0.7916, + "num_input_tokens_seen": 109679275, + "step": 5088 + }, + { + "epoch": 0.6119160704623339, + "flos": 16447788619200.0, + "grad_norm": 3.023508785698706, + "learning_rate": 1.3824105357587152e-06, + "loss": 0.9218, + "num_input_tokens_seen": 109696380, + "step": 5089 + }, + { + "epoch": 0.612036313352973, + "flos": 23881274804160.0, + "grad_norm": 1.8867828800525381, + "learning_rate": 1.381669680327253e-06, + "loss": 0.8261, + "num_input_tokens_seen": 109718060, + "step": 5090 + }, + { + "epoch": 0.6121565562436121, + "flos": 26935921266240.0, + "grad_norm": 2.8174393408540475, + "learning_rate": 1.380928918696008e-06, + "loss": 0.71, + "num_input_tokens_seen": 109736830, + "step": 5091 + }, + { + "epoch": 0.6122767991342511, + "flos": 15646654152000.0, + "grad_norm": 3.369931972281152, + "learning_rate": 1.3801882509773548e-06, + "loss": 0.7142, + "num_input_tokens_seen": 109754690, + "step": 5092 + }, + { + "epoch": 0.6123970420248903, + "flos": 27924196973280.0, + "grad_norm": 2.6612742759131316, + "learning_rate": 1.3794476772836503e-06, + "loss": 0.8202, + "num_input_tokens_seen": 109775790, + "step": 5093 + }, + { + "epoch": 0.6125172849155294, + "flos": 21438019686240.0, + "grad_norm": 2.1413564058979246, + "learning_rate": 1.3787071977272402e-06, + "loss": 0.846, + "num_input_tokens_seen": 109795050, + "step": 5094 + }, + { + "epoch": 0.6126375278061684, + "flos": 16225107355200.0, + "grad_norm": 3.418035434617348, + "learning_rate": 1.3779668124204535e-06, + "loss": 0.7202, + "num_input_tokens_seen": 109811465, + "step": 5095 + }, + { + "epoch": 0.6127577706968076, + "flos": 20419705079520.0, + "grad_norm": 2.3642009364286403, + "learning_rate": 1.3772265214756074e-06, + "loss": 0.8082, + "num_input_tokens_seen": 109831225, + "step": 5096 + }, + { + "epoch": 0.6128780135874466, + "flos": 18234448231200.0, + "grad_norm": 3.3243527234129484, + "learning_rate": 1.3764863250050025e-06, + "loss": 0.7558, + "num_input_tokens_seen": 109849465, + "step": 5097 + }, + { + "epoch": 0.6129982564780857, + "flos": 24901410729120.0, + "grad_norm": 2.9593570889955116, + "learning_rate": 1.3757462231209272e-06, + "loss": 0.8045, + "num_input_tokens_seen": 109869770, + "step": 5098 + }, + { + "epoch": 0.6131184993687249, + "flos": 22459902589920.0, + "grad_norm": 3.0234762972922553, + "learning_rate": 1.3750062159356525e-06, + "loss": 0.8919, + "num_input_tokens_seen": 109889120, + "step": 5099 + }, + { + "epoch": 0.6132387422593639, + "flos": 15863239575360.0, + "grad_norm": 2.9855163947993524, + "learning_rate": 1.3742663035614382e-06, + "loss": 0.8304, + "num_input_tokens_seen": 109906490, + "step": 5100 + }, + { + "epoch": 0.613358985150003, + "flos": 25375176687360.0, + "grad_norm": 2.2539572354637336, + "learning_rate": 1.3735264861105283e-06, + "loss": 0.7989, + "num_input_tokens_seen": 109927130, + "step": 5101 + }, + { + "epoch": 0.6134792280406421, + "flos": 21330804897600.0, + "grad_norm": 7.558049934006235, + "learning_rate": 1.372786763695152e-06, + "loss": 0.787, + "num_input_tokens_seen": 109945890, + "step": 5102 + }, + { + "epoch": 0.6135994709312812, + "flos": 21180839151360.0, + "grad_norm": 2.6763130494859673, + "learning_rate": 1.3720471364275257e-06, + "loss": 0.7783, + "num_input_tokens_seen": 109965535, + "step": 5103 + }, + { + "epoch": 0.6137197138219203, + "flos": 14773138694880.0, + "grad_norm": 3.383856561601302, + "learning_rate": 1.3713076044198486e-06, + "loss": 0.7798, + "num_input_tokens_seen": 109982345, + "step": 5104 + }, + { + "epoch": 0.6138399567125594, + "flos": 20055904472160.0, + "grad_norm": 3.494432653566599, + "learning_rate": 1.3705681677843086e-06, + "loss": 0.8104, + "num_input_tokens_seen": 110000940, + "step": 5105 + }, + { + "epoch": 0.6139601996031985, + "flos": 60038785754400.0, + "grad_norm": 0.7769715841564074, + "learning_rate": 1.3698288266330768e-06, + "loss": 0.6045, + "num_input_tokens_seen": 110061565, + "step": 5106 + }, + { + "epoch": 0.6140804424938375, + "flos": 23552977051200.0, + "grad_norm": 6.590264289430357, + "learning_rate": 1.3690895810783113e-06, + "loss": 0.7223, + "num_input_tokens_seen": 110080360, + "step": 5107 + }, + { + "epoch": 0.6142006853844767, + "flos": 21367683033120.0, + "grad_norm": 2.741250772054262, + "learning_rate": 1.3683504312321543e-06, + "loss": 0.7126, + "num_input_tokens_seen": 110100695, + "step": 5108 + }, + { + "epoch": 0.6143209282751158, + "flos": 12040433956800.0, + "grad_norm": 2.9646478520385315, + "learning_rate": 1.3676113772067355e-06, + "loss": 0.8048, + "num_input_tokens_seen": 110117750, + "step": 5109 + }, + { + "epoch": 0.6144411711657548, + "flos": 25046358557760.0, + "grad_norm": 2.906046378272952, + "learning_rate": 1.3668724191141671e-06, + "loss": 0.7286, + "num_input_tokens_seen": 110137020, + "step": 5110 + }, + { + "epoch": 0.6145614140563939, + "flos": 20085162806880.0, + "grad_norm": 2.7631041681711963, + "learning_rate": 1.3661335570665493e-06, + "loss": 0.6676, + "num_input_tokens_seen": 110154930, + "step": 5111 + }, + { + "epoch": 0.614681656947033, + "flos": 16776532409280.0, + "grad_norm": 4.141043703034231, + "learning_rate": 1.3653947911759676e-06, + "loss": 0.7002, + "num_input_tokens_seen": 110172480, + "step": 5112 + }, + { + "epoch": 0.6148018998376721, + "flos": 38745719630400.0, + "grad_norm": 3.2739095779916894, + "learning_rate": 1.3646561215544904e-06, + "loss": 0.7464, + "num_input_tokens_seen": 110197120, + "step": 5113 + }, + { + "epoch": 0.6149221427283111, + "flos": 23293529160960.0, + "grad_norm": 2.6181152521326685, + "learning_rate": 1.363917548314176e-06, + "loss": 0.7928, + "num_input_tokens_seen": 110216385, + "step": 5114 + }, + { + "epoch": 0.6150423856189503, + "flos": 22347409695360.0, + "grad_norm": 2.9592399461618455, + "learning_rate": 1.3631790715670626e-06, + "loss": 0.7287, + "num_input_tokens_seen": 110234790, + "step": 5115 + }, + { + "epoch": 0.6151626285095894, + "flos": 18665277383040.0, + "grad_norm": 2.449348409770707, + "learning_rate": 1.3624406914251783e-06, + "loss": 0.8607, + "num_input_tokens_seen": 110251910, + "step": 5116 + }, + { + "epoch": 0.6152828714002284, + "flos": 15829223511360.0, + "grad_norm": 2.825057078097155, + "learning_rate": 1.3617024080005335e-06, + "loss": 0.885, + "num_input_tokens_seen": 110268810, + "step": 5117 + }, + { + "epoch": 0.6154031142908676, + "flos": 24834419354400.0, + "grad_norm": 1.8525645445829149, + "learning_rate": 1.3609642214051266e-06, + "loss": 0.7464, + "num_input_tokens_seen": 110293035, + "step": 5118 + }, + { + "epoch": 0.6155233571815066, + "flos": 19217334323040.0, + "grad_norm": 2.6189834126533094, + "learning_rate": 1.3602261317509385e-06, + "loss": 0.6612, + "num_input_tokens_seen": 110309695, + "step": 5119 + }, + { + "epoch": 0.6156436000721457, + "flos": 18743345346240.0, + "grad_norm": 3.634496679917731, + "learning_rate": 1.3594881391499387e-06, + "loss": 0.832, + "num_input_tokens_seen": 110328610, + "step": 5120 + }, + { + "epoch": 0.6157638429627849, + "flos": 18015558282720.0, + "grad_norm": 6.551707375742106, + "learning_rate": 1.3587502437140778e-06, + "loss": 0.788, + "num_input_tokens_seen": 110346775, + "step": 5121 + }, + { + "epoch": 0.6158840858534239, + "flos": 25520124516000.0, + "grad_norm": 2.92990559084081, + "learning_rate": 1.3580124455552952e-06, + "loss": 0.8514, + "num_input_tokens_seen": 110366140, + "step": 5122 + }, + { + "epoch": 0.616004328744063, + "flos": 24605382061440.0, + "grad_norm": 2.05959184399307, + "learning_rate": 1.3572747447855148e-06, + "loss": 0.8746, + "num_input_tokens_seen": 110386550, + "step": 5123 + }, + { + "epoch": 0.6161245716347021, + "flos": 21658730952960.0, + "grad_norm": 2.238238429114281, + "learning_rate": 1.356537141516644e-06, + "loss": 0.6953, + "num_input_tokens_seen": 110403970, + "step": 5124 + }, + { + "epoch": 0.6162448145253412, + "flos": 35138607360960.0, + "grad_norm": 2.9521351926596697, + "learning_rate": 1.3557996358605775e-06, + "loss": 0.6173, + "num_input_tokens_seen": 110423890, + "step": 5125 + }, + { + "epoch": 0.6163650574159802, + "flos": 21584379965760.0, + "grad_norm": 5.101168046221474, + "learning_rate": 1.3550622279291941e-06, + "loss": 0.7047, + "num_input_tokens_seen": 110442035, + "step": 5126 + }, + { + "epoch": 0.6164853003066194, + "flos": 24537015405600.0, + "grad_norm": 1.610581473553226, + "learning_rate": 1.354324917834358e-06, + "loss": 0.8325, + "num_input_tokens_seen": 110463280, + "step": 5127 + }, + { + "epoch": 0.6166055431972585, + "flos": 21804831044160.0, + "grad_norm": 3.0495198043778107, + "learning_rate": 1.353587705687918e-06, + "loss": 0.767, + "num_input_tokens_seen": 110481455, + "step": 5128 + }, + { + "epoch": 0.6167257860878975, + "flos": 17759976047520.0, + "grad_norm": 4.098529151767222, + "learning_rate": 1.3528505916017096e-06, + "loss": 0.7247, + "num_input_tokens_seen": 110499155, + "step": 5129 + }, + { + "epoch": 0.6168460289785367, + "flos": 23181370794240.0, + "grad_norm": 3.8022169673201383, + "learning_rate": 1.3521135756875514e-06, + "loss": 0.8878, + "num_input_tokens_seen": 110515470, + "step": 5130 + }, + { + "epoch": 0.6169662718691757, + "flos": 26175716438400.0, + "grad_norm": 2.0409635254265526, + "learning_rate": 1.3513766580572496e-06, + "loss": 0.865, + "num_input_tokens_seen": 110538290, + "step": 5131 + }, + { + "epoch": 0.6170865147598148, + "flos": 18999039090720.0, + "grad_norm": 2.693776630172621, + "learning_rate": 1.3506398388225924e-06, + "loss": 0.7761, + "num_input_tokens_seen": 110555610, + "step": 5132 + }, + { + "epoch": 0.617206757650454, + "flos": 18234411061440.0, + "grad_norm": 2.1544710643964975, + "learning_rate": 1.349903118095355e-06, + "loss": 0.7184, + "num_input_tokens_seen": 110574745, + "step": 5133 + }, + { + "epoch": 0.617327000541093, + "flos": 18161100827520.0, + "grad_norm": 2.515839751181599, + "learning_rate": 1.349166495987298e-06, + "loss": 0.7337, + "num_input_tokens_seen": 110593310, + "step": 5134 + }, + { + "epoch": 0.6174472434317321, + "flos": 61746076461600.0, + "grad_norm": 0.8286879083839709, + "learning_rate": 1.348429972610166e-06, + "loss": 0.6079, + "num_input_tokens_seen": 110657615, + "step": 5135 + }, + { + "epoch": 0.6175674863223712, + "flos": 71128980135840.0, + "grad_norm": 0.8593647349351418, + "learning_rate": 1.3476935480756897e-06, + "loss": 0.5778, + "num_input_tokens_seen": 110714365, + "step": 5136 + }, + { + "epoch": 0.6176877292130103, + "flos": 21804273497760.0, + "grad_norm": 2.8276125617519527, + "learning_rate": 1.346957222495583e-06, + "loss": 0.752, + "num_input_tokens_seen": 110732160, + "step": 5137 + }, + { + "epoch": 0.6178079721036493, + "flos": 17716258676160.0, + "grad_norm": 3.3128189797037138, + "learning_rate": 1.3462209959815466e-06, + "loss": 0.7143, + "num_input_tokens_seen": 110746900, + "step": 5138 + }, + { + "epoch": 0.6179282149942885, + "flos": 22600613065920.0, + "grad_norm": 2.6091410823295855, + "learning_rate": 1.345484868645265e-06, + "loss": 0.7441, + "num_input_tokens_seen": 110765710, + "step": 5139 + }, + { + "epoch": 0.6180484578849276, + "flos": 22278485493120.0, + "grad_norm": 2.5194533698295274, + "learning_rate": 1.3447488405984088e-06, + "loss": 0.7852, + "num_input_tokens_seen": 110783970, + "step": 5140 + }, + { + "epoch": 0.6181687007755666, + "flos": 35174779271040.0, + "grad_norm": 2.7200348547497453, + "learning_rate": 1.3440129119526322e-06, + "loss": 0.7, + "num_input_tokens_seen": 110806395, + "step": 5141 + }, + { + "epoch": 0.6182889436662057, + "flos": 61458937100160.0, + "grad_norm": 0.8031344194268042, + "learning_rate": 1.3432770828195762e-06, + "loss": 0.511, + "num_input_tokens_seen": 110867380, + "step": 5142 + }, + { + "epoch": 0.6184091865568448, + "flos": 19582064174400.0, + "grad_norm": 3.671731369780618, + "learning_rate": 1.3425413533108635e-06, + "loss": 0.7083, + "num_input_tokens_seen": 110885980, + "step": 5143 + }, + { + "epoch": 0.6185294294474839, + "flos": 23553162900000.0, + "grad_norm": 3.0867211228160962, + "learning_rate": 1.341805723538105e-06, + "loss": 0.7077, + "num_input_tokens_seen": 110906085, + "step": 5144 + }, + { + "epoch": 0.618649672338123, + "flos": 26723721874560.0, + "grad_norm": 2.1028209449483977, + "learning_rate": 1.3410701936128948e-06, + "loss": 0.7755, + "num_input_tokens_seen": 110928865, + "step": 5145 + }, + { + "epoch": 0.6187699152287621, + "flos": 14435883029760.0, + "grad_norm": 3.4304019123082736, + "learning_rate": 1.340334763646812e-06, + "loss": 0.8535, + "num_input_tokens_seen": 110943000, + "step": 5146 + }, + { + "epoch": 0.6188901581194012, + "flos": 20056164660480.0, + "grad_norm": 2.576249984008256, + "learning_rate": 1.3395994337514218e-06, + "loss": 0.7427, + "num_input_tokens_seen": 110963170, + "step": 5147 + }, + { + "epoch": 0.6190104010100402, + "flos": 25663845742560.0, + "grad_norm": 2.1977251038680046, + "learning_rate": 1.3388642040382725e-06, + "loss": 0.7845, + "num_input_tokens_seen": 110983595, + "step": 5148 + }, + { + "epoch": 0.6191306439006794, + "flos": 30398531744160.0, + "grad_norm": 12.229085702102795, + "learning_rate": 1.3381290746188975e-06, + "loss": 0.842, + "num_input_tokens_seen": 111002965, + "step": 5149 + }, + { + "epoch": 0.6192508867913185, + "flos": 26647475229600.0, + "grad_norm": 9.210923418935378, + "learning_rate": 1.3373940456048152e-06, + "loss": 0.6721, + "num_input_tokens_seen": 111022990, + "step": 5150 + }, + { + "epoch": 0.6193711296819575, + "flos": 36669201530880.0, + "grad_norm": 2.2059051071190185, + "learning_rate": 1.3366591171075299e-06, + "loss": 0.5953, + "num_input_tokens_seen": 111046495, + "step": 5151 + }, + { + "epoch": 0.6194913725725967, + "flos": 25156509757440.0, + "grad_norm": 4.061025498637744, + "learning_rate": 1.335924289238529e-06, + "loss": 0.908, + "num_input_tokens_seen": 111065705, + "step": 5152 + }, + { + "epoch": 0.6196116154632357, + "flos": 21148607235840.0, + "grad_norm": 2.962649532358985, + "learning_rate": 1.3351895621092859e-06, + "loss": 0.7708, + "num_input_tokens_seen": 111086050, + "step": 5153 + }, + { + "epoch": 0.6197318583538748, + "flos": 16230719988960.0, + "grad_norm": 2.3702994560632153, + "learning_rate": 1.3344549358312567e-06, + "loss": 0.7631, + "num_input_tokens_seen": 111104450, + "step": 5154 + }, + { + "epoch": 0.619852101244514, + "flos": 24390655126080.0, + "grad_norm": 2.6716258507282222, + "learning_rate": 1.3337204105158852e-06, + "loss": 0.7833, + "num_input_tokens_seen": 111123320, + "step": 5155 + }, + { + "epoch": 0.619972344135153, + "flos": 16703556703200.0, + "grad_norm": 2.5239042907378506, + "learning_rate": 1.332985986274597e-06, + "loss": 0.7275, + "num_input_tokens_seen": 111140950, + "step": 5156 + }, + { + "epoch": 0.6200925870257921, + "flos": 12477619137600.0, + "grad_norm": 2.5026310427383978, + "learning_rate": 1.3322516632188047e-06, + "loss": 0.7557, + "num_input_tokens_seen": 111157845, + "step": 5157 + }, + { + "epoch": 0.6202128299164312, + "flos": 26501226459360.0, + "grad_norm": 2.6594243548214216, + "learning_rate": 1.3315174414599045e-06, + "loss": 0.6699, + "num_input_tokens_seen": 111179165, + "step": 5158 + }, + { + "epoch": 0.6203330728070703, + "flos": 18743791383360.0, + "grad_norm": 4.753450044230827, + "learning_rate": 1.3307833211092768e-06, + "loss": 0.7545, + "num_input_tokens_seen": 111197830, + "step": 5159 + }, + { + "epoch": 0.6204533156977093, + "flos": 20600304441600.0, + "grad_norm": 2.225958062831677, + "learning_rate": 1.3300493022782873e-06, + "loss": 0.7516, + "num_input_tokens_seen": 111218400, + "step": 5160 + }, + { + "epoch": 0.6205735585883485, + "flos": 17323422752640.0, + "grad_norm": 2.678050083105174, + "learning_rate": 1.3293153850782855e-06, + "loss": 0.7283, + "num_input_tokens_seen": 111236720, + "step": 5161 + }, + { + "epoch": 0.6206938014789876, + "flos": 22932144588000.0, + "grad_norm": 5.225721990175474, + "learning_rate": 1.3285815696206069e-06, + "loss": 0.7129, + "num_input_tokens_seen": 111258265, + "step": 5162 + }, + { + "epoch": 0.6208140443696266, + "flos": 23948712216000.0, + "grad_norm": 2.925044497942109, + "learning_rate": 1.32784785601657e-06, + "loss": 0.7688, + "num_input_tokens_seen": 111277675, + "step": 5163 + }, + { + "epoch": 0.6209342872602658, + "flos": 35026746352320.0, + "grad_norm": 2.1663345913843997, + "learning_rate": 1.3271142443774798e-06, + "loss": 0.7441, + "num_input_tokens_seen": 111299910, + "step": 5164 + }, + { + "epoch": 0.6210545301509048, + "flos": 26940641825760.0, + "grad_norm": 2.734000131251256, + "learning_rate": 1.3263807348146228e-06, + "loss": 0.8164, + "num_input_tokens_seen": 111319765, + "step": 5165 + }, + { + "epoch": 0.6211747730415439, + "flos": 33569834113920.0, + "grad_norm": 3.300847123972851, + "learning_rate": 1.3256473274392733e-06, + "loss": 0.7359, + "num_input_tokens_seen": 111341110, + "step": 5166 + }, + { + "epoch": 0.6212950159321831, + "flos": 34117951059360.0, + "grad_norm": 2.1534228380496, + "learning_rate": 1.3249140223626873e-06, + "loss": 0.7035, + "num_input_tokens_seen": 111362730, + "step": 5167 + }, + { + "epoch": 0.6214152588228221, + "flos": 27925683763680.0, + "grad_norm": 2.7988830688790944, + "learning_rate": 1.3241808196961077e-06, + "loss": 0.7512, + "num_input_tokens_seen": 111383855, + "step": 5168 + }, + { + "epoch": 0.6215355017134612, + "flos": 20201372677440.0, + "grad_norm": 2.553954599692414, + "learning_rate": 1.3234477195507608e-06, + "loss": 0.7115, + "num_input_tokens_seen": 111402685, + "step": 5169 + }, + { + "epoch": 0.6216557446041003, + "flos": 41368421847840.0, + "grad_norm": 6.8077513216818755, + "learning_rate": 1.322714722037857e-06, + "loss": 0.6254, + "num_input_tokens_seen": 111424130, + "step": 5170 + }, + { + "epoch": 0.6217759874947394, + "flos": 27888731288640.0, + "grad_norm": 2.595692653936273, + "learning_rate": 1.321981827268591e-06, + "loss": 0.7741, + "num_input_tokens_seen": 111444785, + "step": 5171 + }, + { + "epoch": 0.6218962303853784, + "flos": 21734717409600.0, + "grad_norm": 2.1950898515737194, + "learning_rate": 1.3212490353541426e-06, + "loss": 0.8156, + "num_input_tokens_seen": 111467045, + "step": 5172 + }, + { + "epoch": 0.6220164732760175, + "flos": 21215747289600.0, + "grad_norm": 2.590129025312703, + "learning_rate": 1.3205163464056762e-06, + "loss": 0.8066, + "num_input_tokens_seen": 111483650, + "step": 5173 + }, + { + "epoch": 0.6221367161666567, + "flos": 26098763568000.0, + "grad_norm": 2.7042811572775123, + "learning_rate": 1.319783760534339e-06, + "loss": 0.7288, + "num_input_tokens_seen": 111502895, + "step": 5174 + }, + { + "epoch": 0.6222569590572957, + "flos": 16260907567680.0, + "grad_norm": 4.444617567399312, + "learning_rate": 1.319051277851266e-06, + "loss": 0.7553, + "num_input_tokens_seen": 111519180, + "step": 5175 + }, + { + "epoch": 0.6223772019479348, + "flos": 18197941793280.0, + "grad_norm": 3.28503374195624, + "learning_rate": 1.3183188984675716e-06, + "loss": 0.8443, + "num_input_tokens_seen": 111537545, + "step": 5176 + }, + { + "epoch": 0.6224974448385739, + "flos": 27450356675520.0, + "grad_norm": 3.11822416042735, + "learning_rate": 1.3175866224943586e-06, + "loss": 0.71, + "num_input_tokens_seen": 111556265, + "step": 5177 + }, + { + "epoch": 0.622617687729213, + "flos": 19763741459520.0, + "grad_norm": 3.7100656870658977, + "learning_rate": 1.316854450042712e-06, + "loss": 0.7356, + "num_input_tokens_seen": 111574205, + "step": 5178 + }, + { + "epoch": 0.622737930619852, + "flos": 23005938028800.0, + "grad_norm": 2.3626235929585557, + "learning_rate": 1.3161223812237024e-06, + "loss": 0.7426, + "num_input_tokens_seen": 111593560, + "step": 5179 + }, + { + "epoch": 0.6228581735104912, + "flos": 12616248107040.0, + "grad_norm": 3.303485191721959, + "learning_rate": 1.3153904161483842e-06, + "loss": 0.8542, + "num_input_tokens_seen": 111608860, + "step": 5180 + }, + { + "epoch": 0.6229784164011303, + "flos": 23768670400320.0, + "grad_norm": 2.510790996158701, + "learning_rate": 1.3146585549277953e-06, + "loss": 0.8529, + "num_input_tokens_seen": 111627500, + "step": 5181 + }, + { + "epoch": 0.6230986592917693, + "flos": 22382020475520.0, + "grad_norm": 3.0581900962389317, + "learning_rate": 1.3139267976729591e-06, + "loss": 0.7879, + "num_input_tokens_seen": 111647690, + "step": 5182 + }, + { + "epoch": 0.6232189021824085, + "flos": 34478406388320.0, + "grad_norm": 2.593173870767161, + "learning_rate": 1.3131951444948815e-06, + "loss": 0.7155, + "num_input_tokens_seen": 111667090, + "step": 5183 + }, + { + "epoch": 0.6233391450730476, + "flos": 22205249598720.0, + "grad_norm": 2.1254709507283644, + "learning_rate": 1.3124635955045546e-06, + "loss": 0.7615, + "num_input_tokens_seen": 111686905, + "step": 5184 + }, + { + "epoch": 0.6234593879636866, + "flos": 20303458039200.0, + "grad_norm": 2.1319332336506553, + "learning_rate": 1.3117321508129537e-06, + "loss": 0.843, + "num_input_tokens_seen": 111704985, + "step": 5185 + }, + { + "epoch": 0.6235796308543258, + "flos": 20635324089120.0, + "grad_norm": 1.8213202179808419, + "learning_rate": 1.3110008105310388e-06, + "loss": 0.7678, + "num_input_tokens_seen": 111724805, + "step": 5186 + }, + { + "epoch": 0.6236998737449648, + "flos": 26580632533920.0, + "grad_norm": 2.0995189842461444, + "learning_rate": 1.3102695747697526e-06, + "loss": 0.7809, + "num_input_tokens_seen": 111747675, + "step": 5187 + }, + { + "epoch": 0.6238201166356039, + "flos": 12656657369760.0, + "grad_norm": 3.4113342417892887, + "learning_rate": 1.3095384436400237e-06, + "loss": 0.9059, + "num_input_tokens_seen": 111759205, + "step": 5188 + }, + { + "epoch": 0.623940359526243, + "flos": 10437570306240.0, + "grad_norm": 2.9286567197069657, + "learning_rate": 1.3088074172527633e-06, + "loss": 0.8196, + "num_input_tokens_seen": 111776335, + "step": 5189 + }, + { + "epoch": 0.6240606024168821, + "flos": 29017828980960.0, + "grad_norm": 2.5189763840631807, + "learning_rate": 1.3080764957188684e-06, + "loss": 0.7111, + "num_input_tokens_seen": 111796580, + "step": 5190 + }, + { + "epoch": 0.6241808453075212, + "flos": 21986657008320.0, + "grad_norm": 2.7049566689227227, + "learning_rate": 1.3073456791492192e-06, + "loss": 0.7079, + "num_input_tokens_seen": 111816290, + "step": 5191 + }, + { + "epoch": 0.6243010881981603, + "flos": 21108420991680.0, + "grad_norm": 2.4916576432114668, + "learning_rate": 1.3066149676546801e-06, + "loss": 0.7845, + "num_input_tokens_seen": 111834470, + "step": 5192 + }, + { + "epoch": 0.6244213310887994, + "flos": 22312984764000.0, + "grad_norm": 2.339140262470591, + "learning_rate": 1.3058843613460985e-06, + "loss": 0.6635, + "num_input_tokens_seen": 111852410, + "step": 5193 + }, + { + "epoch": 0.6245415739794384, + "flos": 15210435384960.0, + "grad_norm": 2.3529324517368795, + "learning_rate": 1.3051538603343075e-06, + "loss": 0.7447, + "num_input_tokens_seen": 111870340, + "step": 5194 + }, + { + "epoch": 0.6246618168700776, + "flos": 18853013339040.0, + "grad_norm": 9.199703493730578, + "learning_rate": 1.3044234647301235e-06, + "loss": 0.6792, + "num_input_tokens_seen": 111888365, + "step": 5195 + }, + { + "epoch": 0.6247820597607167, + "flos": 14298963869280.0, + "grad_norm": 2.2810392935653394, + "learning_rate": 1.303693174644347e-06, + "loss": 0.7229, + "num_input_tokens_seen": 111905840, + "step": 5196 + }, + { + "epoch": 0.6249023026513557, + "flos": 22605630983520.0, + "grad_norm": 2.1443005491589306, + "learning_rate": 1.3029629901877625e-06, + "loss": 0.8063, + "num_input_tokens_seen": 111925215, + "step": 5197 + }, + { + "epoch": 0.6250225455419949, + "flos": 20237804775840.0, + "grad_norm": 3.4473567788497252, + "learning_rate": 1.3022329114711376e-06, + "loss": 0.7745, + "num_input_tokens_seen": 111943925, + "step": 5198 + }, + { + "epoch": 0.6251427884326339, + "flos": 23404237907040.0, + "grad_norm": 2.768473047738093, + "learning_rate": 1.3015029386052256e-06, + "loss": 0.7018, + "num_input_tokens_seen": 111964095, + "step": 5199 + }, + { + "epoch": 0.625263031323273, + "flos": 31677483673440.0, + "grad_norm": 3.3363900156894117, + "learning_rate": 1.3007730717007622e-06, + "loss": 0.7316, + "num_input_tokens_seen": 111984910, + "step": 5200 + }, + { + "epoch": 0.6253832742139122, + "flos": 24099830224800.0, + "grad_norm": 2.4510946436413676, + "learning_rate": 1.3000433108684676e-06, + "loss": 0.7565, + "num_input_tokens_seen": 112005410, + "step": 5201 + }, + { + "epoch": 0.6255035171045512, + "flos": 27628986040320.0, + "grad_norm": 3.560216874792411, + "learning_rate": 1.2993136562190467e-06, + "loss": 0.8065, + "num_input_tokens_seen": 112024530, + "step": 5202 + }, + { + "epoch": 0.6256237599951903, + "flos": 20198882303520.0, + "grad_norm": 1.8367542803729915, + "learning_rate": 1.2985841078631871e-06, + "loss": 0.7023, + "num_input_tokens_seen": 112043850, + "step": 5203 + }, + { + "epoch": 0.6257440028858293, + "flos": 24136002134880.0, + "grad_norm": 4.307483575029104, + "learning_rate": 1.2978546659115608e-06, + "loss": 0.7812, + "num_input_tokens_seen": 112061930, + "step": 5204 + }, + { + "epoch": 0.6258642457764685, + "flos": 15829483699680.0, + "grad_norm": 4.657314511678738, + "learning_rate": 1.2971253304748228e-06, + "loss": 0.8534, + "num_input_tokens_seen": 112079645, + "step": 5205 + }, + { + "epoch": 0.6259844886671075, + "flos": 11894631223680.0, + "grad_norm": 2.4046174493393404, + "learning_rate": 1.296396101663614e-06, + "loss": 0.7527, + "num_input_tokens_seen": 112096205, + "step": 5206 + }, + { + "epoch": 0.6261047315577466, + "flos": 15865432591200.0, + "grad_norm": 3.0617424164110676, + "learning_rate": 1.2956669795885565e-06, + "loss": 0.8455, + "num_input_tokens_seen": 112112835, + "step": 5207 + }, + { + "epoch": 0.6262249744483858, + "flos": 31204052243040.0, + "grad_norm": 2.9112995538879645, + "learning_rate": 1.294937964360259e-06, + "loss": 0.6846, + "num_input_tokens_seen": 112133105, + "step": 5208 + }, + { + "epoch": 0.6263452173390248, + "flos": 27159606113760.0, + "grad_norm": 6.463472876744976, + "learning_rate": 1.2942090560893108e-06, + "loss": 0.7149, + "num_input_tokens_seen": 112152510, + "step": 5209 + }, + { + "epoch": 0.6264654602296639, + "flos": 37288324185120.0, + "grad_norm": 2.6566054162495574, + "learning_rate": 1.2934802548862882e-06, + "loss": 0.6042, + "num_input_tokens_seen": 112175295, + "step": 5210 + }, + { + "epoch": 0.626585703120303, + "flos": 14736074710560.0, + "grad_norm": 4.102379243457792, + "learning_rate": 1.292751560861749e-06, + "loss": 0.8278, + "num_input_tokens_seen": 112190200, + "step": 5211 + }, + { + "epoch": 0.6267059460109421, + "flos": 22315437968160.0, + "grad_norm": 2.213979724098548, + "learning_rate": 1.2920229741262354e-06, + "loss": 0.7962, + "num_input_tokens_seen": 112208205, + "step": 5212 + }, + { + "epoch": 0.6268261889015811, + "flos": 17723581118880.0, + "grad_norm": 4.306870697865201, + "learning_rate": 1.2912944947902739e-06, + "loss": 0.7553, + "num_input_tokens_seen": 112224690, + "step": 5213 + }, + { + "epoch": 0.6269464317922203, + "flos": 32798998734720.0, + "grad_norm": 2.9228876164561335, + "learning_rate": 1.2905661229643742e-06, + "loss": 0.7124, + "num_input_tokens_seen": 112244450, + "step": 5214 + }, + { + "epoch": 0.6270666746828594, + "flos": 17903622934560.0, + "grad_norm": 2.7806008256378845, + "learning_rate": 1.2898378587590299e-06, + "loss": 0.8481, + "num_input_tokens_seen": 112261050, + "step": 5215 + }, + { + "epoch": 0.6271869175734984, + "flos": 17432533199040.0, + "grad_norm": 2.2367726513027026, + "learning_rate": 1.2891097022847173e-06, + "loss": 0.8739, + "num_input_tokens_seen": 112278395, + "step": 5216 + }, + { + "epoch": 0.6273071604641376, + "flos": 26830044588960.0, + "grad_norm": 4.002371561823875, + "learning_rate": 1.2883816536518978e-06, + "loss": 0.668, + "num_input_tokens_seen": 112299535, + "step": 5217 + }, + { + "epoch": 0.6274274033547766, + "flos": 26024858617920.0, + "grad_norm": 2.5973043332462047, + "learning_rate": 1.2876537129710155e-06, + "loss": 0.8175, + "num_input_tokens_seen": 112317265, + "step": 5218 + }, + { + "epoch": 0.6275476462454157, + "flos": 20237358738720.0, + "grad_norm": 4.03962134031398, + "learning_rate": 1.286925880352499e-06, + "loss": 0.7562, + "num_input_tokens_seen": 112336125, + "step": 5219 + }, + { + "epoch": 0.6276678891360549, + "flos": 26282596699200.0, + "grad_norm": 2.1690501772122857, + "learning_rate": 1.2861981559067592e-06, + "loss": 0.7132, + "num_input_tokens_seen": 112356730, + "step": 5220 + }, + { + "epoch": 0.6277881320266939, + "flos": 13892672492640.0, + "grad_norm": 5.432700132360465, + "learning_rate": 1.2854705397441917e-06, + "loss": 0.803, + "num_input_tokens_seen": 112372270, + "step": 5221 + }, + { + "epoch": 0.627908374917333, + "flos": 27010012065120.0, + "grad_norm": 3.211587881782485, + "learning_rate": 1.2847430319751747e-06, + "loss": 0.7816, + "num_input_tokens_seen": 112390365, + "step": 5222 + }, + { + "epoch": 0.6280286178079721, + "flos": 23735211882720.0, + "grad_norm": 5.913108436295967, + "learning_rate": 1.2840156327100712e-06, + "loss": 0.6786, + "num_input_tokens_seen": 112407490, + "step": 5223 + }, + { + "epoch": 0.6281488606986112, + "flos": 26321370492480.0, + "grad_norm": 2.5953804314970856, + "learning_rate": 1.2832883420592272e-06, + "loss": 0.727, + "num_input_tokens_seen": 112426385, + "step": 5224 + }, + { + "epoch": 0.6282691035892503, + "flos": 36085172863680.0, + "grad_norm": 2.9575728571761966, + "learning_rate": 1.282561160132972e-06, + "loss": 0.648, + "num_input_tokens_seen": 112446905, + "step": 5225 + }, + { + "epoch": 0.6283893464798894, + "flos": 26499033443520.0, + "grad_norm": 1.9922605014964168, + "learning_rate": 1.2818340870416186e-06, + "loss": 0.8118, + "num_input_tokens_seen": 112468040, + "step": 5226 + }, + { + "epoch": 0.6285095893705285, + "flos": 22205472617280.0, + "grad_norm": 2.6340259506291215, + "learning_rate": 1.2811071228954626e-06, + "loss": 0.7599, + "num_input_tokens_seen": 112486675, + "step": 5227 + }, + { + "epoch": 0.6286298322611675, + "flos": 26504237209920.0, + "grad_norm": 2.63650570609222, + "learning_rate": 1.2803802678047846e-06, + "loss": 0.8069, + "num_input_tokens_seen": 112506825, + "step": 5228 + }, + { + "epoch": 0.6287500751518067, + "flos": 21763975744320.0, + "grad_norm": 2.2913294688593524, + "learning_rate": 1.279653521879848e-06, + "loss": 0.7395, + "num_input_tokens_seen": 112526890, + "step": 5229 + }, + { + "epoch": 0.6288703180424458, + "flos": 19981033108320.0, + "grad_norm": 2.967841079508432, + "learning_rate": 1.2789268852308997e-06, + "loss": 0.8451, + "num_input_tokens_seen": 112542100, + "step": 5230 + }, + { + "epoch": 0.6289905609330848, + "flos": 22093239911040.0, + "grad_norm": 2.591416101721598, + "learning_rate": 1.2782003579681688e-06, + "loss": 0.704, + "num_input_tokens_seen": 112561630, + "step": 5231 + }, + { + "epoch": 0.629110803823724, + "flos": 25481871099360.0, + "grad_norm": 2.749136182190686, + "learning_rate": 1.2774739402018701e-06, + "loss": 0.7454, + "num_input_tokens_seen": 112582465, + "step": 5232 + }, + { + "epoch": 0.629231046714363, + "flos": 20856815920800.0, + "grad_norm": 4.7202136062521, + "learning_rate": 1.2767476320422002e-06, + "loss": 0.733, + "num_input_tokens_seen": 112602390, + "step": 5233 + }, + { + "epoch": 0.6293512896050021, + "flos": 64957825263840.0, + "grad_norm": 0.6972831061386427, + "learning_rate": 1.2760214335993392e-06, + "loss": 0.5712, + "num_input_tokens_seen": 112669035, + "step": 5234 + }, + { + "epoch": 0.6294715324956413, + "flos": 34627963267200.0, + "grad_norm": 3.051457016426039, + "learning_rate": 1.2752953449834514e-06, + "loss": 0.5862, + "num_input_tokens_seen": 112691485, + "step": 5235 + }, + { + "epoch": 0.6295917753862803, + "flos": 22751656735200.0, + "grad_norm": 2.3226096345218625, + "learning_rate": 1.2745693663046836e-06, + "loss": 0.8059, + "num_input_tokens_seen": 112710555, + "step": 5236 + }, + { + "epoch": 0.6297120182769194, + "flos": 20820718350240.0, + "grad_norm": 2.4263816987209954, + "learning_rate": 1.2738434976731662e-06, + "loss": 0.8128, + "num_input_tokens_seen": 112728740, + "step": 5237 + }, + { + "epoch": 0.6298322611675584, + "flos": 19469794298400.0, + "grad_norm": 4.373110553913318, + "learning_rate": 1.2731177391990125e-06, + "loss": 0.7522, + "num_input_tokens_seen": 112748665, + "step": 5238 + }, + { + "epoch": 0.6299525040581976, + "flos": 12586729584000.0, + "grad_norm": 4.674761336655327, + "learning_rate": 1.2723920909923203e-06, + "loss": 0.8181, + "num_input_tokens_seen": 112764410, + "step": 5239 + }, + { + "epoch": 0.6300727469488366, + "flos": 57643156566240.0, + "grad_norm": 0.8693742073872708, + "learning_rate": 1.2716665531631688e-06, + "loss": 0.6042, + "num_input_tokens_seen": 112818695, + "step": 5240 + }, + { + "epoch": 0.6301929898394757, + "flos": 22495070916480.0, + "grad_norm": 3.1265308445248285, + "learning_rate": 1.270941125821623e-06, + "loss": 0.7732, + "num_input_tokens_seen": 112839120, + "step": 5241 + }, + { + "epoch": 0.6303132327301149, + "flos": 28252977933120.0, + "grad_norm": 2.261530682217539, + "learning_rate": 1.2702158090777278e-06, + "loss": 0.7558, + "num_input_tokens_seen": 112860210, + "step": 5242 + }, + { + "epoch": 0.6304334756207539, + "flos": 25228333200960.0, + "grad_norm": 2.2625839344283456, + "learning_rate": 1.2694906030415148e-06, + "loss": 0.7472, + "num_input_tokens_seen": 112877955, + "step": 5243 + }, + { + "epoch": 0.630553718511393, + "flos": 18007641123840.0, + "grad_norm": 4.161948520912129, + "learning_rate": 1.2687655078229958e-06, + "loss": 0.8276, + "num_input_tokens_seen": 112892285, + "step": 5244 + }, + { + "epoch": 0.6306739614020321, + "flos": 27264813735360.0, + "grad_norm": 2.700923486240925, + "learning_rate": 1.2680405235321678e-06, + "loss": 0.697, + "num_input_tokens_seen": 112913620, + "step": 5245 + }, + { + "epoch": 0.6307942042926712, + "flos": 15319731680160.0, + "grad_norm": 3.5108306067186246, + "learning_rate": 1.267315650279011e-06, + "loss": 0.7834, + "num_input_tokens_seen": 112932090, + "step": 5246 + }, + { + "epoch": 0.6309144471833102, + "flos": 19578347198400.0, + "grad_norm": 3.9015577286464618, + "learning_rate": 1.2665908881734874e-06, + "loss": 0.742, + "num_input_tokens_seen": 112950925, + "step": 5247 + }, + { + "epoch": 0.6310346900739494, + "flos": 17468482090560.0, + "grad_norm": 3.3381116156639177, + "learning_rate": 1.2658662373255432e-06, + "loss": 0.8504, + "num_input_tokens_seen": 112969315, + "step": 5248 + }, + { + "epoch": 0.6311549329645885, + "flos": 54991790730240.0, + "grad_norm": 0.7258951957945151, + "learning_rate": 1.2651416978451063e-06, + "loss": 0.5215, + "num_input_tokens_seen": 113034700, + "step": 5249 + }, + { + "epoch": 0.6312751758552275, + "flos": 41842150636320.0, + "grad_norm": 2.7273596005619773, + "learning_rate": 1.2644172698420903e-06, + "loss": 0.6514, + "num_input_tokens_seen": 113056805, + "step": 5250 + }, + { + "epoch": 0.6313954187458667, + "flos": 19618830800640.0, + "grad_norm": 3.065817301373784, + "learning_rate": 1.2636929534263892e-06, + "loss": 0.8494, + "num_input_tokens_seen": 113075790, + "step": 5251 + }, + { + "epoch": 0.6315156616365057, + "flos": 22861584916320.0, + "grad_norm": 3.339654000229299, + "learning_rate": 1.2629687487078821e-06, + "loss": 0.7748, + "num_input_tokens_seen": 113094600, + "step": 5252 + }, + { + "epoch": 0.6316359045271448, + "flos": 23692089227520.0, + "grad_norm": 3.916193642696276, + "learning_rate": 1.2622446557964293e-06, + "loss": 0.7673, + "num_input_tokens_seen": 113112605, + "step": 5253 + }, + { + "epoch": 0.631756147417784, + "flos": 33060788319840.0, + "grad_norm": 2.3278798223852237, + "learning_rate": 1.261520674801876e-06, + "loss": 0.714, + "num_input_tokens_seen": 113133200, + "step": 5254 + }, + { + "epoch": 0.631876390308423, + "flos": 31203531866400.0, + "grad_norm": 3.1838097247485315, + "learning_rate": 1.2607968058340488e-06, + "loss": 0.7286, + "num_input_tokens_seen": 113152895, + "step": 5255 + }, + { + "epoch": 0.6319966331990621, + "flos": 24645828493920.0, + "grad_norm": 3.296228137444681, + "learning_rate": 1.2600730490027583e-06, + "loss": 0.734, + "num_input_tokens_seen": 113173135, + "step": 5256 + }, + { + "epoch": 0.6321168760897012, + "flos": 17466846621120.0, + "grad_norm": 2.8480142334357352, + "learning_rate": 1.2593494044177984e-06, + "loss": 0.8056, + "num_input_tokens_seen": 113191440, + "step": 5257 + }, + { + "epoch": 0.6322371189803403, + "flos": 18269319199680.0, + "grad_norm": 2.804946164925213, + "learning_rate": 1.2586258721889448e-06, + "loss": 0.807, + "num_input_tokens_seen": 113208790, + "step": 5258 + }, + { + "epoch": 0.6323573618709794, + "flos": 20128545650400.0, + "grad_norm": 2.390780375299555, + "learning_rate": 1.2579024524259573e-06, + "loss": 0.8205, + "num_input_tokens_seen": 113225050, + "step": 5259 + }, + { + "epoch": 0.6324776047616185, + "flos": 20013636721440.0, + "grad_norm": 2.798096970585422, + "learning_rate": 1.2571791452385768e-06, + "loss": 0.912, + "num_input_tokens_seen": 113242315, + "step": 5260 + }, + { + "epoch": 0.6325978476522576, + "flos": 30803968216320.0, + "grad_norm": 2.205836257826948, + "learning_rate": 1.2564559507365301e-06, + "loss": 0.773, + "num_input_tokens_seen": 113264720, + "step": 5261 + }, + { + "epoch": 0.6327180905428966, + "flos": 24499988591040.0, + "grad_norm": 2.9186340962499853, + "learning_rate": 1.2557328690295244e-06, + "loss": 0.7928, + "num_input_tokens_seen": 113282910, + "step": 5262 + }, + { + "epoch": 0.6328383334335358, + "flos": 21544305230880.0, + "grad_norm": 5.530394062260547, + "learning_rate": 1.255009900227251e-06, + "loss": 0.7659, + "num_input_tokens_seen": 113301935, + "step": 5263 + }, + { + "epoch": 0.6329585763241748, + "flos": 22897087770720.0, + "grad_norm": 2.506010840365869, + "learning_rate": 1.254287044439383e-06, + "loss": 0.7982, + "num_input_tokens_seen": 113321540, + "step": 5264 + }, + { + "epoch": 0.6330788192148139, + "flos": 70835255993280.0, + "grad_norm": 0.7814485297676719, + "learning_rate": 1.2535643017755776e-06, + "loss": 0.5439, + "num_input_tokens_seen": 113383730, + "step": 5265 + }, + { + "epoch": 0.6331990621054531, + "flos": 21213925971360.0, + "grad_norm": 2.869796023679958, + "learning_rate": 1.2528416723454737e-06, + "loss": 0.7223, + "num_input_tokens_seen": 113400400, + "step": 5266 + }, + { + "epoch": 0.6333193049960921, + "flos": 34409853883680.0, + "grad_norm": 1.746699831783614, + "learning_rate": 1.2521191562586945e-06, + "loss": 0.7132, + "num_input_tokens_seen": 113424050, + "step": 5267 + }, + { + "epoch": 0.6334395478867312, + "flos": 18303706961280.0, + "grad_norm": 2.525649008104628, + "learning_rate": 1.2513967536248445e-06, + "loss": 0.7676, + "num_input_tokens_seen": 113440370, + "step": 5268 + }, + { + "epoch": 0.6335597907773702, + "flos": 23589371979840.0, + "grad_norm": 2.126680719255439, + "learning_rate": 1.2506744645535117e-06, + "loss": 0.8104, + "num_input_tokens_seen": 113460800, + "step": 5269 + }, + { + "epoch": 0.6336800336680094, + "flos": 22678458010560.0, + "grad_norm": 2.39805376294519, + "learning_rate": 1.249952289154267e-06, + "loss": 0.5996, + "num_input_tokens_seen": 113480840, + "step": 5270 + }, + { + "epoch": 0.6338002765586485, + "flos": 23588963112480.0, + "grad_norm": 2.5804672236586748, + "learning_rate": 1.2492302275366635e-06, + "loss": 0.7673, + "num_input_tokens_seen": 113500515, + "step": 5271 + }, + { + "epoch": 0.6339205194492875, + "flos": 26467470583680.0, + "grad_norm": 3.5527363882536216, + "learning_rate": 1.2485082798102377e-06, + "loss": 0.6568, + "num_input_tokens_seen": 113520930, + "step": 5272 + }, + { + "epoch": 0.6340407623399267, + "flos": 18517653331680.0, + "grad_norm": 3.003153415192215, + "learning_rate": 1.2477864460845084e-06, + "loss": 0.6876, + "num_input_tokens_seen": 113537330, + "step": 5273 + }, + { + "epoch": 0.6341610052305657, + "flos": 17687409208800.0, + "grad_norm": 8.58210236941401, + "learning_rate": 1.2470647264689776e-06, + "loss": 0.7352, + "num_input_tokens_seen": 113555810, + "step": 5274 + }, + { + "epoch": 0.6342812481212048, + "flos": 23553311579040.0, + "grad_norm": 25.735733969127196, + "learning_rate": 1.2463431210731282e-06, + "loss": 0.7162, + "num_input_tokens_seen": 113575395, + "step": 5275 + }, + { + "epoch": 0.634401491011844, + "flos": 17796593994720.0, + "grad_norm": 3.6861772912824082, + "learning_rate": 1.2456216300064289e-06, + "loss": 0.7603, + "num_input_tokens_seen": 113592945, + "step": 5276 + }, + { + "epoch": 0.634521733902483, + "flos": 21327571128480.0, + "grad_norm": 2.832388177938777, + "learning_rate": 1.244900253378328e-06, + "loss": 0.7855, + "num_input_tokens_seen": 113613000, + "step": 5277 + }, + { + "epoch": 0.6346419767931221, + "flos": 16521582060000.0, + "grad_norm": 3.5171185521028288, + "learning_rate": 1.2441789912982583e-06, + "loss": 0.6935, + "num_input_tokens_seen": 113630085, + "step": 5278 + }, + { + "epoch": 0.6347622196837612, + "flos": 24316081120320.0, + "grad_norm": 2.4485985854471553, + "learning_rate": 1.2434578438756346e-06, + "loss": 0.6481, + "num_input_tokens_seen": 113650515, + "step": 5279 + }, + { + "epoch": 0.6348824625744003, + "flos": 64429828625760.0, + "grad_norm": 2.953143084435368, + "learning_rate": 1.242736811219855e-06, + "loss": 0.7823, + "num_input_tokens_seen": 113676475, + "step": 5280 + }, + { + "epoch": 0.6350027054650393, + "flos": 28581089837280.0, + "grad_norm": 2.5802672762267007, + "learning_rate": 1.2420158934402988e-06, + "loss": 0.8191, + "num_input_tokens_seen": 113697090, + "step": 5281 + }, + { + "epoch": 0.6351229483556785, + "flos": 22969654609440.0, + "grad_norm": 2.573577291540125, + "learning_rate": 1.2412950906463286e-06, + "loss": 0.845, + "num_input_tokens_seen": 113714395, + "step": 5282 + }, + { + "epoch": 0.6352431912463176, + "flos": 21907771310400.0, + "grad_norm": 3.200304231486792, + "learning_rate": 1.2405744029472902e-06, + "loss": 0.9004, + "num_input_tokens_seen": 113733880, + "step": 5283 + }, + { + "epoch": 0.6353634341369566, + "flos": 13715864446080.0, + "grad_norm": 2.7843053087617173, + "learning_rate": 1.2398538304525108e-06, + "loss": 0.762, + "num_input_tokens_seen": 113752505, + "step": 5284 + }, + { + "epoch": 0.6354836770275958, + "flos": 19288600220160.0, + "grad_norm": 3.244482785989777, + "learning_rate": 1.2391333732713016e-06, + "loss": 0.7591, + "num_input_tokens_seen": 113770310, + "step": 5285 + }, + { + "epoch": 0.6356039199182348, + "flos": 21582409968480.0, + "grad_norm": 3.3602678398826353, + "learning_rate": 1.2384130315129543e-06, + "loss": 0.7846, + "num_input_tokens_seen": 113788635, + "step": 5286 + }, + { + "epoch": 0.6357241628088739, + "flos": 18086155124160.0, + "grad_norm": 4.731823727933695, + "learning_rate": 1.2376928052867447e-06, + "loss": 0.7321, + "num_input_tokens_seen": 113807755, + "step": 5287 + }, + { + "epoch": 0.6358444056995131, + "flos": 24899515071360.0, + "grad_norm": 2.954139172022988, + "learning_rate": 1.2369726947019299e-06, + "loss": 0.7764, + "num_input_tokens_seen": 113828230, + "step": 5288 + }, + { + "epoch": 0.6359646485901521, + "flos": 23259996303840.0, + "grad_norm": 3.3330148226834426, + "learning_rate": 1.2362526998677511e-06, + "loss": 0.6712, + "num_input_tokens_seen": 113844595, + "step": 5289 + }, + { + "epoch": 0.6360848914807912, + "flos": 20857670825280.0, + "grad_norm": 2.2732296353775374, + "learning_rate": 1.2355328208934301e-06, + "loss": 0.8457, + "num_input_tokens_seen": 113863470, + "step": 5290 + }, + { + "epoch": 0.6362051343714303, + "flos": 18453524028480.0, + "grad_norm": 2.1772526092671347, + "learning_rate": 1.2348130578881728e-06, + "loss": 0.7258, + "num_input_tokens_seen": 113881690, + "step": 5291 + }, + { + "epoch": 0.6363253772620694, + "flos": 24354706234560.0, + "grad_norm": 2.7888055250247716, + "learning_rate": 1.2340934109611664e-06, + "loss": 0.7616, + "num_input_tokens_seen": 113902450, + "step": 5292 + }, + { + "epoch": 0.6364456201527084, + "flos": 25921063447200.0, + "grad_norm": 4.45712411233782, + "learning_rate": 1.2333738802215798e-06, + "loss": 0.6861, + "num_input_tokens_seen": 113922670, + "step": 5293 + }, + { + "epoch": 0.6365658630433476, + "flos": 20711124696960.0, + "grad_norm": 2.5978070437611396, + "learning_rate": 1.2326544657785668e-06, + "loss": 0.8109, + "num_input_tokens_seen": 113942360, + "step": 5294 + }, + { + "epoch": 0.6366861059339867, + "flos": 21403557585120.0, + "grad_norm": 3.556093042862094, + "learning_rate": 1.2319351677412608e-06, + "loss": 0.7412, + "num_input_tokens_seen": 113959405, + "step": 5295 + }, + { + "epoch": 0.6368063488246257, + "flos": 22235065479840.0, + "grad_norm": 4.140364728700833, + "learning_rate": 1.2312159862187796e-06, + "loss": 0.7418, + "num_input_tokens_seen": 113977815, + "step": 5296 + }, + { + "epoch": 0.6369265917152649, + "flos": 22388748202080.0, + "grad_norm": 3.609238262806644, + "learning_rate": 1.2304969213202217e-06, + "loss": 0.7637, + "num_input_tokens_seen": 113999075, + "step": 5297 + }, + { + "epoch": 0.6370468346059039, + "flos": 24682818138720.0, + "grad_norm": 3.715733434636477, + "learning_rate": 1.2297779731546692e-06, + "loss": 0.7958, + "num_input_tokens_seen": 114018170, + "step": 5298 + }, + { + "epoch": 0.637167077496543, + "flos": 25260193418880.0, + "grad_norm": 4.32059018448132, + "learning_rate": 1.2290591418311853e-06, + "loss": 0.7776, + "num_input_tokens_seen": 114035565, + "step": 5299 + }, + { + "epoch": 0.637287320387182, + "flos": 27631364904960.0, + "grad_norm": 2.5876641631456017, + "learning_rate": 1.2283404274588172e-06, + "loss": 0.7247, + "num_input_tokens_seen": 114054510, + "step": 5300 + }, + { + "epoch": 0.6374075632778212, + "flos": 63083359211520.0, + "grad_norm": 0.747674794699539, + "learning_rate": 1.227621830146592e-06, + "loss": 0.5267, + "num_input_tokens_seen": 114109875, + "step": 5301 + }, + { + "epoch": 0.6375278061684603, + "flos": 25521574136640.0, + "grad_norm": 3.0469374430199982, + "learning_rate": 1.2269033500035217e-06, + "loss": 0.7931, + "num_input_tokens_seen": 114130010, + "step": 5302 + }, + { + "epoch": 0.6376480490590993, + "flos": 25629606660000.0, + "grad_norm": 2.8648029774931323, + "learning_rate": 1.2261849871385988e-06, + "loss": 0.7382, + "num_input_tokens_seen": 114151115, + "step": 5303 + }, + { + "epoch": 0.6377682919497385, + "flos": 31492275261120.0, + "grad_norm": 3.104931287020558, + "learning_rate": 1.2254667416607972e-06, + "loss": 0.6212, + "num_input_tokens_seen": 114172715, + "step": 5304 + }, + { + "epoch": 0.6378885348403776, + "flos": 23006272556640.0, + "grad_norm": 3.1099005020269255, + "learning_rate": 1.2247486136790756e-06, + "loss": 0.8293, + "num_input_tokens_seen": 114192195, + "step": 5305 + }, + { + "epoch": 0.6380087777310166, + "flos": 18670443979680.0, + "grad_norm": 3.4651191197736897, + "learning_rate": 1.2240306033023726e-06, + "loss": 0.8092, + "num_input_tokens_seen": 114210020, + "step": 5306 + }, + { + "epoch": 0.6381290206216558, + "flos": 23298212550720.0, + "grad_norm": 2.2276674975303705, + "learning_rate": 1.223312710639611e-06, + "loss": 0.72, + "num_input_tokens_seen": 114228740, + "step": 5307 + }, + { + "epoch": 0.6382492635122948, + "flos": 18853608055200.0, + "grad_norm": 3.152290504506951, + "learning_rate": 1.2225949357996928e-06, + "loss": 0.8741, + "num_input_tokens_seen": 114246865, + "step": 5308 + }, + { + "epoch": 0.6383695064029339, + "flos": 27779249144640.0, + "grad_norm": 2.1242897866406394, + "learning_rate": 1.221877278891505e-06, + "loss": 0.8039, + "num_input_tokens_seen": 114266120, + "step": 5309 + }, + { + "epoch": 0.638489749293573, + "flos": 26357802590880.0, + "grad_norm": 3.763952239857491, + "learning_rate": 1.221159740023915e-06, + "loss": 0.7137, + "num_input_tokens_seen": 114285950, + "step": 5310 + }, + { + "epoch": 0.6386099921842121, + "flos": 23954027491680.0, + "grad_norm": 3.873645668852827, + "learning_rate": 1.2204423193057735e-06, + "loss": 0.7237, + "num_input_tokens_seen": 114304780, + "step": 5311 + }, + { + "epoch": 0.6387302350748512, + "flos": 71628064434240.0, + "grad_norm": 0.8806705653123329, + "learning_rate": 1.2197250168459122e-06, + "loss": 0.6324, + "num_input_tokens_seen": 114361180, + "step": 5312 + }, + { + "epoch": 0.6388504779654903, + "flos": 14515140425280.0, + "grad_norm": 2.3400648518892297, + "learning_rate": 1.2190078327531454e-06, + "loss": 0.7453, + "num_input_tokens_seen": 114377425, + "step": 5313 + }, + { + "epoch": 0.6389707208561294, + "flos": 22314211366080.0, + "grad_norm": 2.6719463851940524, + "learning_rate": 1.2182907671362697e-06, + "loss": 0.7288, + "num_input_tokens_seen": 114398120, + "step": 5314 + }, + { + "epoch": 0.6390909637467684, + "flos": 19399197456960.0, + "grad_norm": 2.330165915741893, + "learning_rate": 1.2175738201040626e-06, + "loss": 0.7879, + "num_input_tokens_seen": 114415160, + "step": 5315 + }, + { + "epoch": 0.6392112066374076, + "flos": 24056150023200.0, + "grad_norm": 2.478810250016067, + "learning_rate": 1.2168569917652855e-06, + "loss": 0.7874, + "num_input_tokens_seen": 114435015, + "step": 5316 + }, + { + "epoch": 0.6393314495280467, + "flos": 26756660015520.0, + "grad_norm": 2.0578751006340426, + "learning_rate": 1.2161402822286797e-06, + "loss": 0.6391, + "num_input_tokens_seen": 114455700, + "step": 5317 + }, + { + "epoch": 0.6394516924186857, + "flos": 20231597425920.0, + "grad_norm": 3.2900806581761315, + "learning_rate": 1.2154236916029703e-06, + "loss": 0.7868, + "num_input_tokens_seen": 114473675, + "step": 5318 + }, + { + "epoch": 0.6395719353093249, + "flos": 18342703773120.0, + "grad_norm": 3.854459528063436, + "learning_rate": 1.2147072199968627e-06, + "loss": 0.7337, + "num_input_tokens_seen": 114492310, + "step": 5319 + }, + { + "epoch": 0.6396921781999639, + "flos": 17541346287360.0, + "grad_norm": 3.08496912960886, + "learning_rate": 1.2139908675190454e-06, + "loss": 0.7175, + "num_input_tokens_seen": 114511520, + "step": 5320 + }, + { + "epoch": 0.639812421090603, + "flos": 21220913886240.0, + "grad_norm": 5.171230126135952, + "learning_rate": 1.2132746342781883e-06, + "loss": 0.7538, + "num_input_tokens_seen": 114532680, + "step": 5321 + }, + { + "epoch": 0.6399326639812422, + "flos": 11165283030240.0, + "grad_norm": 3.5268878832292274, + "learning_rate": 1.2125585203829442e-06, + "loss": 0.794, + "num_input_tokens_seen": 114548320, + "step": 5322 + }, + { + "epoch": 0.6400529068718812, + "flos": 23876702923680.0, + "grad_norm": 2.5265997096561366, + "learning_rate": 1.211842525941946e-06, + "loss": 0.7436, + "num_input_tokens_seen": 114568115, + "step": 5323 + }, + { + "epoch": 0.6401731497625203, + "flos": 43955249513280.0, + "grad_norm": 2.4766236306717477, + "learning_rate": 1.2111266510638105e-06, + "loss": 0.7934, + "num_input_tokens_seen": 114591040, + "step": 5324 + }, + { + "epoch": 0.6402933926531594, + "flos": 20632536357120.0, + "grad_norm": 2.134609597132841, + "learning_rate": 1.2104108958571346e-06, + "loss": 0.8003, + "num_input_tokens_seen": 114609310, + "step": 5325 + }, + { + "epoch": 0.6404136355437985, + "flos": 24827728797600.0, + "grad_norm": 1.8949135306164193, + "learning_rate": 1.2096952604304975e-06, + "loss": 0.7563, + "num_input_tokens_seen": 114629740, + "step": 5326 + }, + { + "epoch": 0.6405338784344375, + "flos": 40421224459200.0, + "grad_norm": 3.066221817204327, + "learning_rate": 1.2089797448924616e-06, + "loss": 0.7023, + "num_input_tokens_seen": 114653615, + "step": 5327 + }, + { + "epoch": 0.6406541213250767, + "flos": 20856221204640.0, + "grad_norm": 6.167483851269013, + "learning_rate": 1.2082643493515692e-06, + "loss": 0.6572, + "num_input_tokens_seen": 114671935, + "step": 5328 + }, + { + "epoch": 0.6407743642157158, + "flos": 23262709696320.0, + "grad_norm": 2.944823530573726, + "learning_rate": 1.207549073916346e-06, + "loss": 0.8162, + "num_input_tokens_seen": 114692870, + "step": 5329 + }, + { + "epoch": 0.6408946071063548, + "flos": 14991396757440.0, + "grad_norm": 3.9604428074572846, + "learning_rate": 1.2068339186952976e-06, + "loss": 0.7776, + "num_input_tokens_seen": 114710410, + "step": 5330 + }, + { + "epoch": 0.6410148499969939, + "flos": 22496409027840.0, + "grad_norm": 3.8179703536243945, + "learning_rate": 1.2061188837969136e-06, + "loss": 0.7312, + "num_input_tokens_seen": 114730020, + "step": 5331 + }, + { + "epoch": 0.641135092887633, + "flos": 12404531922240.0, + "grad_norm": 4.579031271953312, + "learning_rate": 1.2054039693296631e-06, + "loss": 0.8412, + "num_input_tokens_seen": 114748015, + "step": 5332 + }, + { + "epoch": 0.6412553357782721, + "flos": 22095730284960.0, + "grad_norm": 3.1061903397109574, + "learning_rate": 1.2046891754019992e-06, + "loss": 0.8183, + "num_input_tokens_seen": 114768625, + "step": 5333 + }, + { + "epoch": 0.6413755786689112, + "flos": 15866101646880.0, + "grad_norm": 2.5628078554890714, + "learning_rate": 1.2039745021223548e-06, + "loss": 0.8314, + "num_input_tokens_seen": 114786045, + "step": 5334 + }, + { + "epoch": 0.6414958215595503, + "flos": 67942506770400.0, + "grad_norm": 0.8038161291205574, + "learning_rate": 1.2032599495991456e-06, + "loss": 0.5698, + "num_input_tokens_seen": 114850785, + "step": 5335 + }, + { + "epoch": 0.6416160644501894, + "flos": 44027704842720.0, + "grad_norm": 3.018950373157999, + "learning_rate": 1.2025455179407685e-06, + "loss": 0.6946, + "num_input_tokens_seen": 114871945, + "step": 5336 + }, + { + "epoch": 0.6417363073408284, + "flos": 20929940305920.0, + "grad_norm": 6.6522095644863555, + "learning_rate": 1.2018312072556022e-06, + "loss": 0.7376, + "num_input_tokens_seen": 114890445, + "step": 5337 + }, + { + "epoch": 0.6418565502314676, + "flos": 22423024454400.0, + "grad_norm": 3.0566006886368706, + "learning_rate": 1.2011170176520077e-06, + "loss": 0.7469, + "num_input_tokens_seen": 114911360, + "step": 5338 + }, + { + "epoch": 0.6419767931221066, + "flos": 25009629101280.0, + "grad_norm": 3.3572068425721815, + "learning_rate": 1.2004029492383256e-06, + "loss": 0.8127, + "num_input_tokens_seen": 114932700, + "step": 5339 + }, + { + "epoch": 0.6420970360127457, + "flos": 19435964083200.0, + "grad_norm": 2.5043498344099473, + "learning_rate": 1.1996890021228814e-06, + "loss": 0.7344, + "num_input_tokens_seen": 114949475, + "step": 5340 + }, + { + "epoch": 0.6422172789033849, + "flos": 40347914225280.0, + "grad_norm": 2.1014889842508655, + "learning_rate": 1.1989751764139785e-06, + "loss": 0.7003, + "num_input_tokens_seen": 114973125, + "step": 5341 + }, + { + "epoch": 0.6423375217940239, + "flos": 27633186223200.0, + "grad_norm": 2.2626441024497392, + "learning_rate": 1.1982614722199044e-06, + "loss": 0.8317, + "num_input_tokens_seen": 114994300, + "step": 5342 + }, + { + "epoch": 0.642457764684663, + "flos": 18343112640480.0, + "grad_norm": 5.448974603905655, + "learning_rate": 1.1975478896489276e-06, + "loss": 0.781, + "num_input_tokens_seen": 115012135, + "step": 5343 + }, + { + "epoch": 0.6425780075753021, + "flos": 19722179934240.0, + "grad_norm": 2.8364489385809075, + "learning_rate": 1.1968344288092981e-06, + "loss": 0.7686, + "num_input_tokens_seen": 115028715, + "step": 5344 + }, + { + "epoch": 0.6426982504659412, + "flos": 20529224393280.0, + "grad_norm": 2.215602307546433, + "learning_rate": 1.1961210898092468e-06, + "loss": 0.6457, + "num_input_tokens_seen": 115047665, + "step": 5345 + }, + { + "epoch": 0.6428184933565803, + "flos": 17826372706080.0, + "grad_norm": 2.643710168128111, + "learning_rate": 1.1954078727569874e-06, + "loss": 0.7924, + "num_input_tokens_seen": 115064965, + "step": 5346 + }, + { + "epoch": 0.6429387362472194, + "flos": 22424920112160.0, + "grad_norm": 2.0444631563012194, + "learning_rate": 1.1946947777607141e-06, + "loss": 0.7809, + "num_input_tokens_seen": 115086100, + "step": 5347 + }, + { + "epoch": 0.6430589791378585, + "flos": 24717094391040.0, + "grad_norm": 3.284211339239388, + "learning_rate": 1.1939818049286024e-06, + "loss": 0.8057, + "num_input_tokens_seen": 115104260, + "step": 5348 + }, + { + "epoch": 0.6431792220284975, + "flos": 24865647686400.0, + "grad_norm": 2.3136626748825186, + "learning_rate": 1.1932689543688101e-06, + "loss": 0.7561, + "num_input_tokens_seen": 115125365, + "step": 5349 + }, + { + "epoch": 0.6432994649191367, + "flos": 21002284126080.0, + "grad_norm": 2.3894939863756286, + "learning_rate": 1.1925562261894756e-06, + "loss": 0.7261, + "num_input_tokens_seen": 115144445, + "step": 5350 + }, + { + "epoch": 0.6434197078097758, + "flos": 30840102956640.0, + "grad_norm": 2.4296127258766362, + "learning_rate": 1.1918436204987207e-06, + "loss": 0.7763, + "num_input_tokens_seen": 115166305, + "step": 5351 + }, + { + "epoch": 0.6435399507004148, + "flos": 14991248078400.0, + "grad_norm": 3.242379571212087, + "learning_rate": 1.191131137404645e-06, + "loss": 0.8144, + "num_input_tokens_seen": 115183045, + "step": 5352 + }, + { + "epoch": 0.643660193591054, + "flos": 19873632470880.0, + "grad_norm": 3.347559247564498, + "learning_rate": 1.190418777015333e-06, + "loss": 0.7722, + "num_input_tokens_seen": 115201955, + "step": 5353 + }, + { + "epoch": 0.643780436481693, + "flos": 24099086829600.0, + "grad_norm": 1.6880372216608988, + "learning_rate": 1.1897065394388487e-06, + "loss": 0.7394, + "num_input_tokens_seen": 115222395, + "step": 5354 + }, + { + "epoch": 0.6439006793723321, + "flos": 23115717530880.0, + "grad_norm": 2.1659884369419116, + "learning_rate": 1.1889944247832385e-06, + "loss": 0.7652, + "num_input_tokens_seen": 115242635, + "step": 5355 + }, + { + "epoch": 0.6440209222629713, + "flos": 23583945194880.0, + "grad_norm": 2.382156131582394, + "learning_rate": 1.1882824331565283e-06, + "loss": 0.7073, + "num_input_tokens_seen": 115262095, + "step": 5356 + }, + { + "epoch": 0.6441411651536103, + "flos": 16521396211200.0, + "grad_norm": 3.256811193918375, + "learning_rate": 1.1875705646667287e-06, + "loss": 0.8889, + "num_input_tokens_seen": 115279985, + "step": 5357 + }, + { + "epoch": 0.6442614080442494, + "flos": 25374581971200.0, + "grad_norm": 3.221505180320382, + "learning_rate": 1.1868588194218282e-06, + "loss": 0.7542, + "num_input_tokens_seen": 115300365, + "step": 5358 + }, + { + "epoch": 0.6443816509348885, + "flos": 28253461140000.0, + "grad_norm": 2.0539050334754148, + "learning_rate": 1.1861471975297979e-06, + "loss": 0.7426, + "num_input_tokens_seen": 115322060, + "step": 5359 + }, + { + "epoch": 0.6445018938255276, + "flos": 36637452822240.0, + "grad_norm": 2.1175260941700262, + "learning_rate": 1.185435699098591e-06, + "loss": 0.7101, + "num_input_tokens_seen": 115348255, + "step": 5360 + }, + { + "epoch": 0.6446221367161666, + "flos": 14480901342720.0, + "grad_norm": 3.576048402514153, + "learning_rate": 1.1847243242361403e-06, + "loss": 0.781, + "num_input_tokens_seen": 115366845, + "step": 5361 + }, + { + "epoch": 0.6447423796068057, + "flos": 24574190899200.0, + "grad_norm": 2.035208381087662, + "learning_rate": 1.1840130730503624e-06, + "loss": 0.7763, + "num_input_tokens_seen": 115388125, + "step": 5362 + }, + { + "epoch": 0.6448626224974449, + "flos": 25011487589280.0, + "grad_norm": 2.911274494151634, + "learning_rate": 1.1833019456491518e-06, + "loss": 0.7492, + "num_input_tokens_seen": 115409655, + "step": 5363 + }, + { + "epoch": 0.6449828653880839, + "flos": 22498230346080.0, + "grad_norm": 9.97273887304199, + "learning_rate": 1.1825909421403871e-06, + "loss": 0.7884, + "num_input_tokens_seen": 115428750, + "step": 5364 + }, + { + "epoch": 0.645103108278723, + "flos": 25658679145920.0, + "grad_norm": 2.435141180124609, + "learning_rate": 1.181880062631926e-06, + "loss": 0.7681, + "num_input_tokens_seen": 115448085, + "step": 5365 + }, + { + "epoch": 0.6452233511693621, + "flos": 27410839487040.0, + "grad_norm": 4.109238306825278, + "learning_rate": 1.1811693072316093e-06, + "loss": 0.8456, + "num_input_tokens_seen": 115465765, + "step": 5366 + }, + { + "epoch": 0.6453435940600012, + "flos": 19181273922240.0, + "grad_norm": 4.040055468339, + "learning_rate": 1.1804586760472574e-06, + "loss": 0.8429, + "num_input_tokens_seen": 115482230, + "step": 5367 + }, + { + "epoch": 0.6454638369506402, + "flos": 25700761047840.0, + "grad_norm": 3.2975688764596134, + "learning_rate": 1.1797481691866736e-06, + "loss": 0.8069, + "num_input_tokens_seen": 115499455, + "step": 5368 + }, + { + "epoch": 0.6455840798412794, + "flos": 20958975622080.0, + "grad_norm": 2.7143211685102324, + "learning_rate": 1.1790377867576393e-06, + "loss": 0.8274, + "num_input_tokens_seen": 115517205, + "step": 5369 + }, + { + "epoch": 0.6457043227319185, + "flos": 26030211063360.0, + "grad_norm": 2.671963963505445, + "learning_rate": 1.1783275288679203e-06, + "loss": 0.7706, + "num_input_tokens_seen": 115534370, + "step": 5370 + }, + { + "epoch": 0.6458245656225575, + "flos": 60285112719360.0, + "grad_norm": 1.017483053324022, + "learning_rate": 1.177617395625262e-06, + "loss": 0.571, + "num_input_tokens_seen": 115592345, + "step": 5371 + }, + { + "epoch": 0.6459448085131967, + "flos": 23043819747840.0, + "grad_norm": 2.6537911197820527, + "learning_rate": 1.1769073871373908e-06, + "loss": 0.7529, + "num_input_tokens_seen": 115612550, + "step": 5372 + }, + { + "epoch": 0.6460650514038357, + "flos": 22565779267200.0, + "grad_norm": 2.4770839394926116, + "learning_rate": 1.176197503512015e-06, + "loss": 0.8396, + "num_input_tokens_seen": 115632265, + "step": 5373 + }, + { + "epoch": 0.6461852942944748, + "flos": 20237841945600.0, + "grad_norm": 5.756813987580079, + "learning_rate": 1.1754877448568223e-06, + "loss": 0.8233, + "num_input_tokens_seen": 115651720, + "step": 5374 + }, + { + "epoch": 0.646305537185114, + "flos": 23333678235360.0, + "grad_norm": 4.2986783768925845, + "learning_rate": 1.1747781112794837e-06, + "loss": 0.9026, + "num_input_tokens_seen": 115668215, + "step": 5375 + }, + { + "epoch": 0.646425780075753, + "flos": 24242659377120.0, + "grad_norm": 1.9395456282422987, + "learning_rate": 1.1740686028876487e-06, + "loss": 0.8329, + "num_input_tokens_seen": 115687080, + "step": 5376 + }, + { + "epoch": 0.6465460229663921, + "flos": 20784286251840.0, + "grad_norm": 3.09100111783568, + "learning_rate": 1.1733592197889507e-06, + "loss": 0.7469, + "num_input_tokens_seen": 115703990, + "step": 5377 + }, + { + "epoch": 0.6466662658570312, + "flos": 22820506597920.0, + "grad_norm": 2.364700874622503, + "learning_rate": 1.1726499620910014e-06, + "loss": 0.7272, + "num_input_tokens_seen": 115724270, + "step": 5378 + }, + { + "epoch": 0.6467865087476703, + "flos": 15282890714400.0, + "grad_norm": 2.6689440618708193, + "learning_rate": 1.1719408299013955e-06, + "loss": 0.7821, + "num_input_tokens_seen": 115742995, + "step": 5379 + }, + { + "epoch": 0.6469067516383094, + "flos": 19545780755040.0, + "grad_norm": 3.163500926727116, + "learning_rate": 1.1712318233277067e-06, + "loss": 0.7595, + "num_input_tokens_seen": 115762015, + "step": 5380 + }, + { + "epoch": 0.6470269945289485, + "flos": 65004819307680.0, + "grad_norm": 0.7716769979937431, + "learning_rate": 1.1705229424774916e-06, + "loss": 0.578, + "num_input_tokens_seen": 115816285, + "step": 5381 + }, + { + "epoch": 0.6471472374195876, + "flos": 30652701528480.0, + "grad_norm": 3.1883273079538186, + "learning_rate": 1.1698141874582867e-06, + "loss": 0.6416, + "num_input_tokens_seen": 115837330, + "step": 5382 + }, + { + "epoch": 0.6472674803102266, + "flos": 20492569276320.0, + "grad_norm": 2.996193855533063, + "learning_rate": 1.169105558377609e-06, + "loss": 0.7242, + "num_input_tokens_seen": 115857420, + "step": 5383 + }, + { + "epoch": 0.6473877232008658, + "flos": 24679993236960.0, + "grad_norm": 2.050065560571862, + "learning_rate": 1.1683970553429587e-06, + "loss": 0.7855, + "num_input_tokens_seen": 115878875, + "step": 5384 + }, + { + "epoch": 0.6475079660915048, + "flos": 15862942217280.0, + "grad_norm": 2.2206496582036483, + "learning_rate": 1.1676886784618128e-06, + "loss": 0.8228, + "num_input_tokens_seen": 115895540, + "step": 5385 + }, + { + "epoch": 0.6476282089821439, + "flos": 17356806930720.0, + "grad_norm": 3.6572729782601963, + "learning_rate": 1.1669804278416332e-06, + "loss": 0.8379, + "num_input_tokens_seen": 115910265, + "step": 5386 + }, + { + "epoch": 0.6477484518727831, + "flos": 20165312276640.0, + "grad_norm": 2.7268625875781134, + "learning_rate": 1.1662723035898602e-06, + "loss": 0.7181, + "num_input_tokens_seen": 115930025, + "step": 5387 + }, + { + "epoch": 0.6478686947634221, + "flos": 25374024424800.0, + "grad_norm": 2.02934751181259, + "learning_rate": 1.165564305813915e-06, + "loss": 0.8187, + "num_input_tokens_seen": 115949025, + "step": 5388 + }, + { + "epoch": 0.6479889376540612, + "flos": 20054938058400.0, + "grad_norm": 2.2383682740022506, + "learning_rate": 1.1648564346212019e-06, + "loss": 0.8155, + "num_input_tokens_seen": 115968145, + "step": 5389 + }, + { + "epoch": 0.6481091805447003, + "flos": 26719633200960.0, + "grad_norm": 2.852619633619051, + "learning_rate": 1.164148690119104e-06, + "loss": 0.7591, + "num_input_tokens_seen": 115989425, + "step": 5390 + }, + { + "epoch": 0.6482294234353394, + "flos": 23917855581600.0, + "grad_norm": 9.414427994720308, + "learning_rate": 1.163441072414985e-06, + "loss": 0.739, + "num_input_tokens_seen": 116009630, + "step": 5391 + }, + { + "epoch": 0.6483496663259785, + "flos": 26172148141440.0, + "grad_norm": 3.273974681833254, + "learning_rate": 1.16273358161619e-06, + "loss": 0.6999, + "num_input_tokens_seen": 116029965, + "step": 5392 + }, + { + "epoch": 0.6484699092166175, + "flos": 20892355944960.0, + "grad_norm": 3.17103016693962, + "learning_rate": 1.1620262178300446e-06, + "loss": 0.8341, + "num_input_tokens_seen": 116048580, + "step": 5393 + }, + { + "epoch": 0.6485901521072567, + "flos": 33025322635200.0, + "grad_norm": 2.2548468625328244, + "learning_rate": 1.1613189811638563e-06, + "loss": 0.7585, + "num_input_tokens_seen": 116070305, + "step": 5394 + }, + { + "epoch": 0.6487103949978957, + "flos": 22242276413280.0, + "grad_norm": 2.2620838773170147, + "learning_rate": 1.1606118717249117e-06, + "loss": 0.7792, + "num_input_tokens_seen": 116090405, + "step": 5395 + }, + { + "epoch": 0.6488306378885348, + "flos": 22897645317120.0, + "grad_norm": 2.2474630676561773, + "learning_rate": 1.1599048896204787e-06, + "loss": 0.6807, + "num_input_tokens_seen": 116110285, + "step": 5396 + }, + { + "epoch": 0.648950880779174, + "flos": 20347063901280.0, + "grad_norm": 2.5654650955457616, + "learning_rate": 1.1591980349578061e-06, + "loss": 0.8095, + "num_input_tokens_seen": 116128955, + "step": 5397 + }, + { + "epoch": 0.649071123669813, + "flos": 59845734522720.0, + "grad_norm": 0.744815230274694, + "learning_rate": 1.158491307844123e-06, + "loss": 0.5431, + "num_input_tokens_seen": 116188875, + "step": 5398 + }, + { + "epoch": 0.6491913665604521, + "flos": 20417326214880.0, + "grad_norm": 2.241144768562983, + "learning_rate": 1.1577847083866387e-06, + "loss": 0.8407, + "num_input_tokens_seen": 116207910, + "step": 5399 + }, + { + "epoch": 0.6493116094510912, + "flos": 16922781179520.0, + "grad_norm": 2.6112221216272165, + "learning_rate": 1.1570782366925453e-06, + "loss": 0.7224, + "num_input_tokens_seen": 116226425, + "step": 5400 + }, + { + "epoch": 0.6494318523417303, + "flos": 18776023298880.0, + "grad_norm": 18.07254836470306, + "learning_rate": 1.1563718928690132e-06, + "loss": 0.7569, + "num_input_tokens_seen": 116243860, + "step": 5401 + }, + { + "epoch": 0.6495520952323693, + "flos": 18955581907680.0, + "grad_norm": 4.124277224941824, + "learning_rate": 1.1556656770231942e-06, + "loss": 0.7166, + "num_input_tokens_seen": 116260055, + "step": 5402 + }, + { + "epoch": 0.6496723381230085, + "flos": 22712808602400.0, + "grad_norm": 3.962753571781433, + "learning_rate": 1.1549595892622207e-06, + "loss": 0.7611, + "num_input_tokens_seen": 116278825, + "step": 5403 + }, + { + "epoch": 0.6497925810136476, + "flos": 62056495560000.0, + "grad_norm": 0.8482642739846331, + "learning_rate": 1.1542536296932047e-06, + "loss": 0.5896, + "num_input_tokens_seen": 116342360, + "step": 5404 + }, + { + "epoch": 0.6499128239042866, + "flos": 20127988104000.0, + "grad_norm": 4.000918044597385, + "learning_rate": 1.1535477984232414e-06, + "loss": 0.7036, + "num_input_tokens_seen": 116362235, + "step": 5405 + }, + { + "epoch": 0.6500330667949258, + "flos": 24427421752320.0, + "grad_norm": 3.9589054831881114, + "learning_rate": 1.152842095559404e-06, + "loss": 0.7736, + "num_input_tokens_seen": 116382895, + "step": 5406 + }, + { + "epoch": 0.6501533096855648, + "flos": 25440681271680.0, + "grad_norm": 2.3604497419589268, + "learning_rate": 1.1521365212087474e-06, + "loss": 0.7688, + "num_input_tokens_seen": 116402880, + "step": 5407 + }, + { + "epoch": 0.6502735525762039, + "flos": 44754525492480.0, + "grad_norm": 2.760988303441933, + "learning_rate": 1.1514310754783062e-06, + "loss": 0.706, + "num_input_tokens_seen": 116425830, + "step": 5408 + }, + { + "epoch": 0.6503937954668431, + "flos": 28617447596160.0, + "grad_norm": 2.9414039804315615, + "learning_rate": 1.1507257584750964e-06, + "loss": 0.735, + "num_input_tokens_seen": 116446010, + "step": 5409 + }, + { + "epoch": 0.6505140383574821, + "flos": 20892690472800.0, + "grad_norm": 3.2332757381015607, + "learning_rate": 1.150020570306113e-06, + "loss": 0.777, + "num_input_tokens_seen": 116465150, + "step": 5410 + }, + { + "epoch": 0.6506342812481212, + "flos": 20565879510240.0, + "grad_norm": 2.7513487716061733, + "learning_rate": 1.1493155110783338e-06, + "loss": 0.7504, + "num_input_tokens_seen": 116483675, + "step": 5411 + }, + { + "epoch": 0.6507545241387603, + "flos": 30183061413600.0, + "grad_norm": 3.701587599302502, + "learning_rate": 1.1486105808987155e-06, + "loss": 0.7072, + "num_input_tokens_seen": 116505165, + "step": 5412 + }, + { + "epoch": 0.6508747670293994, + "flos": 17103900918240.0, + "grad_norm": 2.7540928987058786, + "learning_rate": 1.1479057798741947e-06, + "loss": 0.8131, + "num_input_tokens_seen": 116523220, + "step": 5413 + }, + { + "epoch": 0.6509950099200384, + "flos": 68461445454240.0, + "grad_norm": 0.7918575374661965, + "learning_rate": 1.14720110811169e-06, + "loss": 0.5324, + "num_input_tokens_seen": 116580450, + "step": 5414 + }, + { + "epoch": 0.6511152528106776, + "flos": 22315586647200.0, + "grad_norm": 6.75801373661719, + "learning_rate": 1.146496565718098e-06, + "loss": 0.7691, + "num_input_tokens_seen": 116601020, + "step": 5415 + }, + { + "epoch": 0.6512354957013167, + "flos": 20492903804160.0, + "grad_norm": 2.6715723055879237, + "learning_rate": 1.1457921528002996e-06, + "loss": 0.7562, + "num_input_tokens_seen": 116619455, + "step": 5416 + }, + { + "epoch": 0.6513557385919557, + "flos": 32290659166080.0, + "grad_norm": 5.247270510354682, + "learning_rate": 1.1450878694651522e-06, + "loss": 0.7222, + "num_input_tokens_seen": 116640020, + "step": 5417 + }, + { + "epoch": 0.6514759814825949, + "flos": 12076196999520.0, + "grad_norm": 3.2760909440175987, + "learning_rate": 1.1443837158194954e-06, + "loss": 0.6326, + "num_input_tokens_seen": 116655165, + "step": 5418 + }, + { + "epoch": 0.651596224373234, + "flos": 22494662049120.0, + "grad_norm": 2.879134977822335, + "learning_rate": 1.1436796919701484e-06, + "loss": 0.7468, + "num_input_tokens_seen": 116673880, + "step": 5419 + }, + { + "epoch": 0.651716467263873, + "flos": 27779063295840.0, + "grad_norm": 4.986607736113518, + "learning_rate": 1.1429757980239115e-06, + "loss": 0.6207, + "num_input_tokens_seen": 116694305, + "step": 5420 + }, + { + "epoch": 0.6518367101545122, + "flos": 24281507509920.0, + "grad_norm": 3.4399611938860435, + "learning_rate": 1.1422720340875636e-06, + "loss": 0.8193, + "num_input_tokens_seen": 116713055, + "step": 5421 + }, + { + "epoch": 0.6519569530451512, + "flos": 19983449142720.0, + "grad_norm": 2.5213297730791737, + "learning_rate": 1.1415684002678671e-06, + "loss": 0.7922, + "num_input_tokens_seen": 116731815, + "step": 5422 + }, + { + "epoch": 0.6520771959357903, + "flos": 21545122965600.0, + "grad_norm": 3.971450727779297, + "learning_rate": 1.1408648966715617e-06, + "loss": 0.7776, + "num_input_tokens_seen": 116749335, + "step": 5423 + }, + { + "epoch": 0.6521974388264293, + "flos": 22678569519840.0, + "grad_norm": 3.095477614025911, + "learning_rate": 1.1401615234053683e-06, + "loss": 0.7252, + "num_input_tokens_seen": 116768155, + "step": 5424 + }, + { + "epoch": 0.6523176817170685, + "flos": 22970063476800.0, + "grad_norm": 4.73897215986315, + "learning_rate": 1.1394582805759885e-06, + "loss": 0.76, + "num_input_tokens_seen": 116787435, + "step": 5425 + }, + { + "epoch": 0.6524379246077076, + "flos": 21657244162560.0, + "grad_norm": 2.268010849780415, + "learning_rate": 1.1387551682901022e-06, + "loss": 0.762, + "num_input_tokens_seen": 116808040, + "step": 5426 + }, + { + "epoch": 0.6525581674983466, + "flos": 19363322904960.0, + "grad_norm": 2.498907422692954, + "learning_rate": 1.138052186654373e-06, + "loss": 0.7099, + "num_input_tokens_seen": 116825510, + "step": 5427 + }, + { + "epoch": 0.6526784103889858, + "flos": 17141076411840.0, + "grad_norm": 2.791840865740148, + "learning_rate": 1.1373493357754417e-06, + "loss": 0.8849, + "num_input_tokens_seen": 116844415, + "step": 5428 + }, + { + "epoch": 0.6527986532796248, + "flos": 18962123785440.0, + "grad_norm": 2.2764383024270787, + "learning_rate": 1.1366466157599303e-06, + "loss": 0.7698, + "num_input_tokens_seen": 116863690, + "step": 5429 + }, + { + "epoch": 0.6529188961702639, + "flos": 14218182513600.0, + "grad_norm": 2.925126984277009, + "learning_rate": 1.1359440267144412e-06, + "loss": 0.7632, + "num_input_tokens_seen": 116881145, + "step": 5430 + }, + { + "epoch": 0.653039139060903, + "flos": 36630056040000.0, + "grad_norm": 4.453478561553628, + "learning_rate": 1.1352415687455556e-06, + "loss": 0.7487, + "num_input_tokens_seen": 116902405, + "step": 5431 + }, + { + "epoch": 0.6531593819515421, + "flos": 25340008360800.0, + "grad_norm": 3.1243456346139507, + "learning_rate": 1.1345392419598362e-06, + "loss": 0.6421, + "num_input_tokens_seen": 116921285, + "step": 5432 + }, + { + "epoch": 0.6532796248421812, + "flos": 21148309877760.0, + "grad_norm": 2.6856825249831746, + "learning_rate": 1.1338370464638263e-06, + "loss": 0.7192, + "num_input_tokens_seen": 116940685, + "step": 5433 + }, + { + "epoch": 0.6533998677328203, + "flos": 17650828431360.0, + "grad_norm": 6.544408073964689, + "learning_rate": 1.1331349823640474e-06, + "loss": 0.6409, + "num_input_tokens_seen": 116958115, + "step": 5434 + }, + { + "epoch": 0.6535201106234594, + "flos": 28359560835840.0, + "grad_norm": 3.7856385796681096, + "learning_rate": 1.132433049767003e-06, + "loss": 0.7806, + "num_input_tokens_seen": 116976030, + "step": 5435 + }, + { + "epoch": 0.6536403535140984, + "flos": 23552939881440.0, + "grad_norm": 2.1481701493811283, + "learning_rate": 1.1317312487791748e-06, + "loss": 0.8119, + "num_input_tokens_seen": 116997680, + "step": 5436 + }, + { + "epoch": 0.6537605964047376, + "flos": 21548654092800.0, + "grad_norm": 2.7218174784874813, + "learning_rate": 1.1310295795070253e-06, + "loss": 0.7339, + "num_input_tokens_seen": 117017620, + "step": 5437 + }, + { + "epoch": 0.6538808392953767, + "flos": 26795545318080.0, + "grad_norm": 3.2189270267137315, + "learning_rate": 1.1303280420569982e-06, + "loss": 0.8088, + "num_input_tokens_seen": 117039505, + "step": 5438 + }, + { + "epoch": 0.6540010821860157, + "flos": 30694114374720.0, + "grad_norm": 14.960141950127637, + "learning_rate": 1.1296266365355158e-06, + "loss": 0.772, + "num_input_tokens_seen": 117062890, + "step": 5439 + }, + { + "epoch": 0.6541213250766549, + "flos": 26868595363680.0, + "grad_norm": 5.9483264396686675, + "learning_rate": 1.1289253630489806e-06, + "loss": 0.7346, + "num_input_tokens_seen": 117083940, + "step": 5440 + }, + { + "epoch": 0.6542415679672939, + "flos": 19145102012160.0, + "grad_norm": 2.761774691132511, + "learning_rate": 1.1282242217037753e-06, + "loss": 0.723, + "num_input_tokens_seen": 117101440, + "step": 5441 + }, + { + "epoch": 0.654361810857933, + "flos": 48103825341120.0, + "grad_norm": 2.892843518906256, + "learning_rate": 1.127523212606262e-06, + "loss": 0.6194, + "num_input_tokens_seen": 117124265, + "step": 5442 + }, + { + "epoch": 0.6544820537485722, + "flos": 26904544255200.0, + "grad_norm": 4.619308268628024, + "learning_rate": 1.1268223358627835e-06, + "loss": 0.7323, + "num_input_tokens_seen": 117146755, + "step": 5443 + }, + { + "epoch": 0.6546022966392112, + "flos": 20856518562720.0, + "grad_norm": 2.564169828131752, + "learning_rate": 1.126121591579663e-06, + "loss": 0.7195, + "num_input_tokens_seen": 117165960, + "step": 5444 + }, + { + "epoch": 0.6547225395298503, + "flos": 24901001861760.0, + "grad_norm": 2.28053188778915, + "learning_rate": 1.1254209798632018e-06, + "loss": 0.6918, + "num_input_tokens_seen": 117186980, + "step": 5445 + }, + { + "epoch": 0.6548427824204894, + "flos": 22533733200480.0, + "grad_norm": 2.992865503054315, + "learning_rate": 1.124720500819683e-06, + "loss": 0.8497, + "num_input_tokens_seen": 117205135, + "step": 5446 + }, + { + "epoch": 0.6549630253111285, + "flos": 18416274195360.0, + "grad_norm": 2.29043735708776, + "learning_rate": 1.1240201545553682e-06, + "loss": 0.8202, + "num_input_tokens_seen": 117222935, + "step": 5447 + }, + { + "epoch": 0.6550832682017675, + "flos": 25151082972480.0, + "grad_norm": 2.441579486041303, + "learning_rate": 1.1233199411764987e-06, + "loss": 0.7302, + "num_input_tokens_seen": 117242370, + "step": 5448 + }, + { + "epoch": 0.6552035110924067, + "flos": 22715670673920.0, + "grad_norm": 2.2674881213296563, + "learning_rate": 1.1226198607892978e-06, + "loss": 0.6856, + "num_input_tokens_seen": 117262930, + "step": 5449 + }, + { + "epoch": 0.6553237539830458, + "flos": 21768064417920.0, + "grad_norm": 2.1414092917352874, + "learning_rate": 1.1219199134999664e-06, + "loss": 0.8012, + "num_input_tokens_seen": 117281430, + "step": 5450 + }, + { + "epoch": 0.6554439968736848, + "flos": 20857447806720.0, + "grad_norm": 3.717940939718891, + "learning_rate": 1.1212200994146863e-06, + "loss": 0.7873, + "num_input_tokens_seen": 117299185, + "step": 5451 + }, + { + "epoch": 0.655564239764324, + "flos": 16115996908800.0, + "grad_norm": 2.4087488649738766, + "learning_rate": 1.120520418639618e-06, + "loss": 0.7584, + "num_input_tokens_seen": 117317720, + "step": 5452 + }, + { + "epoch": 0.655684482654963, + "flos": 29527915528320.0, + "grad_norm": 3.6989147863006355, + "learning_rate": 1.119820871280903e-06, + "loss": 0.8326, + "num_input_tokens_seen": 117338795, + "step": 5453 + }, + { + "epoch": 0.6558047255456021, + "flos": 29746656797760.0, + "grad_norm": 2.782213451372012, + "learning_rate": 1.1191214574446614e-06, + "loss": 0.7328, + "num_input_tokens_seen": 117359040, + "step": 5454 + }, + { + "epoch": 0.6559249684362413, + "flos": 29017457283360.0, + "grad_norm": 2.009330205581212, + "learning_rate": 1.118422177236995e-06, + "loss": 0.8029, + "num_input_tokens_seen": 117380865, + "step": 5455 + }, + { + "epoch": 0.6560452113268803, + "flos": 20196466269120.0, + "grad_norm": 2.8254928564043116, + "learning_rate": 1.1177230307639835e-06, + "loss": 0.8606, + "num_input_tokens_seen": 117398405, + "step": 5456 + }, + { + "epoch": 0.6561654542175194, + "flos": 25009666271040.0, + "grad_norm": 2.794241832477189, + "learning_rate": 1.1170240181316865e-06, + "loss": 0.7878, + "num_input_tokens_seen": 117419850, + "step": 5457 + }, + { + "epoch": 0.6562856971081584, + "flos": 22824037725120.0, + "grad_norm": 2.788587000728563, + "learning_rate": 1.1163251394461442e-06, + "loss": 0.7967, + "num_input_tokens_seen": 117438330, + "step": 5458 + }, + { + "epoch": 0.6564059399987976, + "flos": 18845430708000.0, + "grad_norm": 2.3900251446283085, + "learning_rate": 1.1156263948133746e-06, + "loss": 0.8247, + "num_input_tokens_seen": 117454985, + "step": 5459 + }, + { + "epoch": 0.6565261828894366, + "flos": 25447706356320.0, + "grad_norm": 3.5120118966129206, + "learning_rate": 1.1149277843393787e-06, + "loss": 0.777, + "num_input_tokens_seen": 117476145, + "step": 5460 + }, + { + "epoch": 0.6566464257800757, + "flos": 19655485917600.0, + "grad_norm": 5.682017943598643, + "learning_rate": 1.1142293081301342e-06, + "loss": 0.6351, + "num_input_tokens_seen": 117494025, + "step": 5461 + }, + { + "epoch": 0.6567666686707149, + "flos": 23480521721760.0, + "grad_norm": 2.9057426419959618, + "learning_rate": 1.1135309662915995e-06, + "loss": 0.6823, + "num_input_tokens_seen": 117514190, + "step": 5462 + }, + { + "epoch": 0.6568869115613539, + "flos": 32734237545600.0, + "grad_norm": 2.8561765365001177, + "learning_rate": 1.112832758929712e-06, + "loss": 0.6037, + "num_input_tokens_seen": 117536800, + "step": 5463 + }, + { + "epoch": 0.657007154451993, + "flos": 18416199855840.0, + "grad_norm": 2.356927979571079, + "learning_rate": 1.11213468615039e-06, + "loss": 0.7477, + "num_input_tokens_seen": 117556230, + "step": 5464 + }, + { + "epoch": 0.6571273973426321, + "flos": 25120746714720.0, + "grad_norm": 2.226938958717659, + "learning_rate": 1.1114367480595292e-06, + "loss": 0.7541, + "num_input_tokens_seen": 117577310, + "step": 5465 + }, + { + "epoch": 0.6572476402332712, + "flos": 17504914188960.0, + "grad_norm": 3.0505124021264596, + "learning_rate": 1.1107389447630086e-06, + "loss": 0.8121, + "num_input_tokens_seen": 117596010, + "step": 5466 + }, + { + "epoch": 0.6573678831239103, + "flos": 16989921233280.0, + "grad_norm": 2.7969668069614957, + "learning_rate": 1.1100412763666818e-06, + "loss": 0.7884, + "num_input_tokens_seen": 117611270, + "step": 5467 + }, + { + "epoch": 0.6574881260145494, + "flos": 23875699340160.0, + "grad_norm": 2.9104682089506664, + "learning_rate": 1.1093437429763865e-06, + "loss": 0.8035, + "num_input_tokens_seen": 117631535, + "step": 5468 + }, + { + "epoch": 0.6576083689051885, + "flos": 11202755881920.0, + "grad_norm": 3.473425111735668, + "learning_rate": 1.1086463446979361e-06, + "loss": 0.7337, + "num_input_tokens_seen": 117649885, + "step": 5469 + }, + { + "epoch": 0.6577286117958275, + "flos": 22423470491520.0, + "grad_norm": 3.2450293162568964, + "learning_rate": 1.1079490816371277e-06, + "loss": 0.7769, + "num_input_tokens_seen": 117669650, + "step": 5470 + }, + { + "epoch": 0.6578488546864667, + "flos": 21840965784480.0, + "grad_norm": 2.4938140631456407, + "learning_rate": 1.1072519538997352e-06, + "loss": 0.7457, + "num_input_tokens_seen": 117688790, + "step": 5471 + }, + { + "epoch": 0.6579690975771058, + "flos": 23509854396000.0, + "grad_norm": 3.17026437541445, + "learning_rate": 1.1065549615915095e-06, + "loss": 0.823, + "num_input_tokens_seen": 117708620, + "step": 5472 + }, + { + "epoch": 0.6580893404677448, + "flos": 32696913372960.0, + "grad_norm": 4.200258319721845, + "learning_rate": 1.105858104818187e-06, + "loss": 0.7835, + "num_input_tokens_seen": 117729775, + "step": 5473 + }, + { + "epoch": 0.658209583358384, + "flos": 15865618440000.0, + "grad_norm": 5.119279599768455, + "learning_rate": 1.105161383685478e-06, + "loss": 0.7467, + "num_input_tokens_seen": 117746160, + "step": 5474 + }, + { + "epoch": 0.658329826249023, + "flos": 62606433823680.0, + "grad_norm": 0.7330245729883803, + "learning_rate": 1.1044647982990771e-06, + "loss": 0.562, + "num_input_tokens_seen": 117808045, + "step": 5475 + }, + { + "epoch": 0.6584500691396621, + "flos": 31677037636320.0, + "grad_norm": 3.8386506781590146, + "learning_rate": 1.1037683487646536e-06, + "loss": 0.6465, + "num_input_tokens_seen": 117828330, + "step": 5476 + }, + { + "epoch": 0.6585703120303013, + "flos": 18379916436480.0, + "grad_norm": 2.7354314888414337, + "learning_rate": 1.1030720351878583e-06, + "loss": 0.7713, + "num_input_tokens_seen": 117846450, + "step": 5477 + }, + { + "epoch": 0.6586905549209403, + "flos": 58226516177760.0, + "grad_norm": 0.8672962636014087, + "learning_rate": 1.102375857674323e-06, + "loss": 0.5758, + "num_input_tokens_seen": 117908365, + "step": 5478 + }, + { + "epoch": 0.6588107978115794, + "flos": 22750021265760.0, + "grad_norm": 4.945792658487814, + "learning_rate": 1.1016798163296561e-06, + "loss": 0.9045, + "num_input_tokens_seen": 117927565, + "step": 5479 + }, + { + "epoch": 0.6589310407022185, + "flos": 20638037481600.0, + "grad_norm": 2.351802877009115, + "learning_rate": 1.1009839112594471e-06, + "loss": 0.6627, + "num_input_tokens_seen": 117945590, + "step": 5480 + }, + { + "epoch": 0.6590512835928576, + "flos": 25593917956800.0, + "grad_norm": 2.7848317945765544, + "learning_rate": 1.1002881425692638e-06, + "loss": 0.7199, + "num_input_tokens_seen": 117966020, + "step": 5481 + }, + { + "epoch": 0.6591715264834966, + "flos": 23691940548480.0, + "grad_norm": 2.3599624407737494, + "learning_rate": 1.0995925103646532e-06, + "loss": 0.7535, + "num_input_tokens_seen": 117984620, + "step": 5482 + }, + { + "epoch": 0.6592917693741358, + "flos": 35722413009600.0, + "grad_norm": 2.378977406507129, + "learning_rate": 1.0988970147511437e-06, + "loss": 0.6686, + "num_input_tokens_seen": 118006500, + "step": 5483 + }, + { + "epoch": 0.6594120122647749, + "flos": 21366270582240.0, + "grad_norm": 5.028321545293838, + "learning_rate": 1.0982016558342405e-06, + "loss": 0.8052, + "num_input_tokens_seen": 118025470, + "step": 5484 + }, + { + "epoch": 0.6595322551554139, + "flos": 19324177414080.0, + "grad_norm": 3.5372789617166087, + "learning_rate": 1.0975064337194291e-06, + "loss": 0.7097, + "num_input_tokens_seen": 118043515, + "step": 5485 + }, + { + "epoch": 0.6596524980460531, + "flos": 16813224696000.0, + "grad_norm": 2.2201355382423174, + "learning_rate": 1.0968113485121743e-06, + "loss": 0.7035, + "num_input_tokens_seen": 118063105, + "step": 5486 + }, + { + "epoch": 0.6597727409366921, + "flos": 21767432532000.0, + "grad_norm": 2.115517496056945, + "learning_rate": 1.0961164003179185e-06, + "loss": 0.8034, + "num_input_tokens_seen": 118081615, + "step": 5487 + }, + { + "epoch": 0.6598929838273312, + "flos": 23696698277760.0, + "grad_norm": 2.6430931519839964, + "learning_rate": 1.0954215892420884e-06, + "loss": 0.8452, + "num_input_tokens_seen": 118102315, + "step": 5488 + }, + { + "epoch": 0.6600132267179702, + "flos": 19946050630560.0, + "grad_norm": 2.1134698751886853, + "learning_rate": 1.094726915390082e-06, + "loss": 0.7064, + "num_input_tokens_seen": 118120650, + "step": 5489 + }, + { + "epoch": 0.6601334696086094, + "flos": 22310605899360.0, + "grad_norm": 2.702510344902798, + "learning_rate": 1.0940323788672836e-06, + "loss": 0.6945, + "num_input_tokens_seen": 118139825, + "step": 5490 + }, + { + "epoch": 0.6602537124992485, + "flos": 25666298946720.0, + "grad_norm": 2.0114102321452876, + "learning_rate": 1.0933379797790522e-06, + "loss": 0.7391, + "num_input_tokens_seen": 118159240, + "step": 5491 + }, + { + "epoch": 0.6603739553898875, + "flos": 25811544133440.0, + "grad_norm": 3.2077421824482117, + "learning_rate": 1.0926437182307293e-06, + "loss": 0.7074, + "num_input_tokens_seen": 118178050, + "step": 5492 + }, + { + "epoch": 0.6604941982805267, + "flos": 24535082578080.0, + "grad_norm": 2.461306977162263, + "learning_rate": 1.0919495943276338e-06, + "loss": 0.7788, + "num_input_tokens_seen": 118199070, + "step": 5493 + }, + { + "epoch": 0.6606144411711657, + "flos": 13243102071360.0, + "grad_norm": 4.338191146093489, + "learning_rate": 1.0912556081750611e-06, + "loss": 0.7656, + "num_input_tokens_seen": 118217000, + "step": 5494 + }, + { + "epoch": 0.6607346840618048, + "flos": 25119259924320.0, + "grad_norm": 2.7052306865541738, + "learning_rate": 1.0905617598782909e-06, + "loss": 0.7614, + "num_input_tokens_seen": 118237205, + "step": 5495 + }, + { + "epoch": 0.660854926952444, + "flos": 17612909542560.0, + "grad_norm": 2.6412731932575433, + "learning_rate": 1.0898680495425775e-06, + "loss": 0.8194, + "num_input_tokens_seen": 118255495, + "step": 5496 + }, + { + "epoch": 0.660975169843083, + "flos": 16812072433440.0, + "grad_norm": 2.229078510953709, + "learning_rate": 1.0891744772731594e-06, + "loss": 0.806, + "num_input_tokens_seen": 118273310, + "step": 5497 + }, + { + "epoch": 0.6610954127337221, + "flos": 26831903076960.0, + "grad_norm": 2.9543139773838076, + "learning_rate": 1.088481043175248e-06, + "loss": 0.6648, + "num_input_tokens_seen": 118293880, + "step": 5498 + }, + { + "epoch": 0.6612156556243612, + "flos": 26427656037120.0, + "grad_norm": 2.6280153303816918, + "learning_rate": 1.0877877473540368e-06, + "loss": 0.7592, + "num_input_tokens_seen": 118314465, + "step": 5499 + }, + { + "epoch": 0.6613358985150003, + "flos": 19763704289760.0, + "grad_norm": 2.99023389660553, + "learning_rate": 1.0870945899147002e-06, + "loss": 0.7241, + "num_input_tokens_seen": 118331110, + "step": 5500 + }, + { + "epoch": 0.6614561414056394, + "flos": 26793092113920.0, + "grad_norm": 2.187019186311626, + "learning_rate": 1.0864015709623879e-06, + "loss": 0.7631, + "num_input_tokens_seen": 118351980, + "step": 5501 + }, + { + "epoch": 0.6615763842962785, + "flos": 22861919444160.0, + "grad_norm": 4.263782368729744, + "learning_rate": 1.0857086906022313e-06, + "loss": 0.8016, + "num_input_tokens_seen": 118370315, + "step": 5502 + }, + { + "epoch": 0.6616966271869176, + "flos": 24754827431040.0, + "grad_norm": 3.2755015564287238, + "learning_rate": 1.0850159489393388e-06, + "loss": 0.7308, + "num_input_tokens_seen": 118389575, + "step": 5503 + }, + { + "epoch": 0.6618168700775566, + "flos": 17177694359040.0, + "grad_norm": 2.3481891554068453, + "learning_rate": 1.0843233460787992e-06, + "loss": 0.8236, + "num_input_tokens_seen": 118406790, + "step": 5504 + }, + { + "epoch": 0.6619371129681958, + "flos": 25410642372000.0, + "grad_norm": 2.714615231783027, + "learning_rate": 1.0836308821256805e-06, + "loss": 0.7795, + "num_input_tokens_seen": 118427590, + "step": 5505 + }, + { + "epoch": 0.6620573558588349, + "flos": 18014777717760.0, + "grad_norm": 3.625880956480961, + "learning_rate": 1.0829385571850282e-06, + "loss": 0.7777, + "num_input_tokens_seen": 118444570, + "step": 5506 + }, + { + "epoch": 0.6621775987494739, + "flos": 17760310575360.0, + "grad_norm": 3.15862524718404, + "learning_rate": 1.0822463713618679e-06, + "loss": 0.8347, + "num_input_tokens_seen": 118461425, + "step": 5507 + }, + { + "epoch": 0.6622978416401131, + "flos": 17467701525600.0, + "grad_norm": 2.923199847953304, + "learning_rate": 1.0815543247612034e-06, + "loss": 0.8527, + "num_input_tokens_seen": 118478495, + "step": 5508 + }, + { + "epoch": 0.6624180845307521, + "flos": 21617280936960.0, + "grad_norm": 2.76172755562378, + "learning_rate": 1.0808624174880168e-06, + "loss": 0.8315, + "num_input_tokens_seen": 118499145, + "step": 5509 + }, + { + "epoch": 0.6625383274213912, + "flos": 23771866999680.0, + "grad_norm": 2.1521803556755446, + "learning_rate": 1.080170649647272e-06, + "loss": 0.8016, + "num_input_tokens_seen": 118518950, + "step": 5510 + }, + { + "epoch": 0.6626585703120303, + "flos": 33215400286080.0, + "grad_norm": 2.19383032236016, + "learning_rate": 1.0794790213439068e-06, + "loss": 0.6735, + "num_input_tokens_seen": 118545850, + "step": 5511 + }, + { + "epoch": 0.6627788132026694, + "flos": 22054057250400.0, + "grad_norm": 2.41301825693609, + "learning_rate": 1.078787532682843e-06, + "loss": 0.7863, + "num_input_tokens_seen": 118563325, + "step": 5512 + }, + { + "epoch": 0.6628990560933085, + "flos": 36121530622560.0, + "grad_norm": 2.8664177217251363, + "learning_rate": 1.0780961837689773e-06, + "loss": 0.7582, + "num_input_tokens_seen": 118582835, + "step": 5513 + }, + { + "epoch": 0.6630192989839476, + "flos": 18487168394880.0, + "grad_norm": 1.9364394700183432, + "learning_rate": 1.0774049747071883e-06, + "loss": 0.7025, + "num_input_tokens_seen": 118600715, + "step": 5514 + }, + { + "epoch": 0.6631395418745867, + "flos": 35757990203520.0, + "grad_norm": 1.9319686071202906, + "learning_rate": 1.076713905602332e-06, + "loss": 0.6808, + "num_input_tokens_seen": 118621290, + "step": 5515 + }, + { + "epoch": 0.6632597847652257, + "flos": 20019175015680.0, + "grad_norm": 2.020235097631196, + "learning_rate": 1.07602297655924e-06, + "loss": 0.8161, + "num_input_tokens_seen": 118639610, + "step": 5516 + }, + { + "epoch": 0.6633800276558649, + "flos": 21184184429760.0, + "grad_norm": 2.657284075343465, + "learning_rate": 1.0753321876827292e-06, + "loss": 0.812, + "num_input_tokens_seen": 118658895, + "step": 5517 + }, + { + "epoch": 0.663500270546504, + "flos": 23953432775520.0, + "grad_norm": 2.6999711176422414, + "learning_rate": 1.0746415390775893e-06, + "loss": 0.7411, + "num_input_tokens_seen": 118677025, + "step": 5518 + }, + { + "epoch": 0.663620513437143, + "flos": 17906931043200.0, + "grad_norm": 3.294214992527383, + "learning_rate": 1.0739510308485939e-06, + "loss": 0.7619, + "num_input_tokens_seen": 118694240, + "step": 5519 + }, + { + "epoch": 0.6637407563277821, + "flos": 57757730967360.0, + "grad_norm": 0.818012792354345, + "learning_rate": 1.07326066310049e-06, + "loss": 0.6246, + "num_input_tokens_seen": 118758365, + "step": 5520 + }, + { + "epoch": 0.6638609992184212, + "flos": 27266783732640.0, + "grad_norm": 2.619696209017195, + "learning_rate": 1.0725704359380059e-06, + "loss": 0.7929, + "num_input_tokens_seen": 118778220, + "step": 5521 + }, + { + "epoch": 0.6639812421090603, + "flos": 18598360347840.0, + "grad_norm": 2.5241568753478787, + "learning_rate": 1.0718803494658497e-06, + "loss": 0.7175, + "num_input_tokens_seen": 118797110, + "step": 5522 + }, + { + "epoch": 0.6641014849996993, + "flos": 15901901859360.0, + "grad_norm": 3.8516225442597185, + "learning_rate": 1.071190403788707e-06, + "loss": 0.8411, + "num_input_tokens_seen": 118812415, + "step": 5523 + }, + { + "epoch": 0.6642217278903385, + "flos": 26467507753440.0, + "grad_norm": 2.8204434435766137, + "learning_rate": 1.0705005990112415e-06, + "loss": 0.7605, + "num_input_tokens_seen": 118832195, + "step": 5524 + }, + { + "epoch": 0.6643419707809776, + "flos": 15355531892640.0, + "grad_norm": 3.128756287370705, + "learning_rate": 1.0698109352380957e-06, + "loss": 0.7492, + "num_input_tokens_seen": 118849795, + "step": 5525 + }, + { + "epoch": 0.6644622136716166, + "flos": 25081601223840.0, + "grad_norm": 2.4692838067566365, + "learning_rate": 1.0691214125738909e-06, + "loss": 0.7823, + "num_input_tokens_seen": 118870000, + "step": 5526 + }, + { + "epoch": 0.6645824565622558, + "flos": 66107260548480.0, + "grad_norm": 0.8397939977768878, + "learning_rate": 1.0684320311232287e-06, + "loss": 0.5741, + "num_input_tokens_seen": 118932905, + "step": 5527 + }, + { + "epoch": 0.6647026994528948, + "flos": 25046209878720.0, + "grad_norm": 2.6091555200430263, + "learning_rate": 1.0677427909906865e-06, + "loss": 0.8129, + "num_input_tokens_seen": 118953355, + "step": 5528 + }, + { + "epoch": 0.6648229423435339, + "flos": 18196231984320.0, + "grad_norm": 5.021116868662255, + "learning_rate": 1.0670536922808216e-06, + "loss": 0.7239, + "num_input_tokens_seen": 118973480, + "step": 5529 + }, + { + "epoch": 0.6649431852341731, + "flos": 18270880329600.0, + "grad_norm": 3.397141623890378, + "learning_rate": 1.06636473509817e-06, + "loss": 0.7213, + "num_input_tokens_seen": 118989860, + "step": 5530 + }, + { + "epoch": 0.6650634281248121, + "flos": 16995422357760.0, + "grad_norm": 3.383030556220817, + "learning_rate": 1.0656759195472447e-06, + "loss": 0.8075, + "num_input_tokens_seen": 119007150, + "step": 5531 + }, + { + "epoch": 0.6651836710154512, + "flos": 69195551376960.0, + "grad_norm": 0.7758905831209114, + "learning_rate": 1.0649872457325414e-06, + "loss": 0.5964, + "num_input_tokens_seen": 119068435, + "step": 5532 + }, + { + "epoch": 0.6653039139060903, + "flos": 66787576094880.0, + "grad_norm": 0.9336047949995516, + "learning_rate": 1.0642987137585278e-06, + "loss": 0.5502, + "num_input_tokens_seen": 119127960, + "step": 5533 + }, + { + "epoch": 0.6654241567967294, + "flos": 21439394967360.0, + "grad_norm": 2.126276194003506, + "learning_rate": 1.0636103237296561e-06, + "loss": 0.8268, + "num_input_tokens_seen": 119146885, + "step": 5534 + }, + { + "epoch": 0.6655443996873684, + "flos": 25083571221120.0, + "grad_norm": 3.215533677297438, + "learning_rate": 1.062922075750353e-06, + "loss": 0.8465, + "num_input_tokens_seen": 119166900, + "step": 5535 + }, + { + "epoch": 0.6656646425780076, + "flos": 17432533199040.0, + "grad_norm": 5.022301529133444, + "learning_rate": 1.0622339699250267e-06, + "loss": 0.7248, + "num_input_tokens_seen": 119184820, + "step": 5536 + }, + { + "epoch": 0.6657848854686467, + "flos": 23400446591520.0, + "grad_norm": 2.505458594582046, + "learning_rate": 1.0615460063580624e-06, + "loss": 0.7926, + "num_input_tokens_seen": 119203295, + "step": 5537 + }, + { + "epoch": 0.6659051283592857, + "flos": 11493766632000.0, + "grad_norm": 3.4536071716884496, + "learning_rate": 1.060858185153821e-06, + "loss": 0.7328, + "num_input_tokens_seen": 119221395, + "step": 5538 + }, + { + "epoch": 0.6660253712499249, + "flos": 20565173284800.0, + "grad_norm": 2.89363296305069, + "learning_rate": 1.0601705064166474e-06, + "loss": 0.7582, + "num_input_tokens_seen": 119239790, + "step": 5539 + }, + { + "epoch": 0.666145614140564, + "flos": 21221174074560.0, + "grad_norm": 2.700406951612143, + "learning_rate": 1.0594829702508596e-06, + "loss": 0.7341, + "num_input_tokens_seen": 119257505, + "step": 5540 + }, + { + "epoch": 0.666265857031203, + "flos": 33677569279200.0, + "grad_norm": 2.4769475499168454, + "learning_rate": 1.0587955767607592e-06, + "loss": 0.5481, + "num_input_tokens_seen": 119279365, + "step": 5541 + }, + { + "epoch": 0.6663860999218422, + "flos": 17431826973600.0, + "grad_norm": 4.602374434283853, + "learning_rate": 1.0581083260506206e-06, + "loss": 0.7714, + "num_input_tokens_seen": 119296150, + "step": 5542 + }, + { + "epoch": 0.6665063428124812, + "flos": 17651646166080.0, + "grad_norm": 3.1094030637768566, + "learning_rate": 1.0574212182246993e-06, + "loss": 0.7693, + "num_input_tokens_seen": 119314845, + "step": 5543 + }, + { + "epoch": 0.6666265857031203, + "flos": 27633595090560.0, + "grad_norm": 4.066766933601719, + "learning_rate": 1.0567342533872303e-06, + "loss": 0.7588, + "num_input_tokens_seen": 119334875, + "step": 5544 + }, + { + "epoch": 0.6667468285937594, + "flos": 25011004382400.0, + "grad_norm": 2.305390952131371, + "learning_rate": 1.0560474316424255e-06, + "loss": 0.8066, + "num_input_tokens_seen": 119354635, + "step": 5545 + }, + { + "epoch": 0.6668670714843985, + "flos": 22747791080160.0, + "grad_norm": 3.0328949859580714, + "learning_rate": 1.0553607530944746e-06, + "loss": 0.74, + "num_input_tokens_seen": 119372690, + "step": 5546 + }, + { + "epoch": 0.6669873143750376, + "flos": 22132236722880.0, + "grad_norm": 2.620623436928921, + "learning_rate": 1.0546742178475463e-06, + "loss": 0.8977, + "num_input_tokens_seen": 119391685, + "step": 5547 + }, + { + "epoch": 0.6671075572656767, + "flos": 20485321173120.0, + "grad_norm": 2.51503135358253, + "learning_rate": 1.0539878260057868e-06, + "loss": 0.8689, + "num_input_tokens_seen": 119410725, + "step": 5548 + }, + { + "epoch": 0.6672278001563158, + "flos": 17905704441120.0, + "grad_norm": 3.867014131609475, + "learning_rate": 1.0533015776733226e-06, + "loss": 0.6857, + "num_input_tokens_seen": 119426875, + "step": 5549 + }, + { + "epoch": 0.6673480430469548, + "flos": 22310085522720.0, + "grad_norm": 4.410148032681671, + "learning_rate": 1.0526154729542566e-06, + "loss": 0.7841, + "num_input_tokens_seen": 119446935, + "step": 5550 + }, + { + "epoch": 0.6674682859375939, + "flos": 20674283731200.0, + "grad_norm": 3.604190025180396, + "learning_rate": 1.0519295119526699e-06, + "loss": 0.8048, + "num_input_tokens_seen": 119463240, + "step": 5551 + }, + { + "epoch": 0.667588528828233, + "flos": 26168877202560.0, + "grad_norm": 3.2677929266666004, + "learning_rate": 1.0512436947726227e-06, + "loss": 0.8297, + "num_input_tokens_seen": 119484655, + "step": 5552 + }, + { + "epoch": 0.6677087717188721, + "flos": 23037872586240.0, + "grad_norm": 2.7500399673331897, + "learning_rate": 1.0505580215181517e-06, + "loss": 0.658, + "num_input_tokens_seen": 119502895, + "step": 5553 + }, + { + "epoch": 0.6678290146095112, + "flos": 70839827873760.0, + "grad_norm": 0.7877520263897623, + "learning_rate": 1.0498724922932753e-06, + "loss": 0.5649, + "num_input_tokens_seen": 119561010, + "step": 5554 + }, + { + "epoch": 0.6679492575001503, + "flos": 18634532257920.0, + "grad_norm": 6.181869256731674, + "learning_rate": 1.0491871072019851e-06, + "loss": 0.8637, + "num_input_tokens_seen": 119578900, + "step": 5555 + }, + { + "epoch": 0.6680695003907894, + "flos": 29669183550720.0, + "grad_norm": 2.8776349142281687, + "learning_rate": 1.0485018663482555e-06, + "loss": 0.6433, + "num_input_tokens_seen": 119598920, + "step": 5556 + }, + { + "epoch": 0.6681897432814284, + "flos": 28178441097120.0, + "grad_norm": 2.9918219041705028, + "learning_rate": 1.0478167698360354e-06, + "loss": 0.7025, + "num_input_tokens_seen": 119618220, + "step": 5557 + }, + { + "epoch": 0.6683099861720676, + "flos": 25010707024320.0, + "grad_norm": 2.911865387896811, + "learning_rate": 1.0471318177692556e-06, + "loss": 0.7047, + "num_input_tokens_seen": 119638520, + "step": 5558 + }, + { + "epoch": 0.6684302290627067, + "flos": 22963893296640.0, + "grad_norm": 2.6637037824383123, + "learning_rate": 1.046447010251821e-06, + "loss": 0.7624, + "num_input_tokens_seen": 119655850, + "step": 5559 + }, + { + "epoch": 0.6685504719533457, + "flos": 26535577051200.0, + "grad_norm": 3.127946608846284, + "learning_rate": 1.0457623473876157e-06, + "loss": 0.7618, + "num_input_tokens_seen": 119675355, + "step": 5560 + }, + { + "epoch": 0.6686707148439849, + "flos": 28944481577280.0, + "grad_norm": 2.542390209666968, + "learning_rate": 1.0450778292805046e-06, + "loss": 0.7108, + "num_input_tokens_seen": 119695340, + "step": 5561 + }, + { + "epoch": 0.6687909577346239, + "flos": 23589371979840.0, + "grad_norm": 1.963078337087161, + "learning_rate": 1.0443934560343267e-06, + "loss": 0.7855, + "num_input_tokens_seen": 119716750, + "step": 5562 + }, + { + "epoch": 0.668911200625263, + "flos": 23115494512320.0, + "grad_norm": 2.4622923981347857, + "learning_rate": 1.0437092277529034e-06, + "loss": 0.7821, + "num_input_tokens_seen": 119736005, + "step": 5563 + }, + { + "epoch": 0.6690314435159022, + "flos": 18525533320800.0, + "grad_norm": 3.25796120576377, + "learning_rate": 1.0430251445400292e-06, + "loss": 0.7348, + "num_input_tokens_seen": 119754050, + "step": 5564 + }, + { + "epoch": 0.6691516864065412, + "flos": 31713915771840.0, + "grad_norm": 4.773550978090596, + "learning_rate": 1.0423412064994787e-06, + "loss": 0.6254, + "num_input_tokens_seen": 119774820, + "step": 5565 + }, + { + "epoch": 0.6692719292971803, + "flos": 34883619841920.0, + "grad_norm": 2.5001540136580584, + "learning_rate": 1.0416574137350064e-06, + "loss": 0.7436, + "num_input_tokens_seen": 119795525, + "step": 5566 + }, + { + "epoch": 0.6693921721878194, + "flos": 20420039607360.0, + "grad_norm": 5.20184255380628, + "learning_rate": 1.0409737663503428e-06, + "loss": 0.8066, + "num_input_tokens_seen": 119813905, + "step": 5567 + }, + { + "epoch": 0.6695124150784585, + "flos": 16590840790080.0, + "grad_norm": 2.3036711957413494, + "learning_rate": 1.040290264449196e-06, + "loss": 0.8267, + "num_input_tokens_seen": 119832010, + "step": 5568 + }, + { + "epoch": 0.6696326579690975, + "flos": 26613979542240.0, + "grad_norm": 2.3856612389690075, + "learning_rate": 1.0396069081352532e-06, + "loss": 0.6426, + "num_input_tokens_seen": 119851165, + "step": 5569 + }, + { + "epoch": 0.6697529008597367, + "flos": 66868914996960.0, + "grad_norm": 0.8735499969808191, + "learning_rate": 1.0389236975121782e-06, + "loss": 0.5592, + "num_input_tokens_seen": 119906015, + "step": 5570 + }, + { + "epoch": 0.6698731437503758, + "flos": 20857150448640.0, + "grad_norm": 2.614594498771343, + "learning_rate": 1.0382406326836147e-06, + "loss": 0.7131, + "num_input_tokens_seen": 119925160, + "step": 5571 + }, + { + "epoch": 0.6699933866410148, + "flos": 20380262230560.0, + "grad_norm": 2.6946565133849942, + "learning_rate": 1.0375577137531828e-06, + "loss": 0.7536, + "num_input_tokens_seen": 119943720, + "step": 5572 + }, + { + "epoch": 0.670113629531654, + "flos": 28981434052320.0, + "grad_norm": 1.8581250260728719, + "learning_rate": 1.0368749408244802e-06, + "loss": 0.7204, + "num_input_tokens_seen": 119966640, + "step": 5573 + }, + { + "epoch": 0.670233872422293, + "flos": 19763481271200.0, + "grad_norm": 2.193218929859031, + "learning_rate": 1.0361923140010836e-06, + "loss": 0.7855, + "num_input_tokens_seen": 119985125, + "step": 5574 + }, + { + "epoch": 0.6703541153129321, + "flos": 24209015010720.0, + "grad_norm": 2.848454534884608, + "learning_rate": 1.0355098333865455e-06, + "loss": 0.637, + "num_input_tokens_seen": 120004355, + "step": 5575 + }, + { + "epoch": 0.6704743582035713, + "flos": 26650300131360.0, + "grad_norm": 2.346499343654135, + "learning_rate": 1.0348274990844006e-06, + "loss": 0.6943, + "num_input_tokens_seen": 120027115, + "step": 5576 + }, + { + "epoch": 0.6705946010942103, + "flos": 23480781910080.0, + "grad_norm": 3.676052622654018, + "learning_rate": 1.034145311198155e-06, + "loss": 0.7259, + "num_input_tokens_seen": 120047130, + "step": 5577 + }, + { + "epoch": 0.6707148439848494, + "flos": 24026594330400.0, + "grad_norm": 2.2870934255354953, + "learning_rate": 1.0334632698312989e-06, + "loss": 0.6375, + "num_input_tokens_seen": 120067925, + "step": 5578 + }, + { + "epoch": 0.6708350868754885, + "flos": 22496483367360.0, + "grad_norm": 2.438977893874621, + "learning_rate": 1.032781375087295e-06, + "loss": 0.7528, + "num_input_tokens_seen": 120087825, + "step": 5579 + }, + { + "epoch": 0.6709553297661276, + "flos": 25191603744480.0, + "grad_norm": 1.7138278880599322, + "learning_rate": 1.0320996270695891e-06, + "loss": 0.6759, + "num_input_tokens_seen": 120108895, + "step": 5580 + }, + { + "epoch": 0.6710755726567667, + "flos": 20419593570240.0, + "grad_norm": 2.1508093486021544, + "learning_rate": 1.0314180258815998e-06, + "loss": 0.7385, + "num_input_tokens_seen": 120127535, + "step": 5581 + }, + { + "epoch": 0.6711958155474057, + "flos": 25958201771040.0, + "grad_norm": 2.8286978631434194, + "learning_rate": 1.0307365716267247e-06, + "loss": 0.7431, + "num_input_tokens_seen": 120147980, + "step": 5582 + }, + { + "epoch": 0.6713160584380449, + "flos": 19909284004320.0, + "grad_norm": 5.19493838232395, + "learning_rate": 1.0300552644083423e-06, + "loss": 0.785, + "num_input_tokens_seen": 120166905, + "step": 5583 + }, + { + "epoch": 0.6714363013286839, + "flos": 18197458586400.0, + "grad_norm": 3.3311603297891907, + "learning_rate": 1.0293741043298036e-06, + "loss": 0.7248, + "num_input_tokens_seen": 120185255, + "step": 5584 + }, + { + "epoch": 0.671556544219323, + "flos": 25775409393120.0, + "grad_norm": 2.3946433608925357, + "learning_rate": 1.0286930914944436e-06, + "loss": 0.7161, + "num_input_tokens_seen": 120205070, + "step": 5585 + }, + { + "epoch": 0.6716767871099621, + "flos": 15827922569760.0, + "grad_norm": 2.9362356833364287, + "learning_rate": 1.0280122260055684e-06, + "loss": 0.7734, + "num_input_tokens_seen": 120220780, + "step": 5586 + }, + { + "epoch": 0.6717970300006012, + "flos": 19727978416800.0, + "grad_norm": 3.7150733788535204, + "learning_rate": 1.0273315079664652e-06, + "loss": 0.8247, + "num_input_tokens_seen": 120238735, + "step": 5587 + }, + { + "epoch": 0.6719172728912403, + "flos": 25448598430560.0, + "grad_norm": 2.9181177798288274, + "learning_rate": 1.0266509374803992e-06, + "loss": 0.7451, + "num_input_tokens_seen": 120259895, + "step": 5588 + }, + { + "epoch": 0.6720375157818794, + "flos": 15857552602080.0, + "grad_norm": 3.76164827053004, + "learning_rate": 1.0259705146506123e-06, + "loss": 0.8431, + "num_input_tokens_seen": 120274790, + "step": 5589 + }, + { + "epoch": 0.6721577586725185, + "flos": 31964703108000.0, + "grad_norm": 2.706762963854182, + "learning_rate": 1.025290239580324e-06, + "loss": 0.7757, + "num_input_tokens_seen": 120295460, + "step": 5590 + }, + { + "epoch": 0.6722780015631575, + "flos": 20708188285920.0, + "grad_norm": 2.038596030478373, + "learning_rate": 1.0246101123727313e-06, + "loss": 0.7535, + "num_input_tokens_seen": 120314440, + "step": 5591 + }, + { + "epoch": 0.6723982444537967, + "flos": 16885605685920.0, + "grad_norm": 2.631941563569178, + "learning_rate": 1.0239301331310085e-06, + "loss": 0.7917, + "num_input_tokens_seen": 120332060, + "step": 5592 + }, + { + "epoch": 0.6725184873444358, + "flos": 20638334839680.0, + "grad_norm": 2.4086792808488893, + "learning_rate": 1.0232503019583088e-06, + "loss": 0.8827, + "num_input_tokens_seen": 120351670, + "step": 5593 + }, + { + "epoch": 0.6726387302350748, + "flos": 23693501678400.0, + "grad_norm": 2.3226694793786336, + "learning_rate": 1.0225706189577619e-06, + "loss": 0.6949, + "num_input_tokens_seen": 120370195, + "step": 5594 + }, + { + "epoch": 0.672758973125714, + "flos": 15166532164800.0, + "grad_norm": 7.425744262795838, + "learning_rate": 1.021891084232475e-06, + "loss": 0.7491, + "num_input_tokens_seen": 120388130, + "step": 5595 + }, + { + "epoch": 0.672879216016353, + "flos": 18051544344000.0, + "grad_norm": 3.2240011102470967, + "learning_rate": 1.0212116978855325e-06, + "loss": 0.8022, + "num_input_tokens_seen": 120406145, + "step": 5596 + }, + { + "epoch": 0.6729994589069921, + "flos": 23442751512000.0, + "grad_norm": 2.175512656818343, + "learning_rate": 1.020532460019997e-06, + "loss": 0.7908, + "num_input_tokens_seen": 120425395, + "step": 5597 + }, + { + "epoch": 0.6731197017976313, + "flos": 26284269338400.0, + "grad_norm": 2.121117129820503, + "learning_rate": 1.0198533707389096e-06, + "loss": 0.7102, + "num_input_tokens_seen": 120446270, + "step": 5598 + }, + { + "epoch": 0.6732399446882703, + "flos": 21585569398080.0, + "grad_norm": 2.824739062656005, + "learning_rate": 1.0191744301452853e-06, + "loss": 0.7327, + "num_input_tokens_seen": 120465570, + "step": 5599 + }, + { + "epoch": 0.6733601875789094, + "flos": 25842921144480.0, + "grad_norm": 2.2795845118288165, + "learning_rate": 1.0184956383421208e-06, + "loss": 0.7039, + "num_input_tokens_seen": 120484220, + "step": 5600 + }, + { + "epoch": 0.6734804304695485, + "flos": 22896790412640.0, + "grad_norm": 2.7922813156411572, + "learning_rate": 1.017816995432387e-06, + "loss": 0.6566, + "num_input_tokens_seen": 120503320, + "step": 5601 + }, + { + "epoch": 0.6736006733601876, + "flos": 18671373223680.0, + "grad_norm": 2.2976936386578837, + "learning_rate": 1.0171385015190353e-06, + "loss": 0.7434, + "num_input_tokens_seen": 120523180, + "step": 5602 + }, + { + "epoch": 0.6737209162508266, + "flos": 19400089531200.0, + "grad_norm": 3.1115855884528676, + "learning_rate": 1.0164601567049908e-06, + "loss": 0.7333, + "num_input_tokens_seen": 120541710, + "step": 5603 + }, + { + "epoch": 0.6738411591414658, + "flos": 20129363385120.0, + "grad_norm": 2.072466726369613, + "learning_rate": 1.015781961093158e-06, + "loss": 0.8043, + "num_input_tokens_seen": 120561030, + "step": 5604 + }, + { + "epoch": 0.6739614020321049, + "flos": 21622967910240.0, + "grad_norm": 2.4345793220852143, + "learning_rate": 1.0151039147864197e-06, + "loss": 0.7725, + "num_input_tokens_seen": 120581005, + "step": 5605 + }, + { + "epoch": 0.6740816449227439, + "flos": 19144395786720.0, + "grad_norm": 4.067151267981744, + "learning_rate": 1.0144260178876336e-06, + "loss": 0.6592, + "num_input_tokens_seen": 120600350, + "step": 5606 + }, + { + "epoch": 0.6742018878133831, + "flos": 21066599278080.0, + "grad_norm": 3.3063402304208376, + "learning_rate": 1.0137482704996388e-06, + "loss": 0.6696, + "num_input_tokens_seen": 120614700, + "step": 5607 + }, + { + "epoch": 0.6743221307040221, + "flos": 23516284764480.0, + "grad_norm": 2.3844172842693583, + "learning_rate": 1.0130706727252461e-06, + "loss": 0.7903, + "num_input_tokens_seen": 120631755, + "step": 5608 + }, + { + "epoch": 0.6744423735946612, + "flos": 16226371127040.0, + "grad_norm": 3.5851099060747433, + "learning_rate": 1.0123932246672468e-06, + "loss": 0.6812, + "num_input_tokens_seen": 120645900, + "step": 5609 + }, + { + "epoch": 0.6745626164853004, + "flos": 57760853227200.0, + "grad_norm": 0.7556579084779257, + "learning_rate": 1.0117159264284114e-06, + "loss": 0.5571, + "num_input_tokens_seen": 120709070, + "step": 5610 + }, + { + "epoch": 0.6746828593759394, + "flos": 20455802650080.0, + "grad_norm": 2.665538569096738, + "learning_rate": 1.0110387781114837e-06, + "loss": 0.7731, + "num_input_tokens_seen": 120727685, + "step": 5611 + }, + { + "epoch": 0.6748031022665785, + "flos": 19180865054880.0, + "grad_norm": 3.9804509006208604, + "learning_rate": 1.0103617798191872e-06, + "loss": 0.7727, + "num_input_tokens_seen": 120747160, + "step": 5612 + }, + { + "epoch": 0.6749233451572175, + "flos": 15173520079680.0, + "grad_norm": 3.8296774694030464, + "learning_rate": 1.0096849316542217e-06, + "loss": 0.8262, + "num_input_tokens_seen": 120763710, + "step": 5613 + }, + { + "epoch": 0.6750435880478567, + "flos": 26461300403520.0, + "grad_norm": 2.973244407105195, + "learning_rate": 1.0090082337192643e-06, + "loss": 0.7465, + "num_input_tokens_seen": 120783355, + "step": 5614 + }, + { + "epoch": 0.6751638309384957, + "flos": 23370853728960.0, + "grad_norm": 2.564880567897009, + "learning_rate": 1.0083316861169705e-06, + "loss": 0.7843, + "num_input_tokens_seen": 120802090, + "step": 5615 + }, + { + "epoch": 0.6752840738291348, + "flos": 23407917713280.0, + "grad_norm": 2.5312321038261074, + "learning_rate": 1.0076552889499713e-06, + "loss": 0.714, + "num_input_tokens_seen": 120822855, + "step": 5616 + }, + { + "epoch": 0.675404316719774, + "flos": 30292246199520.0, + "grad_norm": 2.381240376130128, + "learning_rate": 1.006979042320876e-06, + "loss": 0.741, + "num_input_tokens_seen": 120843070, + "step": 5617 + }, + { + "epoch": 0.675524559610413, + "flos": 23588814433440.0, + "grad_norm": 3.4019336475178727, + "learning_rate": 1.0063029463322702e-06, + "loss": 0.632, + "num_input_tokens_seen": 120863345, + "step": 5618 + }, + { + "epoch": 0.6756448025010521, + "flos": 21218200493760.0, + "grad_norm": 3.8330825778934368, + "learning_rate": 1.0056270010867164e-06, + "loss": 0.7564, + "num_input_tokens_seen": 120880915, + "step": 5619 + }, + { + "epoch": 0.6757650453916912, + "flos": 21615385279200.0, + "grad_norm": 3.7046362774805965, + "learning_rate": 1.004951206686758e-06, + "loss": 0.7865, + "num_input_tokens_seen": 120899190, + "step": 5620 + }, + { + "epoch": 0.6758852882823303, + "flos": 21764421781440.0, + "grad_norm": 2.280519655972216, + "learning_rate": 1.0042755632349087e-06, + "loss": 0.7158, + "num_input_tokens_seen": 120916080, + "step": 5621 + }, + { + "epoch": 0.6760055311729694, + "flos": 27049938120960.0, + "grad_norm": 3.5156580944622875, + "learning_rate": 1.0036000708336653e-06, + "loss": 0.6277, + "num_input_tokens_seen": 120935210, + "step": 5622 + }, + { + "epoch": 0.6761257740636085, + "flos": 17974033927200.0, + "grad_norm": 3.1590101806824356, + "learning_rate": 1.0029247295854984e-06, + "loss": 0.7984, + "num_input_tokens_seen": 120951830, + "step": 5623 + }, + { + "epoch": 0.6762460169542476, + "flos": 15100507203840.0, + "grad_norm": 2.671014301360307, + "learning_rate": 1.0022495395928588e-06, + "loss": 0.7186, + "num_input_tokens_seen": 120970310, + "step": 5624 + }, + { + "epoch": 0.6763662598448866, + "flos": 67790310838560.0, + "grad_norm": 0.7935577234787059, + "learning_rate": 1.0015745009581697e-06, + "loss": 0.6222, + "num_input_tokens_seen": 121031915, + "step": 5625 + }, + { + "epoch": 0.6764865027355258, + "flos": 20602423117920.0, + "grad_norm": 4.2223152072326995, + "learning_rate": 1.0008996137838343e-06, + "loss": 0.6697, + "num_input_tokens_seen": 121050890, + "step": 5626 + }, + { + "epoch": 0.6766067456261649, + "flos": 21185559710880.0, + "grad_norm": 2.486284302042853, + "learning_rate": 1.000224878172234e-06, + "loss": 0.7993, + "num_input_tokens_seen": 121070015, + "step": 5627 + }, + { + "epoch": 0.6767269885168039, + "flos": 19909841550720.0, + "grad_norm": 4.041662445590513, + "learning_rate": 9.99550294225724e-07, + "loss": 0.7289, + "num_input_tokens_seen": 121089170, + "step": 5628 + }, + { + "epoch": 0.6768472314074431, + "flos": 20784397761120.0, + "grad_norm": 2.8053492028503086, + "learning_rate": 9.988758620466402e-07, + "loss": 0.7253, + "num_input_tokens_seen": 121108040, + "step": 5629 + }, + { + "epoch": 0.6769674742980821, + "flos": 23152893024480.0, + "grad_norm": 2.0497025622933855, + "learning_rate": 9.982015817372917e-07, + "loss": 0.7601, + "num_input_tokens_seen": 121128480, + "step": 5630 + }, + { + "epoch": 0.6770877171887212, + "flos": 24208234445760.0, + "grad_norm": 2.8109811708409334, + "learning_rate": 9.975274533999657e-07, + "loss": 0.8228, + "num_input_tokens_seen": 121148010, + "step": 5631 + }, + { + "epoch": 0.6772079600793603, + "flos": 18115450628640.0, + "grad_norm": 3.293673873108487, + "learning_rate": 9.96853477136929e-07, + "loss": 0.8417, + "num_input_tokens_seen": 121162755, + "step": 5632 + }, + { + "epoch": 0.6773282029699994, + "flos": 22419865024800.0, + "grad_norm": 2.880564073848197, + "learning_rate": 9.96179653050422e-07, + "loss": 0.7521, + "num_input_tokens_seen": 121182915, + "step": 5633 + }, + { + "epoch": 0.6774484458606385, + "flos": 18666987192000.0, + "grad_norm": 4.544166413631371, + "learning_rate": 9.955059812426635e-07, + "loss": 0.7396, + "num_input_tokens_seen": 121200445, + "step": 5634 + }, + { + "epoch": 0.6775686887512776, + "flos": 25956789320160.0, + "grad_norm": 2.643013410109159, + "learning_rate": 9.948324618158493e-07, + "loss": 0.8288, + "num_input_tokens_seen": 121220785, + "step": 5635 + }, + { + "epoch": 0.6776889316419167, + "flos": 13568760771360.0, + "grad_norm": 4.450561428015137, + "learning_rate": 9.941590948721502e-07, + "loss": 0.7796, + "num_input_tokens_seen": 121237985, + "step": 5636 + }, + { + "epoch": 0.6778091745325557, + "flos": 27561585798240.0, + "grad_norm": 2.8468902352990044, + "learning_rate": 9.934858805137188e-07, + "loss": 0.7643, + "num_input_tokens_seen": 121258635, + "step": 5637 + }, + { + "epoch": 0.6779294174231949, + "flos": 18707396454720.0, + "grad_norm": 2.3605622594856936, + "learning_rate": 9.92812818842677e-07, + "loss": 0.812, + "num_input_tokens_seen": 121277810, + "step": 5638 + }, + { + "epoch": 0.678049660313834, + "flos": 45807413709600.0, + "grad_norm": 3.169857492222023, + "learning_rate": 9.921399099611306e-07, + "loss": 0.643, + "num_input_tokens_seen": 121298975, + "step": 5639 + }, + { + "epoch": 0.678169903204473, + "flos": 19946087800320.0, + "grad_norm": 2.7253173783290032, + "learning_rate": 9.914671539711588e-07, + "loss": 0.6928, + "num_input_tokens_seen": 121318495, + "step": 5640 + }, + { + "epoch": 0.6782901460951122, + "flos": 21365006810400.0, + "grad_norm": 3.4031490945094443, + "learning_rate": 9.90794550974817e-07, + "loss": 0.7862, + "num_input_tokens_seen": 121338890, + "step": 5641 + }, + { + "epoch": 0.6784103889857512, + "flos": 21403408906080.0, + "grad_norm": 4.738609107641652, + "learning_rate": 9.901221010741407e-07, + "loss": 0.8186, + "num_input_tokens_seen": 121358210, + "step": 5642 + }, + { + "epoch": 0.6785306318763903, + "flos": 32624978420160.0, + "grad_norm": 5.099194758554427, + "learning_rate": 9.894498043711375e-07, + "loss": 0.7476, + "num_input_tokens_seen": 121379955, + "step": 5643 + }, + { + "epoch": 0.6786508747670293, + "flos": 25595516256480.0, + "grad_norm": 3.8375986390996304, + "learning_rate": 9.887776609677962e-07, + "loss": 0.6945, + "num_input_tokens_seen": 121401325, + "step": 5644 + }, + { + "epoch": 0.6787711176576685, + "flos": 19144581635520.0, + "grad_norm": 2.310763557702192, + "learning_rate": 9.88105670966079e-07, + "loss": 0.7258, + "num_input_tokens_seen": 121419785, + "step": 5645 + }, + { + "epoch": 0.6788913605483076, + "flos": 13965202161600.0, + "grad_norm": 2.613568163347593, + "learning_rate": 9.874338344679283e-07, + "loss": 0.7917, + "num_input_tokens_seen": 121435630, + "step": 5646 + }, + { + "epoch": 0.6790116034389466, + "flos": 21986136631680.0, + "grad_norm": 2.214176891658874, + "learning_rate": 9.86762151575259e-07, + "loss": 0.7377, + "num_input_tokens_seen": 121455625, + "step": 5647 + }, + { + "epoch": 0.6791318463295858, + "flos": 20892578963520.0, + "grad_norm": 1.7855240924911466, + "learning_rate": 9.860906223899651e-07, + "loss": 0.8033, + "num_input_tokens_seen": 121475020, + "step": 5648 + }, + { + "epoch": 0.6792520892202248, + "flos": 28471867881600.0, + "grad_norm": 2.3100841378085955, + "learning_rate": 9.854192470139184e-07, + "loss": 0.7547, + "num_input_tokens_seen": 121496500, + "step": 5649 + }, + { + "epoch": 0.6793723321108639, + "flos": 19983411972960.0, + "grad_norm": 3.6488393688855636, + "learning_rate": 9.847480255489645e-07, + "loss": 0.7169, + "num_input_tokens_seen": 121515525, + "step": 5650 + }, + { + "epoch": 0.6794925750015031, + "flos": 26611228980000.0, + "grad_norm": 2.3594452298050808, + "learning_rate": 9.840769580969295e-07, + "loss": 0.6902, + "num_input_tokens_seen": 121535965, + "step": 5651 + }, + { + "epoch": 0.6796128178921421, + "flos": 21549583336800.0, + "grad_norm": 2.9285429079215484, + "learning_rate": 9.834060447596114e-07, + "loss": 0.7974, + "num_input_tokens_seen": 121555235, + "step": 5652 + }, + { + "epoch": 0.6797330607827812, + "flos": 22459865420160.0, + "grad_norm": 2.7824308481731803, + "learning_rate": 9.827352856387868e-07, + "loss": 0.7814, + "num_input_tokens_seen": 121574945, + "step": 5653 + }, + { + "epoch": 0.6798533036734203, + "flos": 66211873453920.0, + "grad_norm": 0.7892567612627558, + "learning_rate": 9.820646808362118e-07, + "loss": 0.6416, + "num_input_tokens_seen": 121641200, + "step": 5654 + }, + { + "epoch": 0.6799735465640594, + "flos": 16157223906240.0, + "grad_norm": 3.9104681814733424, + "learning_rate": 9.813942304536154e-07, + "loss": 0.7291, + "num_input_tokens_seen": 121659170, + "step": 5655 + }, + { + "epoch": 0.6800937894546984, + "flos": 22094094815520.0, + "grad_norm": 2.1958452440357434, + "learning_rate": 9.807239345927043e-07, + "loss": 0.6413, + "num_input_tokens_seen": 121679180, + "step": 5656 + }, + { + "epoch": 0.6802140323453376, + "flos": 31567518322560.0, + "grad_norm": 3.430175746496886, + "learning_rate": 9.80053793355162e-07, + "loss": 0.7217, + "num_input_tokens_seen": 121697875, + "step": 5657 + }, + { + "epoch": 0.6803342752359767, + "flos": 17687780906400.0, + "grad_norm": 2.3608360566638447, + "learning_rate": 9.793838068426472e-07, + "loss": 0.7486, + "num_input_tokens_seen": 121715570, + "step": 5658 + }, + { + "epoch": 0.6804545181266157, + "flos": 11310899914560.0, + "grad_norm": 2.80323392556468, + "learning_rate": 9.78713975156799e-07, + "loss": 0.6137, + "num_input_tokens_seen": 121731435, + "step": 5659 + }, + { + "epoch": 0.6805747610172549, + "flos": 29308839731040.0, + "grad_norm": 2.5359901032600676, + "learning_rate": 9.780442983992273e-07, + "loss": 0.715, + "num_input_tokens_seen": 121749950, + "step": 5660 + }, + { + "epoch": 0.680695003907894, + "flos": 37577625126240.0, + "grad_norm": 2.5436093040753263, + "learning_rate": 9.773747766715238e-07, + "loss": 0.7218, + "num_input_tokens_seen": 121770770, + "step": 5661 + }, + { + "epoch": 0.680815246798533, + "flos": 22095767454720.0, + "grad_norm": 2.3688555087544283, + "learning_rate": 9.767054100752536e-07, + "loss": 0.803, + "num_input_tokens_seen": 121790720, + "step": 5662 + }, + { + "epoch": 0.6809354896891722, + "flos": 17177397000960.0, + "grad_norm": 3.6453662827469318, + "learning_rate": 9.760361987119584e-07, + "loss": 0.8164, + "num_input_tokens_seen": 121808455, + "step": 5663 + }, + { + "epoch": 0.6810557325798112, + "flos": 12440629492800.0, + "grad_norm": 4.893410763729925, + "learning_rate": 9.753671426831592e-07, + "loss": 0.6787, + "num_input_tokens_seen": 121824470, + "step": 5664 + }, + { + "epoch": 0.6811759754704503, + "flos": 22124245224480.0, + "grad_norm": 2.696289150924709, + "learning_rate": 9.746982420903483e-07, + "loss": 0.7958, + "num_input_tokens_seen": 121842665, + "step": 5665 + }, + { + "epoch": 0.6812962183610894, + "flos": 17500305138720.0, + "grad_norm": 1.7971136183486085, + "learning_rate": 9.740294970349993e-07, + "loss": 0.7449, + "num_input_tokens_seen": 121859080, + "step": 5666 + }, + { + "epoch": 0.6814164612517285, + "flos": 60188051008800.0, + "grad_norm": 0.9416052263746844, + "learning_rate": 9.733609076185594e-07, + "loss": 0.6081, + "num_input_tokens_seen": 121915485, + "step": 5667 + }, + { + "epoch": 0.6815367041423676, + "flos": 19290235689600.0, + "grad_norm": 3.708744854899387, + "learning_rate": 9.72692473942455e-07, + "loss": 0.8416, + "num_input_tokens_seen": 121932710, + "step": 5668 + }, + { + "epoch": 0.6816569470330067, + "flos": 22129969367520.0, + "grad_norm": 2.2360416412092876, + "learning_rate": 9.720241961080849e-07, + "loss": 0.7786, + "num_input_tokens_seen": 121952025, + "step": 5669 + }, + { + "epoch": 0.6817771899236458, + "flos": 41403701683680.0, + "grad_norm": 3.407487836831453, + "learning_rate": 9.713560742168259e-07, + "loss": 0.7295, + "num_input_tokens_seen": 121974650, + "step": 5670 + }, + { + "epoch": 0.6818974328142848, + "flos": 21076040397120.0, + "grad_norm": 3.4342372321698003, + "learning_rate": 9.706881083700333e-07, + "loss": 0.7131, + "num_input_tokens_seen": 121994335, + "step": 5671 + }, + { + "epoch": 0.682017675704924, + "flos": 20412494146080.0, + "grad_norm": 2.9941782093613036, + "learning_rate": 9.700202986690357e-07, + "loss": 0.8248, + "num_input_tokens_seen": 122012635, + "step": 5672 + }, + { + "epoch": 0.682137918595563, + "flos": 20015755397760.0, + "grad_norm": 2.720122769053475, + "learning_rate": 9.693526452151413e-07, + "loss": 0.6671, + "num_input_tokens_seen": 122031685, + "step": 5673 + }, + { + "epoch": 0.6822581614862021, + "flos": 31638970068480.0, + "grad_norm": 2.016935292262756, + "learning_rate": 9.686851481096305e-07, + "loss": 0.7533, + "num_input_tokens_seen": 122052995, + "step": 5674 + }, + { + "epoch": 0.6823784043768413, + "flos": 23444015283840.0, + "grad_norm": 8.953637978868736, + "learning_rate": 9.68017807453762e-07, + "loss": 0.7209, + "num_input_tokens_seen": 122071740, + "step": 5675 + }, + { + "epoch": 0.6824986472674803, + "flos": 14117175074880.0, + "grad_norm": 2.968574759681959, + "learning_rate": 9.673506233487721e-07, + "loss": 0.7349, + "num_input_tokens_seen": 122089705, + "step": 5676 + }, + { + "epoch": 0.6826188901581194, + "flos": 21474228766080.0, + "grad_norm": 2.227349821394851, + "learning_rate": 9.666835958958717e-07, + "loss": 0.8615, + "num_input_tokens_seen": 122109025, + "step": 5677 + }, + { + "epoch": 0.6827391330487584, + "flos": 20780309087520.0, + "grad_norm": 2.427699511546987, + "learning_rate": 9.660167251962484e-07, + "loss": 0.8089, + "num_input_tokens_seen": 122127385, + "step": 5678 + }, + { + "epoch": 0.6828593759393976, + "flos": 21657801708960.0, + "grad_norm": 2.0493791710106417, + "learning_rate": 9.653500113510654e-07, + "loss": 0.7794, + "num_input_tokens_seen": 122146500, + "step": 5679 + }, + { + "epoch": 0.6829796188300367, + "flos": 25301420416320.0, + "grad_norm": 4.224855653001211, + "learning_rate": 9.646834544614627e-07, + "loss": 0.6737, + "num_input_tokens_seen": 122167000, + "step": 5680 + }, + { + "epoch": 0.6830998617206757, + "flos": 20675287314720.0, + "grad_norm": 2.621227459972068, + "learning_rate": 9.64017054628558e-07, + "loss": 0.7641, + "num_input_tokens_seen": 122185825, + "step": 5681 + }, + { + "epoch": 0.6832201046113149, + "flos": 21695646258240.0, + "grad_norm": 2.0911612377494055, + "learning_rate": 9.63350811953441e-07, + "loss": 0.7858, + "num_input_tokens_seen": 122206200, + "step": 5682 + }, + { + "epoch": 0.6833403475019539, + "flos": 19508716770720.0, + "grad_norm": 2.936282128636933, + "learning_rate": 9.626847265371826e-07, + "loss": 0.7065, + "num_input_tokens_seen": 122225520, + "step": 5683 + }, + { + "epoch": 0.683460590392593, + "flos": 19324511941920.0, + "grad_norm": 2.9809142813401652, + "learning_rate": 9.620187984808262e-07, + "loss": 0.787, + "num_input_tokens_seen": 122244320, + "step": 5684 + }, + { + "epoch": 0.6835808332832322, + "flos": 23254978386240.0, + "grad_norm": 2.2192312444023474, + "learning_rate": 9.613530278853919e-07, + "loss": 0.8589, + "num_input_tokens_seen": 122264295, + "step": 5685 + }, + { + "epoch": 0.6837010761738712, + "flos": 21622521873120.0, + "grad_norm": 5.734820023816476, + "learning_rate": 9.60687414851879e-07, + "loss": 0.7376, + "num_input_tokens_seen": 122285300, + "step": 5686 + }, + { + "epoch": 0.6838213190645103, + "flos": 17541420626880.0, + "grad_norm": 5.563201860017551, + "learning_rate": 9.600219594812575e-07, + "loss": 0.7745, + "num_input_tokens_seen": 122303240, + "step": 5687 + }, + { + "epoch": 0.6839415619551494, + "flos": 23079508451040.0, + "grad_norm": 2.2649318292392544, + "learning_rate": 9.593566618744786e-07, + "loss": 0.7302, + "num_input_tokens_seen": 122323390, + "step": 5688 + }, + { + "epoch": 0.6840618048457885, + "flos": 22096287831360.0, + "grad_norm": 2.3216676190719605, + "learning_rate": 9.58691522132466e-07, + "loss": 0.7445, + "num_input_tokens_seen": 122342200, + "step": 5689 + }, + { + "epoch": 0.6841820477364275, + "flos": 21984389652960.0, + "grad_norm": 2.5148115086688505, + "learning_rate": 9.58026540356123e-07, + "loss": 0.8477, + "num_input_tokens_seen": 122360465, + "step": 5690 + }, + { + "epoch": 0.6843022906270667, + "flos": 24864904291200.0, + "grad_norm": 2.7893648384416863, + "learning_rate": 9.573617166463246e-07, + "loss": 0.8658, + "num_input_tokens_seen": 122381680, + "step": 5691 + }, + { + "epoch": 0.6844225335177058, + "flos": 19941292901280.0, + "grad_norm": 2.909220500248715, + "learning_rate": 9.56697051103924e-07, + "loss": 0.6034, + "num_input_tokens_seen": 122399120, + "step": 5692 + }, + { + "epoch": 0.6845427764083448, + "flos": 25846192083360.0, + "grad_norm": 3.275411879869169, + "learning_rate": 9.560325438297522e-07, + "loss": 0.8165, + "num_input_tokens_seen": 122417430, + "step": 5693 + }, + { + "epoch": 0.684663019298984, + "flos": 18853273527360.0, + "grad_norm": 3.3610895084355934, + "learning_rate": 9.553681949246127e-07, + "loss": 0.8723, + "num_input_tokens_seen": 122435055, + "step": 5694 + }, + { + "epoch": 0.684783262189623, + "flos": 54115270256160.0, + "grad_norm": 2.622702819574392, + "learning_rate": 9.547040044892886e-07, + "loss": 0.7523, + "num_input_tokens_seen": 122462570, + "step": 5695 + }, + { + "epoch": 0.6849035050802621, + "flos": 63878843875200.0, + "grad_norm": 0.8755774295690506, + "learning_rate": 9.540399726245354e-07, + "loss": 0.5995, + "num_input_tokens_seen": 122519275, + "step": 5696 + }, + { + "epoch": 0.6850237479709013, + "flos": 25187849598720.0, + "grad_norm": 2.11526300440863, + "learning_rate": 9.533760994310859e-07, + "loss": 0.6916, + "num_input_tokens_seen": 122539675, + "step": 5697 + }, + { + "epoch": 0.6851439908615403, + "flos": 19326556278720.0, + "grad_norm": 2.608249992851053, + "learning_rate": 9.527123850096508e-07, + "loss": 0.7509, + "num_input_tokens_seen": 122558035, + "step": 5698 + }, + { + "epoch": 0.6852642337521794, + "flos": 23148878690400.0, + "grad_norm": 3.2292585954604163, + "learning_rate": 9.520488294609142e-07, + "loss": 0.7178, + "num_input_tokens_seen": 122576815, + "step": 5699 + }, + { + "epoch": 0.6853844766428185, + "flos": 62557364006880.0, + "grad_norm": 0.7482673180765365, + "learning_rate": 9.513854328855368e-07, + "loss": 0.5376, + "num_input_tokens_seen": 122634690, + "step": 5700 + }, + { + "epoch": 0.6855047195334576, + "flos": 23403606021120.0, + "grad_norm": 2.5906600931365285, + "learning_rate": 9.507221953841558e-07, + "loss": 0.8143, + "num_input_tokens_seen": 122652320, + "step": 5701 + }, + { + "epoch": 0.6856249624240967, + "flos": 20635249749600.0, + "grad_norm": 4.64182931191593, + "learning_rate": 9.500591170573824e-07, + "loss": 0.7793, + "num_input_tokens_seen": 122672340, + "step": 5702 + }, + { + "epoch": 0.6857452053147358, + "flos": 17062859769600.0, + "grad_norm": 2.2312726444409625, + "learning_rate": 9.493961980058078e-07, + "loss": 0.7453, + "num_input_tokens_seen": 122689935, + "step": 5703 + }, + { + "epoch": 0.6858654482053749, + "flos": 30802964632800.0, + "grad_norm": 2.5522679061059743, + "learning_rate": 9.48733438329993e-07, + "loss": 0.6726, + "num_input_tokens_seen": 122710200, + "step": 5704 + }, + { + "epoch": 0.6859856910960139, + "flos": 28835891507520.0, + "grad_norm": 2.4704749925556366, + "learning_rate": 9.480708381304807e-07, + "loss": 0.7442, + "num_input_tokens_seen": 122731495, + "step": 5705 + }, + { + "epoch": 0.6861059339866531, + "flos": 19326965146080.0, + "grad_norm": 4.166446425657132, + "learning_rate": 9.474083975077858e-07, + "loss": 0.8408, + "num_input_tokens_seen": 122748620, + "step": 5706 + }, + { + "epoch": 0.6862261768772921, + "flos": 22168259953920.0, + "grad_norm": 3.8054128343189797, + "learning_rate": 9.467461165623994e-07, + "loss": 0.8014, + "num_input_tokens_seen": 122767670, + "step": 5707 + }, + { + "epoch": 0.6863464197679312, + "flos": 26247985919040.0, + "grad_norm": 3.1640305505717183, + "learning_rate": 9.46083995394791e-07, + "loss": 0.7955, + "num_input_tokens_seen": 122785480, + "step": 5708 + }, + { + "epoch": 0.6864666626585703, + "flos": 37760752032000.0, + "grad_norm": 3.0085120242263534, + "learning_rate": 9.454220341054012e-07, + "loss": 0.637, + "num_input_tokens_seen": 122810265, + "step": 5709 + }, + { + "epoch": 0.6865869055492094, + "flos": 19363545923520.0, + "grad_norm": 2.9661338120446605, + "learning_rate": 9.447602327946512e-07, + "loss": 0.811, + "num_input_tokens_seen": 122828905, + "step": 5710 + }, + { + "epoch": 0.6867071484398485, + "flos": 20346915222240.0, + "grad_norm": 2.642632438062046, + "learning_rate": 9.440985915629338e-07, + "loss": 0.7627, + "num_input_tokens_seen": 122846235, + "step": 5711 + }, + { + "epoch": 0.6868273913304875, + "flos": 15866659193280.0, + "grad_norm": 2.3211962112004048, + "learning_rate": 9.434371105106223e-07, + "loss": 0.7343, + "num_input_tokens_seen": 122863835, + "step": 5712 + }, + { + "epoch": 0.6869476342211267, + "flos": 24427496091840.0, + "grad_norm": 2.2304048278521003, + "learning_rate": 9.427757897380602e-07, + "loss": 0.7101, + "num_input_tokens_seen": 122883235, + "step": 5713 + }, + { + "epoch": 0.6870678771117658, + "flos": 18416348534880.0, + "grad_norm": 3.1527672616687075, + "learning_rate": 9.421146293455695e-07, + "loss": 0.8496, + "num_input_tokens_seen": 122898975, + "step": 5714 + }, + { + "epoch": 0.6871881200024048, + "flos": 22168668821280.0, + "grad_norm": 8.508865221256922, + "learning_rate": 9.414536294334489e-07, + "loss": 0.6886, + "num_input_tokens_seen": 122918995, + "step": 5715 + }, + { + "epoch": 0.687308362893044, + "flos": 22096139152320.0, + "grad_norm": 2.834713842626814, + "learning_rate": 9.407927901019708e-07, + "loss": 0.6976, + "num_input_tokens_seen": 122938125, + "step": 5716 + }, + { + "epoch": 0.687428605783683, + "flos": 25004499674400.0, + "grad_norm": 2.5830480561609788, + "learning_rate": 9.401321114513854e-07, + "loss": 0.767, + "num_input_tokens_seen": 122957020, + "step": 5717 + }, + { + "epoch": 0.6875488486743221, + "flos": 23736103956960.0, + "grad_norm": 2.0817503319148263, + "learning_rate": 9.394715935819155e-07, + "loss": 0.7551, + "num_input_tokens_seen": 122977410, + "step": 5718 + }, + { + "epoch": 0.6876690915649613, + "flos": 25480272799680.0, + "grad_norm": 2.7084528285343556, + "learning_rate": 9.388112365937608e-07, + "loss": 0.626, + "num_input_tokens_seen": 122996875, + "step": 5719 + }, + { + "epoch": 0.6877893344556003, + "flos": 19400349719520.0, + "grad_norm": 2.832639852439835, + "learning_rate": 9.381510405870985e-07, + "loss": 0.8292, + "num_input_tokens_seen": 123015890, + "step": 5720 + }, + { + "epoch": 0.6879095773462394, + "flos": 18634755276480.0, + "grad_norm": 6.6228991362355805, + "learning_rate": 9.374910056620791e-07, + "loss": 0.7746, + "num_input_tokens_seen": 123034955, + "step": 5721 + }, + { + "epoch": 0.6880298202368785, + "flos": 20853210454080.0, + "grad_norm": 2.233649929817152, + "learning_rate": 9.368311319188293e-07, + "loss": 0.8138, + "num_input_tokens_seen": 123052770, + "step": 5722 + }, + { + "epoch": 0.6881500631275176, + "flos": 30109974198240.0, + "grad_norm": 2.363474839533553, + "learning_rate": 9.361714194574515e-07, + "loss": 0.7955, + "num_input_tokens_seen": 123075105, + "step": 5723 + }, + { + "epoch": 0.6882703060181566, + "flos": 66087703521120.0, + "grad_norm": 0.7467042113200069, + "learning_rate": 9.355118683780228e-07, + "loss": 0.5821, + "num_input_tokens_seen": 123145175, + "step": 5724 + }, + { + "epoch": 0.6883905489087958, + "flos": 18188426334720.0, + "grad_norm": 2.5895995057987924, + "learning_rate": 9.348524787805987e-07, + "loss": 0.7896, + "num_input_tokens_seen": 123160365, + "step": 5725 + }, + { + "epoch": 0.6885107917994349, + "flos": 14036133530880.0, + "grad_norm": 4.807878926103111, + "learning_rate": 9.341932507652053e-07, + "loss": 0.8511, + "num_input_tokens_seen": 123174855, + "step": 5726 + }, + { + "epoch": 0.6886310346900739, + "flos": 28649419323360.0, + "grad_norm": 3.5622012364443676, + "learning_rate": 9.335341844318489e-07, + "loss": 0.787, + "num_input_tokens_seen": 123194995, + "step": 5727 + }, + { + "epoch": 0.6887512775807131, + "flos": 24500211609600.0, + "grad_norm": 2.121979530242933, + "learning_rate": 9.328752798805091e-07, + "loss": 0.7318, + "num_input_tokens_seen": 123213465, + "step": 5728 + }, + { + "epoch": 0.6888715204713521, + "flos": 22382169154560.0, + "grad_norm": 2.806960690909743, + "learning_rate": 9.322165372111399e-07, + "loss": 0.7624, + "num_input_tokens_seen": 123231610, + "step": 5729 + }, + { + "epoch": 0.6889917633619912, + "flos": 22022828918400.0, + "grad_norm": 2.4007181761997063, + "learning_rate": 9.315579565236747e-07, + "loss": 0.7553, + "num_input_tokens_seen": 123250715, + "step": 5730 + }, + { + "epoch": 0.6891120062526304, + "flos": 23915402377440.0, + "grad_norm": 2.1790870194865724, + "learning_rate": 9.308995379180162e-07, + "loss": 0.7474, + "num_input_tokens_seen": 123270270, + "step": 5731 + }, + { + "epoch": 0.6892322491432694, + "flos": 64025761701120.0, + "grad_norm": 0.7809106089590043, + "learning_rate": 9.302412814940488e-07, + "loss": 0.5943, + "num_input_tokens_seen": 123333045, + "step": 5732 + }, + { + "epoch": 0.6893524920339085, + "flos": 22969543100160.0, + "grad_norm": 2.8472734358815552, + "learning_rate": 9.295831873516276e-07, + "loss": 0.7107, + "num_input_tokens_seen": 123352320, + "step": 5733 + }, + { + "epoch": 0.6894727349245476, + "flos": 21365675866080.0, + "grad_norm": 1.999758109723775, + "learning_rate": 9.289252555905873e-07, + "loss": 0.7603, + "num_input_tokens_seen": 123372400, + "step": 5734 + }, + { + "epoch": 0.6895929778151867, + "flos": 19837014523680.0, + "grad_norm": 3.6965002364957433, + "learning_rate": 9.282674863107334e-07, + "loss": 0.763, + "num_input_tokens_seen": 123390215, + "step": 5735 + }, + { + "epoch": 0.6897132207058257, + "flos": 18153034989600.0, + "grad_norm": 3.9687629759568877, + "learning_rate": 9.276098796118488e-07, + "loss": 0.7646, + "num_input_tokens_seen": 123406700, + "step": 5736 + }, + { + "epoch": 0.6898334635964649, + "flos": 32514827220480.0, + "grad_norm": 2.542789669733695, + "learning_rate": 9.269524355936938e-07, + "loss": 0.6592, + "num_input_tokens_seen": 123429880, + "step": 5737 + }, + { + "epoch": 0.689953706487104, + "flos": 22786341854880.0, + "grad_norm": 2.624890643067159, + "learning_rate": 9.262951543560002e-07, + "loss": 0.85, + "num_input_tokens_seen": 123449105, + "step": 5738 + }, + { + "epoch": 0.690073949377743, + "flos": 18489026882880.0, + "grad_norm": 6.960650334284281, + "learning_rate": 9.256380359984795e-07, + "loss": 0.8701, + "num_input_tokens_seen": 123466215, + "step": 5739 + }, + { + "epoch": 0.6901941922683821, + "flos": 34806927159840.0, + "grad_norm": 2.865106227429972, + "learning_rate": 9.249810806208139e-07, + "loss": 0.7499, + "num_input_tokens_seen": 123485480, + "step": 5740 + }, + { + "epoch": 0.6903144351590212, + "flos": 16230645649440.0, + "grad_norm": 2.4241857518449863, + "learning_rate": 9.243242883226627e-07, + "loss": 0.8043, + "num_input_tokens_seen": 123504575, + "step": 5741 + }, + { + "epoch": 0.6904346780496603, + "flos": 27994830984480.0, + "grad_norm": 2.9377899026868994, + "learning_rate": 9.236676592036628e-07, + "loss": 0.7006, + "num_input_tokens_seen": 123524250, + "step": 5742 + }, + { + "epoch": 0.6905549209402994, + "flos": 23590524242400.0, + "grad_norm": 2.0990515247289183, + "learning_rate": 9.230111933634228e-07, + "loss": 0.7372, + "num_input_tokens_seen": 123545845, + "step": 5743 + }, + { + "epoch": 0.6906751638309385, + "flos": 23081998824960.0, + "grad_norm": 1.8488750027069882, + "learning_rate": 9.223548909015288e-07, + "loss": 0.8104, + "num_input_tokens_seen": 123567535, + "step": 5744 + }, + { + "epoch": 0.6907954067215776, + "flos": 27266226186240.0, + "grad_norm": 3.5993776790173735, + "learning_rate": 9.216987519175407e-07, + "loss": 0.7218, + "num_input_tokens_seen": 123587710, + "step": 5745 + }, + { + "epoch": 0.6909156496122166, + "flos": 21658433594880.0, + "grad_norm": 2.058443183387241, + "learning_rate": 9.210427765109942e-07, + "loss": 0.6818, + "num_input_tokens_seen": 123607385, + "step": 5746 + }, + { + "epoch": 0.6910358925028558, + "flos": 22529086980480.0, + "grad_norm": 2.959881838424724, + "learning_rate": 9.20386964781402e-07, + "loss": 0.816, + "num_input_tokens_seen": 123625405, + "step": 5747 + }, + { + "epoch": 0.6911561353934949, + "flos": 22022605899840.0, + "grad_norm": 2.9676841026987093, + "learning_rate": 9.197313168282472e-07, + "loss": 0.8416, + "num_input_tokens_seen": 123642850, + "step": 5748 + }, + { + "epoch": 0.6912763782841339, + "flos": 24171913856640.0, + "grad_norm": 4.257049100141508, + "learning_rate": 9.190758327509935e-07, + "loss": 0.7245, + "num_input_tokens_seen": 123661595, + "step": 5749 + }, + { + "epoch": 0.6913966211747731, + "flos": 52255108827840.0, + "grad_norm": 0.9316396333381285, + "learning_rate": 9.184205126490767e-07, + "loss": 0.6434, + "num_input_tokens_seen": 123710490, + "step": 5750 + }, + { + "epoch": 0.6915168640654121, + "flos": 66645750526080.0, + "grad_norm": 1.135859895807466, + "learning_rate": 9.177653566219075e-07, + "loss": 0.5956, + "num_input_tokens_seen": 123765215, + "step": 5751 + }, + { + "epoch": 0.6916371069560512, + "flos": 18270508632000.0, + "grad_norm": 2.9080037393873854, + "learning_rate": 9.171103647688744e-07, + "loss": 0.7604, + "num_input_tokens_seen": 123783955, + "step": 5752 + }, + { + "epoch": 0.6917573498466904, + "flos": 19617790047360.0, + "grad_norm": 2.422025319018275, + "learning_rate": 9.164555371893367e-07, + "loss": 0.6907, + "num_input_tokens_seen": 123803885, + "step": 5753 + }, + { + "epoch": 0.6918775927373294, + "flos": 14190262290240.0, + "grad_norm": 4.904956039558357, + "learning_rate": 9.158008739826333e-07, + "loss": 0.7558, + "num_input_tokens_seen": 123821485, + "step": 5754 + }, + { + "epoch": 0.6919978356279685, + "flos": 23951053910880.0, + "grad_norm": 2.257948399450009, + "learning_rate": 9.151463752480744e-07, + "loss": 0.8673, + "num_input_tokens_seen": 123840215, + "step": 5755 + }, + { + "epoch": 0.6921180785186076, + "flos": 23589148961280.0, + "grad_norm": 2.5479122928094218, + "learning_rate": 9.144920410849493e-07, + "loss": 0.8055, + "num_input_tokens_seen": 123861450, + "step": 5756 + }, + { + "epoch": 0.6922383214092467, + "flos": 21112472495520.0, + "grad_norm": 2.432555146959523, + "learning_rate": 9.138378715925176e-07, + "loss": 0.8069, + "num_input_tokens_seen": 123880545, + "step": 5757 + }, + { + "epoch": 0.6923585642998857, + "flos": 21440026853280.0, + "grad_norm": 2.1855703289556434, + "learning_rate": 9.131838668700167e-07, + "loss": 0.8086, + "num_input_tokens_seen": 123900615, + "step": 5758 + }, + { + "epoch": 0.6924788071905249, + "flos": 21075073983360.0, + "grad_norm": 2.6251731088870187, + "learning_rate": 9.125300270166598e-07, + "loss": 0.864, + "num_input_tokens_seen": 123921220, + "step": 5759 + }, + { + "epoch": 0.692599050081164, + "flos": 26212408725120.0, + "grad_norm": 2.349833210646851, + "learning_rate": 9.118763521316324e-07, + "loss": 0.8576, + "num_input_tokens_seen": 123941795, + "step": 5760 + }, + { + "epoch": 0.692719292971803, + "flos": 20855366300160.0, + "grad_norm": 1.9368216034482846, + "learning_rate": 9.112228423140987e-07, + "loss": 0.76, + "num_input_tokens_seen": 123960715, + "step": 5761 + }, + { + "epoch": 0.6928395358624422, + "flos": 25884817197600.0, + "grad_norm": 4.938081344954209, + "learning_rate": 9.105694976631932e-07, + "loss": 0.8603, + "num_input_tokens_seen": 123978625, + "step": 5762 + }, + { + "epoch": 0.6929597787530812, + "flos": 23552828372160.0, + "grad_norm": 3.6532320788429535, + "learning_rate": 9.099163182780283e-07, + "loss": 0.7269, + "num_input_tokens_seen": 123996780, + "step": 5763 + }, + { + "epoch": 0.6930800216437203, + "flos": 18229541822880.0, + "grad_norm": 3.5785598157526213, + "learning_rate": 9.092633042576916e-07, + "loss": 0.4955, + "num_input_tokens_seen": 124014045, + "step": 5764 + }, + { + "epoch": 0.6932002645343595, + "flos": 29126753578560.0, + "grad_norm": 2.368388459890425, + "learning_rate": 9.086104557012446e-07, + "loss": 0.5662, + "num_input_tokens_seen": 124034615, + "step": 5765 + }, + { + "epoch": 0.6933205074249985, + "flos": 23808670795680.0, + "grad_norm": 2.9240248008125036, + "learning_rate": 9.079577727077239e-07, + "loss": 0.6606, + "num_input_tokens_seen": 124053445, + "step": 5766 + }, + { + "epoch": 0.6934407503156376, + "flos": 24131987800800.0, + "grad_norm": 2.631017236993133, + "learning_rate": 9.073052553761404e-07, + "loss": 0.7229, + "num_input_tokens_seen": 124072810, + "step": 5767 + }, + { + "epoch": 0.6935609932062767, + "flos": 20601902741280.0, + "grad_norm": 1.974249953282638, + "learning_rate": 9.066529038054805e-07, + "loss": 0.7836, + "num_input_tokens_seen": 124092875, + "step": 5768 + }, + { + "epoch": 0.6936812360969158, + "flos": 18228017862720.0, + "grad_norm": 2.6712781451415606, + "learning_rate": 9.060007180947071e-07, + "loss": 0.7417, + "num_input_tokens_seen": 124110930, + "step": 5769 + }, + { + "epoch": 0.6938014789875548, + "flos": 31272195880320.0, + "grad_norm": 2.6445417981522, + "learning_rate": 9.053486983427534e-07, + "loss": 0.7323, + "num_input_tokens_seen": 124132180, + "step": 5770 + }, + { + "epoch": 0.6939217218781939, + "flos": 17503464568320.0, + "grad_norm": 3.8121202623704553, + "learning_rate": 9.046968446485326e-07, + "loss": 0.7075, + "num_input_tokens_seen": 124150585, + "step": 5771 + }, + { + "epoch": 0.6940419647688331, + "flos": 18525384641760.0, + "grad_norm": 3.7710790115046304, + "learning_rate": 9.040451571109295e-07, + "loss": 0.7022, + "num_input_tokens_seen": 124166205, + "step": 5772 + }, + { + "epoch": 0.6941622076594721, + "flos": 66830587240800.0, + "grad_norm": 0.8823195575027332, + "learning_rate": 9.033936358288042e-07, + "loss": 0.6031, + "num_input_tokens_seen": 124219940, + "step": 5773 + }, + { + "epoch": 0.6942824505501112, + "flos": 26539963082880.0, + "grad_norm": 2.9771506862260253, + "learning_rate": 9.027422809009937e-07, + "loss": 0.8245, + "num_input_tokens_seen": 124239885, + "step": 5774 + }, + { + "epoch": 0.6944026934407503, + "flos": 21217865965920.0, + "grad_norm": 2.6536725273449346, + "learning_rate": 9.020910924263054e-07, + "loss": 0.8319, + "num_input_tokens_seen": 124258410, + "step": 5775 + }, + { + "epoch": 0.6945229363313894, + "flos": 70575770933280.0, + "grad_norm": 0.8228077313393269, + "learning_rate": 9.014400705035261e-07, + "loss": 0.5809, + "num_input_tokens_seen": 124315315, + "step": 5776 + }, + { + "epoch": 0.6946431792220285, + "flos": 18925951875360.0, + "grad_norm": 2.7180866002089323, + "learning_rate": 9.00789215231414e-07, + "loss": 0.7698, + "num_input_tokens_seen": 124333710, + "step": 5777 + }, + { + "epoch": 0.6947634221126676, + "flos": 20309814068160.0, + "grad_norm": 2.8294295363482136, + "learning_rate": 9.001385267087056e-07, + "loss": 0.822, + "num_input_tokens_seen": 124352855, + "step": 5778 + }, + { + "epoch": 0.6948836650033067, + "flos": 21802489349280.0, + "grad_norm": 2.0718600265399143, + "learning_rate": 8.994880050341072e-07, + "loss": 0.7028, + "num_input_tokens_seen": 124372960, + "step": 5779 + }, + { + "epoch": 0.6950039078939457, + "flos": 23623611062400.0, + "grad_norm": 2.24954833689604, + "learning_rate": 8.988376503063026e-07, + "loss": 0.7705, + "num_input_tokens_seen": 124394220, + "step": 5780 + }, + { + "epoch": 0.6951241507845849, + "flos": 21761113672800.0, + "grad_norm": 2.9413373516063386, + "learning_rate": 8.981874626239521e-07, + "loss": 0.8169, + "num_input_tokens_seen": 124412795, + "step": 5781 + }, + { + "epoch": 0.695244393675224, + "flos": 14626852754880.0, + "grad_norm": 2.863207640488847, + "learning_rate": 8.975374420856872e-07, + "loss": 0.8836, + "num_input_tokens_seen": 124429690, + "step": 5782 + }, + { + "epoch": 0.695364636565863, + "flos": 16849247927040.0, + "grad_norm": 2.6541787542989126, + "learning_rate": 8.968875887901157e-07, + "loss": 0.7249, + "num_input_tokens_seen": 124447070, + "step": 5783 + }, + { + "epoch": 0.6954848794565022, + "flos": 19326965146080.0, + "grad_norm": 2.671650124171638, + "learning_rate": 8.9623790283582e-07, + "loss": 0.631, + "num_input_tokens_seen": 124464950, + "step": 5784 + }, + { + "epoch": 0.6956051223471412, + "flos": 18963201708480.0, + "grad_norm": 5.303363797819407, + "learning_rate": 8.955883843213561e-07, + "loss": 0.7668, + "num_input_tokens_seen": 124483965, + "step": 5785 + }, + { + "epoch": 0.6957253652377803, + "flos": 16084396879200.0, + "grad_norm": 2.6445404458273463, + "learning_rate": 8.949390333452569e-07, + "loss": 0.8696, + "num_input_tokens_seen": 124501910, + "step": 5786 + }, + { + "epoch": 0.6958456081284194, + "flos": 29346498431520.0, + "grad_norm": 2.0671452316841483, + "learning_rate": 8.942898500060279e-07, + "loss": 0.6792, + "num_input_tokens_seen": 124521625, + "step": 5787 + }, + { + "epoch": 0.6959658510190585, + "flos": 25118813887200.0, + "grad_norm": 6.135307203092471, + "learning_rate": 8.936408344021493e-07, + "loss": 0.7189, + "num_input_tokens_seen": 124538540, + "step": 5788 + }, + { + "epoch": 0.6960860939096976, + "flos": 42752655738240.0, + "grad_norm": 2.7872171574945908, + "learning_rate": 8.929919866320765e-07, + "loss": 0.7149, + "num_input_tokens_seen": 124559355, + "step": 5789 + }, + { + "epoch": 0.6962063368003367, + "flos": 17541606475680.0, + "grad_norm": 3.0605392407271568, + "learning_rate": 8.923433067942385e-07, + "loss": 0.8133, + "num_input_tokens_seen": 124577920, + "step": 5790 + }, + { + "epoch": 0.6963265796909758, + "flos": 21221322753600.0, + "grad_norm": 2.2999040003105633, + "learning_rate": 8.916947949870417e-07, + "loss": 0.6855, + "num_input_tokens_seen": 124597585, + "step": 5791 + }, + { + "epoch": 0.6964468225816148, + "flos": 68729999935680.0, + "grad_norm": 0.7428063472866915, + "learning_rate": 8.910464513088615e-07, + "loss": 0.5802, + "num_input_tokens_seen": 124661625, + "step": 5792 + }, + { + "epoch": 0.696567065472254, + "flos": 18923833199040.0, + "grad_norm": 2.4569516798274313, + "learning_rate": 8.903982758580542e-07, + "loss": 0.7891, + "num_input_tokens_seen": 124680565, + "step": 5793 + }, + { + "epoch": 0.696687308362893, + "flos": 22824000555360.0, + "grad_norm": 2.8926798347348672, + "learning_rate": 8.897502687329457e-07, + "loss": 0.8067, + "num_input_tokens_seen": 124700365, + "step": 5794 + }, + { + "epoch": 0.6968075512535321, + "flos": 24937173771840.0, + "grad_norm": 2.8010229115527956, + "learning_rate": 8.891024300318382e-07, + "loss": 0.8, + "num_input_tokens_seen": 124718935, + "step": 5795 + }, + { + "epoch": 0.6969277941441713, + "flos": 21000351298560.0, + "grad_norm": 1.714894287476948, + "learning_rate": 8.884547598530103e-07, + "loss": 0.7632, + "num_input_tokens_seen": 124739505, + "step": 5796 + }, + { + "epoch": 0.6970480370348103, + "flos": 21548579753280.0, + "grad_norm": 2.276263705552227, + "learning_rate": 8.8780725829471e-07, + "loss": 0.7535, + "num_input_tokens_seen": 124757410, + "step": 5797 + }, + { + "epoch": 0.6971682799254494, + "flos": 22387149902400.0, + "grad_norm": 2.732802631061547, + "learning_rate": 8.87159925455165e-07, + "loss": 0.7791, + "num_input_tokens_seen": 124777240, + "step": 5798 + }, + { + "epoch": 0.6972885228160886, + "flos": 19977130283520.0, + "grad_norm": 3.3966870941156815, + "learning_rate": 8.865127614325738e-07, + "loss": 0.7343, + "num_input_tokens_seen": 124795670, + "step": 5799 + }, + { + "epoch": 0.6974087657067276, + "flos": 37798968278880.0, + "grad_norm": 2.4100965000950705, + "learning_rate": 8.85865766325113e-07, + "loss": 0.6661, + "num_input_tokens_seen": 124819600, + "step": 5800 + }, + { + "epoch": 0.6975290085973667, + "flos": 29446650965760.0, + "grad_norm": 3.5377070713432217, + "learning_rate": 8.852189402309287e-07, + "loss": 0.7197, + "num_input_tokens_seen": 124838785, + "step": 5801 + }, + { + "epoch": 0.6976492514880057, + "flos": 12877442976000.0, + "grad_norm": 3.335102628990364, + "learning_rate": 8.845722832481441e-07, + "loss": 0.7433, + "num_input_tokens_seen": 124854215, + "step": 5802 + }, + { + "epoch": 0.6977694943786449, + "flos": 24318088287360.0, + "grad_norm": 2.605498764116854, + "learning_rate": 8.83925795474858e-07, + "loss": 0.7752, + "num_input_tokens_seen": 124874340, + "step": 5803 + }, + { + "epoch": 0.6978897372692839, + "flos": 29855544225600.0, + "grad_norm": 2.9331124186413624, + "learning_rate": 8.832794770091414e-07, + "loss": 0.5921, + "num_input_tokens_seen": 124895175, + "step": 5804 + }, + { + "epoch": 0.698009980159923, + "flos": 21730480056960.0, + "grad_norm": 2.6674733630551293, + "learning_rate": 8.826333279490401e-07, + "loss": 0.828, + "num_input_tokens_seen": 124915810, + "step": 5805 + }, + { + "epoch": 0.6981302230505622, + "flos": 19828576988160.0, + "grad_norm": 2.7400732339692397, + "learning_rate": 8.819873483925748e-07, + "loss": 0.6817, + "num_input_tokens_seen": 124932930, + "step": 5806 + }, + { + "epoch": 0.6982504659412012, + "flos": 22166884672800.0, + "grad_norm": 2.9192808334265514, + "learning_rate": 8.81341538437739e-07, + "loss": 0.7434, + "num_input_tokens_seen": 124951220, + "step": 5807 + }, + { + "epoch": 0.6983707088318403, + "flos": 35537204597280.0, + "grad_norm": 2.7696559861774888, + "learning_rate": 8.80695898182503e-07, + "loss": 0.687, + "num_input_tokens_seen": 124972200, + "step": 5808 + }, + { + "epoch": 0.6984909517224794, + "flos": 65346869871840.0, + "grad_norm": 0.8656969061226126, + "learning_rate": 8.800504277248093e-07, + "loss": 0.6496, + "num_input_tokens_seen": 125036950, + "step": 5809 + }, + { + "epoch": 0.6986111946131185, + "flos": 18520366724160.0, + "grad_norm": 2.1842247557917376, + "learning_rate": 8.794051271625753e-07, + "loss": 0.7574, + "num_input_tokens_seen": 125054585, + "step": 5810 + }, + { + "epoch": 0.6987314375037575, + "flos": 23006235386880.0, + "grad_norm": 2.8538718709433537, + "learning_rate": 8.787599965936925e-07, + "loss": 0.8351, + "num_input_tokens_seen": 125075515, + "step": 5811 + }, + { + "epoch": 0.6988516803943967, + "flos": 38345115227040.0, + "grad_norm": 5.932504901843773, + "learning_rate": 8.781150361160261e-07, + "loss": 0.7219, + "num_input_tokens_seen": 125097425, + "step": 5812 + }, + { + "epoch": 0.6989719232850358, + "flos": 24062506052160.0, + "grad_norm": 2.0960211890219766, + "learning_rate": 8.774702458274181e-07, + "loss": 0.7413, + "num_input_tokens_seen": 125117830, + "step": 5813 + }, + { + "epoch": 0.6990921661756748, + "flos": 14845817042880.0, + "grad_norm": 3.137443031310567, + "learning_rate": 8.768256258256799e-07, + "loss": 0.7095, + "num_input_tokens_seen": 125134455, + "step": 5814 + }, + { + "epoch": 0.699212409066314, + "flos": 20164940579040.0, + "grad_norm": 2.565383777127741, + "learning_rate": 8.76181176208602e-07, + "loss": 0.7384, + "num_input_tokens_seen": 125152555, + "step": 5815 + }, + { + "epoch": 0.699332651956953, + "flos": 19400238210240.0, + "grad_norm": 2.580231332092496, + "learning_rate": 8.755368970739461e-07, + "loss": 0.7326, + "num_input_tokens_seen": 125171915, + "step": 5816 + }, + { + "epoch": 0.6994528948475921, + "flos": 16120828977600.0, + "grad_norm": 3.3898881384024087, + "learning_rate": 8.748927885194479e-07, + "loss": 0.6169, + "num_input_tokens_seen": 125190220, + "step": 5817 + }, + { + "epoch": 0.6995731377382313, + "flos": 64859462611680.0, + "grad_norm": 0.7972329401683267, + "learning_rate": 8.742488506428209e-07, + "loss": 0.5727, + "num_input_tokens_seen": 125249310, + "step": 5818 + }, + { + "epoch": 0.6996933806288703, + "flos": 24864495423840.0, + "grad_norm": 3.03285803861395, + "learning_rate": 8.736050835417466e-07, + "loss": 0.7815, + "num_input_tokens_seen": 125269350, + "step": 5819 + }, + { + "epoch": 0.6998136235195094, + "flos": 20748188681280.0, + "grad_norm": 2.5183316472584574, + "learning_rate": 8.729614873138862e-07, + "loss": 0.612, + "num_input_tokens_seen": 125286985, + "step": 5820 + }, + { + "epoch": 0.6999338664101485, + "flos": 23699560349280.0, + "grad_norm": 2.404818418631002, + "learning_rate": 8.723180620568716e-07, + "loss": 0.7758, + "num_input_tokens_seen": 125306240, + "step": 5821 + }, + { + "epoch": 0.7000541093007876, + "flos": 19836271128480.0, + "grad_norm": 3.4255961212628003, + "learning_rate": 8.716748078683116e-07, + "loss": 0.8508, + "num_input_tokens_seen": 125323015, + "step": 5822 + }, + { + "epoch": 0.7001743521914267, + "flos": 29636431258560.0, + "grad_norm": 2.966742228142207, + "learning_rate": 8.710317248457855e-07, + "loss": 0.69, + "num_input_tokens_seen": 125342630, + "step": 5823 + }, + { + "epoch": 0.7002945950820658, + "flos": 27450319505760.0, + "grad_norm": 2.4855886889580643, + "learning_rate": 8.703888130868482e-07, + "loss": 0.7255, + "num_input_tokens_seen": 125364480, + "step": 5824 + }, + { + "epoch": 0.7004148379727049, + "flos": 22126624089120.0, + "grad_norm": 2.3557531586821896, + "learning_rate": 8.697460726890307e-07, + "loss": 0.8232, + "num_input_tokens_seen": 125381625, + "step": 5825 + }, + { + "epoch": 0.7005350808633439, + "flos": 19395963687840.0, + "grad_norm": 2.929982895264156, + "learning_rate": 8.691035037498354e-07, + "loss": 0.9041, + "num_input_tokens_seen": 125397615, + "step": 5826 + }, + { + "epoch": 0.7006553237539831, + "flos": 23443123209600.0, + "grad_norm": 1.862261927193281, + "learning_rate": 8.684611063667391e-07, + "loss": 0.7241, + "num_input_tokens_seen": 125418080, + "step": 5827 + }, + { + "epoch": 0.7007755666446221, + "flos": 31168066181760.0, + "grad_norm": 2.7423813797535606, + "learning_rate": 8.678188806371935e-07, + "loss": 0.7702, + "num_input_tokens_seen": 125440115, + "step": 5828 + }, + { + "epoch": 0.7008958095352612, + "flos": 18124074012960.0, + "grad_norm": 2.081833248492787, + "learning_rate": 8.671768266586228e-07, + "loss": 0.8547, + "num_input_tokens_seen": 125458240, + "step": 5829 + }, + { + "epoch": 0.7010160524259004, + "flos": 27412883823840.0, + "grad_norm": 2.1658814425651953, + "learning_rate": 8.665349445284275e-07, + "loss": 0.7815, + "num_input_tokens_seen": 125477980, + "step": 5830 + }, + { + "epoch": 0.7011362953165394, + "flos": 23807964570240.0, + "grad_norm": 17.6692982758937, + "learning_rate": 8.658932343439799e-07, + "loss": 0.8095, + "num_input_tokens_seen": 125497765, + "step": 5831 + }, + { + "epoch": 0.7012565382071785, + "flos": 24787765572000.0, + "grad_norm": 3.1352973653374248, + "learning_rate": 8.65251696202627e-07, + "loss": 0.7809, + "num_input_tokens_seen": 125514145, + "step": 5832 + }, + { + "epoch": 0.7013767810978175, + "flos": 21366902468160.0, + "grad_norm": 3.4171621537107977, + "learning_rate": 8.646103302016896e-07, + "loss": 0.8744, + "num_input_tokens_seen": 125533115, + "step": 5833 + }, + { + "epoch": 0.7014970239884567, + "flos": 16663333289280.0, + "grad_norm": 3.8991950451499693, + "learning_rate": 8.639691364384614e-07, + "loss": 0.886, + "num_input_tokens_seen": 125550740, + "step": 5834 + }, + { + "epoch": 0.7016172668790958, + "flos": 12550594843680.0, + "grad_norm": 2.452719588097697, + "learning_rate": 8.633281150102136e-07, + "loss": 0.7283, + "num_input_tokens_seen": 125567590, + "step": 5835 + }, + { + "epoch": 0.7017375097697348, + "flos": 17427775469760.0, + "grad_norm": 2.6816011496775856, + "learning_rate": 8.626872660141855e-07, + "loss": 0.682, + "num_input_tokens_seen": 125585500, + "step": 5836 + }, + { + "epoch": 0.701857752660374, + "flos": 18486610848480.0, + "grad_norm": 2.005350822901718, + "learning_rate": 8.620465895475957e-07, + "loss": 0.7493, + "num_input_tokens_seen": 125603720, + "step": 5837 + }, + { + "epoch": 0.701977995551013, + "flos": 24391026823680.0, + "grad_norm": 2.541550576315611, + "learning_rate": 8.614060857076333e-07, + "loss": 0.7508, + "num_input_tokens_seen": 125624390, + "step": 5838 + }, + { + "epoch": 0.7020982384416521, + "flos": 22969877628000.0, + "grad_norm": 2.492879167678854, + "learning_rate": 8.60765754591462e-07, + "loss": 0.7479, + "num_input_tokens_seen": 125644085, + "step": 5839 + }, + { + "epoch": 0.7022184813322913, + "flos": 20419928098080.0, + "grad_norm": 2.8932796351405683, + "learning_rate": 8.601255962962211e-07, + "loss": 0.729, + "num_input_tokens_seen": 125663095, + "step": 5840 + }, + { + "epoch": 0.7023387242229303, + "flos": 19762254669120.0, + "grad_norm": 3.8140755821544383, + "learning_rate": 8.594856109190194e-07, + "loss": 0.7278, + "num_input_tokens_seen": 125680125, + "step": 5841 + }, + { + "epoch": 0.7024589671135694, + "flos": 33211608970560.0, + "grad_norm": 2.4511214900822575, + "learning_rate": 8.588457985569446e-07, + "loss": 0.6883, + "num_input_tokens_seen": 125703035, + "step": 5842 + }, + { + "epoch": 0.7025792100042085, + "flos": 19071791778240.0, + "grad_norm": 3.4245455697974205, + "learning_rate": 8.582061593070542e-07, + "loss": 0.7205, + "num_input_tokens_seen": 125723765, + "step": 5843 + }, + { + "epoch": 0.7026994528948476, + "flos": 18925617347520.0, + "grad_norm": 2.7754469777882034, + "learning_rate": 8.57566693266383e-07, + "loss": 0.7634, + "num_input_tokens_seen": 125741455, + "step": 5844 + }, + { + "epoch": 0.7028196957854866, + "flos": 19508828280000.0, + "grad_norm": 4.43717869414721, + "learning_rate": 8.569274005319354e-07, + "loss": 0.6955, + "num_input_tokens_seen": 125759855, + "step": 5845 + }, + { + "epoch": 0.7029399386761258, + "flos": 20815403074560.0, + "grad_norm": 2.800853605945165, + "learning_rate": 8.562882812006913e-07, + "loss": 0.7985, + "num_input_tokens_seen": 125777345, + "step": 5846 + }, + { + "epoch": 0.7030601815667649, + "flos": 22023535143840.0, + "grad_norm": 2.4194881367622174, + "learning_rate": 8.556493353696066e-07, + "loss": 0.7768, + "num_input_tokens_seen": 125796345, + "step": 5847 + }, + { + "epoch": 0.7031804244574039, + "flos": 27159754792800.0, + "grad_norm": 3.7853018749399405, + "learning_rate": 8.550105631356077e-07, + "loss": 0.6836, + "num_input_tokens_seen": 125816070, + "step": 5848 + }, + { + "epoch": 0.7033006673480431, + "flos": 22347781392960.0, + "grad_norm": 3.3335694388376322, + "learning_rate": 8.543719645955961e-07, + "loss": 0.7759, + "num_input_tokens_seen": 125834400, + "step": 5849 + }, + { + "epoch": 0.7034209102386821, + "flos": 24680550783360.0, + "grad_norm": 2.577707337736104, + "learning_rate": 8.537335398464467e-07, + "loss": 0.7436, + "num_input_tokens_seen": 125854720, + "step": 5850 + }, + { + "epoch": 0.7035411531293212, + "flos": 22523623025760.0, + "grad_norm": 3.8958165105500684, + "learning_rate": 8.53095288985007e-07, + "loss": 0.8549, + "num_input_tokens_seen": 125868455, + "step": 5851 + }, + { + "epoch": 0.7036613960199604, + "flos": 22642583458560.0, + "grad_norm": 2.0841353843483796, + "learning_rate": 8.524572121081009e-07, + "loss": 0.8228, + "num_input_tokens_seen": 125888555, + "step": 5852 + }, + { + "epoch": 0.7037816389105994, + "flos": 22460014099200.0, + "grad_norm": 2.8355670613454387, + "learning_rate": 8.518193093125232e-07, + "loss": 0.6228, + "num_input_tokens_seen": 125907610, + "step": 5853 + }, + { + "epoch": 0.7039018818012385, + "flos": 27048600009600.0, + "grad_norm": 4.048733435488267, + "learning_rate": 8.511815806950436e-07, + "loss": 0.8077, + "num_input_tokens_seen": 125928640, + "step": 5854 + }, + { + "epoch": 0.7040221246918776, + "flos": 17724621872160.0, + "grad_norm": 1.9816105126503634, + "learning_rate": 8.505440263524044e-07, + "loss": 0.778, + "num_input_tokens_seen": 125947485, + "step": 5855 + }, + { + "epoch": 0.7041423675825167, + "flos": 16256410026720.0, + "grad_norm": 4.52852448675058, + "learning_rate": 8.49906646381322e-07, + "loss": 0.8911, + "num_input_tokens_seen": 125960320, + "step": 5856 + }, + { + "epoch": 0.7042626104731557, + "flos": 25446665603040.0, + "grad_norm": 3.522130618089191, + "learning_rate": 8.492694408784884e-07, + "loss": 0.7192, + "num_input_tokens_seen": 125980575, + "step": 5857 + }, + { + "epoch": 0.7043828533637949, + "flos": 17832208358400.0, + "grad_norm": 3.292123271965584, + "learning_rate": 8.486324099405642e-07, + "loss": 0.6243, + "num_input_tokens_seen": 125997420, + "step": 5858 + }, + { + "epoch": 0.704503096254434, + "flos": 29451817562400.0, + "grad_norm": 2.4291638365042565, + "learning_rate": 8.479955536641887e-07, + "loss": 0.7501, + "num_input_tokens_seen": 126018915, + "step": 5859 + }, + { + "epoch": 0.704623339145073, + "flos": 30876349206240.0, + "grad_norm": 4.9309778716179355, + "learning_rate": 8.473588721459716e-07, + "loss": 0.6604, + "num_input_tokens_seen": 126038825, + "step": 5860 + }, + { + "epoch": 0.7047435820357122, + "flos": 23880531408960.0, + "grad_norm": 2.389130867178814, + "learning_rate": 8.467223654824967e-07, + "loss": 0.7082, + "num_input_tokens_seen": 126058280, + "step": 5861 + }, + { + "epoch": 0.7048638249263512, + "flos": 46426722212640.0, + "grad_norm": 2.422312641594881, + "learning_rate": 8.460860337703233e-07, + "loss": 0.6269, + "num_input_tokens_seen": 126078885, + "step": 5862 + }, + { + "epoch": 0.7049840678169903, + "flos": 21658247746080.0, + "grad_norm": 2.666492774379918, + "learning_rate": 8.454498771059797e-07, + "loss": 0.7073, + "num_input_tokens_seen": 126098260, + "step": 5863 + }, + { + "epoch": 0.7051043107076294, + "flos": 18378875683200.0, + "grad_norm": 2.7308379920948447, + "learning_rate": 8.448138955859725e-07, + "loss": 0.8347, + "num_input_tokens_seen": 126114845, + "step": 5864 + }, + { + "epoch": 0.7052245535982685, + "flos": 19291573800960.0, + "grad_norm": 2.4769728648517066, + "learning_rate": 8.44178089306778e-07, + "loss": 0.9037, + "num_input_tokens_seen": 126132780, + "step": 5865 + }, + { + "epoch": 0.7053447964889076, + "flos": 19034987982240.0, + "grad_norm": 2.8329053069291503, + "learning_rate": 8.4354245836485e-07, + "loss": 0.7731, + "num_input_tokens_seen": 126151225, + "step": 5866 + }, + { + "epoch": 0.7054650393795466, + "flos": 27339908117760.0, + "grad_norm": 1.9733371277184135, + "learning_rate": 8.429070028566108e-07, + "loss": 0.7299, + "num_input_tokens_seen": 126172535, + "step": 5867 + }, + { + "epoch": 0.7055852822701858, + "flos": 16079193112800.0, + "grad_norm": 2.6098498980158493, + "learning_rate": 8.422717228784586e-07, + "loss": 0.7512, + "num_input_tokens_seen": 126189410, + "step": 5868 + }, + { + "epoch": 0.7057055251608249, + "flos": 11675889954240.0, + "grad_norm": 2.562667841182973, + "learning_rate": 8.416366185267663e-07, + "loss": 0.6937, + "num_input_tokens_seen": 126206910, + "step": 5869 + }, + { + "epoch": 0.7058257680514639, + "flos": 22679127066240.0, + "grad_norm": 2.25757763970952, + "learning_rate": 8.410016898978778e-07, + "loss": 0.7772, + "num_input_tokens_seen": 126224385, + "step": 5870 + }, + { + "epoch": 0.7059460109421031, + "flos": 17504616830880.0, + "grad_norm": 2.5530084814520233, + "learning_rate": 8.403669370881115e-07, + "loss": 0.7935, + "num_input_tokens_seen": 126243120, + "step": 5871 + }, + { + "epoch": 0.7060662538327421, + "flos": 23510820809760.0, + "grad_norm": 2.092255015918409, + "learning_rate": 8.397323601937587e-07, + "loss": 0.7874, + "num_input_tokens_seen": 126263020, + "step": 5872 + }, + { + "epoch": 0.7061864967233812, + "flos": 30217188986880.0, + "grad_norm": 2.2825865292311995, + "learning_rate": 8.390979593110838e-07, + "loss": 0.7714, + "num_input_tokens_seen": 126285150, + "step": 5873 + }, + { + "epoch": 0.7063067396140204, + "flos": 20671830527040.0, + "grad_norm": 2.293880041110557, + "learning_rate": 8.384637345363262e-07, + "loss": 0.8169, + "num_input_tokens_seen": 126304340, + "step": 5874 + }, + { + "epoch": 0.7064269825046594, + "flos": 32219653457280.0, + "grad_norm": 2.5471306420959356, + "learning_rate": 8.378296859656964e-07, + "loss": 0.7682, + "num_input_tokens_seen": 126325495, + "step": 5875 + }, + { + "epoch": 0.7065472253952985, + "flos": 30183730469280.0, + "grad_norm": 5.162527436190431, + "learning_rate": 8.371958136953792e-07, + "loss": 0.6813, + "num_input_tokens_seen": 126345525, + "step": 5876 + }, + { + "epoch": 0.7066674682859376, + "flos": 16193730344160.0, + "grad_norm": 4.584906753041116, + "learning_rate": 8.365621178215326e-07, + "loss": 0.6671, + "num_input_tokens_seen": 126361995, + "step": 5877 + }, + { + "epoch": 0.7067877111765767, + "flos": 14809496453760.0, + "grad_norm": 2.9926830708045706, + "learning_rate": 8.359285984402871e-07, + "loss": 0.7532, + "num_input_tokens_seen": 126379260, + "step": 5878 + }, + { + "epoch": 0.7069079540672157, + "flos": 25403505778080.0, + "grad_norm": 4.838123819508499, + "learning_rate": 8.352952556477489e-07, + "loss": 0.7416, + "num_input_tokens_seen": 126397170, + "step": 5879 + }, + { + "epoch": 0.7070281969578549, + "flos": 24572518260000.0, + "grad_norm": 2.376069380283466, + "learning_rate": 8.34662089539993e-07, + "loss": 0.7623, + "num_input_tokens_seen": 126416680, + "step": 5880 + }, + { + "epoch": 0.707148439848494, + "flos": 26686174683360.0, + "grad_norm": 2.5777208416047794, + "learning_rate": 8.340291002130722e-07, + "loss": 0.7909, + "num_input_tokens_seen": 126435870, + "step": 5881 + }, + { + "epoch": 0.707268682739133, + "flos": 15064409633280.0, + "grad_norm": 4.72479991993812, + "learning_rate": 8.3339628776301e-07, + "loss": 0.7993, + "num_input_tokens_seen": 126454010, + "step": 5882 + }, + { + "epoch": 0.7073889256297722, + "flos": 34263790962240.0, + "grad_norm": 2.3908960525336513, + "learning_rate": 8.327636522858033e-07, + "loss": 0.5731, + "num_input_tokens_seen": 126473615, + "step": 5883 + }, + { + "epoch": 0.7075091685204112, + "flos": 20055309756000.0, + "grad_norm": 2.213055139067589, + "learning_rate": 8.321311938774225e-07, + "loss": 0.7777, + "num_input_tokens_seen": 126492705, + "step": 5884 + }, + { + "epoch": 0.7076294114110503, + "flos": 20747482455840.0, + "grad_norm": 10.0043912782654, + "learning_rate": 8.314989126338104e-07, + "loss": 0.7935, + "num_input_tokens_seen": 126512715, + "step": 5885 + }, + { + "epoch": 0.7077496543016895, + "flos": 17942136539520.0, + "grad_norm": 2.0401851442880456, + "learning_rate": 8.308668086508847e-07, + "loss": 0.8433, + "num_input_tokens_seen": 126530795, + "step": 5886 + }, + { + "epoch": 0.7078698971923285, + "flos": 45412496279520.0, + "grad_norm": 4.0006451086544015, + "learning_rate": 8.302348820245342e-07, + "loss": 0.7416, + "num_input_tokens_seen": 126553360, + "step": 5887 + }, + { + "epoch": 0.7079901400829676, + "flos": 26905101801600.0, + "grad_norm": 4.47980119809061, + "learning_rate": 8.296031328506232e-07, + "loss": 0.7133, + "num_input_tokens_seen": 126573110, + "step": 5888 + }, + { + "epoch": 0.7081103829736067, + "flos": 24388164752160.0, + "grad_norm": 2.1174061792479684, + "learning_rate": 8.289715612249857e-07, + "loss": 0.7536, + "num_input_tokens_seen": 126593725, + "step": 5889 + }, + { + "epoch": 0.7082306258642458, + "flos": 18516612578400.0, + "grad_norm": 2.6850337755575833, + "learning_rate": 8.283401672434305e-07, + "loss": 0.7738, + "num_input_tokens_seen": 126608950, + "step": 5890 + }, + { + "epoch": 0.7083508687548848, + "flos": 23443978114080.0, + "grad_norm": 3.4734701397898498, + "learning_rate": 8.277089510017412e-07, + "loss": 0.706, + "num_input_tokens_seen": 126629755, + "step": 5891 + }, + { + "epoch": 0.708471111645524, + "flos": 22387447260480.0, + "grad_norm": 2.1867259154589083, + "learning_rate": 8.270779125956719e-07, + "loss": 0.8213, + "num_input_tokens_seen": 126650135, + "step": 5892 + }, + { + "epoch": 0.7085913545361631, + "flos": 20893025000640.0, + "grad_norm": 3.7285885211738625, + "learning_rate": 8.264470521209505e-07, + "loss": 0.8015, + "num_input_tokens_seen": 126668500, + "step": 5893 + }, + { + "epoch": 0.7087115974268021, + "flos": 14991136569120.0, + "grad_norm": 4.330091780114914, + "learning_rate": 8.258163696732785e-07, + "loss": 0.7664, + "num_input_tokens_seen": 126686090, + "step": 5894 + }, + { + "epoch": 0.7088318403174413, + "flos": 21508133320800.0, + "grad_norm": 2.1525796805951307, + "learning_rate": 8.251858653483288e-07, + "loss": 0.7697, + "num_input_tokens_seen": 126704255, + "step": 5895 + }, + { + "epoch": 0.7089520832080803, + "flos": 15494569729440.0, + "grad_norm": 2.6995153415990822, + "learning_rate": 8.245555392417501e-07, + "loss": 0.8601, + "num_input_tokens_seen": 126718910, + "step": 5896 + }, + { + "epoch": 0.7090723260987194, + "flos": 20383570339200.0, + "grad_norm": 3.447606160145563, + "learning_rate": 8.239253914491613e-07, + "loss": 0.7874, + "num_input_tokens_seen": 126737235, + "step": 5897 + }, + { + "epoch": 0.7091925689893585, + "flos": 25631650996800.0, + "grad_norm": 3.413300942128192, + "learning_rate": 8.232954220661556e-07, + "loss": 0.753, + "num_input_tokens_seen": 126759970, + "step": 5898 + }, + { + "epoch": 0.7093128118799976, + "flos": 24171988196160.0, + "grad_norm": 7.878510434483143, + "learning_rate": 8.226656311882989e-07, + "loss": 0.7031, + "num_input_tokens_seen": 126779280, + "step": 5899 + }, + { + "epoch": 0.7094330547706367, + "flos": 16622886856800.0, + "grad_norm": 2.7850842566902734, + "learning_rate": 8.22036018911129e-07, + "loss": 0.7685, + "num_input_tokens_seen": 126797310, + "step": 5900 + }, + { + "epoch": 0.7095532976612757, + "flos": 16260052663200.0, + "grad_norm": 2.4197133908123827, + "learning_rate": 8.214065853301599e-07, + "loss": 0.8062, + "num_input_tokens_seen": 126812840, + "step": 5901 + }, + { + "epoch": 0.7096735405519149, + "flos": 70620900755520.0, + "grad_norm": 0.8475279469228594, + "learning_rate": 8.207773305408734e-07, + "loss": 0.5815, + "num_input_tokens_seen": 126880060, + "step": 5902 + }, + { + "epoch": 0.709793783442554, + "flos": 23587662170880.0, + "grad_norm": 2.3976111961250983, + "learning_rate": 8.201482546387288e-07, + "loss": 0.7979, + "num_input_tokens_seen": 126899535, + "step": 5903 + }, + { + "epoch": 0.709914026333193, + "flos": 25956194604000.0, + "grad_norm": 13.50490057266395, + "learning_rate": 8.195193577191553e-07, + "loss": 0.9187, + "num_input_tokens_seen": 126921365, + "step": 5904 + }, + { + "epoch": 0.7100342692238322, + "flos": 24825870309600.0, + "grad_norm": 2.2207106403099526, + "learning_rate": 8.188906398775579e-07, + "loss": 0.8473, + "num_input_tokens_seen": 126941910, + "step": 5905 + }, + { + "epoch": 0.7101545121144712, + "flos": 24896355641760.0, + "grad_norm": 3.592386799875798, + "learning_rate": 8.18262101209311e-07, + "loss": 0.6907, + "num_input_tokens_seen": 126961120, + "step": 5906 + }, + { + "epoch": 0.7102747550051103, + "flos": 23734840185120.0, + "grad_norm": 3.0539456220281145, + "learning_rate": 8.176337418097626e-07, + "loss": 0.6997, + "num_input_tokens_seen": 126981590, + "step": 5907 + }, + { + "epoch": 0.7103949978957494, + "flos": 15282221658720.0, + "grad_norm": 2.7097061394855575, + "learning_rate": 8.170055617742364e-07, + "loss": 0.7998, + "num_input_tokens_seen": 126998870, + "step": 5908 + }, + { + "epoch": 0.7105152407863885, + "flos": 22606485888000.0, + "grad_norm": 2.2332922172557335, + "learning_rate": 8.163775611980252e-07, + "loss": 0.7059, + "num_input_tokens_seen": 127017980, + "step": 5909 + }, + { + "epoch": 0.7106354836770276, + "flos": 17214126457440.0, + "grad_norm": 2.981879823961036, + "learning_rate": 8.157497401763982e-07, + "loss": 0.79, + "num_input_tokens_seen": 127035645, + "step": 5910 + }, + { + "epoch": 0.7107557265676667, + "flos": 20164866239520.0, + "grad_norm": 2.311010551200758, + "learning_rate": 8.151220988045935e-07, + "loss": 0.7794, + "num_input_tokens_seen": 127054900, + "step": 5911 + }, + { + "epoch": 0.7108759694583058, + "flos": 21476384612160.0, + "grad_norm": 26.508579232943035, + "learning_rate": 8.144946371778234e-07, + "loss": 0.8279, + "num_input_tokens_seen": 127075010, + "step": 5912 + }, + { + "epoch": 0.7109962123489448, + "flos": 24027523574400.0, + "grad_norm": 2.3712308991687947, + "learning_rate": 8.138673553912751e-07, + "loss": 0.783, + "num_input_tokens_seen": 127095570, + "step": 5913 + }, + { + "epoch": 0.711116455239584, + "flos": 30436636481760.0, + "grad_norm": 6.175257738425687, + "learning_rate": 8.132402535401059e-07, + "loss": 0.566, + "num_input_tokens_seen": 127116825, + "step": 5914 + }, + { + "epoch": 0.711236698130223, + "flos": 25009740610560.0, + "grad_norm": 2.2413122083601467, + "learning_rate": 8.126133317194465e-07, + "loss": 0.7428, + "num_input_tokens_seen": 127137015, + "step": 5915 + }, + { + "epoch": 0.7113569410208621, + "flos": 24172062535680.0, + "grad_norm": 2.3792577023523283, + "learning_rate": 8.11986590024401e-07, + "loss": 0.738, + "num_input_tokens_seen": 127156755, + "step": 5916 + }, + { + "epoch": 0.7114771839115013, + "flos": 35388130925280.0, + "grad_norm": 2.173277038750054, + "learning_rate": 8.113600285500442e-07, + "loss": 0.6916, + "num_input_tokens_seen": 127176965, + "step": 5917 + }, + { + "epoch": 0.7115974268021403, + "flos": 21069126821760.0, + "grad_norm": 2.336857949271466, + "learning_rate": 8.107336473914268e-07, + "loss": 0.747, + "num_input_tokens_seen": 127195595, + "step": 5918 + }, + { + "epoch": 0.7117176696927794, + "flos": 56672201967360.0, + "grad_norm": 0.7898828755242823, + "learning_rate": 8.101074466435694e-07, + "loss": 0.5565, + "num_input_tokens_seen": 127255070, + "step": 5919 + }, + { + "epoch": 0.7118379125834186, + "flos": 15902719594080.0, + "grad_norm": 2.0317981860573573, + "learning_rate": 8.094814264014662e-07, + "loss": 0.6842, + "num_input_tokens_seen": 127273825, + "step": 5920 + }, + { + "epoch": 0.7119581554740576, + "flos": 20164568881440.0, + "grad_norm": 3.253373376047485, + "learning_rate": 8.088555867600844e-07, + "loss": 0.8165, + "num_input_tokens_seen": 127289990, + "step": 5921 + }, + { + "epoch": 0.7120783983646967, + "flos": 34665473288640.0, + "grad_norm": 2.134493881198944, + "learning_rate": 8.08229927814362e-07, + "loss": 0.6019, + "num_input_tokens_seen": 127312880, + "step": 5922 + }, + { + "epoch": 0.7121986412553358, + "flos": 26321333322720.0, + "grad_norm": 1.8620109786581696, + "learning_rate": 8.076044496592134e-07, + "loss": 0.647, + "num_input_tokens_seen": 127334730, + "step": 5923 + }, + { + "epoch": 0.7123188841459749, + "flos": 11129743006080.0, + "grad_norm": 3.9276676030034183, + "learning_rate": 8.069791523895204e-07, + "loss": 0.7851, + "num_input_tokens_seen": 127351180, + "step": 5924 + }, + { + "epoch": 0.7124391270366139, + "flos": 20781907387200.0, + "grad_norm": 2.3893573865907363, + "learning_rate": 8.063540361001422e-07, + "loss": 0.7727, + "num_input_tokens_seen": 127369750, + "step": 5925 + }, + { + "epoch": 0.7125593699272531, + "flos": 17578744799520.0, + "grad_norm": 2.5606235984405488, + "learning_rate": 8.057291008859069e-07, + "loss": 0.7909, + "num_input_tokens_seen": 127387910, + "step": 5926 + }, + { + "epoch": 0.7126796128178922, + "flos": 28612801376160.0, + "grad_norm": 4.017190472228213, + "learning_rate": 8.051043468416187e-07, + "loss": 0.6885, + "num_input_tokens_seen": 127409160, + "step": 5927 + }, + { + "epoch": 0.7127998557085312, + "flos": 16011718531200.0, + "grad_norm": 2.5367561076767324, + "learning_rate": 8.044797740620506e-07, + "loss": 0.8226, + "num_input_tokens_seen": 127427765, + "step": 5928 + }, + { + "epoch": 0.7129200985991703, + "flos": 23369775805920.0, + "grad_norm": 2.4082325286183957, + "learning_rate": 8.038553826419494e-07, + "loss": 0.7889, + "num_input_tokens_seen": 127446475, + "step": 5929 + }, + { + "epoch": 0.7130403414898094, + "flos": 21366902468160.0, + "grad_norm": 2.0619704982893015, + "learning_rate": 8.032311726760364e-07, + "loss": 0.8118, + "num_input_tokens_seen": 127467695, + "step": 5930 + }, + { + "epoch": 0.7131605843804485, + "flos": 74630816177760.0, + "grad_norm": 2.5463053060301575, + "learning_rate": 8.026071442590022e-07, + "loss": 0.6894, + "num_input_tokens_seen": 127494590, + "step": 5931 + }, + { + "epoch": 0.7132808272710875, + "flos": 18342592263840.0, + "grad_norm": 3.981805548839869, + "learning_rate": 8.019832974855134e-07, + "loss": 0.8077, + "num_input_tokens_seen": 127512550, + "step": 5932 + }, + { + "epoch": 0.7134010701617267, + "flos": 23219958738720.0, + "grad_norm": 3.664000032783075, + "learning_rate": 8.013596324502052e-07, + "loss": 0.8253, + "num_input_tokens_seen": 127531015, + "step": 5933 + }, + { + "epoch": 0.7135213130523658, + "flos": 23619076351680.0, + "grad_norm": 2.011828610281506, + "learning_rate": 8.007361492476872e-07, + "loss": 0.7846, + "num_input_tokens_seen": 127550340, + "step": 5934 + }, + { + "epoch": 0.7136415559430048, + "flos": 24755161958880.0, + "grad_norm": 1.9372932569887862, + "learning_rate": 8.001128479725426e-07, + "loss": 0.7922, + "num_input_tokens_seen": 127572245, + "step": 5935 + }, + { + "epoch": 0.713761798833644, + "flos": 18270954669120.0, + "grad_norm": 3.4923270451530763, + "learning_rate": 7.994897287193248e-07, + "loss": 0.8122, + "num_input_tokens_seen": 127591625, + "step": 5936 + }, + { + "epoch": 0.713882041724283, + "flos": 15536168424480.0, + "grad_norm": 8.879635741864817, + "learning_rate": 7.988667915825605e-07, + "loss": 0.8378, + "num_input_tokens_seen": 127608690, + "step": 5937 + }, + { + "epoch": 0.7140022846149221, + "flos": 24026594330400.0, + "grad_norm": 4.580361714967651, + "learning_rate": 7.982440366567491e-07, + "loss": 0.7607, + "num_input_tokens_seen": 127627180, + "step": 5938 + }, + { + "epoch": 0.7141225275055613, + "flos": 27851704474080.0, + "grad_norm": 2.7166798043274665, + "learning_rate": 7.97621464036361e-07, + "loss": 0.7506, + "num_input_tokens_seen": 127648940, + "step": 5939 + }, + { + "epoch": 0.7142427703962003, + "flos": 19654556673600.0, + "grad_norm": 2.3022555693899642, + "learning_rate": 7.969990738158417e-07, + "loss": 0.6803, + "num_input_tokens_seen": 127667350, + "step": 5940 + }, + { + "epoch": 0.7143630132868394, + "flos": 21002358465600.0, + "grad_norm": 2.84617494587205, + "learning_rate": 7.963768660896062e-07, + "loss": 0.8523, + "num_input_tokens_seen": 127685760, + "step": 5941 + }, + { + "epoch": 0.7144832561774785, + "flos": 24095258344320.0, + "grad_norm": 2.89281022237305, + "learning_rate": 7.957548409520432e-07, + "loss": 0.8218, + "num_input_tokens_seen": 127704985, + "step": 5942 + }, + { + "epoch": 0.7146034990681176, + "flos": 16302654941760.0, + "grad_norm": 3.734059340411564, + "learning_rate": 7.951329984975135e-07, + "loss": 0.843, + "num_input_tokens_seen": 127721925, + "step": 5943 + }, + { + "epoch": 0.7147237419587567, + "flos": 69527566105920.0, + "grad_norm": 0.7256594141932268, + "learning_rate": 7.94511338820349e-07, + "loss": 0.5415, + "num_input_tokens_seen": 127784230, + "step": 5944 + }, + { + "epoch": 0.7148439848493958, + "flos": 22234768121760.0, + "grad_norm": 3.329585654796914, + "learning_rate": 7.938898620148575e-07, + "loss": 0.786, + "num_input_tokens_seen": 127801990, + "step": 5945 + }, + { + "epoch": 0.7149642277400349, + "flos": 17905890289920.0, + "grad_norm": 3.036292689381783, + "learning_rate": 7.932685681753135e-07, + "loss": 0.7129, + "num_input_tokens_seen": 127819270, + "step": 5946 + }, + { + "epoch": 0.7150844706306739, + "flos": 31636256676000.0, + "grad_norm": 2.261742151959565, + "learning_rate": 7.92647457395969e-07, + "loss": 0.6283, + "num_input_tokens_seen": 127841095, + "step": 5947 + }, + { + "epoch": 0.7152047135213131, + "flos": 10910927397120.0, + "grad_norm": 3.809179517253752, + "learning_rate": 7.920265297710444e-07, + "loss": 0.7454, + "num_input_tokens_seen": 127858485, + "step": 5948 + }, + { + "epoch": 0.7153249564119522, + "flos": 20965666178880.0, + "grad_norm": 2.4889838695378934, + "learning_rate": 7.914057853947363e-07, + "loss": 0.7357, + "num_input_tokens_seen": 127877665, + "step": 5949 + }, + { + "epoch": 0.7154451993025912, + "flos": 24208420294560.0, + "grad_norm": 2.316609417141646, + "learning_rate": 7.907852243612089e-07, + "loss": 0.627, + "num_input_tokens_seen": 127898070, + "step": 5950 + }, + { + "epoch": 0.7155654421932304, + "flos": 23297580664800.0, + "grad_norm": 2.7966650345910833, + "learning_rate": 7.901648467646009e-07, + "loss": 0.7296, + "num_input_tokens_seen": 127917010, + "step": 5951 + }, + { + "epoch": 0.7156856850838694, + "flos": 22679461594080.0, + "grad_norm": 2.243829531309402, + "learning_rate": 7.895446526990244e-07, + "loss": 0.7276, + "num_input_tokens_seen": 127937025, + "step": 5952 + }, + { + "epoch": 0.7158059279745085, + "flos": 19837126032960.0, + "grad_norm": 2.1612473227824656, + "learning_rate": 7.889246422585609e-07, + "loss": 0.7592, + "num_input_tokens_seen": 127956410, + "step": 5953 + }, + { + "epoch": 0.7159261708651476, + "flos": 24100276261920.0, + "grad_norm": 2.3359191281415916, + "learning_rate": 7.883048155372675e-07, + "loss": 0.7311, + "num_input_tokens_seen": 127974925, + "step": 5954 + }, + { + "epoch": 0.7160464137557867, + "flos": 16959101768640.0, + "grad_norm": 3.112885606771037, + "learning_rate": 7.876851726291698e-07, + "loss": 0.717, + "num_input_tokens_seen": 127993225, + "step": 5955 + }, + { + "epoch": 0.7161666566464258, + "flos": 25192124121120.0, + "grad_norm": 3.707028806020335, + "learning_rate": 7.870657136282666e-07, + "loss": 0.787, + "num_input_tokens_seen": 128012085, + "step": 5956 + }, + { + "epoch": 0.7162868995370649, + "flos": 26430443769120.0, + "grad_norm": 1.8191521347497641, + "learning_rate": 7.86446438628531e-07, + "loss": 0.8201, + "num_input_tokens_seen": 128033155, + "step": 5957 + }, + { + "epoch": 0.716407142427704, + "flos": 69899023683840.0, + "grad_norm": 0.7770403552677297, + "learning_rate": 7.858273477239059e-07, + "loss": 0.5682, + "num_input_tokens_seen": 128101575, + "step": 5958 + }, + { + "epoch": 0.716527385318343, + "flos": 20710864508640.0, + "grad_norm": 2.6035169070584803, + "learning_rate": 7.852084410083067e-07, + "loss": 0.7112, + "num_input_tokens_seen": 128120395, + "step": 5959 + }, + { + "epoch": 0.7166476282089821, + "flos": 25335547989600.0, + "grad_norm": 1.8305867891761645, + "learning_rate": 7.84589718575621e-07, + "loss": 0.6378, + "num_input_tokens_seen": 128140840, + "step": 5960 + }, + { + "epoch": 0.7167678710996213, + "flos": 24099198338880.0, + "grad_norm": 2.3601562015112956, + "learning_rate": 7.83971180519708e-07, + "loss": 0.6887, + "num_input_tokens_seen": 128159695, + "step": 5961 + }, + { + "epoch": 0.7168881139902603, + "flos": 30183172922880.0, + "grad_norm": 4.934609834809816, + "learning_rate": 7.833528269344008e-07, + "loss": 0.754, + "num_input_tokens_seen": 128179600, + "step": 5962 + }, + { + "epoch": 0.7170083568808994, + "flos": 14585588587680.0, + "grad_norm": 2.714867461396003, + "learning_rate": 7.827346579135023e-07, + "loss": 0.7796, + "num_input_tokens_seen": 128196940, + "step": 5963 + }, + { + "epoch": 0.7171285997715385, + "flos": 23297729343840.0, + "grad_norm": 2.3408750066944095, + "learning_rate": 7.821166735507885e-07, + "loss": 0.8341, + "num_input_tokens_seen": 128215970, + "step": 5964 + }, + { + "epoch": 0.7172488426621776, + "flos": 16520020930080.0, + "grad_norm": 3.1418724808873493, + "learning_rate": 7.81498873940007e-07, + "loss": 0.6895, + "num_input_tokens_seen": 128233185, + "step": 5965 + }, + { + "epoch": 0.7173690855528166, + "flos": 26503568154240.0, + "grad_norm": 3.265208094509699, + "learning_rate": 7.808812591748768e-07, + "loss": 0.7728, + "num_input_tokens_seen": 128253565, + "step": 5966 + }, + { + "epoch": 0.7174893284434558, + "flos": 22751545225920.0, + "grad_norm": 2.9825956622479772, + "learning_rate": 7.802638293490915e-07, + "loss": 0.6526, + "num_input_tokens_seen": 128273210, + "step": 5967 + }, + { + "epoch": 0.7176095713340949, + "flos": 23260219322400.0, + "grad_norm": 2.1285872754159034, + "learning_rate": 7.796465845563123e-07, + "loss": 0.7668, + "num_input_tokens_seen": 128292085, + "step": 5968 + }, + { + "epoch": 0.7177298142247339, + "flos": 25554921144960.0, + "grad_norm": 3.724190202799258, + "learning_rate": 7.790295248901766e-07, + "loss": 0.7964, + "num_input_tokens_seen": 128313215, + "step": 5969 + }, + { + "epoch": 0.7178500571153731, + "flos": 31607407208640.0, + "grad_norm": 2.1028424871049496, + "learning_rate": 7.784126504442902e-07, + "loss": 0.6255, + "num_input_tokens_seen": 128336445, + "step": 5970 + }, + { + "epoch": 0.7179703000060121, + "flos": 19399569154560.0, + "grad_norm": 2.0019212078248403, + "learning_rate": 7.777959613122351e-07, + "loss": 0.6772, + "num_input_tokens_seen": 128356270, + "step": 5971 + }, + { + "epoch": 0.7180905428966512, + "flos": 28797823939680.0, + "grad_norm": 2.134791312102433, + "learning_rate": 7.771794575875604e-07, + "loss": 0.7788, + "num_input_tokens_seen": 128378140, + "step": 5972 + }, + { + "epoch": 0.7182107857872904, + "flos": 20018989166880.0, + "grad_norm": 3.4179575892674925, + "learning_rate": 7.765631393637888e-07, + "loss": 0.7755, + "num_input_tokens_seen": 128396335, + "step": 5973 + }, + { + "epoch": 0.7183310286779294, + "flos": 22715187467040.0, + "grad_norm": 8.942587091641968, + "learning_rate": 7.75947006734417e-07, + "loss": 0.4825, + "num_input_tokens_seen": 128414115, + "step": 5974 + }, + { + "epoch": 0.7184512715685685, + "flos": 17133642459840.0, + "grad_norm": 4.323073541330288, + "learning_rate": 7.753310597929101e-07, + "loss": 0.8257, + "num_input_tokens_seen": 128430755, + "step": 5975 + }, + { + "epoch": 0.7185715144592076, + "flos": 65416202941440.0, + "grad_norm": 0.7644767714689574, + "learning_rate": 7.747152986327095e-07, + "loss": 0.549, + "num_input_tokens_seen": 128491300, + "step": 5976 + }, + { + "epoch": 0.7186917573498467, + "flos": 16157595603840.0, + "grad_norm": 2.3798164918821323, + "learning_rate": 7.740997233472228e-07, + "loss": 0.6753, + "num_input_tokens_seen": 128508920, + "step": 5977 + }, + { + "epoch": 0.7188120002404857, + "flos": 29200324000800.0, + "grad_norm": 3.3315271888481086, + "learning_rate": 7.734843340298329e-07, + "loss": 0.7084, + "num_input_tokens_seen": 128528745, + "step": 5978 + }, + { + "epoch": 0.7189322431311249, + "flos": 33353137181280.0, + "grad_norm": 4.789892625660459, + "learning_rate": 7.72869130773895e-07, + "loss": 0.7496, + "num_input_tokens_seen": 128549345, + "step": 5979 + }, + { + "epoch": 0.719052486021764, + "flos": 61263947307360.0, + "grad_norm": 0.7948694610186755, + "learning_rate": 7.722541136727343e-07, + "loss": 0.5929, + "num_input_tokens_seen": 128605360, + "step": 5980 + }, + { + "epoch": 0.719172728912403, + "flos": 15574719199200.0, + "grad_norm": 8.310165651141622, + "learning_rate": 7.716392828196483e-07, + "loss": 0.8062, + "num_input_tokens_seen": 128623160, + "step": 5981 + }, + { + "epoch": 0.7192929718030422, + "flos": 15530890318560.0, + "grad_norm": 3.170918287491765, + "learning_rate": 7.710246383079064e-07, + "loss": 0.7778, + "num_input_tokens_seen": 128638545, + "step": 5982 + }, + { + "epoch": 0.7194132146936812, + "flos": 21830929949280.0, + "grad_norm": 3.5067837232432804, + "learning_rate": 7.704101802307492e-07, + "loss": 0.9266, + "num_input_tokens_seen": 128650845, + "step": 5983 + }, + { + "epoch": 0.7195334575843203, + "flos": 27299498855040.0, + "grad_norm": 2.905543087538872, + "learning_rate": 7.697959086813912e-07, + "loss": 0.8695, + "num_input_tokens_seen": 128667010, + "step": 5984 + }, + { + "epoch": 0.7196537004749595, + "flos": 18743902892640.0, + "grad_norm": 2.7239689337461543, + "learning_rate": 7.691818237530145e-07, + "loss": 0.8034, + "num_input_tokens_seen": 128685870, + "step": 5985 + }, + { + "epoch": 0.7197739433655985, + "flos": 24495751238400.0, + "grad_norm": 2.642366365574277, + "learning_rate": 7.685679255387774e-07, + "loss": 0.7724, + "num_input_tokens_seen": 128704185, + "step": 5986 + }, + { + "epoch": 0.7198941862562376, + "flos": 18014926396800.0, + "grad_norm": 3.368937097566293, + "learning_rate": 7.679542141318065e-07, + "loss": 0.7652, + "num_input_tokens_seen": 128721290, + "step": 5987 + }, + { + "epoch": 0.7200144291468767, + "flos": 28980727826880.0, + "grad_norm": 2.3847970171587596, + "learning_rate": 7.673406896252013e-07, + "loss": 0.7561, + "num_input_tokens_seen": 128742665, + "step": 5988 + }, + { + "epoch": 0.7201346720375158, + "flos": 25337926854240.0, + "grad_norm": 1.9726274847067324, + "learning_rate": 7.667273521120347e-07, + "loss": 0.7858, + "num_input_tokens_seen": 128762225, + "step": 5989 + }, + { + "epoch": 0.7202549149281549, + "flos": 14335321628160.0, + "grad_norm": 2.5424266887886837, + "learning_rate": 7.661142016853468e-07, + "loss": 0.7951, + "num_input_tokens_seen": 128779585, + "step": 5990 + }, + { + "epoch": 0.7203751578187939, + "flos": 22968762535200.0, + "grad_norm": 2.850412967945099, + "learning_rate": 7.655012384381543e-07, + "loss": 0.7473, + "num_input_tokens_seen": 128799070, + "step": 5991 + }, + { + "epoch": 0.7204954007094331, + "flos": 23658407691360.0, + "grad_norm": 5.490133625898792, + "learning_rate": 7.648884624634415e-07, + "loss": 0.8193, + "num_input_tokens_seen": 128817620, + "step": 5992 + }, + { + "epoch": 0.7206156436000721, + "flos": 16229976593760.0, + "grad_norm": 2.1841244402426083, + "learning_rate": 7.642758738541683e-07, + "loss": 0.8888, + "num_input_tokens_seen": 128834200, + "step": 5993 + }, + { + "epoch": 0.7207358864907112, + "flos": 54300472934880.0, + "grad_norm": 0.775171013391425, + "learning_rate": 7.636634727032621e-07, + "loss": 0.6072, + "num_input_tokens_seen": 128891305, + "step": 5994 + }, + { + "epoch": 0.7208561293813504, + "flos": 19108298216160.0, + "grad_norm": 2.413414432978986, + "learning_rate": 7.630512591036231e-07, + "loss": 0.7841, + "num_input_tokens_seen": 128910615, + "step": 5995 + }, + { + "epoch": 0.7209763722719894, + "flos": 17723283760800.0, + "grad_norm": 3.752138672370628, + "learning_rate": 7.624392331481255e-07, + "loss": 0.6516, + "num_input_tokens_seen": 128928270, + "step": 5996 + }, + { + "epoch": 0.7210966151626285, + "flos": 66724041507840.0, + "grad_norm": 0.762492642601663, + "learning_rate": 7.618273949296115e-07, + "loss": 0.5173, + "num_input_tokens_seen": 128987780, + "step": 5997 + }, + { + "epoch": 0.7212168580532676, + "flos": 21111691930560.0, + "grad_norm": 2.6254780651913245, + "learning_rate": 7.612157445408987e-07, + "loss": 0.6881, + "num_input_tokens_seen": 129005590, + "step": 5998 + }, + { + "epoch": 0.7213371009439067, + "flos": 22313802498720.0, + "grad_norm": 3.228344293699498, + "learning_rate": 7.606042820747716e-07, + "loss": 0.7469, + "num_input_tokens_seen": 129021995, + "step": 5999 + }, + { + "epoch": 0.7214573438345457, + "flos": 18489584429280.0, + "grad_norm": 2.053353808701388, + "learning_rate": 7.599930076239889e-07, + "loss": 0.854, + "num_input_tokens_seen": 129039280, + "step": 6000 + }, + { + "epoch": 0.7215775867251849, + "flos": 35684679969600.0, + "grad_norm": 2.973489611956314, + "learning_rate": 7.593819212812818e-07, + "loss": 0.7026, + "num_input_tokens_seen": 129060860, + "step": 6001 + }, + { + "epoch": 0.721697829615824, + "flos": 20343235416000.0, + "grad_norm": 3.5194376299025283, + "learning_rate": 7.587710231393508e-07, + "loss": 0.7159, + "num_input_tokens_seen": 129079215, + "step": 6002 + }, + { + "epoch": 0.721818072506463, + "flos": 20200963810080.0, + "grad_norm": 2.526878632517621, + "learning_rate": 7.581603132908685e-07, + "loss": 0.8358, + "num_input_tokens_seen": 129097185, + "step": 6003 + }, + { + "epoch": 0.7219383153971022, + "flos": 18161026488000.0, + "grad_norm": 3.270098091310784, + "learning_rate": 7.575497918284795e-07, + "loss": 0.787, + "num_input_tokens_seen": 129114730, + "step": 6004 + }, + { + "epoch": 0.7220585582877412, + "flos": 17317289742240.0, + "grad_norm": 2.4036671040683184, + "learning_rate": 7.569394588447984e-07, + "loss": 0.7463, + "num_input_tokens_seen": 129131745, + "step": 6005 + }, + { + "epoch": 0.7221788011783803, + "flos": 16952002344480.0, + "grad_norm": 4.468846460114962, + "learning_rate": 7.563293144324146e-07, + "loss": 0.7832, + "num_input_tokens_seen": 129147295, + "step": 6006 + }, + { + "epoch": 0.7222990440690195, + "flos": 26248692144480.0, + "grad_norm": 2.588957437836551, + "learning_rate": 7.557193586838834e-07, + "loss": 0.8024, + "num_input_tokens_seen": 129162660, + "step": 6007 + }, + { + "epoch": 0.7224192869596585, + "flos": 17576626123200.0, + "grad_norm": 3.1875785369173437, + "learning_rate": 7.551095916917371e-07, + "loss": 0.7109, + "num_input_tokens_seen": 129179990, + "step": 6008 + }, + { + "epoch": 0.7225395298502976, + "flos": 12914098092960.0, + "grad_norm": 5.072501957396181, + "learning_rate": 7.545000135484758e-07, + "loss": 0.6645, + "num_input_tokens_seen": 129197425, + "step": 6009 + }, + { + "epoch": 0.7226597727409367, + "flos": 29601374441280.0, + "grad_norm": 2.6382309935676864, + "learning_rate": 7.538906243465714e-07, + "loss": 0.6324, + "num_input_tokens_seen": 129217560, + "step": 6010 + }, + { + "epoch": 0.7227800156315758, + "flos": 13752185035200.0, + "grad_norm": 2.807173878045247, + "learning_rate": 7.5328142417847e-07, + "loss": 0.7896, + "num_input_tokens_seen": 129234325, + "step": 6011 + }, + { + "epoch": 0.7229002585222148, + "flos": 20272898762880.0, + "grad_norm": 2.5771266969585556, + "learning_rate": 7.526724131365838e-07, + "loss": 0.6932, + "num_input_tokens_seen": 129255280, + "step": 6012 + }, + { + "epoch": 0.723020501412854, + "flos": 16557716800320.0, + "grad_norm": 2.8010848770601138, + "learning_rate": 7.520635913133017e-07, + "loss": 0.7022, + "num_input_tokens_seen": 129273910, + "step": 6013 + }, + { + "epoch": 0.7231407443034931, + "flos": 28507333566240.0, + "grad_norm": 2.5291792323027105, + "learning_rate": 7.514549588009798e-07, + "loss": 0.8226, + "num_input_tokens_seen": 129294785, + "step": 6014 + }, + { + "epoch": 0.7232609871941321, + "flos": 29965472406720.0, + "grad_norm": 8.198914469997119, + "learning_rate": 7.508465156919492e-07, + "loss": 0.7027, + "num_input_tokens_seen": 129318295, + "step": 6015 + }, + { + "epoch": 0.7233812300847713, + "flos": 16630357978560.0, + "grad_norm": 3.226726093717053, + "learning_rate": 7.502382620785083e-07, + "loss": 0.6101, + "num_input_tokens_seen": 129334845, + "step": 6016 + }, + { + "epoch": 0.7235014729754103, + "flos": 67162750648800.0, + "grad_norm": 0.8237434933395309, + "learning_rate": 7.496301980529289e-07, + "loss": 0.6247, + "num_input_tokens_seen": 129398055, + "step": 6017 + }, + { + "epoch": 0.7236217158660494, + "flos": 26904618594720.0, + "grad_norm": 2.3995699616439556, + "learning_rate": 7.490223237074547e-07, + "loss": 0.7426, + "num_input_tokens_seen": 129417765, + "step": 6018 + }, + { + "epoch": 0.7237419587566886, + "flos": 29381109211680.0, + "grad_norm": 2.1664633533258093, + "learning_rate": 7.484146391342989e-07, + "loss": 0.6596, + "num_input_tokens_seen": 129437560, + "step": 6019 + }, + { + "epoch": 0.7238622016473276, + "flos": 17796147957600.0, + "grad_norm": 3.7288467156312977, + "learning_rate": 7.478071444256484e-07, + "loss": 0.5653, + "num_input_tokens_seen": 129455320, + "step": 6020 + }, + { + "epoch": 0.7239824445379667, + "flos": 25702247838240.0, + "grad_norm": 2.8008490092701344, + "learning_rate": 7.471998396736579e-07, + "loss": 0.7915, + "num_input_tokens_seen": 129475700, + "step": 6021 + }, + { + "epoch": 0.7241026874286057, + "flos": 23115829040160.0, + "grad_norm": 3.362089825539335, + "learning_rate": 7.465927249704549e-07, + "loss": 0.7631, + "num_input_tokens_seen": 129493585, + "step": 6022 + }, + { + "epoch": 0.7242229303192449, + "flos": 20238027794400.0, + "grad_norm": 2.8542253994061566, + "learning_rate": 7.459858004081398e-07, + "loss": 0.7744, + "num_input_tokens_seen": 129511555, + "step": 6023 + }, + { + "epoch": 0.724343173209884, + "flos": 62223373547040.0, + "grad_norm": 0.6623382397868146, + "learning_rate": 7.453790660787815e-07, + "loss": 0.5794, + "num_input_tokens_seen": 129579650, + "step": 6024 + }, + { + "epoch": 0.724463416100523, + "flos": 34956038001600.0, + "grad_norm": 2.981610520674535, + "learning_rate": 7.447725220744214e-07, + "loss": 0.6362, + "num_input_tokens_seen": 129601895, + "step": 6025 + }, + { + "epoch": 0.7245836589911622, + "flos": 21840742765920.0, + "grad_norm": 2.2755801666293984, + "learning_rate": 7.441661684870717e-07, + "loss": 0.7702, + "num_input_tokens_seen": 129622150, + "step": 6026 + }, + { + "epoch": 0.7247039018818012, + "flos": 22973743283040.0, + "grad_norm": 2.1000109763085657, + "learning_rate": 7.435600054087152e-07, + "loss": 0.8194, + "num_input_tokens_seen": 129644315, + "step": 6027 + }, + { + "epoch": 0.7248241447724403, + "flos": 31677000466560.0, + "grad_norm": 12.307975751107238, + "learning_rate": 7.42954032931308e-07, + "loss": 0.7425, + "num_input_tokens_seen": 129665355, + "step": 6028 + }, + { + "epoch": 0.7249443876630794, + "flos": 34846778876160.0, + "grad_norm": 2.8647906604068125, + "learning_rate": 7.423482511467733e-07, + "loss": 0.7504, + "num_input_tokens_seen": 129686125, + "step": 6029 + }, + { + "epoch": 0.7250646305537185, + "flos": 26321519171520.0, + "grad_norm": 3.185196287134161, + "learning_rate": 7.417426601470099e-07, + "loss": 0.6512, + "num_input_tokens_seen": 129706485, + "step": 6030 + }, + { + "epoch": 0.7251848734443576, + "flos": 30038448112800.0, + "grad_norm": 3.8572031929706263, + "learning_rate": 7.411372600238841e-07, + "loss": 0.7835, + "num_input_tokens_seen": 129727100, + "step": 6031 + }, + { + "epoch": 0.7253051163349967, + "flos": 17759976047520.0, + "grad_norm": 4.187456728583564, + "learning_rate": 7.405320508692346e-07, + "loss": 0.7404, + "num_input_tokens_seen": 129745840, + "step": 6032 + }, + { + "epoch": 0.7254253592256358, + "flos": 12623570549760.0, + "grad_norm": 2.4213240458244525, + "learning_rate": 7.399270327748727e-07, + "loss": 0.7546, + "num_input_tokens_seen": 129763500, + "step": 6033 + }, + { + "epoch": 0.7255456021162748, + "flos": 27160237999680.0, + "grad_norm": 2.154666595689533, + "learning_rate": 7.39322205832577e-07, + "loss": 0.7414, + "num_input_tokens_seen": 129784390, + "step": 6034 + }, + { + "epoch": 0.725665845006914, + "flos": 21257829191520.0, + "grad_norm": 2.4226691143896195, + "learning_rate": 7.387175701341009e-07, + "loss": 0.8144, + "num_input_tokens_seen": 129803060, + "step": 6035 + }, + { + "epoch": 0.7257860878975531, + "flos": 16010937966240.0, + "grad_norm": 2.822076219585139, + "learning_rate": 7.381131257711659e-07, + "loss": 0.7237, + "num_input_tokens_seen": 129820165, + "step": 6036 + }, + { + "epoch": 0.7259063307881921, + "flos": 12112889286240.0, + "grad_norm": 2.254841258515716, + "learning_rate": 7.375088728354677e-07, + "loss": 0.8371, + "num_input_tokens_seen": 129835195, + "step": 6037 + }, + { + "epoch": 0.7260265736788313, + "flos": 30400055704320.0, + "grad_norm": 2.3024938456135198, + "learning_rate": 7.369048114186691e-07, + "loss": 0.6764, + "num_input_tokens_seen": 129856240, + "step": 6038 + }, + { + "epoch": 0.7261468165694703, + "flos": 21112100797920.0, + "grad_norm": 3.449898295354664, + "learning_rate": 7.363009416124055e-07, + "loss": 0.8343, + "num_input_tokens_seen": 129875565, + "step": 6039 + }, + { + "epoch": 0.7262670594601094, + "flos": 22274322480000.0, + "grad_norm": 4.384404129725487, + "learning_rate": 7.356972635082852e-07, + "loss": 0.6324, + "num_input_tokens_seen": 129894420, + "step": 6040 + }, + { + "epoch": 0.7263873023507486, + "flos": 25298669854080.0, + "grad_norm": 2.664956756584491, + "learning_rate": 7.35093777197884e-07, + "loss": 0.7552, + "num_input_tokens_seen": 129914490, + "step": 6041 + }, + { + "epoch": 0.7265075452413876, + "flos": 23844471008160.0, + "grad_norm": 3.5486749459580422, + "learning_rate": 7.344904827727525e-07, + "loss": 0.8563, + "num_input_tokens_seen": 129931670, + "step": 6042 + }, + { + "epoch": 0.7266277881320267, + "flos": 28683100859520.0, + "grad_norm": 3.595791842035971, + "learning_rate": 7.338873803244076e-07, + "loss": 0.7379, + "num_input_tokens_seen": 129946905, + "step": 6043 + }, + { + "epoch": 0.7267480310226658, + "flos": 24827914646400.0, + "grad_norm": 2.3475688086552116, + "learning_rate": 7.332844699443401e-07, + "loss": 0.8098, + "num_input_tokens_seen": 129965505, + "step": 6044 + }, + { + "epoch": 0.7268682739133049, + "flos": 27159494604480.0, + "grad_norm": 2.266859888742131, + "learning_rate": 7.326817517240121e-07, + "loss": 0.7533, + "num_input_tokens_seen": 129987210, + "step": 6045 + }, + { + "epoch": 0.7269885168039439, + "flos": 33459868763040.0, + "grad_norm": 2.055990060069774, + "learning_rate": 7.320792257548545e-07, + "loss": 0.8341, + "num_input_tokens_seen": 130008385, + "step": 6046 + }, + { + "epoch": 0.7271087596945831, + "flos": 24278868456960.0, + "grad_norm": 2.3204635416827406, + "learning_rate": 7.314768921282704e-07, + "loss": 0.7604, + "num_input_tokens_seen": 130029040, + "step": 6047 + }, + { + "epoch": 0.7272290025852222, + "flos": 23771086434720.0, + "grad_norm": 27.04019563863733, + "learning_rate": 7.30874750935633e-07, + "loss": 0.7199, + "num_input_tokens_seen": 130048725, + "step": 6048 + }, + { + "epoch": 0.7273492454758612, + "flos": 16696382939520.0, + "grad_norm": 2.405479307730718, + "learning_rate": 7.30272802268286e-07, + "loss": 0.7957, + "num_input_tokens_seen": 130065720, + "step": 6049 + }, + { + "epoch": 0.7274694883665004, + "flos": 27990890989920.0, + "grad_norm": 2.058927273741034, + "learning_rate": 7.29671046217547e-07, + "loss": 0.7638, + "num_input_tokens_seen": 130084830, + "step": 6050 + }, + { + "epoch": 0.7275897312571394, + "flos": 30329087165280.0, + "grad_norm": 2.265783896300326, + "learning_rate": 7.290694828746988e-07, + "loss": 0.8188, + "num_input_tokens_seen": 130104495, + "step": 6051 + }, + { + "epoch": 0.7277099741477785, + "flos": 19177296757920.0, + "grad_norm": 2.5583240628980572, + "learning_rate": 7.284681123310004e-07, + "loss": 0.8564, + "num_input_tokens_seen": 130123210, + "step": 6052 + }, + { + "epoch": 0.7278302170384175, + "flos": 20638409179200.0, + "grad_norm": 2.0900185199993873, + "learning_rate": 7.27866934677678e-07, + "loss": 0.7961, + "num_input_tokens_seen": 130142880, + "step": 6053 + }, + { + "epoch": 0.7279504599290567, + "flos": 19065807446880.0, + "grad_norm": 2.102215860333448, + "learning_rate": 7.272659500059297e-07, + "loss": 0.7834, + "num_input_tokens_seen": 130160220, + "step": 6054 + }, + { + "epoch": 0.7280707028196958, + "flos": 19035025152000.0, + "grad_norm": 3.4122097129458604, + "learning_rate": 7.266651584069264e-07, + "loss": 0.8026, + "num_input_tokens_seen": 130177885, + "step": 6055 + }, + { + "epoch": 0.7281909457103348, + "flos": 37142781640320.0, + "grad_norm": 1.9375026456753295, + "learning_rate": 7.260645599718045e-07, + "loss": 0.5722, + "num_input_tokens_seen": 130204240, + "step": 6056 + }, + { + "epoch": 0.728311188600974, + "flos": 20637591444480.0, + "grad_norm": 3.232036537643729, + "learning_rate": 7.254641547916767e-07, + "loss": 0.6697, + "num_input_tokens_seen": 130221735, + "step": 6057 + }, + { + "epoch": 0.728431431491613, + "flos": 28799310730080.0, + "grad_norm": 2.4129655932278773, + "learning_rate": 7.248639429576226e-07, + "loss": 0.6983, + "num_input_tokens_seen": 130241190, + "step": 6058 + }, + { + "epoch": 0.7285516743822521, + "flos": 25954521964800.0, + "grad_norm": 2.2564279079302585, + "learning_rate": 7.242639245606959e-07, + "loss": 0.7228, + "num_input_tokens_seen": 130260980, + "step": 6059 + }, + { + "epoch": 0.7286719172728913, + "flos": 16376485552320.0, + "grad_norm": 2.0130691292230285, + "learning_rate": 7.236640996919168e-07, + "loss": 0.8262, + "num_input_tokens_seen": 130280025, + "step": 6060 + }, + { + "epoch": 0.7287921601635303, + "flos": 21986731347840.0, + "grad_norm": 2.4341272348015495, + "learning_rate": 7.230644684422782e-07, + "loss": 0.7098, + "num_input_tokens_seen": 130300255, + "step": 6061 + }, + { + "epoch": 0.7289124030541694, + "flos": 24565456005600.0, + "grad_norm": 1.9341996910018446, + "learning_rate": 7.224650309027451e-07, + "loss": 0.8186, + "num_input_tokens_seen": 130320005, + "step": 6062 + }, + { + "epoch": 0.7290326459448085, + "flos": 21362962473600.0, + "grad_norm": 2.9991193114457966, + "learning_rate": 7.218657871642506e-07, + "loss": 0.6898, + "num_input_tokens_seen": 130338810, + "step": 6063 + }, + { + "epoch": 0.7291528888354476, + "flos": 18561222024000.0, + "grad_norm": 3.0723708535499363, + "learning_rate": 7.212667373177012e-07, + "loss": 0.6246, + "num_input_tokens_seen": 130353805, + "step": 6064 + }, + { + "epoch": 0.7292731317260867, + "flos": 18924019047840.0, + "grad_norm": 2.103870358392911, + "learning_rate": 7.206678814539704e-07, + "loss": 0.7549, + "num_input_tokens_seen": 130372105, + "step": 6065 + }, + { + "epoch": 0.7293933746167258, + "flos": 21037452452640.0, + "grad_norm": 2.0969338194191254, + "learning_rate": 7.20069219663904e-07, + "loss": 0.7293, + "num_input_tokens_seen": 130391990, + "step": 6066 + }, + { + "epoch": 0.7295136175073649, + "flos": 22421165966400.0, + "grad_norm": 2.313912656106607, + "learning_rate": 7.1947075203832e-07, + "loss": 0.7964, + "num_input_tokens_seen": 130411970, + "step": 6067 + }, + { + "epoch": 0.7296338603980039, + "flos": 56045496682080.0, + "grad_norm": 0.8745710820295157, + "learning_rate": 7.188724786680049e-07, + "loss": 0.6004, + "num_input_tokens_seen": 130472440, + "step": 6068 + }, + { + "epoch": 0.7297541032886431, + "flos": 25192161290880.0, + "grad_norm": 2.6288550996834794, + "learning_rate": 7.182743996437162e-07, + "loss": 0.757, + "num_input_tokens_seen": 130491975, + "step": 6069 + }, + { + "epoch": 0.7298743461792822, + "flos": 26429626034400.0, + "grad_norm": 3.192857466051836, + "learning_rate": 7.176765150561819e-07, + "loss": 0.6911, + "num_input_tokens_seen": 130510580, + "step": 6070 + }, + { + "epoch": 0.7299945890699212, + "flos": 19541023025760.0, + "grad_norm": 2.4974037040345105, + "learning_rate": 7.170788249961002e-07, + "loss": 0.7996, + "num_input_tokens_seen": 130529090, + "step": 6071 + }, + { + "epoch": 0.7301148319605604, + "flos": 22896158526720.0, + "grad_norm": 2.430838946821052, + "learning_rate": 7.164813295541418e-07, + "loss": 0.886, + "num_input_tokens_seen": 130548655, + "step": 6072 + }, + { + "epoch": 0.7302350748511994, + "flos": 25333206294720.0, + "grad_norm": 2.1543167957204536, + "learning_rate": 7.15884028820944e-07, + "loss": 0.704, + "num_input_tokens_seen": 130567710, + "step": 6073 + }, + { + "epoch": 0.7303553177418385, + "flos": 27779286314400.0, + "grad_norm": 2.6769261493378536, + "learning_rate": 7.152869228871185e-07, + "loss": 0.6014, + "num_input_tokens_seen": 130590195, + "step": 6074 + }, + { + "epoch": 0.7304755606324776, + "flos": 24391584370080.0, + "grad_norm": 2.826640629993341, + "learning_rate": 7.146900118432457e-07, + "loss": 0.725, + "num_input_tokens_seen": 130609940, + "step": 6075 + }, + { + "epoch": 0.7305958035231167, + "flos": 23806329100800.0, + "grad_norm": 2.0550347676549787, + "learning_rate": 7.140932957798753e-07, + "loss": 0.8606, + "num_input_tokens_seen": 130628380, + "step": 6076 + }, + { + "epoch": 0.7307160464137558, + "flos": 16703185005600.0, + "grad_norm": 2.902510339158312, + "learning_rate": 7.134967747875309e-07, + "loss": 0.7143, + "num_input_tokens_seen": 130646590, + "step": 6077 + }, + { + "epoch": 0.7308362893043949, + "flos": 21767135173920.0, + "grad_norm": 3.520751470633613, + "learning_rate": 7.129004489567014e-07, + "loss": 0.8172, + "num_input_tokens_seen": 130664070, + "step": 6078 + }, + { + "epoch": 0.730956532195034, + "flos": 10692409146240.0, + "grad_norm": 2.9410914117069877, + "learning_rate": 7.123043183778512e-07, + "loss": 0.7815, + "num_input_tokens_seen": 130681400, + "step": 6079 + }, + { + "epoch": 0.731076775085673, + "flos": 19763778629280.0, + "grad_norm": 3.0216762725907436, + "learning_rate": 7.117083831414114e-07, + "loss": 0.6528, + "num_input_tokens_seen": 130700675, + "step": 6080 + }, + { + "epoch": 0.7311970179763122, + "flos": 20417809421760.0, + "grad_norm": 2.4787693720458304, + "learning_rate": 7.11112643337787e-07, + "loss": 0.6994, + "num_input_tokens_seen": 130719720, + "step": 6081 + }, + { + "epoch": 0.7313172608669513, + "flos": 18487428583200.0, + "grad_norm": 2.7752463875752125, + "learning_rate": 7.10517099057349e-07, + "loss": 0.7651, + "num_input_tokens_seen": 130736670, + "step": 6082 + }, + { + "epoch": 0.7314375037575903, + "flos": 16157818622400.0, + "grad_norm": 3.801316163861764, + "learning_rate": 7.099217503904411e-07, + "loss": 0.612, + "num_input_tokens_seen": 130754525, + "step": 6083 + }, + { + "epoch": 0.7315577466482295, + "flos": 17942396727840.0, + "grad_norm": 2.5442138061587856, + "learning_rate": 7.093265974273788e-07, + "loss": 0.8961, + "num_input_tokens_seen": 130772420, + "step": 6084 + }, + { + "epoch": 0.7316779895388685, + "flos": 18379247380800.0, + "grad_norm": 2.5162767786154054, + "learning_rate": 7.087316402584447e-07, + "loss": 0.7218, + "num_input_tokens_seen": 130791245, + "step": 6085 + }, + { + "epoch": 0.7317982324295076, + "flos": 17902693690560.0, + "grad_norm": 2.1698912198625018, + "learning_rate": 7.081368789738953e-07, + "loss": 0.865, + "num_input_tokens_seen": 130808445, + "step": 6086 + }, + { + "epoch": 0.7319184753201466, + "flos": 27190611427200.0, + "grad_norm": 4.481056394345172, + "learning_rate": 7.075423136639537e-07, + "loss": 0.7817, + "num_input_tokens_seen": 130827700, + "step": 6087 + }, + { + "epoch": 0.7320387182107858, + "flos": 37394535390240.0, + "grad_norm": 3.8528429112196045, + "learning_rate": 7.069479444188149e-07, + "loss": 0.7513, + "num_input_tokens_seen": 130848290, + "step": 6088 + }, + { + "epoch": 0.7321589611014249, + "flos": 17833657979040.0, + "grad_norm": 2.589988094582314, + "learning_rate": 7.063537713286453e-07, + "loss": 0.8248, + "num_input_tokens_seen": 130865970, + "step": 6089 + }, + { + "epoch": 0.7322792039920639, + "flos": 26063074864800.0, + "grad_norm": 3.2022854332733655, + "learning_rate": 7.057597944835803e-07, + "loss": 0.8054, + "num_input_tokens_seen": 130885245, + "step": 6090 + }, + { + "epoch": 0.7323994468827031, + "flos": 25332574408800.0, + "grad_norm": 1.8861413550008153, + "learning_rate": 7.051660139737253e-07, + "loss": 0.746, + "num_input_tokens_seen": 130903465, + "step": 6091 + }, + { + "epoch": 0.7325196897733421, + "flos": 26868521024160.0, + "grad_norm": 2.260799346267163, + "learning_rate": 7.045724298891565e-07, + "loss": 0.7636, + "num_input_tokens_seen": 130922935, + "step": 6092 + }, + { + "epoch": 0.7326399326639812, + "flos": 25738865785440.0, + "grad_norm": 6.266053601300822, + "learning_rate": 7.039790423199192e-07, + "loss": 0.6914, + "num_input_tokens_seen": 130942575, + "step": 6093 + }, + { + "epoch": 0.7327601755546204, + "flos": 21002246956320.0, + "grad_norm": 2.8852370076532763, + "learning_rate": 7.033858513560322e-07, + "loss": 0.7806, + "num_input_tokens_seen": 130958870, + "step": 6094 + }, + { + "epoch": 0.7328804184452594, + "flos": 16266111334080.0, + "grad_norm": 3.0500182542098306, + "learning_rate": 7.027928570874794e-07, + "loss": 0.7674, + "num_input_tokens_seen": 130977060, + "step": 6095 + }, + { + "epoch": 0.7330006613358985, + "flos": 17832840244320.0, + "grad_norm": 2.768113046234015, + "learning_rate": 7.022000596042194e-07, + "loss": 0.8573, + "num_input_tokens_seen": 130994160, + "step": 6096 + }, + { + "epoch": 0.7331209042265376, + "flos": 22459939759680.0, + "grad_norm": 2.7917451325581912, + "learning_rate": 7.016074589961784e-07, + "loss": 0.8195, + "num_input_tokens_seen": 131012725, + "step": 6097 + }, + { + "epoch": 0.7332411471171767, + "flos": 33025211125920.0, + "grad_norm": 2.182753071832729, + "learning_rate": 7.01015055353253e-07, + "loss": 0.6706, + "num_input_tokens_seen": 131035150, + "step": 6098 + }, + { + "epoch": 0.7333613900078157, + "flos": 22710355398240.0, + "grad_norm": 2.6128389129562875, + "learning_rate": 7.004228487653123e-07, + "loss": 0.7834, + "num_input_tokens_seen": 131055955, + "step": 6099 + }, + { + "epoch": 0.7334816328984549, + "flos": 22314211366080.0, + "grad_norm": 4.349101111834149, + "learning_rate": 6.998308393221906e-07, + "loss": 0.7831, + "num_input_tokens_seen": 131074360, + "step": 6100 + }, + { + "epoch": 0.733601875789094, + "flos": 20706404137440.0, + "grad_norm": 3.1564192018384336, + "learning_rate": 6.992390271136977e-07, + "loss": 0.7144, + "num_input_tokens_seen": 131090070, + "step": 6101 + }, + { + "epoch": 0.733722118679733, + "flos": 22532097731040.0, + "grad_norm": 2.31571942547441, + "learning_rate": 6.986474122296094e-07, + "loss": 0.8532, + "num_input_tokens_seen": 131109185, + "step": 6102 + }, + { + "epoch": 0.7338423615703722, + "flos": 20055569944320.0, + "grad_norm": 5.289849677735381, + "learning_rate": 6.980559947596751e-07, + "loss": 0.7234, + "num_input_tokens_seen": 131127725, + "step": 6103 + }, + { + "epoch": 0.7339626044610112, + "flos": 21656649446400.0, + "grad_norm": 3.988681993517427, + "learning_rate": 6.974647747936109e-07, + "loss": 0.7559, + "num_input_tokens_seen": 131146110, + "step": 6104 + }, + { + "epoch": 0.7340828473516503, + "flos": 15246644464800.0, + "grad_norm": 4.466089270171458, + "learning_rate": 6.968737524211039e-07, + "loss": 0.8219, + "num_input_tokens_seen": 131162920, + "step": 6105 + }, + { + "epoch": 0.7342030902422895, + "flos": 22897347959040.0, + "grad_norm": 4.596903438726507, + "learning_rate": 6.962829277318132e-07, + "loss": 0.801, + "num_input_tokens_seen": 131182515, + "step": 6106 + }, + { + "epoch": 0.7343233331329285, + "flos": 25810837908000.0, + "grad_norm": 2.5825623212829827, + "learning_rate": 6.956923008153652e-07, + "loss": 0.8396, + "num_input_tokens_seen": 131202280, + "step": 6107 + }, + { + "epoch": 0.7344435760235676, + "flos": 18452706293760.0, + "grad_norm": 3.673049024828548, + "learning_rate": 6.951018717613593e-07, + "loss": 0.8437, + "num_input_tokens_seen": 131221295, + "step": 6108 + }, + { + "epoch": 0.7345638189142067, + "flos": 17614991049120.0, + "grad_norm": 2.1510929966698944, + "learning_rate": 6.945116406593614e-07, + "loss": 0.7837, + "num_input_tokens_seen": 131240150, + "step": 6109 + }, + { + "epoch": 0.7346840618048458, + "flos": 20231114219040.0, + "grad_norm": 2.957142836608255, + "learning_rate": 6.939216075989089e-07, + "loss": 0.7428, + "num_input_tokens_seen": 131258080, + "step": 6110 + }, + { + "epoch": 0.7348043046954849, + "flos": 28981545561600.0, + "grad_norm": 2.4516704158724036, + "learning_rate": 6.933317726695109e-07, + "loss": 0.6614, + "num_input_tokens_seen": 131279310, + "step": 6111 + }, + { + "epoch": 0.734924547586124, + "flos": 17905704441120.0, + "grad_norm": 4.062961493469032, + "learning_rate": 6.92742135960644e-07, + "loss": 0.7977, + "num_input_tokens_seen": 131297010, + "step": 6112 + }, + { + "epoch": 0.7350447904767631, + "flos": 63498162463200.0, + "grad_norm": 0.8363089404696414, + "learning_rate": 6.921526975617556e-07, + "loss": 0.5556, + "num_input_tokens_seen": 131356470, + "step": 6113 + }, + { + "epoch": 0.7351650333674021, + "flos": 21549211639200.0, + "grad_norm": 2.6145952282719858, + "learning_rate": 6.915634575622631e-07, + "loss": 0.7528, + "num_input_tokens_seen": 131374985, + "step": 6114 + }, + { + "epoch": 0.7352852762580413, + "flos": 18160468941600.0, + "grad_norm": 3.0556543873688313, + "learning_rate": 6.909744160515532e-07, + "loss": 0.7109, + "num_input_tokens_seen": 131393125, + "step": 6115 + }, + { + "epoch": 0.7354055191486804, + "flos": 38854012342080.0, + "grad_norm": 2.5467020923920014, + "learning_rate": 6.903855731189849e-07, + "loss": 0.6958, + "num_input_tokens_seen": 131415760, + "step": 6116 + }, + { + "epoch": 0.7355257620393194, + "flos": 16266891899040.0, + "grad_norm": 3.746291806710389, + "learning_rate": 6.897969288538825e-07, + "loss": 0.8263, + "num_input_tokens_seen": 131433705, + "step": 6117 + }, + { + "epoch": 0.7356460049299585, + "flos": 18088496819040.0, + "grad_norm": 3.519860594113032, + "learning_rate": 6.892084833455452e-07, + "loss": 0.812, + "num_input_tokens_seen": 131453275, + "step": 6118 + }, + { + "epoch": 0.7357662478205976, + "flos": 21294781666560.0, + "grad_norm": 1.7534257872154078, + "learning_rate": 6.886202366832384e-07, + "loss": 0.8369, + "num_input_tokens_seen": 131474960, + "step": 6119 + }, + { + "epoch": 0.7358864907112367, + "flos": 14226731558400.0, + "grad_norm": 2.3606507344922476, + "learning_rate": 6.880321889561987e-07, + "loss": 0.7341, + "num_input_tokens_seen": 131492935, + "step": 6120 + }, + { + "epoch": 0.7360067336018757, + "flos": 22277816437440.0, + "grad_norm": 3.461670808625869, + "learning_rate": 6.874443402536338e-07, + "loss": 0.6552, + "num_input_tokens_seen": 131512025, + "step": 6121 + }, + { + "epoch": 0.7361269764925149, + "flos": 25517931500160.0, + "grad_norm": 2.197747343028553, + "learning_rate": 6.868566906647177e-07, + "loss": 0.8062, + "num_input_tokens_seen": 131532975, + "step": 6122 + }, + { + "epoch": 0.736247219383154, + "flos": 20347472768640.0, + "grad_norm": 2.3924846521919676, + "learning_rate": 6.862692402785984e-07, + "loss": 0.8371, + "num_input_tokens_seen": 131553225, + "step": 6123 + }, + { + "epoch": 0.736367462273793, + "flos": 70239587457600.0, + "grad_norm": 0.7051001886101637, + "learning_rate": 6.856819891843899e-07, + "loss": 0.4956, + "num_input_tokens_seen": 131617930, + "step": 6124 + }, + { + "epoch": 0.7364877051644322, + "flos": 22380459345600.0, + "grad_norm": 2.637625983244048, + "learning_rate": 6.8509493747118e-07, + "loss": 0.7191, + "num_input_tokens_seen": 131636740, + "step": 6125 + }, + { + "epoch": 0.7366079480550712, + "flos": 12113223814080.0, + "grad_norm": 3.00496945459262, + "learning_rate": 6.845080852280221e-07, + "loss": 0.882, + "num_input_tokens_seen": 131653810, + "step": 6126 + }, + { + "epoch": 0.7367281909457103, + "flos": 15028051874400.0, + "grad_norm": 2.0048643799764583, + "learning_rate": 6.839214325439409e-07, + "loss": 0.7451, + "num_input_tokens_seen": 131671505, + "step": 6127 + }, + { + "epoch": 0.7368484338363495, + "flos": 23477027764320.0, + "grad_norm": 1.9070857982968228, + "learning_rate": 6.833349795079327e-07, + "loss": 0.7224, + "num_input_tokens_seen": 131690615, + "step": 6128 + }, + { + "epoch": 0.7369686767269885, + "flos": 27378533232000.0, + "grad_norm": 3.899926644637964, + "learning_rate": 6.827487262089613e-07, + "loss": 0.6871, + "num_input_tokens_seen": 131711120, + "step": 6129 + }, + { + "epoch": 0.7370889196176276, + "flos": 70193119524000.0, + "grad_norm": 0.872825918890061, + "learning_rate": 6.821626727359606e-07, + "loss": 0.5667, + "num_input_tokens_seen": 131776680, + "step": 6130 + }, + { + "epoch": 0.7372091625082667, + "flos": 18015186585120.0, + "grad_norm": 3.7710552396165267, + "learning_rate": 6.815768191778348e-07, + "loss": 0.7779, + "num_input_tokens_seen": 131794760, + "step": 6131 + }, + { + "epoch": 0.7373294053989058, + "flos": 33677717958240.0, + "grad_norm": 2.9187275462951234, + "learning_rate": 6.809911656234569e-07, + "loss": 0.7302, + "num_input_tokens_seen": 131816735, + "step": 6132 + }, + { + "epoch": 0.7374496482895448, + "flos": 21475604047200.0, + "grad_norm": 2.6142680268057834, + "learning_rate": 6.804057121616707e-07, + "loss": 0.785, + "num_input_tokens_seen": 131834940, + "step": 6133 + }, + { + "epoch": 0.737569891180184, + "flos": 24901224880320.0, + "grad_norm": 2.292602717803433, + "learning_rate": 6.798204588812888e-07, + "loss": 0.7196, + "num_input_tokens_seen": 131854355, + "step": 6134 + }, + { + "epoch": 0.7376901340708231, + "flos": 20634952391520.0, + "grad_norm": 2.314320543572056, + "learning_rate": 6.792354058710937e-07, + "loss": 0.7553, + "num_input_tokens_seen": 131871825, + "step": 6135 + }, + { + "epoch": 0.7378103769614621, + "flos": 23771755490400.0, + "grad_norm": 2.1842127487326124, + "learning_rate": 6.786505532198374e-07, + "loss": 0.6479, + "num_input_tokens_seen": 131890770, + "step": 6136 + }, + { + "epoch": 0.7379306198521013, + "flos": 22205472617280.0, + "grad_norm": 2.064103630170138, + "learning_rate": 6.780659010162411e-07, + "loss": 0.8547, + "num_input_tokens_seen": 131909720, + "step": 6137 + }, + { + "epoch": 0.7380508627427403, + "flos": 14882026122720.0, + "grad_norm": 2.0887746197421246, + "learning_rate": 6.774814493489975e-07, + "loss": 0.8336, + "num_input_tokens_seen": 131927395, + "step": 6138 + }, + { + "epoch": 0.7381711056333794, + "flos": 21654753788640.0, + "grad_norm": 2.389567022647638, + "learning_rate": 6.768971983067655e-07, + "loss": 0.6609, + "num_input_tokens_seen": 131947725, + "step": 6139 + }, + { + "epoch": 0.7382913485240186, + "flos": 52329682833600.0, + "grad_norm": 1.015888414702025, + "learning_rate": 6.763131479781772e-07, + "loss": 0.6773, + "num_input_tokens_seen": 131997485, + "step": 6140 + }, + { + "epoch": 0.7384115914146576, + "flos": 21767692720320.0, + "grad_norm": 2.5333599197933427, + "learning_rate": 6.757292984518316e-07, + "loss": 0.7652, + "num_input_tokens_seen": 132016885, + "step": 6141 + }, + { + "epoch": 0.7385318343052967, + "flos": 61406702120160.0, + "grad_norm": 0.7480538868388734, + "learning_rate": 6.751456498162981e-07, + "loss": 0.5629, + "num_input_tokens_seen": 132075920, + "step": 6142 + }, + { + "epoch": 0.7386520771959358, + "flos": 16989400856640.0, + "grad_norm": 2.314954331608743, + "learning_rate": 6.745622021601174e-07, + "loss": 0.8577, + "num_input_tokens_seen": 132092945, + "step": 6143 + }, + { + "epoch": 0.7387723200865749, + "flos": 18744051571680.0, + "grad_norm": 2.6491037616619075, + "learning_rate": 6.739789555717954e-07, + "loss": 0.6949, + "num_input_tokens_seen": 132109920, + "step": 6144 + }, + { + "epoch": 0.738892562977214, + "flos": 22493249598240.0, + "grad_norm": 2.3810465287713454, + "learning_rate": 6.733959101398124e-07, + "loss": 0.773, + "num_input_tokens_seen": 132128050, + "step": 6145 + }, + { + "epoch": 0.7390128058678531, + "flos": 21470586129600.0, + "grad_norm": 11.070957550014306, + "learning_rate": 6.728130659526143e-07, + "loss": 0.8122, + "num_input_tokens_seen": 132145860, + "step": 6146 + }, + { + "epoch": 0.7391330487584922, + "flos": 25739274652800.0, + "grad_norm": 2.692012812636381, + "learning_rate": 6.7223042309862e-07, + "loss": 0.707, + "num_input_tokens_seen": 132166060, + "step": 6147 + }, + { + "epoch": 0.7392532916491312, + "flos": 28326065148480.0, + "grad_norm": 2.3900776155390395, + "learning_rate": 6.716479816662144e-07, + "loss": 0.7357, + "num_input_tokens_seen": 132187790, + "step": 6148 + }, + { + "epoch": 0.7393735345397703, + "flos": 23552122146720.0, + "grad_norm": 2.508808354191835, + "learning_rate": 6.710657417437531e-07, + "loss": 0.7322, + "num_input_tokens_seen": 132207495, + "step": 6149 + }, + { + "epoch": 0.7394937774304094, + "flos": 19946310818880.0, + "grad_norm": 2.6756224647800453, + "learning_rate": 6.704837034195628e-07, + "loss": 0.8062, + "num_input_tokens_seen": 132225960, + "step": 6150 + }, + { + "epoch": 0.7396140203210485, + "flos": 23444498490720.0, + "grad_norm": 2.1932741523959924, + "learning_rate": 6.699018667819376e-07, + "loss": 0.8488, + "num_input_tokens_seen": 132245360, + "step": 6151 + }, + { + "epoch": 0.7397342632116876, + "flos": 25518786404640.0, + "grad_norm": 2.2197078732577715, + "learning_rate": 6.693202319191415e-07, + "loss": 0.729, + "num_input_tokens_seen": 132267095, + "step": 6152 + }, + { + "epoch": 0.7398545061023267, + "flos": 24719621934720.0, + "grad_norm": 2.4150371867726745, + "learning_rate": 6.687387989194084e-07, + "loss": 0.7506, + "num_input_tokens_seen": 132286610, + "step": 6153 + }, + { + "epoch": 0.7399747489929658, + "flos": 16485187131360.0, + "grad_norm": 2.2650047768717196, + "learning_rate": 6.681575678709404e-07, + "loss": 0.7947, + "num_input_tokens_seen": 132305300, + "step": 6154 + }, + { + "epoch": 0.7400949918836048, + "flos": 24062580391680.0, + "grad_norm": 3.1654780482746814, + "learning_rate": 6.67576538861911e-07, + "loss": 0.708, + "num_input_tokens_seen": 132323875, + "step": 6155 + }, + { + "epoch": 0.740215234774244, + "flos": 21771521205600.0, + "grad_norm": 1.6300295243192953, + "learning_rate": 6.669957119804612e-07, + "loss": 0.8215, + "num_input_tokens_seen": 132345510, + "step": 6156 + }, + { + "epoch": 0.7403354776648831, + "flos": 18706504380480.0, + "grad_norm": 5.631108337043039, + "learning_rate": 6.66415087314702e-07, + "loss": 0.7222, + "num_input_tokens_seen": 132360465, + "step": 6157 + }, + { + "epoch": 0.7404557205555221, + "flos": 16885568516160.0, + "grad_norm": 2.7149419313968637, + "learning_rate": 6.65834664952714e-07, + "loss": 0.7333, + "num_input_tokens_seen": 132377915, + "step": 6158 + }, + { + "epoch": 0.7405759634461613, + "flos": 21184110090240.0, + "grad_norm": 2.176955503759859, + "learning_rate": 6.652544449825457e-07, + "loss": 0.7628, + "num_input_tokens_seen": 132398170, + "step": 6159 + }, + { + "epoch": 0.7406962063368003, + "flos": 20450710392960.0, + "grad_norm": 2.340162456122421, + "learning_rate": 6.646744274922182e-07, + "loss": 0.7668, + "num_input_tokens_seen": 132416615, + "step": 6160 + }, + { + "epoch": 0.7408164492274394, + "flos": 19763704289760.0, + "grad_norm": 4.060031316852524, + "learning_rate": 6.640946125697171e-07, + "loss": 0.7553, + "num_input_tokens_seen": 132434145, + "step": 6161 + }, + { + "epoch": 0.7409366921180786, + "flos": 29162925488640.0, + "grad_norm": 2.502349607691249, + "learning_rate": 6.635150003030017e-07, + "loss": 0.7578, + "num_input_tokens_seen": 132452670, + "step": 6162 + }, + { + "epoch": 0.7410569350087176, + "flos": 22897236449760.0, + "grad_norm": 3.0085716181366733, + "learning_rate": 6.629355907799981e-07, + "loss": 0.8615, + "num_input_tokens_seen": 132472475, + "step": 6163 + }, + { + "epoch": 0.7411771778993567, + "flos": 30396487407360.0, + "grad_norm": 2.1068757121068793, + "learning_rate": 6.623563840886015e-07, + "loss": 0.6909, + "num_input_tokens_seen": 132493400, + "step": 6164 + }, + { + "epoch": 0.7412974207899958, + "flos": 20492792294880.0, + "grad_norm": 3.5230632845467573, + "learning_rate": 6.617773803166795e-07, + "loss": 0.6977, + "num_input_tokens_seen": 132511725, + "step": 6165 + }, + { + "epoch": 0.7414176636806349, + "flos": 22059149507520.0, + "grad_norm": 2.3978574964799915, + "learning_rate": 6.611985795520634e-07, + "loss": 0.8183, + "num_input_tokens_seen": 132530270, + "step": 6166 + }, + { + "epoch": 0.7415379065712739, + "flos": 25119222754560.0, + "grad_norm": 2.619172849179564, + "learning_rate": 6.606199818825588e-07, + "loss": 0.7727, + "num_input_tokens_seen": 132550725, + "step": 6167 + }, + { + "epoch": 0.7416581494619131, + "flos": 16847723966880.0, + "grad_norm": 2.4337428553965696, + "learning_rate": 6.600415873959377e-07, + "loss": 0.8187, + "num_input_tokens_seen": 132568600, + "step": 6168 + }, + { + "epoch": 0.7417783923525522, + "flos": 28398074440800.0, + "grad_norm": 2.728105081352238, + "learning_rate": 6.594633961799437e-07, + "loss": 0.6436, + "num_input_tokens_seen": 132587640, + "step": 6169 + }, + { + "epoch": 0.7418986352431912, + "flos": 20055681453600.0, + "grad_norm": 2.1990050623663295, + "learning_rate": 6.588854083222857e-07, + "loss": 0.8182, + "num_input_tokens_seen": 132607075, + "step": 6170 + }, + { + "epoch": 0.7420188781338304, + "flos": 18233593326720.0, + "grad_norm": 3.9069717350824167, + "learning_rate": 6.583076239106444e-07, + "loss": 0.8095, + "num_input_tokens_seen": 132625580, + "step": 6171 + }, + { + "epoch": 0.7421391210244694, + "flos": 13752556732800.0, + "grad_norm": 2.857065070319547, + "learning_rate": 6.577300430326707e-07, + "loss": 0.7586, + "num_input_tokens_seen": 132640525, + "step": 6172 + }, + { + "epoch": 0.7422593639151085, + "flos": 15939077352960.0, + "grad_norm": 2.4566054225345932, + "learning_rate": 6.571526657759821e-07, + "loss": 0.7253, + "num_input_tokens_seen": 132656265, + "step": 6173 + }, + { + "epoch": 0.7423796068057477, + "flos": 30071051725920.0, + "grad_norm": 2.2811748544518133, + "learning_rate": 6.565754922281663e-07, + "loss": 0.7125, + "num_input_tokens_seen": 132678510, + "step": 6174 + }, + { + "epoch": 0.7424998496963867, + "flos": 20492643615840.0, + "grad_norm": 2.2140211835087285, + "learning_rate": 6.559985224767801e-07, + "loss": 0.7828, + "num_input_tokens_seen": 132696385, + "step": 6175 + }, + { + "epoch": 0.7426200925870258, + "flos": 21840333898560.0, + "grad_norm": 3.448410751729916, + "learning_rate": 6.55421756609349e-07, + "loss": 0.7578, + "num_input_tokens_seen": 132714640, + "step": 6176 + }, + { + "epoch": 0.7427403354776649, + "flos": 26394160349760.0, + "grad_norm": 2.6515420817374116, + "learning_rate": 6.54845194713369e-07, + "loss": 0.788, + "num_input_tokens_seen": 132735590, + "step": 6177 + }, + { + "epoch": 0.742860578368304, + "flos": 19869915494880.0, + "grad_norm": 2.358884444617642, + "learning_rate": 6.542688368763034e-07, + "loss": 0.8023, + "num_input_tokens_seen": 132753995, + "step": 6178 + }, + { + "epoch": 0.742980821258943, + "flos": 24791668396800.0, + "grad_norm": 2.2746467651459565, + "learning_rate": 6.536926831855854e-07, + "loss": 0.7689, + "num_input_tokens_seen": 132773160, + "step": 6179 + }, + { + "epoch": 0.7431010641495821, + "flos": 25192532988480.0, + "grad_norm": 3.295865017191671, + "learning_rate": 6.531167337286165e-07, + "loss": 0.7328, + "num_input_tokens_seen": 132793180, + "step": 6180 + }, + { + "epoch": 0.7432213070402213, + "flos": 21731297791680.0, + "grad_norm": 1.6406092325088675, + "learning_rate": 6.52540988592768e-07, + "loss": 0.7981, + "num_input_tokens_seen": 132814200, + "step": 6181 + }, + { + "epoch": 0.7433415499308603, + "flos": 14772915676320.0, + "grad_norm": 2.8576044322877623, + "learning_rate": 6.519654478653814e-07, + "loss": 0.8356, + "num_input_tokens_seen": 132832565, + "step": 6182 + }, + { + "epoch": 0.7434617928214994, + "flos": 67059921891840.0, + "grad_norm": 0.7708573122848451, + "learning_rate": 6.51390111633763e-07, + "loss": 0.5604, + "num_input_tokens_seen": 132897840, + "step": 6183 + }, + { + "epoch": 0.7435820357121385, + "flos": 27338198308800.0, + "grad_norm": 2.7080087809624573, + "learning_rate": 6.508149799851932e-07, + "loss": 0.7611, + "num_input_tokens_seen": 132919505, + "step": 6184 + }, + { + "epoch": 0.7437022786027776, + "flos": 23953321266240.0, + "grad_norm": 2.5463753426609497, + "learning_rate": 6.502400530069183e-07, + "loss": 0.6097, + "num_input_tokens_seen": 132939390, + "step": 6185 + }, + { + "epoch": 0.7438225214934167, + "flos": 21835353150720.0, + "grad_norm": 2.328976059577252, + "learning_rate": 6.496653307861535e-07, + "loss": 0.6855, + "num_input_tokens_seen": 132960060, + "step": 6186 + }, + { + "epoch": 0.7439427643840558, + "flos": 20201261168160.0, + "grad_norm": 3.1336760056246304, + "learning_rate": 6.490908134100857e-07, + "loss": 0.6585, + "num_input_tokens_seen": 132978525, + "step": 6187 + }, + { + "epoch": 0.7440630072746949, + "flos": 20820160803840.0, + "grad_norm": 2.75733515685087, + "learning_rate": 6.48516500965866e-07, + "loss": 0.6887, + "num_input_tokens_seen": 132998460, + "step": 6188 + }, + { + "epoch": 0.7441832501653339, + "flos": 26465723604960.0, + "grad_norm": 1.9814166748022917, + "learning_rate": 6.479423935406192e-07, + "loss": 0.8158, + "num_input_tokens_seen": 133018445, + "step": 6189 + }, + { + "epoch": 0.7443034930559731, + "flos": 68504642448960.0, + "grad_norm": 1.1624421997467553, + "learning_rate": 6.473684912214357e-07, + "loss": 0.6196, + "num_input_tokens_seen": 133082005, + "step": 6190 + }, + { + "epoch": 0.7444237359466122, + "flos": 18628213398720.0, + "grad_norm": 2.51057191951707, + "learning_rate": 6.467947940953778e-07, + "loss": 0.7006, + "num_input_tokens_seen": 133100530, + "step": 6191 + }, + { + "epoch": 0.7445439788372512, + "flos": 22784892234240.0, + "grad_norm": 2.6847923300791545, + "learning_rate": 6.462213022494732e-07, + "loss": 0.7263, + "num_input_tokens_seen": 133119775, + "step": 6192 + }, + { + "epoch": 0.7446642217278904, + "flos": 66950179559520.0, + "grad_norm": 0.776032688848464, + "learning_rate": 6.456480157707201e-07, + "loss": 0.6091, + "num_input_tokens_seen": 133184550, + "step": 6193 + }, + { + "epoch": 0.7447844646185294, + "flos": 17392607143200.0, + "grad_norm": 6.579597507677587, + "learning_rate": 6.450749347460866e-07, + "loss": 0.8509, + "num_input_tokens_seen": 133201525, + "step": 6194 + }, + { + "epoch": 0.7449047075091685, + "flos": 26577770462400.0, + "grad_norm": 2.6945320215663764, + "learning_rate": 6.445020592625083e-07, + "loss": 0.7894, + "num_input_tokens_seen": 133222175, + "step": 6195 + }, + { + "epoch": 0.7450249503998077, + "flos": 14150336234400.0, + "grad_norm": 4.813035216851963, + "learning_rate": 6.4392938940689e-07, + "loss": 0.8041, + "num_input_tokens_seen": 133237590, + "step": 6196 + }, + { + "epoch": 0.7451451932904467, + "flos": 19578644556480.0, + "grad_norm": 3.434659581980694, + "learning_rate": 6.433569252661049e-07, + "loss": 0.7153, + "num_input_tokens_seen": 133255590, + "step": 6197 + }, + { + "epoch": 0.7452654361810858, + "flos": 12477842156160.0, + "grad_norm": 3.151211181788908, + "learning_rate": 6.427846669269952e-07, + "loss": 0.7146, + "num_input_tokens_seen": 133273210, + "step": 6198 + }, + { + "epoch": 0.7453856790717249, + "flos": 22095730284960.0, + "grad_norm": 3.394581679133108, + "learning_rate": 6.422126144763729e-07, + "loss": 0.8252, + "num_input_tokens_seen": 133292600, + "step": 6199 + }, + { + "epoch": 0.745505921962364, + "flos": 19982148201120.0, + "grad_norm": 4.135992488920671, + "learning_rate": 6.416407680010174e-07, + "loss": 0.7695, + "num_input_tokens_seen": 133306960, + "step": 6200 + }, + { + "epoch": 0.745626164853003, + "flos": 24642446045760.0, + "grad_norm": 5.873319344880978, + "learning_rate": 6.410691275876774e-07, + "loss": 0.8146, + "num_input_tokens_seen": 133326380, + "step": 6201 + }, + { + "epoch": 0.7457464077436422, + "flos": 14518299854880.0, + "grad_norm": 2.5820616288930074, + "learning_rate": 6.404976933230704e-07, + "loss": 0.7683, + "num_input_tokens_seen": 133342900, + "step": 6202 + }, + { + "epoch": 0.7458666506342813, + "flos": 33972259835520.0, + "grad_norm": 2.696052604456392, + "learning_rate": 6.399264652938813e-07, + "loss": 0.7288, + "num_input_tokens_seen": 133363805, + "step": 6203 + }, + { + "epoch": 0.7459868935249203, + "flos": 24244257676800.0, + "grad_norm": 2.3177706477899913, + "learning_rate": 6.393554435867679e-07, + "loss": 0.7436, + "num_input_tokens_seen": 133384655, + "step": 6204 + }, + { + "epoch": 0.7461071364155595, + "flos": 21877509392160.0, + "grad_norm": 6.542404359986496, + "learning_rate": 6.387846282883502e-07, + "loss": 0.837, + "num_input_tokens_seen": 133401185, + "step": 6205 + }, + { + "epoch": 0.7462273793061985, + "flos": 22856864356800.0, + "grad_norm": 2.740401596655079, + "learning_rate": 6.38214019485223e-07, + "loss": 0.7701, + "num_input_tokens_seen": 133420010, + "step": 6206 + }, + { + "epoch": 0.7463476221968376, + "flos": 19939768941120.0, + "grad_norm": 2.85442119122495, + "learning_rate": 6.376436172639461e-07, + "loss": 0.7186, + "num_input_tokens_seen": 133437855, + "step": 6207 + }, + { + "epoch": 0.7464678650874768, + "flos": 16812704319360.0, + "grad_norm": 3.621465209212377, + "learning_rate": 6.370734217110487e-07, + "loss": 0.6533, + "num_input_tokens_seen": 133456600, + "step": 6208 + }, + { + "epoch": 0.7465881079781158, + "flos": 48031295672160.0, + "grad_norm": 1.8518760190375798, + "learning_rate": 6.36503432913031e-07, + "loss": 0.6416, + "num_input_tokens_seen": 133479745, + "step": 6209 + }, + { + "epoch": 0.7467083508687549, + "flos": 19649129888640.0, + "grad_norm": 2.67553513780339, + "learning_rate": 6.359336509563569e-07, + "loss": 0.694, + "num_input_tokens_seen": 133495765, + "step": 6210 + }, + { + "epoch": 0.7468285937593939, + "flos": 17870015737920.0, + "grad_norm": 2.7159784155087685, + "learning_rate": 6.353640759274641e-07, + "loss": 0.8071, + "num_input_tokens_seen": 133514655, + "step": 6211 + }, + { + "epoch": 0.7469488366500331, + "flos": 23108097730080.0, + "grad_norm": 3.3191954405394024, + "learning_rate": 6.347947079127556e-07, + "loss": 0.7545, + "num_input_tokens_seen": 133532555, + "step": 6212 + }, + { + "epoch": 0.7470690795406721, + "flos": 16667161774560.0, + "grad_norm": 2.5780757090162574, + "learning_rate": 6.342255469986053e-07, + "loss": 0.7699, + "num_input_tokens_seen": 133551300, + "step": 6213 + }, + { + "epoch": 0.7471893224313112, + "flos": 25156658436480.0, + "grad_norm": 2.2601097590032433, + "learning_rate": 6.336565932713533e-07, + "loss": 0.7638, + "num_input_tokens_seen": 133570725, + "step": 6214 + }, + { + "epoch": 0.7473095653219504, + "flos": 22493955823680.0, + "grad_norm": 7.598387004944161, + "learning_rate": 6.330878468173088e-07, + "loss": 0.7768, + "num_input_tokens_seen": 133590790, + "step": 6215 + }, + { + "epoch": 0.7474298082125894, + "flos": 18088236630720.0, + "grad_norm": 2.1730030889932417, + "learning_rate": 6.32519307722752e-07, + "loss": 0.7299, + "num_input_tokens_seen": 133608685, + "step": 6216 + }, + { + "epoch": 0.7475500511032285, + "flos": 61998350588160.0, + "grad_norm": 0.8777955221107563, + "learning_rate": 6.31950976073929e-07, + "loss": 0.5486, + "num_input_tokens_seen": 133662775, + "step": 6217 + }, + { + "epoch": 0.7476702939938676, + "flos": 17760310575360.0, + "grad_norm": 2.8557619708803066, + "learning_rate": 6.31382851957055e-07, + "loss": 0.8109, + "num_input_tokens_seen": 133679595, + "step": 6218 + }, + { + "epoch": 0.7477905368845067, + "flos": 27887876384160.0, + "grad_norm": 2.9872700494988416, + "learning_rate": 6.308149354583143e-07, + "loss": 0.7131, + "num_input_tokens_seen": 133699000, + "step": 6219 + }, + { + "epoch": 0.7479107797751458, + "flos": 26831865907200.0, + "grad_norm": 3.901579238800697, + "learning_rate": 6.302472266638586e-07, + "loss": 0.8189, + "num_input_tokens_seen": 133719540, + "step": 6220 + }, + { + "epoch": 0.7480310226657849, + "flos": 33894377721120.0, + "grad_norm": 2.349291645995779, + "learning_rate": 6.296797256598101e-07, + "loss": 0.7002, + "num_input_tokens_seen": 133741020, + "step": 6221 + }, + { + "epoch": 0.748151265556424, + "flos": 24790999341120.0, + "grad_norm": 2.6738532701177955, + "learning_rate": 6.291124325322576e-07, + "loss": 0.8163, + "num_input_tokens_seen": 133761145, + "step": 6222 + }, + { + "epoch": 0.748271508447063, + "flos": 38344037304000.0, + "grad_norm": 2.4080375868186215, + "learning_rate": 6.285453473672595e-07, + "loss": 0.6261, + "num_input_tokens_seen": 133783715, + "step": 6223 + }, + { + "epoch": 0.7483917513377022, + "flos": 21511032562080.0, + "grad_norm": 2.6410346983046105, + "learning_rate": 6.279784702508415e-07, + "loss": 0.7549, + "num_input_tokens_seen": 133804465, + "step": 6224 + }, + { + "epoch": 0.7485119942283412, + "flos": 62226347127840.0, + "grad_norm": 0.7925627848427108, + "learning_rate": 6.274118012689979e-07, + "loss": 0.5849, + "num_input_tokens_seen": 133866365, + "step": 6225 + }, + { + "epoch": 0.7486322371189803, + "flos": 29894726886240.0, + "grad_norm": 1.6413866001062336, + "learning_rate": 6.268453405076943e-07, + "loss": 0.686, + "num_input_tokens_seen": 133888760, + "step": 6226 + }, + { + "epoch": 0.7487524800096195, + "flos": 18922309238880.0, + "grad_norm": 3.281498926802202, + "learning_rate": 6.262790880528592e-07, + "loss": 0.8232, + "num_input_tokens_seen": 133906380, + "step": 6227 + }, + { + "epoch": 0.7488727229002585, + "flos": 18670778507520.0, + "grad_norm": 3.31085308087945, + "learning_rate": 6.257130439903951e-07, + "loss": 0.8005, + "num_input_tokens_seen": 133922875, + "step": 6228 + }, + { + "epoch": 0.7489929657908976, + "flos": 23589594998400.0, + "grad_norm": 2.1694307913237303, + "learning_rate": 6.251472084061695e-07, + "loss": 0.8162, + "num_input_tokens_seen": 133941795, + "step": 6229 + }, + { + "epoch": 0.7491132086815367, + "flos": 20521976290080.0, + "grad_norm": 21.78874036154841, + "learning_rate": 6.245815813860191e-07, + "loss": 0.8901, + "num_input_tokens_seen": 133957305, + "step": 6230 + }, + { + "epoch": 0.7492334515721758, + "flos": 22970137816320.0, + "grad_norm": 5.201560945244211, + "learning_rate": 6.240161630157495e-07, + "loss": 0.6994, + "num_input_tokens_seen": 133976660, + "step": 6231 + }, + { + "epoch": 0.7493536944628149, + "flos": 16375556308320.0, + "grad_norm": 2.9684957294995606, + "learning_rate": 6.23450953381133e-07, + "loss": 0.7015, + "num_input_tokens_seen": 133994750, + "step": 6232 + }, + { + "epoch": 0.749473937353454, + "flos": 15316758099360.0, + "grad_norm": 3.537175161591421, + "learning_rate": 6.228859525679131e-07, + "loss": 0.682, + "num_input_tokens_seen": 134009165, + "step": 6233 + }, + { + "epoch": 0.7495941802440931, + "flos": 18924093387360.0, + "grad_norm": 3.0404953116412305, + "learning_rate": 6.223211606617986e-07, + "loss": 0.7962, + "num_input_tokens_seen": 134027585, + "step": 6234 + }, + { + "epoch": 0.7497144231347321, + "flos": 22460348627040.0, + "grad_norm": 2.0140623200670253, + "learning_rate": 6.217565777484701e-07, + "loss": 0.8391, + "num_input_tokens_seen": 134046680, + "step": 6235 + }, + { + "epoch": 0.7498346660253713, + "flos": 24208829161920.0, + "grad_norm": 2.10623873925524, + "learning_rate": 6.211922039135722e-07, + "loss": 0.8055, + "num_input_tokens_seen": 134066825, + "step": 6236 + }, + { + "epoch": 0.7499549089160104, + "flos": 24352104351360.0, + "grad_norm": 2.6407492343186787, + "learning_rate": 6.206280392427201e-07, + "loss": 0.8144, + "num_input_tokens_seen": 134086410, + "step": 6237 + }, + { + "epoch": 0.7500751518066494, + "flos": 34008320236320.0, + "grad_norm": 2.4306761272739448, + "learning_rate": 6.200640838214983e-07, + "loss": 0.7376, + "num_input_tokens_seen": 134109185, + "step": 6238 + }, + { + "epoch": 0.7501953946972886, + "flos": 18816767089440.0, + "grad_norm": 2.326902132370498, + "learning_rate": 6.195003377354578e-07, + "loss": 0.6711, + "num_input_tokens_seen": 134128455, + "step": 6239 + }, + { + "epoch": 0.7503156375879276, + "flos": 20228475166080.0, + "grad_norm": 13.676387648116826, + "learning_rate": 6.189368010701183e-07, + "loss": 0.7344, + "num_input_tokens_seen": 134145515, + "step": 6240 + }, + { + "epoch": 0.7504358804785667, + "flos": 13461025606080.0, + "grad_norm": 2.8526701147788613, + "learning_rate": 6.183734739109683e-07, + "loss": 0.7669, + "num_input_tokens_seen": 134163335, + "step": 6241 + }, + { + "epoch": 0.7505561233692057, + "flos": 29418767912160.0, + "grad_norm": 3.4127476864310133, + "learning_rate": 6.178103563434629e-07, + "loss": 0.6935, + "num_input_tokens_seen": 134182335, + "step": 6242 + }, + { + "epoch": 0.7506763662598449, + "flos": 20273493479040.0, + "grad_norm": 3.8564372754490193, + "learning_rate": 6.172474484530283e-07, + "loss": 0.8403, + "num_input_tokens_seen": 134201070, + "step": 6243 + }, + { + "epoch": 0.750796609150484, + "flos": 37177318080960.0, + "grad_norm": 2.1090679934282157, + "learning_rate": 6.166847503250563e-07, + "loss": 0.7629, + "num_input_tokens_seen": 134223310, + "step": 6244 + }, + { + "epoch": 0.750916852041123, + "flos": 19581469458240.0, + "grad_norm": 5.746266355327361, + "learning_rate": 6.161222620449078e-07, + "loss": 0.7935, + "num_input_tokens_seen": 134242085, + "step": 6245 + }, + { + "epoch": 0.7510370949317622, + "flos": 25076508966720.0, + "grad_norm": 3.482712832462774, + "learning_rate": 6.155599836979117e-07, + "loss": 0.8018, + "num_input_tokens_seen": 134260770, + "step": 6246 + }, + { + "epoch": 0.7511573378224012, + "flos": 19035136661280.0, + "grad_norm": 3.236454244675932, + "learning_rate": 6.149979153693649e-07, + "loss": 0.8184, + "num_input_tokens_seen": 134278025, + "step": 6247 + }, + { + "epoch": 0.7512775807130403, + "flos": 19909246834560.0, + "grad_norm": 2.577877425820027, + "learning_rate": 6.144360571445343e-07, + "loss": 0.769, + "num_input_tokens_seen": 134297170, + "step": 6248 + }, + { + "epoch": 0.7513978236036795, + "flos": 20710306962240.0, + "grad_norm": 2.2842296841137446, + "learning_rate": 6.138744091086509e-07, + "loss": 0.8021, + "num_input_tokens_seen": 134316755, + "step": 6249 + }, + { + "epoch": 0.7515180664943185, + "flos": 27523629739680.0, + "grad_norm": 2.9801749988350754, + "learning_rate": 6.133129713469183e-07, + "loss": 0.731, + "num_input_tokens_seen": 134335960, + "step": 6250 + }, + { + "epoch": 0.7516383093849576, + "flos": 33754596489120.0, + "grad_norm": 3.6810209607825795, + "learning_rate": 6.127517439445053e-07, + "loss": 0.6425, + "num_input_tokens_seen": 134356805, + "step": 6251 + }, + { + "epoch": 0.7517585522755967, + "flos": 29703236784480.0, + "grad_norm": 2.61411553766288, + "learning_rate": 6.121907269865498e-07, + "loss": 0.8232, + "num_input_tokens_seen": 134376295, + "step": 6252 + }, + { + "epoch": 0.7518787951662358, + "flos": 69708053958720.0, + "grad_norm": 0.9311347083595621, + "learning_rate": 6.116299205581577e-07, + "loss": 0.671, + "num_input_tokens_seen": 134431125, + "step": 6253 + }, + { + "epoch": 0.7519990380568748, + "flos": 34154011460160.0, + "grad_norm": 4.29208095489121, + "learning_rate": 6.110693247444018e-07, + "loss": 0.6854, + "num_input_tokens_seen": 134452960, + "step": 6254 + }, + { + "epoch": 0.752119280947514, + "flos": 21694679844480.0, + "grad_norm": 2.2320390775260894, + "learning_rate": 6.105089396303258e-07, + "loss": 0.8282, + "num_input_tokens_seen": 134471350, + "step": 6255 + }, + { + "epoch": 0.7522395238381531, + "flos": 32695649601120.0, + "grad_norm": 2.5258275101847425, + "learning_rate": 6.099487653009383e-07, + "loss": 0.7558, + "num_input_tokens_seen": 134490695, + "step": 6256 + }, + { + "epoch": 0.7523597667287921, + "flos": 23442416984160.0, + "grad_norm": 2.379888723160233, + "learning_rate": 6.093888018412192e-07, + "loss": 0.8359, + "num_input_tokens_seen": 134506885, + "step": 6257 + }, + { + "epoch": 0.7524800096194313, + "flos": 67251114635520.0, + "grad_norm": 0.7187255772297431, + "learning_rate": 6.088290493361125e-07, + "loss": 0.5449, + "num_input_tokens_seen": 134571770, + "step": 6258 + }, + { + "epoch": 0.7526002525100703, + "flos": 12987742854720.0, + "grad_norm": 2.9048432257627548, + "learning_rate": 6.082695078705322e-07, + "loss": 0.7185, + "num_input_tokens_seen": 134589250, + "step": 6259 + }, + { + "epoch": 0.7527204954007094, + "flos": 21366679449600.0, + "grad_norm": 3.2876646283488364, + "learning_rate": 6.077101775293618e-07, + "loss": 0.6862, + "num_input_tokens_seen": 134608075, + "step": 6260 + }, + { + "epoch": 0.7528407382913486, + "flos": 18920004713760.0, + "grad_norm": 13.708414342267922, + "learning_rate": 6.071510583974504e-07, + "loss": 0.8262, + "num_input_tokens_seen": 134623260, + "step": 6261 + }, + { + "epoch": 0.7529609811819876, + "flos": 15210249536160.0, + "grad_norm": 5.7724882798635555, + "learning_rate": 6.065921505596161e-07, + "loss": 0.7188, + "num_input_tokens_seen": 134641540, + "step": 6262 + }, + { + "epoch": 0.7530812240726267, + "flos": 19326927976320.0, + "grad_norm": 2.4289803901907887, + "learning_rate": 6.060334541006445e-07, + "loss": 0.7688, + "num_input_tokens_seen": 134660035, + "step": 6263 + }, + { + "epoch": 0.7532014669632658, + "flos": 27708540793920.0, + "grad_norm": 1.9568487451547127, + "learning_rate": 6.05474969105289e-07, + "loss": 0.6943, + "num_input_tokens_seen": 134683025, + "step": 6264 + }, + { + "epoch": 0.7533217098539049, + "flos": 14117472432960.0, + "grad_norm": 3.8685088395245386, + "learning_rate": 6.049166956582725e-07, + "loss": 0.74, + "num_input_tokens_seen": 134701290, + "step": 6265 + }, + { + "epoch": 0.753441952744544, + "flos": 26391075259680.0, + "grad_norm": 2.5133799136606996, + "learning_rate": 6.043586338442841e-07, + "loss": 0.879, + "num_input_tokens_seen": 134720345, + "step": 6266 + }, + { + "epoch": 0.7535621956351831, + "flos": 23844471008160.0, + "grad_norm": 1.8928125509885896, + "learning_rate": 6.038007837479815e-07, + "loss": 0.7296, + "num_input_tokens_seen": 134741450, + "step": 6267 + }, + { + "epoch": 0.7536824385258222, + "flos": 21764644800000.0, + "grad_norm": 2.5670168076348725, + "learning_rate": 6.032431454539897e-07, + "loss": 0.6431, + "num_input_tokens_seen": 134760295, + "step": 6268 + }, + { + "epoch": 0.7538026814164612, + "flos": 28870093420320.0, + "grad_norm": 2.1481429574375115, + "learning_rate": 6.026857190469014e-07, + "loss": 0.8182, + "num_input_tokens_seen": 134782050, + "step": 6269 + }, + { + "epoch": 0.7539229243071004, + "flos": 21074553606720.0, + "grad_norm": 2.3499712962568715, + "learning_rate": 6.0212850461128e-07, + "loss": 0.7375, + "num_input_tokens_seen": 134801170, + "step": 6270 + }, + { + "epoch": 0.7540431671977395, + "flos": 15137013641760.0, + "grad_norm": 3.302630855442639, + "learning_rate": 6.015715022316516e-07, + "loss": 0.7484, + "num_input_tokens_seen": 134819150, + "step": 6271 + }, + { + "epoch": 0.7541634100883785, + "flos": 18743865722880.0, + "grad_norm": 3.702483689029473, + "learning_rate": 6.010147119925154e-07, + "loss": 0.7769, + "num_input_tokens_seen": 134836905, + "step": 6272 + }, + { + "epoch": 0.7542836529790176, + "flos": 20565396303360.0, + "grad_norm": 2.4884850435027857, + "learning_rate": 6.004581339783348e-07, + "loss": 0.6625, + "num_input_tokens_seen": 134855225, + "step": 6273 + }, + { + "epoch": 0.7544038958696567, + "flos": 19067777444160.0, + "grad_norm": 3.908925597342769, + "learning_rate": 5.999017682735425e-07, + "loss": 0.6839, + "num_input_tokens_seen": 134871615, + "step": 6274 + }, + { + "epoch": 0.7545241387602958, + "flos": 31677743861760.0, + "grad_norm": 2.4948245838093492, + "learning_rate": 5.993456149625387e-07, + "loss": 0.6642, + "num_input_tokens_seen": 134892765, + "step": 6275 + }, + { + "epoch": 0.7546443816509348, + "flos": 20267323298880.0, + "grad_norm": 2.080624332261716, + "learning_rate": 5.987896741296909e-07, + "loss": 0.8289, + "num_input_tokens_seen": 134910505, + "step": 6276 + }, + { + "epoch": 0.754764624541574, + "flos": 23662273346400.0, + "grad_norm": 3.2005959551908125, + "learning_rate": 5.982339458593361e-07, + "loss": 0.7832, + "num_input_tokens_seen": 134930445, + "step": 6277 + }, + { + "epoch": 0.7548848674322131, + "flos": 25300900039680.0, + "grad_norm": 1.8057188579689414, + "learning_rate": 5.976784302357767e-07, + "loss": 0.839, + "num_input_tokens_seen": 134951010, + "step": 6278 + }, + { + "epoch": 0.7550051103228521, + "flos": 19545260378400.0, + "grad_norm": 2.3747077561317878, + "learning_rate": 5.971231273432855e-07, + "loss": 0.7328, + "num_input_tokens_seen": 134970495, + "step": 6279 + }, + { + "epoch": 0.7551253532134913, + "flos": 64058290974720.0, + "grad_norm": 0.8171278397497791, + "learning_rate": 5.965680372661e-07, + "loss": 0.5448, + "num_input_tokens_seen": 135028060, + "step": 6280 + }, + { + "epoch": 0.7552455961041303, + "flos": 26030657100480.0, + "grad_norm": 2.1468543305713865, + "learning_rate": 5.960131600884266e-07, + "loss": 0.5612, + "num_input_tokens_seen": 135047330, + "step": 6281 + }, + { + "epoch": 0.7553658389947694, + "flos": 24463147625280.0, + "grad_norm": 3.5533294431034093, + "learning_rate": 5.954584958944413e-07, + "loss": 0.7638, + "num_input_tokens_seen": 135065995, + "step": 6282 + }, + { + "epoch": 0.7554860818854086, + "flos": 21768361776000.0, + "grad_norm": 2.2738857810491644, + "learning_rate": 5.949040447682854e-07, + "loss": 0.8188, + "num_input_tokens_seen": 135085820, + "step": 6283 + }, + { + "epoch": 0.7556063247760476, + "flos": 16338901191360.0, + "grad_norm": 2.724117811699829, + "learning_rate": 5.943498067940686e-07, + "loss": 0.6849, + "num_input_tokens_seen": 135102845, + "step": 6284 + }, + { + "epoch": 0.7557265676666867, + "flos": 27196112551680.0, + "grad_norm": 2.2080935461987674, + "learning_rate": 5.937957820558686e-07, + "loss": 0.8149, + "num_input_tokens_seen": 135122190, + "step": 6285 + }, + { + "epoch": 0.7558468105573258, + "flos": 62100547459200.0, + "grad_norm": 0.8643008119048137, + "learning_rate": 5.932419706377296e-07, + "loss": 0.6508, + "num_input_tokens_seen": 135180495, + "step": 6286 + }, + { + "epoch": 0.7559670534479649, + "flos": 33201052758720.0, + "grad_norm": 2.5308156523863063, + "learning_rate": 5.92688372623666e-07, + "loss": 0.741, + "num_input_tokens_seen": 135199200, + "step": 6287 + }, + { + "epoch": 0.7560872963386039, + "flos": 14044533896640.0, + "grad_norm": 3.1831711939184926, + "learning_rate": 5.921349880976574e-07, + "loss": 0.7406, + "num_input_tokens_seen": 135217035, + "step": 6288 + }, + { + "epoch": 0.7562075392292431, + "flos": 20383198641600.0, + "grad_norm": 2.292090278545158, + "learning_rate": 5.915818171436515e-07, + "loss": 0.8204, + "num_input_tokens_seen": 135236370, + "step": 6289 + }, + { + "epoch": 0.7563277821198822, + "flos": 20347770126720.0, + "grad_norm": 2.193941495965589, + "learning_rate": 5.910288598455642e-07, + "loss": 0.7471, + "num_input_tokens_seen": 135255720, + "step": 6290 + }, + { + "epoch": 0.7564480250105212, + "flos": 18561816740160.0, + "grad_norm": 5.368923744880786, + "learning_rate": 5.90476116287278e-07, + "loss": 0.7451, + "num_input_tokens_seen": 135273320, + "step": 6291 + }, + { + "epoch": 0.7565682679011604, + "flos": 21185262352800.0, + "grad_norm": 2.2352613278396225, + "learning_rate": 5.899235865526456e-07, + "loss": 0.6827, + "num_input_tokens_seen": 135292615, + "step": 6292 + }, + { + "epoch": 0.7566885107917994, + "flos": 20419853758560.0, + "grad_norm": 3.7290061759866897, + "learning_rate": 5.893712707254825e-07, + "loss": 0.8246, + "num_input_tokens_seen": 135310075, + "step": 6293 + }, + { + "epoch": 0.7568087536824385, + "flos": 18998741732640.0, + "grad_norm": 16.63697968435311, + "learning_rate": 5.888191688895769e-07, + "loss": 0.6597, + "num_input_tokens_seen": 135327335, + "step": 6294 + }, + { + "epoch": 0.7569289965730777, + "flos": 15203670488640.0, + "grad_norm": 2.753084187199709, + "learning_rate": 5.882672811286813e-07, + "loss": 0.6196, + "num_input_tokens_seen": 135343615, + "step": 6295 + }, + { + "epoch": 0.7570492394637167, + "flos": 20739453787680.0, + "grad_norm": 3.418684634851299, + "learning_rate": 5.877156075265166e-07, + "loss": 0.6965, + "num_input_tokens_seen": 135359880, + "step": 6296 + }, + { + "epoch": 0.7571694823543558, + "flos": 15641710573920.0, + "grad_norm": 7.106196973832643, + "learning_rate": 5.871641481667715e-07, + "loss": 0.6967, + "num_input_tokens_seen": 135374325, + "step": 6297 + }, + { + "epoch": 0.7572897252449949, + "flos": 25372797822720.0, + "grad_norm": 2.1339880529711004, + "learning_rate": 5.866129031331011e-07, + "loss": 0.8451, + "num_input_tokens_seen": 135393980, + "step": 6298 + }, + { + "epoch": 0.757409968135634, + "flos": 24244592204640.0, + "grad_norm": 2.909746916996679, + "learning_rate": 5.8606187250913e-07, + "loss": 0.8383, + "num_input_tokens_seen": 135412030, + "step": 6299 + }, + { + "epoch": 0.757530211026273, + "flos": 24099235508640.0, + "grad_norm": 3.0494813384263635, + "learning_rate": 5.855110563784482e-07, + "loss": 0.8409, + "num_input_tokens_seen": 135430565, + "step": 6300 + }, + { + "epoch": 0.7576504539169122, + "flos": 23917558223520.0, + "grad_norm": 2.217305962675584, + "learning_rate": 5.849604548246156e-07, + "loss": 0.6432, + "num_input_tokens_seen": 135451675, + "step": 6301 + }, + { + "epoch": 0.7577706968075513, + "flos": 21221025395520.0, + "grad_norm": 11.819605068057212, + "learning_rate": 5.844100679311565e-07, + "loss": 0.8062, + "num_input_tokens_seen": 135470635, + "step": 6302 + }, + { + "epoch": 0.7578909396981903, + "flos": 18270582971520.0, + "grad_norm": 4.058712176448338, + "learning_rate": 5.838598957815637e-07, + "loss": 0.7626, + "num_input_tokens_seen": 135487865, + "step": 6303 + }, + { + "epoch": 0.7580111825888295, + "flos": 25337517986880.0, + "grad_norm": 2.3424875546901363, + "learning_rate": 5.833099384592996e-07, + "loss": 0.8561, + "num_input_tokens_seen": 135508440, + "step": 6304 + }, + { + "epoch": 0.7581314254794685, + "flos": 23734654336320.0, + "grad_norm": 2.724284359814945, + "learning_rate": 5.827601960477913e-07, + "loss": 0.7182, + "num_input_tokens_seen": 135526365, + "step": 6305 + }, + { + "epoch": 0.7582516683701076, + "flos": 22023126276480.0, + "grad_norm": 2.8516994495688683, + "learning_rate": 5.822106686304344e-07, + "loss": 0.7073, + "num_input_tokens_seen": 135545655, + "step": 6306 + }, + { + "epoch": 0.7583719112607467, + "flos": 31604247779040.0, + "grad_norm": 2.6469306372477686, + "learning_rate": 5.816613562905919e-07, + "loss": 0.5783, + "num_input_tokens_seen": 135566840, + "step": 6307 + }, + { + "epoch": 0.7584921541513858, + "flos": 33020676415200.0, + "grad_norm": 2.6068064216880025, + "learning_rate": 5.811122591115933e-07, + "loss": 0.6996, + "num_input_tokens_seen": 135587825, + "step": 6308 + }, + { + "epoch": 0.7586123970420249, + "flos": 23292822935520.0, + "grad_norm": 3.1040149284672363, + "learning_rate": 5.805633771767376e-07, + "loss": 0.7139, + "num_input_tokens_seen": 135605220, + "step": 6309 + }, + { + "epoch": 0.7587326399326639, + "flos": 18308204502240.0, + "grad_norm": 2.291096379384531, + "learning_rate": 5.800147105692888e-07, + "loss": 0.7777, + "num_input_tokens_seen": 135624795, + "step": 6310 + }, + { + "epoch": 0.7588528828233031, + "flos": 17250558555840.0, + "grad_norm": 11.898377867340717, + "learning_rate": 5.794662593724795e-07, + "loss": 0.7909, + "num_input_tokens_seen": 135643800, + "step": 6311 + }, + { + "epoch": 0.7589731257139422, + "flos": 17688487131840.0, + "grad_norm": 3.500540010482702, + "learning_rate": 5.789180236695091e-07, + "loss": 0.7507, + "num_input_tokens_seen": 135660655, + "step": 6312 + }, + { + "epoch": 0.7590933686045812, + "flos": 15938445467040.0, + "grad_norm": 2.2997395824118043, + "learning_rate": 5.78370003543544e-07, + "loss": 0.8519, + "num_input_tokens_seen": 135679410, + "step": 6313 + }, + { + "epoch": 0.7592136114952204, + "flos": 21037972829280.0, + "grad_norm": 2.677006484139605, + "learning_rate": 5.778221990777203e-07, + "loss": 0.8398, + "num_input_tokens_seen": 135697150, + "step": 6314 + }, + { + "epoch": 0.7593338543858594, + "flos": 25261308511680.0, + "grad_norm": 6.665904891678424, + "learning_rate": 5.772746103551372e-07, + "loss": 0.8301, + "num_input_tokens_seen": 135712545, + "step": 6315 + }, + { + "epoch": 0.7594540972764985, + "flos": 31786631289600.0, + "grad_norm": 3.1301910010612772, + "learning_rate": 5.767272374588648e-07, + "loss": 0.718, + "num_input_tokens_seen": 135733950, + "step": 6316 + }, + { + "epoch": 0.7595743401671377, + "flos": 37543125855360.0, + "grad_norm": 1.8428299758235163, + "learning_rate": 5.76180080471939e-07, + "loss": 0.7787, + "num_input_tokens_seen": 135757120, + "step": 6317 + }, + { + "epoch": 0.7596945830577767, + "flos": 18261402040800.0, + "grad_norm": 3.0923591338680865, + "learning_rate": 5.756331394773631e-07, + "loss": 0.7223, + "num_input_tokens_seen": 135773335, + "step": 6318 + }, + { + "epoch": 0.7598148259484158, + "flos": 22201235264640.0, + "grad_norm": 2.6666681965187773, + "learning_rate": 5.750864145581071e-07, + "loss": 0.7644, + "num_input_tokens_seen": 135792555, + "step": 6319 + }, + { + "epoch": 0.7599350688390549, + "flos": 27122765148000.0, + "grad_norm": 2.7985364358645266, + "learning_rate": 5.745399057971085e-07, + "loss": 0.8702, + "num_input_tokens_seen": 135813690, + "step": 6320 + }, + { + "epoch": 0.760055311729694, + "flos": 15538695968160.0, + "grad_norm": 2.787164977601025, + "learning_rate": 5.739936132772738e-07, + "loss": 0.7552, + "num_input_tokens_seen": 135832255, + "step": 6321 + }, + { + "epoch": 0.760175554620333, + "flos": 25119482942880.0, + "grad_norm": 2.614881723912902, + "learning_rate": 5.734475370814733e-07, + "loss": 0.7446, + "num_input_tokens_seen": 135851935, + "step": 6322 + }, + { + "epoch": 0.7602957975109722, + "flos": 24318980361600.0, + "grad_norm": 2.05601452809704, + "learning_rate": 5.729016772925483e-07, + "loss": 0.7861, + "num_input_tokens_seen": 135873510, + "step": 6323 + }, + { + "epoch": 0.7604160404016113, + "flos": 25156621266720.0, + "grad_norm": 4.159480395511503, + "learning_rate": 5.723560339933038e-07, + "loss": 0.709, + "num_input_tokens_seen": 135893845, + "step": 6324 + }, + { + "epoch": 0.7605362832922503, + "flos": 29818963448160.0, + "grad_norm": 2.615265371012171, + "learning_rate": 5.71810607266513e-07, + "loss": 0.6545, + "num_input_tokens_seen": 135912430, + "step": 6325 + }, + { + "epoch": 0.7606565261828895, + "flos": 13898285126400.0, + "grad_norm": 2.930218701842066, + "learning_rate": 5.712653971949184e-07, + "loss": 0.6041, + "num_input_tokens_seen": 135930550, + "step": 6326 + }, + { + "epoch": 0.7607767690735285, + "flos": 18524864265120.0, + "grad_norm": 3.098940262087322, + "learning_rate": 5.707204038612268e-07, + "loss": 0.7524, + "num_input_tokens_seen": 135947980, + "step": 6327 + }, + { + "epoch": 0.7608970119641676, + "flos": 20893136509920.0, + "grad_norm": 4.683704110850519, + "learning_rate": 5.701756273481138e-07, + "loss": 0.7357, + "num_input_tokens_seen": 135965400, + "step": 6328 + }, + { + "epoch": 0.7610172548548068, + "flos": 23773316620320.0, + "grad_norm": 5.476502770548482, + "learning_rate": 5.696310677382212e-07, + "loss": 0.7408, + "num_input_tokens_seen": 135986795, + "step": 6329 + }, + { + "epoch": 0.7611374977454458, + "flos": 66401802425760.0, + "grad_norm": 0.8675253957279702, + "learning_rate": 5.690867251141576e-07, + "loss": 0.616, + "num_input_tokens_seen": 136053450, + "step": 6330 + }, + { + "epoch": 0.7612577406360849, + "flos": 15611002618560.0, + "grad_norm": 2.6688910882538264, + "learning_rate": 5.685425995585013e-07, + "loss": 0.9235, + "num_input_tokens_seen": 136071375, + "step": 6331 + }, + { + "epoch": 0.761377983526724, + "flos": 60440653929600.0, + "grad_norm": 0.7624991307488161, + "learning_rate": 5.679986911537935e-07, + "loss": 0.5888, + "num_input_tokens_seen": 136138905, + "step": 6332 + }, + { + "epoch": 0.7614982264173631, + "flos": 35721037728480.0, + "grad_norm": 2.4045894887877686, + "learning_rate": 5.674549999825462e-07, + "loss": 0.6746, + "num_input_tokens_seen": 136161720, + "step": 6333 + }, + { + "epoch": 0.7616184693080021, + "flos": 67828861613280.0, + "grad_norm": 0.9173505933763316, + "learning_rate": 5.669115261272363e-07, + "loss": 0.7132, + "num_input_tokens_seen": 136222040, + "step": 6334 + }, + { + "epoch": 0.7617387121986413, + "flos": 20492717955360.0, + "grad_norm": 2.801179099455978, + "learning_rate": 5.663682696703081e-07, + "loss": 0.7256, + "num_input_tokens_seen": 136240305, + "step": 6335 + }, + { + "epoch": 0.7618589550892804, + "flos": 18597988650240.0, + "grad_norm": 4.522079553834821, + "learning_rate": 5.658252306941746e-07, + "loss": 0.8224, + "num_input_tokens_seen": 136259625, + "step": 6336 + }, + { + "epoch": 0.7619791979799194, + "flos": 17428816223040.0, + "grad_norm": 2.5917079194476726, + "learning_rate": 5.65282409281212e-07, + "loss": 0.7574, + "num_input_tokens_seen": 136277090, + "step": 6337 + }, + { + "epoch": 0.7620994408705585, + "flos": 14117472432960.0, + "grad_norm": 2.6149774586160057, + "learning_rate": 5.64739805513768e-07, + "loss": 0.7055, + "num_input_tokens_seen": 136294635, + "step": 6338 + }, + { + "epoch": 0.7622196837611976, + "flos": 70607147944320.0, + "grad_norm": 0.8029511616107974, + "learning_rate": 5.641974194741541e-07, + "loss": 0.5566, + "num_input_tokens_seen": 136350320, + "step": 6339 + }, + { + "epoch": 0.7623399266518367, + "flos": 60597427475520.0, + "grad_norm": 0.786295581339375, + "learning_rate": 5.636552512446502e-07, + "loss": 0.6363, + "num_input_tokens_seen": 136411375, + "step": 6340 + }, + { + "epoch": 0.7624601695424758, + "flos": 26430964145760.0, + "grad_norm": 2.2941745791268193, + "learning_rate": 5.631133009075027e-07, + "loss": 0.7812, + "num_input_tokens_seen": 136430560, + "step": 6341 + }, + { + "epoch": 0.7625804124331149, + "flos": 19108521234720.0, + "grad_norm": 2.8395102618623653, + "learning_rate": 5.625715685449242e-07, + "loss": 0.6896, + "num_input_tokens_seen": 136448665, + "step": 6342 + }, + { + "epoch": 0.762700655323754, + "flos": 26175530589600.0, + "grad_norm": 2.0930933136020955, + "learning_rate": 5.620300542390966e-07, + "loss": 0.719, + "num_input_tokens_seen": 136469710, + "step": 6343 + }, + { + "epoch": 0.762820898214393, + "flos": 22350606294720.0, + "grad_norm": 3.670189523402751, + "learning_rate": 5.614887580721659e-07, + "loss": 0.8553, + "num_input_tokens_seen": 136489855, + "step": 6344 + }, + { + "epoch": 0.7629411411050322, + "flos": 15678068332800.0, + "grad_norm": 3.687455904422195, + "learning_rate": 5.609476801262481e-07, + "loss": 0.7437, + "num_input_tokens_seen": 136504715, + "step": 6345 + }, + { + "epoch": 0.7630613839956712, + "flos": 13750846923840.0, + "grad_norm": 5.530730487111927, + "learning_rate": 5.604068204834223e-07, + "loss": 0.6412, + "num_input_tokens_seen": 136521610, + "step": 6346 + }, + { + "epoch": 0.7631816268863103, + "flos": 14548784791680.0, + "grad_norm": 3.890662871027091, + "learning_rate": 5.598661792257367e-07, + "loss": 0.7652, + "num_input_tokens_seen": 136538655, + "step": 6347 + }, + { + "epoch": 0.7633018697769495, + "flos": 19035062321760.0, + "grad_norm": 2.165141583718064, + "learning_rate": 5.593257564352071e-07, + "loss": 0.7596, + "num_input_tokens_seen": 136557095, + "step": 6348 + }, + { + "epoch": 0.7634221126675885, + "flos": 22021416467520.0, + "grad_norm": 1.709471749700738, + "learning_rate": 5.58785552193815e-07, + "loss": 0.7566, + "num_input_tokens_seen": 136577340, + "step": 6349 + }, + { + "epoch": 0.7635423555582276, + "flos": 29340365421120.0, + "grad_norm": 2.4470127032861813, + "learning_rate": 5.582455665835086e-07, + "loss": 0.758, + "num_input_tokens_seen": 136597635, + "step": 6350 + }, + { + "epoch": 0.7636625984488667, + "flos": 17759604349920.0, + "grad_norm": 4.4994400523745295, + "learning_rate": 5.577057996862036e-07, + "loss": 0.7264, + "num_input_tokens_seen": 136615260, + "step": 6351 + }, + { + "epoch": 0.7637828413395058, + "flos": 23700712611840.0, + "grad_norm": 1.9104022470288504, + "learning_rate": 5.571662515837814e-07, + "loss": 0.7605, + "num_input_tokens_seen": 136637220, + "step": 6352 + }, + { + "epoch": 0.7639030842301449, + "flos": 36231049936320.0, + "grad_norm": 2.2970546729988492, + "learning_rate": 5.566269223580926e-07, + "loss": 0.8379, + "num_input_tokens_seen": 136658930, + "step": 6353 + }, + { + "epoch": 0.764023327120784, + "flos": 28836709242240.0, + "grad_norm": 1.904839853979455, + "learning_rate": 5.560878120909511e-07, + "loss": 0.7532, + "num_input_tokens_seen": 136681530, + "step": 6354 + }, + { + "epoch": 0.7641435700114231, + "flos": 64697708317920.0, + "grad_norm": 0.8543295743708, + "learning_rate": 5.55548920864141e-07, + "loss": 0.585, + "num_input_tokens_seen": 136742185, + "step": 6355 + }, + { + "epoch": 0.7642638129020621, + "flos": 16811849414880.0, + "grad_norm": 2.5831454321502916, + "learning_rate": 5.550102487594113e-07, + "loss": 0.7805, + "num_input_tokens_seen": 136760855, + "step": 6356 + }, + { + "epoch": 0.7643840557927013, + "flos": 30365259075360.0, + "grad_norm": 1.91307210406082, + "learning_rate": 5.54471795858477e-07, + "loss": 0.7142, + "num_input_tokens_seen": 136780925, + "step": 6357 + }, + { + "epoch": 0.7645042986833404, + "flos": 16959138938400.0, + "grad_norm": 3.177621681170529, + "learning_rate": 5.539335622430235e-07, + "loss": 0.8323, + "num_input_tokens_seen": 136799545, + "step": 6358 + }, + { + "epoch": 0.7646245415739794, + "flos": 17286916314720.0, + "grad_norm": 3.2690602901076766, + "learning_rate": 5.533955479946975e-07, + "loss": 0.7519, + "num_input_tokens_seen": 136818325, + "step": 6359 + }, + { + "epoch": 0.7647447844646186, + "flos": 70301932612320.0, + "grad_norm": 0.9103655922280655, + "learning_rate": 5.528577531951173e-07, + "loss": 0.6563, + "num_input_tokens_seen": 136878730, + "step": 6360 + }, + { + "epoch": 0.7648650273552576, + "flos": 17650791261600.0, + "grad_norm": 2.876106215243927, + "learning_rate": 5.523201779258653e-07, + "loss": 0.7412, + "num_input_tokens_seen": 136897705, + "step": 6361 + }, + { + "epoch": 0.7649852702458967, + "flos": 22130303895360.0, + "grad_norm": 3.163880745647766, + "learning_rate": 5.517828222684912e-07, + "loss": 0.8428, + "num_input_tokens_seen": 136918360, + "step": 6362 + }, + { + "epoch": 0.7651055131365359, + "flos": 69748500391200.0, + "grad_norm": 0.7745529762162128, + "learning_rate": 5.512456863045117e-07, + "loss": 0.5895, + "num_input_tokens_seen": 136979050, + "step": 6363 + }, + { + "epoch": 0.7652257560271749, + "flos": 19436298611040.0, + "grad_norm": 2.515681769636765, + "learning_rate": 5.507087701154089e-07, + "loss": 0.7431, + "num_input_tokens_seen": 136998970, + "step": 6364 + }, + { + "epoch": 0.765345998917814, + "flos": 15938482636800.0, + "grad_norm": 2.238996301645391, + "learning_rate": 5.50172073782634e-07, + "loss": 0.754, + "num_input_tokens_seen": 137016950, + "step": 6365 + }, + { + "epoch": 0.7654662418084531, + "flos": 23626027096800.0, + "grad_norm": 3.2373717058013955, + "learning_rate": 5.496355973876023e-07, + "loss": 0.8805, + "num_input_tokens_seen": 137036205, + "step": 6366 + }, + { + "epoch": 0.7655864846990922, + "flos": 41403813192960.0, + "grad_norm": 2.1658035050623323, + "learning_rate": 5.490993410116984e-07, + "loss": 0.7076, + "num_input_tokens_seen": 137059195, + "step": 6367 + }, + { + "epoch": 0.7657067275897312, + "flos": 43080470284320.0, + "grad_norm": 1.9338558751461812, + "learning_rate": 5.485633047362704e-07, + "loss": 0.697, + "num_input_tokens_seen": 137081200, + "step": 6368 + }, + { + "epoch": 0.7658269704803703, + "flos": 17287213672800.0, + "grad_norm": 4.365540659696671, + "learning_rate": 5.480274886426341e-07, + "loss": 0.7846, + "num_input_tokens_seen": 137097840, + "step": 6369 + }, + { + "epoch": 0.7659472133710095, + "flos": 12550483334400.0, + "grad_norm": 2.60453338075551, + "learning_rate": 5.474918928120744e-07, + "loss": 0.776, + "num_input_tokens_seen": 137114335, + "step": 6370 + }, + { + "epoch": 0.7660674562616485, + "flos": 22674852543840.0, + "grad_norm": 1.9096320745595912, + "learning_rate": 5.469565173258392e-07, + "loss": 0.8724, + "num_input_tokens_seen": 137132850, + "step": 6371 + }, + { + "epoch": 0.7661876991522876, + "flos": 17032263323520.0, + "grad_norm": 10.720429669779449, + "learning_rate": 5.464213622651454e-07, + "loss": 0.6377, + "num_input_tokens_seen": 137150665, + "step": 6372 + }, + { + "epoch": 0.7663079420429267, + "flos": 20055607114080.0, + "grad_norm": 2.680031801575962, + "learning_rate": 5.458864277111753e-07, + "loss": 0.8461, + "num_input_tokens_seen": 137168500, + "step": 6373 + }, + { + "epoch": 0.7664281849335658, + "flos": 12659630950560.0, + "grad_norm": 3.665480609726474, + "learning_rate": 5.453517137450769e-07, + "loss": 0.6944, + "num_input_tokens_seen": 137185075, + "step": 6374 + }, + { + "epoch": 0.7665484278242048, + "flos": 22313319291840.0, + "grad_norm": 2.183844864110757, + "learning_rate": 5.448172204479684e-07, + "loss": 0.7623, + "num_input_tokens_seen": 137204355, + "step": 6375 + }, + { + "epoch": 0.766668670714844, + "flos": 23583945194880.0, + "grad_norm": 2.209637047147552, + "learning_rate": 5.442829479009294e-07, + "loss": 0.7472, + "num_input_tokens_seen": 137223135, + "step": 6376 + }, + { + "epoch": 0.7667889136054831, + "flos": 19399457645280.0, + "grad_norm": 2.156163974096339, + "learning_rate": 5.437488961850103e-07, + "loss": 0.7173, + "num_input_tokens_seen": 137242935, + "step": 6377 + }, + { + "epoch": 0.7669091564961221, + "flos": 26827702894080.0, + "grad_norm": 2.1222624255106566, + "learning_rate": 5.432150653812258e-07, + "loss": 0.754, + "num_input_tokens_seen": 137262970, + "step": 6378 + }, + { + "epoch": 0.7670293993867613, + "flos": 12367876805280.0, + "grad_norm": 2.5697263843628804, + "learning_rate": 5.42681455570557e-07, + "loss": 0.8238, + "num_input_tokens_seen": 137279500, + "step": 6379 + }, + { + "epoch": 0.7671496422774003, + "flos": 21731892507840.0, + "grad_norm": 2.843649726189706, + "learning_rate": 5.42148066833954e-07, + "loss": 0.6489, + "num_input_tokens_seen": 137298745, + "step": 6380 + }, + { + "epoch": 0.7672698851680394, + "flos": 21039162261600.0, + "grad_norm": 2.7384660035412653, + "learning_rate": 5.416148992523289e-07, + "loss": 0.754, + "num_input_tokens_seen": 137317320, + "step": 6381 + }, + { + "epoch": 0.7673901280586786, + "flos": 16952374042080.0, + "grad_norm": 2.7577226748440236, + "learning_rate": 5.410819529065644e-07, + "loss": 0.7825, + "num_input_tokens_seen": 137335275, + "step": 6382 + }, + { + "epoch": 0.7675103709493176, + "flos": 29200732868160.0, + "grad_norm": 2.549054857013183, + "learning_rate": 5.405492278775079e-07, + "loss": 0.6529, + "num_input_tokens_seen": 137355055, + "step": 6383 + }, + { + "epoch": 0.7676306138399567, + "flos": 28981359712800.0, + "grad_norm": 2.465893215197058, + "learning_rate": 5.400167242459732e-07, + "loss": 0.801, + "num_input_tokens_seen": 137374015, + "step": 6384 + }, + { + "epoch": 0.7677508567305958, + "flos": 22532878296000.0, + "grad_norm": 2.329130633990475, + "learning_rate": 5.394844420927405e-07, + "loss": 0.8074, + "num_input_tokens_seen": 137393625, + "step": 6385 + }, + { + "epoch": 0.7678710996212349, + "flos": 25375065178080.0, + "grad_norm": 3.4674974534204837, + "learning_rate": 5.389523814985562e-07, + "loss": 0.7367, + "num_input_tokens_seen": 137413045, + "step": 6386 + }, + { + "epoch": 0.767991342511874, + "flos": 26718332259360.0, + "grad_norm": 3.1689609615368615, + "learning_rate": 5.384205425441344e-07, + "loss": 0.7647, + "num_input_tokens_seen": 137433955, + "step": 6387 + }, + { + "epoch": 0.7681115854025131, + "flos": 26321742190080.0, + "grad_norm": 2.61321601729238, + "learning_rate": 5.378889253101537e-07, + "loss": 0.8435, + "num_input_tokens_seen": 137454510, + "step": 6388 + }, + { + "epoch": 0.7682318282931522, + "flos": 23223750054240.0, + "grad_norm": 2.0601201053901717, + "learning_rate": 5.373575298772617e-07, + "loss": 0.81, + "num_input_tokens_seen": 137473630, + "step": 6389 + }, + { + "epoch": 0.7683520711837912, + "flos": 69972482596800.0, + "grad_norm": 0.8011843155406965, + "learning_rate": 5.368263563260689e-07, + "loss": 0.6122, + "num_input_tokens_seen": 137538765, + "step": 6390 + }, + { + "epoch": 0.7684723140744304, + "flos": 18598397517600.0, + "grad_norm": 2.15496743799378, + "learning_rate": 5.362954047371537e-07, + "loss": 0.645, + "num_input_tokens_seen": 137557875, + "step": 6391 + }, + { + "epoch": 0.7685925569650695, + "flos": 27413590049280.0, + "grad_norm": 2.899598416114108, + "learning_rate": 5.357646751910627e-07, + "loss": 0.7242, + "num_input_tokens_seen": 137579055, + "step": 6392 + }, + { + "epoch": 0.7687127998557085, + "flos": 24500620476960.0, + "grad_norm": 4.079860975477429, + "learning_rate": 5.352341677683061e-07, + "loss": 0.7947, + "num_input_tokens_seen": 137600355, + "step": 6393 + }, + { + "epoch": 0.7688330427463477, + "flos": 25119966149760.0, + "grad_norm": 2.5194369986004794, + "learning_rate": 5.347038825493617e-07, + "loss": 0.7913, + "num_input_tokens_seen": 137621885, + "step": 6394 + }, + { + "epoch": 0.7689532856369867, + "flos": 21180876321120.0, + "grad_norm": 3.3428899782033015, + "learning_rate": 5.341738196146732e-07, + "loss": 0.6879, + "num_input_tokens_seen": 137640230, + "step": 6395 + }, + { + "epoch": 0.7690735285276258, + "flos": 25083125184000.0, + "grad_norm": 2.8789806175202237, + "learning_rate": 5.336439790446503e-07, + "loss": 0.7316, + "num_input_tokens_seen": 137659330, + "step": 6396 + }, + { + "epoch": 0.769193771418265, + "flos": 54664873992000.0, + "grad_norm": 3.2497385360007294, + "learning_rate": 5.331143609196711e-07, + "loss": 0.6236, + "num_input_tokens_seen": 137683145, + "step": 6397 + }, + { + "epoch": 0.769314014308904, + "flos": 37288138336320.0, + "grad_norm": 2.3833317425417984, + "learning_rate": 5.325849653200758e-07, + "loss": 0.7734, + "num_input_tokens_seen": 137707095, + "step": 6398 + }, + { + "epoch": 0.7694342571995431, + "flos": 20602274438880.0, + "grad_norm": 2.1542008102731383, + "learning_rate": 5.32055792326175e-07, + "loss": 0.7663, + "num_input_tokens_seen": 137725520, + "step": 6399 + }, + { + "epoch": 0.7695545000901821, + "flos": 24173103288960.0, + "grad_norm": 2.1873913329245114, + "learning_rate": 5.315268420182437e-07, + "loss": 0.7306, + "num_input_tokens_seen": 137744195, + "step": 6400 + }, + { + "epoch": 0.7696747429808213, + "flos": 27961112278560.0, + "grad_norm": 2.271646954662453, + "learning_rate": 5.309981144765221e-07, + "loss": 0.7648, + "num_input_tokens_seen": 137764340, + "step": 6401 + }, + { + "epoch": 0.7697949858714603, + "flos": 11493394934400.0, + "grad_norm": 6.778458684408313, + "learning_rate": 5.304696097812196e-07, + "loss": 0.7514, + "num_input_tokens_seen": 137780940, + "step": 6402 + }, + { + "epoch": 0.7699152287620994, + "flos": 26650039943040.0, + "grad_norm": 4.338199939374539, + "learning_rate": 5.299413280125078e-07, + "loss": 0.6038, + "num_input_tokens_seen": 137799250, + "step": 6403 + }, + { + "epoch": 0.7700354716527386, + "flos": 16521470550720.0, + "grad_norm": 2.4309086812551315, + "learning_rate": 5.294132692505284e-07, + "loss": 0.729, + "num_input_tokens_seen": 137817660, + "step": 6404 + }, + { + "epoch": 0.7701557145433776, + "flos": 19215252816480.0, + "grad_norm": 3.631794711045925, + "learning_rate": 5.288854335753861e-07, + "loss": 0.7928, + "num_input_tokens_seen": 137835920, + "step": 6405 + }, + { + "epoch": 0.7702759574340167, + "flos": 31640196670560.0, + "grad_norm": 1.9837273569623584, + "learning_rate": 5.283578210671551e-07, + "loss": 0.7569, + "num_input_tokens_seen": 137858160, + "step": 6406 + }, + { + "epoch": 0.7703962003246558, + "flos": 16776235051200.0, + "grad_norm": 2.7713512902166335, + "learning_rate": 5.278304318058719e-07, + "loss": 0.764, + "num_input_tokens_seen": 137876015, + "step": 6407 + }, + { + "epoch": 0.7705164432152949, + "flos": 35684642799840.0, + "grad_norm": 2.8532124801544625, + "learning_rate": 5.273032658715411e-07, + "loss": 0.7916, + "num_input_tokens_seen": 137898825, + "step": 6408 + }, + { + "epoch": 0.7706366861059339, + "flos": 23332488803040.0, + "grad_norm": 2.4821812762965596, + "learning_rate": 5.267763233441347e-07, + "loss": 0.768, + "num_input_tokens_seen": 137919455, + "step": 6409 + }, + { + "epoch": 0.7707569289965731, + "flos": 22897124940480.0, + "grad_norm": 2.879840590515771, + "learning_rate": 5.26249604303588e-07, + "loss": 0.6989, + "num_input_tokens_seen": 137937960, + "step": 6410 + }, + { + "epoch": 0.7708771718872122, + "flos": 17395915251840.0, + "grad_norm": 3.42093523611738, + "learning_rate": 5.257231088298057e-07, + "loss": 0.7838, + "num_input_tokens_seen": 137956370, + "step": 6411 + }, + { + "epoch": 0.7709974147778512, + "flos": 72137444756160.0, + "grad_norm": 0.817088448992998, + "learning_rate": 5.25196837002655e-07, + "loss": 0.5378, + "num_input_tokens_seen": 138016080, + "step": 6412 + }, + { + "epoch": 0.7711176576684904, + "flos": 39802287653760.0, + "grad_norm": 2.706255120323218, + "learning_rate": 5.24670788901971e-07, + "loss": 0.6851, + "num_input_tokens_seen": 138039170, + "step": 6413 + }, + { + "epoch": 0.7712379005591294, + "flos": 36923148296640.0, + "grad_norm": 2.906299877764197, + "learning_rate": 5.241449646075557e-07, + "loss": 0.6909, + "num_input_tokens_seen": 138062395, + "step": 6414 + }, + { + "epoch": 0.7713581434497685, + "flos": 22743962594880.0, + "grad_norm": 3.472572488707911, + "learning_rate": 5.236193641991762e-07, + "loss": 0.7245, + "num_input_tokens_seen": 138080325, + "step": 6415 + }, + { + "epoch": 0.7714783863404077, + "flos": 24062989259040.0, + "grad_norm": 2.6321209188819976, + "learning_rate": 5.23093987756565e-07, + "loss": 0.7009, + "num_input_tokens_seen": 138099020, + "step": 6416 + }, + { + "epoch": 0.7715986292310467, + "flos": 21033400948800.0, + "grad_norm": 2.376160689375746, + "learning_rate": 5.225688353594217e-07, + "loss": 0.7534, + "num_input_tokens_seen": 138118650, + "step": 6417 + }, + { + "epoch": 0.7717188721216858, + "flos": 20565507812640.0, + "grad_norm": 3.6067923681761593, + "learning_rate": 5.220439070874108e-07, + "loss": 0.7807, + "num_input_tokens_seen": 138137890, + "step": 6418 + }, + { + "epoch": 0.7718391150123249, + "flos": 26213375138880.0, + "grad_norm": 3.0744962264090705, + "learning_rate": 5.215192030201652e-07, + "loss": 0.7129, + "num_input_tokens_seen": 138158880, + "step": 6419 + }, + { + "epoch": 0.771959357902964, + "flos": 22018108358880.0, + "grad_norm": 3.2609795195244344, + "learning_rate": 5.209947232372798e-07, + "loss": 0.8647, + "num_input_tokens_seen": 138176370, + "step": 6420 + }, + { + "epoch": 0.772079600793603, + "flos": 30402285889920.0, + "grad_norm": 2.2962118216441616, + "learning_rate": 5.204704678183196e-07, + "loss": 0.8148, + "num_input_tokens_seen": 138195105, + "step": 6421 + }, + { + "epoch": 0.7721998436842422, + "flos": 12951608114400.0, + "grad_norm": 3.734086029491952, + "learning_rate": 5.19946436842813e-07, + "loss": 0.8528, + "num_input_tokens_seen": 138212235, + "step": 6422 + }, + { + "epoch": 0.7723200865748813, + "flos": 32588620661280.0, + "grad_norm": 1.9664611832982366, + "learning_rate": 5.194226303902546e-07, + "loss": 0.6841, + "num_input_tokens_seen": 138231970, + "step": 6423 + }, + { + "epoch": 0.7724403294655203, + "flos": 21075594360000.0, + "grad_norm": 2.243749094673335, + "learning_rate": 5.188990485401072e-07, + "loss": 0.7098, + "num_input_tokens_seen": 138251175, + "step": 6424 + }, + { + "epoch": 0.7725605723561595, + "flos": 22059186677280.0, + "grad_norm": 2.625307882509461, + "learning_rate": 5.183756913717954e-07, + "loss": 0.8621, + "num_input_tokens_seen": 138270950, + "step": 6425 + }, + { + "epoch": 0.7726808152467985, + "flos": 34445988624000.0, + "grad_norm": 3.1154475889516093, + "learning_rate": 5.178525589647136e-07, + "loss": 0.7343, + "num_input_tokens_seen": 138292590, + "step": 6426 + }, + { + "epoch": 0.7728010581374376, + "flos": 22274359649760.0, + "grad_norm": 2.442294955437002, + "learning_rate": 5.173296513982197e-07, + "loss": 0.7897, + "num_input_tokens_seen": 138311115, + "step": 6427 + }, + { + "epoch": 0.7729213010280768, + "flos": 27087596821440.0, + "grad_norm": 3.0040149842661905, + "learning_rate": 5.168069687516398e-07, + "loss": 0.6481, + "num_input_tokens_seen": 138330885, + "step": 6428 + }, + { + "epoch": 0.7730415439187158, + "flos": 18124408540800.0, + "grad_norm": 3.80369702239913, + "learning_rate": 5.16284511104263e-07, + "loss": 0.7219, + "num_input_tokens_seen": 138350020, + "step": 6429 + }, + { + "epoch": 0.7731617868093549, + "flos": 11930691624480.0, + "grad_norm": 6.614878858637785, + "learning_rate": 5.157622785353457e-07, + "loss": 0.8106, + "num_input_tokens_seen": 138368135, + "step": 6430 + }, + { + "epoch": 0.7732820296999939, + "flos": 64109702486400.0, + "grad_norm": 0.654834810859034, + "learning_rate": 5.152402711241113e-07, + "loss": 0.6029, + "num_input_tokens_seen": 138436040, + "step": 6431 + }, + { + "epoch": 0.7734022725906331, + "flos": 25265991901440.0, + "grad_norm": 2.101448027602306, + "learning_rate": 5.147184889497465e-07, + "loss": 0.8334, + "num_input_tokens_seen": 138455510, + "step": 6432 + }, + { + "epoch": 0.7735225154812722, + "flos": 17322790866720.0, + "grad_norm": 3.3783496199749896, + "learning_rate": 5.141969320914072e-07, + "loss": 0.8007, + "num_input_tokens_seen": 138473015, + "step": 6433 + }, + { + "epoch": 0.7736427583719112, + "flos": 32583008027520.0, + "grad_norm": 3.3547503198534954, + "learning_rate": 5.136756006282113e-07, + "loss": 0.631, + "num_input_tokens_seen": 138491680, + "step": 6434 + }, + { + "epoch": 0.7737630012625504, + "flos": 19831587738720.0, + "grad_norm": 3.2286051272190104, + "learning_rate": 5.131544946392446e-07, + "loss": 0.8537, + "num_input_tokens_seen": 138506960, + "step": 6435 + }, + { + "epoch": 0.7738832441531894, + "flos": 35970598462560.0, + "grad_norm": 2.73697665015098, + "learning_rate": 5.126336142035592e-07, + "loss": 0.6395, + "num_input_tokens_seen": 138526995, + "step": 6436 + }, + { + "epoch": 0.7740034870438285, + "flos": 13386302921280.0, + "grad_norm": 3.20063081722554, + "learning_rate": 5.121129594001721e-07, + "loss": 0.7221, + "num_input_tokens_seen": 138541260, + "step": 6437 + }, + { + "epoch": 0.7741237299344677, + "flos": 22055023664160.0, + "grad_norm": 2.7294894796777713, + "learning_rate": 5.115925303080661e-07, + "loss": 0.8135, + "num_input_tokens_seen": 138560970, + "step": 6438 + }, + { + "epoch": 0.7742439728251067, + "flos": 19836456977280.0, + "grad_norm": 2.45550654761813, + "learning_rate": 5.110723270061899e-07, + "loss": 0.7957, + "num_input_tokens_seen": 138577460, + "step": 6439 + }, + { + "epoch": 0.7743642157157458, + "flos": 16666455549120.0, + "grad_norm": 2.156649345237985, + "learning_rate": 5.105523495734572e-07, + "loss": 0.7992, + "num_input_tokens_seen": 138594860, + "step": 6440 + }, + { + "epoch": 0.7744844586063849, + "flos": 20275463476320.0, + "grad_norm": 1.9939429832502436, + "learning_rate": 5.100325980887499e-07, + "loss": 0.7551, + "num_input_tokens_seen": 138614785, + "step": 6441 + }, + { + "epoch": 0.774604701497024, + "flos": 22933891566720.0, + "grad_norm": 2.426192047669115, + "learning_rate": 5.095130726309116e-07, + "loss": 0.8358, + "num_input_tokens_seen": 138634960, + "step": 6442 + }, + { + "epoch": 0.774724944387663, + "flos": 60202584384960.0, + "grad_norm": 0.7947272735015021, + "learning_rate": 5.089937732787559e-07, + "loss": 0.5889, + "num_input_tokens_seen": 138699520, + "step": 6443 + }, + { + "epoch": 0.7748451872783022, + "flos": 26722681121280.0, + "grad_norm": 3.070182816967428, + "learning_rate": 5.084747001110592e-07, + "loss": 0.6681, + "num_input_tokens_seen": 138719145, + "step": 6444 + }, + { + "epoch": 0.7749654301689413, + "flos": 30295145440800.0, + "grad_norm": 2.9745875737874754, + "learning_rate": 5.07955853206564e-07, + "loss": 0.7027, + "num_input_tokens_seen": 138743850, + "step": 6445 + }, + { + "epoch": 0.7750856730595803, + "flos": 43117608608160.0, + "grad_norm": 2.0409172654977796, + "learning_rate": 5.074372326439807e-07, + "loss": 0.7138, + "num_input_tokens_seen": 138766860, + "step": 6446 + }, + { + "epoch": 0.7752059159502195, + "flos": 17614991049120.0, + "grad_norm": 2.904374861038957, + "learning_rate": 5.069188385019814e-07, + "loss": 0.7379, + "num_input_tokens_seen": 138784560, + "step": 6447 + }, + { + "epoch": 0.7753261588408585, + "flos": 12659853969120.0, + "grad_norm": 3.6863512739186537, + "learning_rate": 5.064006708592077e-07, + "loss": 0.6063, + "num_input_tokens_seen": 138800805, + "step": 6448 + }, + { + "epoch": 0.7754464017314976, + "flos": 16667050265280.0, + "grad_norm": 3.298947229939839, + "learning_rate": 5.058827297942641e-07, + "loss": 0.7568, + "num_input_tokens_seen": 138815910, + "step": 6449 + }, + { + "epoch": 0.7755666446221368, + "flos": 19690840092960.0, + "grad_norm": 4.787898409129778, + "learning_rate": 5.053650153857237e-07, + "loss": 0.753, + "num_input_tokens_seen": 138834920, + "step": 6450 + }, + { + "epoch": 0.7756868875127758, + "flos": 18666875682720.0, + "grad_norm": 3.307669230058174, + "learning_rate": 5.048475277121214e-07, + "loss": 0.6993, + "num_input_tokens_seen": 138852135, + "step": 6451 + }, + { + "epoch": 0.7758071304034149, + "flos": 28363500830400.0, + "grad_norm": 3.381045119870496, + "learning_rate": 5.043302668519598e-07, + "loss": 0.7711, + "num_input_tokens_seen": 138871470, + "step": 6452 + }, + { + "epoch": 0.775927373294054, + "flos": 20565991019520.0, + "grad_norm": 2.197274739564429, + "learning_rate": 5.038132328837079e-07, + "loss": 0.7219, + "num_input_tokens_seen": 138889860, + "step": 6453 + }, + { + "epoch": 0.7760476161846931, + "flos": 22494104502720.0, + "grad_norm": 2.665136215994351, + "learning_rate": 5.032964258857993e-07, + "loss": 0.7422, + "num_input_tokens_seen": 138909955, + "step": 6454 + }, + { + "epoch": 0.7761678590753321, + "flos": 48581233935840.0, + "grad_norm": 2.100597741569577, + "learning_rate": 5.027798459366329e-07, + "loss": 0.6874, + "num_input_tokens_seen": 138935320, + "step": 6455 + }, + { + "epoch": 0.7762881019659713, + "flos": 26139544528320.0, + "grad_norm": 2.1523652395847583, + "learning_rate": 5.02263493114573e-07, + "loss": 0.6393, + "num_input_tokens_seen": 138957115, + "step": 6456 + }, + { + "epoch": 0.7764083448566104, + "flos": 20558742916320.0, + "grad_norm": 4.355046574270188, + "learning_rate": 5.017473674979502e-07, + "loss": 0.769, + "num_input_tokens_seen": 138973140, + "step": 6457 + }, + { + "epoch": 0.7765285877472494, + "flos": 67196915391840.0, + "grad_norm": 0.7547090831379323, + "learning_rate": 5.01231469165061e-07, + "loss": 0.5824, + "num_input_tokens_seen": 139028965, + "step": 6458 + }, + { + "epoch": 0.7766488306378886, + "flos": 61256922222720.0, + "grad_norm": 0.8320821827087954, + "learning_rate": 5.007157981941663e-07, + "loss": 0.5672, + "num_input_tokens_seen": 139094325, + "step": 6459 + }, + { + "epoch": 0.7767690735285276, + "flos": 62856180406560.0, + "grad_norm": 0.8828651277502441, + "learning_rate": 5.002003546634928e-07, + "loss": 0.6734, + "num_input_tokens_seen": 139150425, + "step": 6460 + }, + { + "epoch": 0.7768893164191667, + "flos": 20856592902240.0, + "grad_norm": 2.576291258665273, + "learning_rate": 4.996851386512331e-07, + "loss": 0.7624, + "num_input_tokens_seen": 139169130, + "step": 6461 + }, + { + "epoch": 0.7770095593098058, + "flos": 20675064296160.0, + "grad_norm": 2.2156709609759324, + "learning_rate": 4.991701502355444e-07, + "loss": 0.832, + "num_input_tokens_seen": 139189595, + "step": 6462 + }, + { + "epoch": 0.7771298022004449, + "flos": 24682334931840.0, + "grad_norm": 2.128699504367636, + "learning_rate": 4.986553894945518e-07, + "loss": 0.7625, + "num_input_tokens_seen": 139210805, + "step": 6463 + }, + { + "epoch": 0.777250045091084, + "flos": 24973420021440.0, + "grad_norm": 2.650376591824492, + "learning_rate": 4.981408565063416e-07, + "loss": 0.865, + "num_input_tokens_seen": 139230750, + "step": 6464 + }, + { + "epoch": 0.777370287981723, + "flos": 20091184308000.0, + "grad_norm": 2.30386803492806, + "learning_rate": 4.976265513489701e-07, + "loss": 0.7593, + "num_input_tokens_seen": 139250720, + "step": 6465 + }, + { + "epoch": 0.7774905308723622, + "flos": 21687543250560.0, + "grad_norm": 2.819581335927749, + "learning_rate": 4.971124741004562e-07, + "loss": 0.8091, + "num_input_tokens_seen": 139269720, + "step": 6466 + }, + { + "epoch": 0.7776107737630013, + "flos": 16011384003360.0, + "grad_norm": 2.6409359359510884, + "learning_rate": 4.965986248387846e-07, + "loss": 0.7668, + "num_input_tokens_seen": 139288035, + "step": 6467 + }, + { + "epoch": 0.7777310166536403, + "flos": 24755422147200.0, + "grad_norm": 2.210320484347048, + "learning_rate": 4.960850036419073e-07, + "loss": 0.7747, + "num_input_tokens_seen": 139307135, + "step": 6468 + }, + { + "epoch": 0.7778512595442795, + "flos": 17247919502880.0, + "grad_norm": 19.422032914935247, + "learning_rate": 4.955716105877378e-07, + "loss": 0.7877, + "num_input_tokens_seen": 139325905, + "step": 6469 + }, + { + "epoch": 0.7779715024349185, + "flos": 17723172251520.0, + "grad_norm": 2.204436757126004, + "learning_rate": 4.950584457541598e-07, + "loss": 0.832, + "num_input_tokens_seen": 139344370, + "step": 6470 + }, + { + "epoch": 0.7780917453255576, + "flos": 24282102226080.0, + "grad_norm": 1.6282524425102292, + "learning_rate": 4.945455092190183e-07, + "loss": 0.8186, + "num_input_tokens_seen": 139365625, + "step": 6471 + }, + { + "epoch": 0.7782119882161967, + "flos": 56284978677600.0, + "grad_norm": 0.6959740566866495, + "learning_rate": 4.940328010601271e-07, + "loss": 0.559, + "num_input_tokens_seen": 139430540, + "step": 6472 + }, + { + "epoch": 0.7783322311068358, + "flos": 46722416352480.0, + "grad_norm": 3.158371467326565, + "learning_rate": 4.935203213552621e-07, + "loss": 0.7686, + "num_input_tokens_seen": 139454280, + "step": 6473 + }, + { + "epoch": 0.7784524739974749, + "flos": 19030564780800.0, + "grad_norm": 2.4525070870051127, + "learning_rate": 4.930080701821662e-07, + "loss": 0.67, + "num_input_tokens_seen": 139471745, + "step": 6474 + }, + { + "epoch": 0.778572716888114, + "flos": 24755645165760.0, + "grad_norm": 5.429470930651714, + "learning_rate": 4.92496047618548e-07, + "loss": 0.772, + "num_input_tokens_seen": 139491505, + "step": 6475 + }, + { + "epoch": 0.7786929597787531, + "flos": 20049251085120.0, + "grad_norm": 1.926830508856508, + "learning_rate": 4.919842537420811e-07, + "loss": 0.7795, + "num_input_tokens_seen": 139508620, + "step": 6476 + }, + { + "epoch": 0.7788132026693921, + "flos": 21841002954240.0, + "grad_norm": 2.75216242993968, + "learning_rate": 4.91472688630404e-07, + "loss": 0.7931, + "num_input_tokens_seen": 139529360, + "step": 6477 + }, + { + "epoch": 0.7789334455600313, + "flos": 11166286613760.0, + "grad_norm": 2.0592184837961143, + "learning_rate": 4.909613523611202e-07, + "loss": 0.7392, + "num_input_tokens_seen": 139546240, + "step": 6478 + }, + { + "epoch": 0.7790536884506704, + "flos": 28654288561920.0, + "grad_norm": 2.1579314600301243, + "learning_rate": 4.904502450117991e-07, + "loss": 0.7456, + "num_input_tokens_seen": 139567200, + "step": 6479 + }, + { + "epoch": 0.7791739313413094, + "flos": 11056804469760.0, + "grad_norm": 3.2809359445594555, + "learning_rate": 4.899393666599762e-07, + "loss": 0.7205, + "num_input_tokens_seen": 139583775, + "step": 6480 + }, + { + "epoch": 0.7792941742319486, + "flos": 14658341275200.0, + "grad_norm": 3.3173930179505224, + "learning_rate": 4.894287173831506e-07, + "loss": 0.7289, + "num_input_tokens_seen": 139599735, + "step": 6481 + }, + { + "epoch": 0.7794144171225876, + "flos": 23225645712000.0, + "grad_norm": 3.3919682105482103, + "learning_rate": 4.889182972587877e-07, + "loss": 0.8459, + "num_input_tokens_seen": 139619030, + "step": 6482 + }, + { + "epoch": 0.7795346600132267, + "flos": 21476384612160.0, + "grad_norm": 2.1352844445456562, + "learning_rate": 4.884081063643177e-07, + "loss": 0.6657, + "num_input_tokens_seen": 139637690, + "step": 6483 + }, + { + "epoch": 0.7796549029038659, + "flos": 69951593191680.0, + "grad_norm": 0.8584204834764102, + "learning_rate": 4.878981447771353e-07, + "loss": 0.5239, + "num_input_tokens_seen": 139692070, + "step": 6484 + }, + { + "epoch": 0.7797751457945049, + "flos": 23955142584480.0, + "grad_norm": 2.013771683654922, + "learning_rate": 4.873884125746035e-07, + "loss": 0.7322, + "num_input_tokens_seen": 139713745, + "step": 6485 + }, + { + "epoch": 0.779895388685144, + "flos": 22642360440000.0, + "grad_norm": 2.994136484321175, + "learning_rate": 4.868789098340456e-07, + "loss": 0.7282, + "num_input_tokens_seen": 139731650, + "step": 6486 + }, + { + "epoch": 0.7800156315757831, + "flos": 23734691506080.0, + "grad_norm": 3.207129380540391, + "learning_rate": 4.863696366327543e-07, + "loss": 0.7372, + "num_input_tokens_seen": 139750820, + "step": 6487 + }, + { + "epoch": 0.7801358744664222, + "flos": 26391744315360.0, + "grad_norm": 2.1050202806814187, + "learning_rate": 4.85860593047986e-07, + "loss": 0.7799, + "num_input_tokens_seen": 139770885, + "step": 6488 + }, + { + "epoch": 0.7802561173570612, + "flos": 26284418017440.0, + "grad_norm": 3.174452462630367, + "learning_rate": 4.853517791569613e-07, + "loss": 0.753, + "num_input_tokens_seen": 139793065, + "step": 6489 + }, + { + "epoch": 0.7803763602477004, + "flos": 40275644744640.0, + "grad_norm": 3.2694993102998615, + "learning_rate": 4.848431950368684e-07, + "loss": 0.6589, + "num_input_tokens_seen": 139815625, + "step": 6490 + }, + { + "epoch": 0.7804966031383395, + "flos": 66905718792960.0, + "grad_norm": 0.712809226209589, + "learning_rate": 4.843348407648569e-07, + "loss": 0.5568, + "num_input_tokens_seen": 139876905, + "step": 6491 + }, + { + "epoch": 0.7806168460289785, + "flos": 17715217922880.0, + "grad_norm": 3.787607948099415, + "learning_rate": 4.838267164180457e-07, + "loss": 0.8307, + "num_input_tokens_seen": 139892575, + "step": 6492 + }, + { + "epoch": 0.7807370889196176, + "flos": 23911796910720.0, + "grad_norm": 2.5489773176822013, + "learning_rate": 4.833188220735156e-07, + "loss": 0.8385, + "num_input_tokens_seen": 139912245, + "step": 6493 + }, + { + "epoch": 0.7808573318102567, + "flos": 18961975106400.0, + "grad_norm": 3.0197537477149323, + "learning_rate": 4.828111578083152e-07, + "loss": 0.7461, + "num_input_tokens_seen": 139929150, + "step": 6494 + }, + { + "epoch": 0.7809775747008958, + "flos": 23946184672320.0, + "grad_norm": 2.436417167960571, + "learning_rate": 4.823037236994556e-07, + "loss": 0.8143, + "num_input_tokens_seen": 139947315, + "step": 6495 + }, + { + "epoch": 0.7810978175915348, + "flos": 68437985602080.0, + "grad_norm": 0.7277508523399423, + "learning_rate": 4.817965198239136e-07, + "loss": 0.5628, + "num_input_tokens_seen": 140013775, + "step": 6496 + }, + { + "epoch": 0.781218060482174, + "flos": 19613998731840.0, + "grad_norm": 2.4676261951522287, + "learning_rate": 4.812895462586331e-07, + "loss": 0.7474, + "num_input_tokens_seen": 140031600, + "step": 6497 + }, + { + "epoch": 0.7813383033728131, + "flos": 25591093055040.0, + "grad_norm": 2.2098006166301043, + "learning_rate": 4.807828030805207e-07, + "loss": 0.8219, + "num_input_tokens_seen": 140050590, + "step": 6498 + }, + { + "epoch": 0.7814585462634521, + "flos": 20456731894080.0, + "grad_norm": 2.301377949304645, + "learning_rate": 4.802762903664495e-07, + "loss": 0.6779, + "num_input_tokens_seen": 140069770, + "step": 6499 + }, + { + "epoch": 0.7815787891540913, + "flos": 22272240973440.0, + "grad_norm": 7.974213765291256, + "learning_rate": 4.797700081932565e-07, + "loss": 0.7413, + "num_input_tokens_seen": 140087635, + "step": 6500 + }, + { + "epoch": 0.7816990320447303, + "flos": 22568343980640.0, + "grad_norm": 4.19450566722211, + "learning_rate": 4.792639566377442e-07, + "loss": 0.8244, + "num_input_tokens_seen": 140105045, + "step": 6501 + }, + { + "epoch": 0.7818192749353694, + "flos": 24900258466560.0, + "grad_norm": 2.155420669069181, + "learning_rate": 4.78758135776681e-07, + "loss": 0.7776, + "num_input_tokens_seen": 140124900, + "step": 6502 + }, + { + "epoch": 0.7819395178260086, + "flos": 23699114312160.0, + "grad_norm": 4.497947436685813, + "learning_rate": 4.782525456867989e-07, + "loss": 0.7904, + "num_input_tokens_seen": 140143755, + "step": 6503 + }, + { + "epoch": 0.7820597607166476, + "flos": 23188470218400.0, + "grad_norm": 1.89962561833981, + "learning_rate": 4.777471864447959e-07, + "loss": 0.8327, + "num_input_tokens_seen": 140164495, + "step": 6504 + }, + { + "epoch": 0.7821800036072867, + "flos": 22277667758400.0, + "grad_norm": 2.5479127606798384, + "learning_rate": 4.772420581273344e-07, + "loss": 0.8036, + "num_input_tokens_seen": 140181980, + "step": 6505 + }, + { + "epoch": 0.7823002464979258, + "flos": 21513448596480.0, + "grad_norm": 2.7352211651130736, + "learning_rate": 4.7673716081104134e-07, + "loss": 0.7607, + "num_input_tokens_seen": 140202155, + "step": 6506 + }, + { + "epoch": 0.7824204893885649, + "flos": 24500546137440.0, + "grad_norm": 1.9335475106923896, + "learning_rate": 4.762324945725109e-07, + "loss": 0.8459, + "num_input_tokens_seen": 140221600, + "step": 6507 + }, + { + "epoch": 0.782540732279204, + "flos": 27375819839520.0, + "grad_norm": 2.2395821031670557, + "learning_rate": 4.7572805948829844e-07, + "loss": 0.7594, + "num_input_tokens_seen": 140241860, + "step": 6508 + }, + { + "epoch": 0.7826609751698431, + "flos": 24318385645440.0, + "grad_norm": 2.368810921642991, + "learning_rate": 4.7522385563492795e-07, + "loss": 0.7118, + "num_input_tokens_seen": 140262160, + "step": 6509 + }, + { + "epoch": 0.7827812180604822, + "flos": 23954250510240.0, + "grad_norm": 2.520691502472043, + "learning_rate": 4.747198830888863e-07, + "loss": 0.7024, + "num_input_tokens_seen": 140282030, + "step": 6510 + }, + { + "epoch": 0.7829014609511212, + "flos": 27414965330400.0, + "grad_norm": 3.6638118438550924, + "learning_rate": 4.742161419266251e-07, + "loss": 0.6879, + "num_input_tokens_seen": 140301190, + "step": 6511 + }, + { + "epoch": 0.7830217038417604, + "flos": 29162590960800.0, + "grad_norm": 3.2694761210069974, + "learning_rate": 4.7371263222456304e-07, + "loss": 0.6484, + "num_input_tokens_seen": 140318510, + "step": 6512 + }, + { + "epoch": 0.7831419467323995, + "flos": 60864197808480.0, + "grad_norm": 0.8075999542943896, + "learning_rate": 4.7320935405908004e-07, + "loss": 0.6119, + "num_input_tokens_seen": 140379380, + "step": 6513 + }, + { + "epoch": 0.7832621896230385, + "flos": 19654482334080.0, + "grad_norm": 4.944623227167321, + "learning_rate": 4.7270630750652475e-07, + "loss": 0.837, + "num_input_tokens_seen": 140395335, + "step": 6514 + }, + { + "epoch": 0.7833824325136777, + "flos": 24973420021440.0, + "grad_norm": 1.992884196211583, + "learning_rate": 4.7220349264320746e-07, + "loss": 0.8042, + "num_input_tokens_seen": 140414420, + "step": 6515 + }, + { + "epoch": 0.7835026754043167, + "flos": 68702191221600.0, + "grad_norm": 0.7456207773451973, + "learning_rate": 4.71700909545407e-07, + "loss": 0.5481, + "num_input_tokens_seen": 140477365, + "step": 6516 + }, + { + "epoch": 0.7836229182949558, + "flos": 19836271128480.0, + "grad_norm": 2.1967599244454123, + "learning_rate": 4.711985582893627e-07, + "loss": 0.7723, + "num_input_tokens_seen": 140495885, + "step": 6517 + }, + { + "epoch": 0.783743161185595, + "flos": 22932999492480.0, + "grad_norm": 2.403008990718232, + "learning_rate": 4.706964389512811e-07, + "loss": 0.7189, + "num_input_tokens_seen": 140515920, + "step": 6518 + }, + { + "epoch": 0.783863404076234, + "flos": 12440964020640.0, + "grad_norm": 2.149339632152071, + "learning_rate": 4.701945516073345e-07, + "loss": 0.8779, + "num_input_tokens_seen": 140533395, + "step": 6519 + }, + { + "epoch": 0.7839836469668731, + "flos": 24209275199040.0, + "grad_norm": 2.295586672327145, + "learning_rate": 4.696928963336577e-07, + "loss": 0.752, + "num_input_tokens_seen": 140552825, + "step": 6520 + }, + { + "epoch": 0.7841038898575122, + "flos": 62033704763520.0, + "grad_norm": 0.8585007121639153, + "learning_rate": 4.6919147320635224e-07, + "loss": 0.6072, + "num_input_tokens_seen": 140615535, + "step": 6521 + }, + { + "epoch": 0.7842241327481513, + "flos": 20165126427840.0, + "grad_norm": 11.22586046756794, + "learning_rate": 4.6869028230148286e-07, + "loss": 0.7343, + "num_input_tokens_seen": 140633330, + "step": 6522 + }, + { + "epoch": 0.7843443756387903, + "flos": 28034273833440.0, + "grad_norm": 3.8570947745015682, + "learning_rate": 4.6818932369507957e-07, + "loss": 0.5959, + "num_input_tokens_seen": 140652830, + "step": 6523 + }, + { + "epoch": 0.7844646185294295, + "flos": 21292923178560.0, + "grad_norm": 2.527573352507723, + "learning_rate": 4.676885974631386e-07, + "loss": 0.8918, + "num_input_tokens_seen": 140671190, + "step": 6524 + }, + { + "epoch": 0.7845848614200686, + "flos": 23622644648640.0, + "grad_norm": 2.5182441676790033, + "learning_rate": 4.67188103681619e-07, + "loss": 0.8127, + "num_input_tokens_seen": 140690975, + "step": 6525 + }, + { + "epoch": 0.7847051043107076, + "flos": 23368660713120.0, + "grad_norm": 2.7440223749822223, + "learning_rate": 4.666878424264453e-07, + "loss": 0.6913, + "num_input_tokens_seen": 140710930, + "step": 6526 + }, + { + "epoch": 0.7848253472013467, + "flos": 18995061926400.0, + "grad_norm": 2.659138533840797, + "learning_rate": 4.661878137735069e-07, + "loss": 0.7408, + "num_input_tokens_seen": 140729120, + "step": 6527 + }, + { + "epoch": 0.7849455900919858, + "flos": 21148978933440.0, + "grad_norm": 2.3363421545737957, + "learning_rate": 4.656880177986571e-07, + "loss": 0.7517, + "num_input_tokens_seen": 140747665, + "step": 6528 + }, + { + "epoch": 0.7850658329826249, + "flos": 19508642431200.0, + "grad_norm": 4.785190529213953, + "learning_rate": 4.6518845457771607e-07, + "loss": 0.8193, + "num_input_tokens_seen": 140766525, + "step": 6529 + }, + { + "epoch": 0.7851860758732639, + "flos": 12477767816640.0, + "grad_norm": 2.5988103382322425, + "learning_rate": 4.646891241864652e-07, + "loss": 0.7939, + "num_input_tokens_seen": 140785090, + "step": 6530 + }, + { + "epoch": 0.7853063187639031, + "flos": 22927944405120.0, + "grad_norm": 4.240713407850091, + "learning_rate": 4.6419002670065397e-07, + "loss": 0.7373, + "num_input_tokens_seen": 140804670, + "step": 6531 + }, + { + "epoch": 0.7854265616545422, + "flos": 17322605017920.0, + "grad_norm": 2.438912691431278, + "learning_rate": 4.6369116219599445e-07, + "loss": 0.8683, + "num_input_tokens_seen": 140820655, + "step": 6532 + }, + { + "epoch": 0.7855468045451812, + "flos": 23804061745440.0, + "grad_norm": 2.470227246197739, + "learning_rate": 4.631925307481637e-07, + "loss": 0.7939, + "num_input_tokens_seen": 140840470, + "step": 6533 + }, + { + "epoch": 0.7856670474358204, + "flos": 25630052697120.0, + "grad_norm": 2.639172668145064, + "learning_rate": 4.6269413243280533e-07, + "loss": 0.7557, + "num_input_tokens_seen": 140859890, + "step": 6534 + }, + { + "epoch": 0.7857872903264594, + "flos": 18118647228000.0, + "grad_norm": 3.903527371486129, + "learning_rate": 4.621959673255236e-07, + "loss": 0.7433, + "num_input_tokens_seen": 140876190, + "step": 6535 + }, + { + "epoch": 0.7859075332170985, + "flos": 14365583546400.0, + "grad_norm": 3.4231743283536997, + "learning_rate": 4.6169803550189135e-07, + "loss": 0.9047, + "num_input_tokens_seen": 140891885, + "step": 6536 + }, + { + "epoch": 0.7860277761077377, + "flos": 19835713582080.0, + "grad_norm": 2.2048710606735162, + "learning_rate": 4.6120033703744355e-07, + "loss": 0.7777, + "num_input_tokens_seen": 140910780, + "step": 6537 + }, + { + "epoch": 0.7861480189983767, + "flos": 26358360137280.0, + "grad_norm": 2.2617944365439473, + "learning_rate": 4.607028720076822e-07, + "loss": 0.7835, + "num_input_tokens_seen": 140927460, + "step": 6538 + }, + { + "epoch": 0.7862682618890158, + "flos": 24201655398240.0, + "grad_norm": 2.5025543514907977, + "learning_rate": 4.6020564048807074e-07, + "loss": 0.734, + "num_input_tokens_seen": 140945135, + "step": 6539 + }, + { + "epoch": 0.7863885047796549, + "flos": 47482658350080.0, + "grad_norm": 3.1529180483392616, + "learning_rate": 4.5970864255403883e-07, + "loss": 0.7211, + "num_input_tokens_seen": 140966660, + "step": 6540 + }, + { + "epoch": 0.786508747670294, + "flos": 24354334536960.0, + "grad_norm": 3.117152316628669, + "learning_rate": 4.59211878280982e-07, + "loss": 0.8259, + "num_input_tokens_seen": 140986175, + "step": 6541 + }, + { + "epoch": 0.786628990560933, + "flos": 18015595452480.0, + "grad_norm": 2.6140166703315026, + "learning_rate": 4.587153477442578e-07, + "loss": 0.6999, + "num_input_tokens_seen": 141002490, + "step": 6542 + }, + { + "epoch": 0.7867492334515722, + "flos": 25811841491520.0, + "grad_norm": 5.00565361823528, + "learning_rate": 4.582190510191899e-07, + "loss": 0.8139, + "num_input_tokens_seen": 141021150, + "step": 6543 + }, + { + "epoch": 0.7868694763422113, + "flos": 16557270763200.0, + "grad_norm": 2.9073382565355272, + "learning_rate": 4.5772298818106625e-07, + "loss": 0.8758, + "num_input_tokens_seen": 141036940, + "step": 6544 + }, + { + "epoch": 0.7869897192328503, + "flos": 29344045227360.0, + "grad_norm": 4.291544323013756, + "learning_rate": 4.572271593051384e-07, + "loss": 0.7222, + "num_input_tokens_seen": 141054295, + "step": 6545 + }, + { + "epoch": 0.7871099621234895, + "flos": 17104607143680.0, + "grad_norm": 1.9926138627317957, + "learning_rate": 4.567315644666245e-07, + "loss": 0.7841, + "num_input_tokens_seen": 141073090, + "step": 6546 + }, + { + "epoch": 0.7872302050141285, + "flos": 23406653941440.0, + "grad_norm": 2.8003727869100756, + "learning_rate": 4.5623620374070507e-07, + "loss": 0.8522, + "num_input_tokens_seen": 141092405, + "step": 6547 + }, + { + "epoch": 0.7873504479047676, + "flos": 65866137349920.0, + "grad_norm": 0.7662645899268854, + "learning_rate": 4.557410772025263e-07, + "loss": 0.5821, + "num_input_tokens_seen": 141154355, + "step": 6548 + }, + { + "epoch": 0.7874706907954068, + "flos": 23225088165600.0, + "grad_norm": 3.1350771360018626, + "learning_rate": 4.5524618492719803e-07, + "loss": 0.6651, + "num_input_tokens_seen": 141173575, + "step": 6549 + }, + { + "epoch": 0.7875909336860458, + "flos": 28726520872800.0, + "grad_norm": 1.9542693180975519, + "learning_rate": 4.54751526989795e-07, + "loss": 0.7906, + "num_input_tokens_seen": 141195415, + "step": 6550 + }, + { + "epoch": 0.7877111765766849, + "flos": 18670666998240.0, + "grad_norm": 2.539152642630633, + "learning_rate": 4.5425710346535775e-07, + "loss": 0.7952, + "num_input_tokens_seen": 141213360, + "step": 6551 + }, + { + "epoch": 0.787831419467324, + "flos": 27553148262720.0, + "grad_norm": 2.7508976078315226, + "learning_rate": 4.537629144288877e-07, + "loss": 0.8211, + "num_input_tokens_seen": 141230815, + "step": 6552 + }, + { + "epoch": 0.7879516623579631, + "flos": 18124371371040.0, + "grad_norm": 2.7020976148319598, + "learning_rate": 4.5326895995535477e-07, + "loss": 0.7468, + "num_input_tokens_seen": 141249715, + "step": 6553 + }, + { + "epoch": 0.7880719052486022, + "flos": 20310185765760.0, + "grad_norm": 3.1224331804491237, + "learning_rate": 4.527752401196907e-07, + "loss": 0.8467, + "num_input_tokens_seen": 141267730, + "step": 6554 + }, + { + "epoch": 0.7881921481392413, + "flos": 21622336024320.0, + "grad_norm": 3.8504753512921237, + "learning_rate": 4.5228175499679254e-07, + "loss": 0.6683, + "num_input_tokens_seen": 141287315, + "step": 6555 + }, + { + "epoch": 0.7883123910298804, + "flos": 68467689973920.0, + "grad_norm": 0.8424062979007207, + "learning_rate": 4.5178850466152174e-07, + "loss": 0.5445, + "num_input_tokens_seen": 141346145, + "step": 6556 + }, + { + "epoch": 0.7884326339205194, + "flos": 19290718896480.0, + "grad_norm": 2.323232233026961, + "learning_rate": 4.512954891887031e-07, + "loss": 0.8171, + "num_input_tokens_seen": 141364445, + "step": 6557 + }, + { + "epoch": 0.7885528768111585, + "flos": 17759604349920.0, + "grad_norm": 2.797862555197646, + "learning_rate": 4.5080270865312806e-07, + "loss": 0.8334, + "num_input_tokens_seen": 141381640, + "step": 6558 + }, + { + "epoch": 0.7886731197017977, + "flos": 18780892537440.0, + "grad_norm": 5.36346901983761, + "learning_rate": 4.5031016312954985e-07, + "loss": 0.7073, + "num_input_tokens_seen": 141399955, + "step": 6559 + }, + { + "epoch": 0.7887933625924367, + "flos": 33317411308320.0, + "grad_norm": 3.244610499207099, + "learning_rate": 4.498178526926886e-07, + "loss": 0.75, + "num_input_tokens_seen": 141420820, + "step": 6560 + }, + { + "epoch": 0.7889136054830758, + "flos": 16995199339200.0, + "grad_norm": 3.884700323891833, + "learning_rate": 4.4932577741722635e-07, + "loss": 0.7243, + "num_input_tokens_seen": 141439350, + "step": 6561 + }, + { + "epoch": 0.7890338483737149, + "flos": 29382410153280.0, + "grad_norm": 3.4464005808678353, + "learning_rate": 4.4883393737780985e-07, + "loss": 0.7445, + "num_input_tokens_seen": 141460300, + "step": 6562 + }, + { + "epoch": 0.789154091264354, + "flos": 19943002710240.0, + "grad_norm": 2.363664781959796, + "learning_rate": 4.4834233264905254e-07, + "loss": 0.7852, + "num_input_tokens_seen": 141477315, + "step": 6563 + }, + { + "epoch": 0.789274334154993, + "flos": 14517259101600.0, + "grad_norm": 3.37843204764961, + "learning_rate": 4.478509633055294e-07, + "loss": 0.7148, + "num_input_tokens_seen": 141495025, + "step": 6564 + }, + { + "epoch": 0.7893945770456322, + "flos": 21795947471520.0, + "grad_norm": 3.266979465949031, + "learning_rate": 4.473598294217813e-07, + "loss": 0.8031, + "num_input_tokens_seen": 141510450, + "step": 6565 + }, + { + "epoch": 0.7895148199362713, + "flos": 20711050357440.0, + "grad_norm": 3.148981468594374, + "learning_rate": 4.468689310723124e-07, + "loss": 0.7153, + "num_input_tokens_seen": 141528265, + "step": 6566 + }, + { + "epoch": 0.7896350628269103, + "flos": 16666975925760.0, + "grad_norm": 2.1293028855715708, + "learning_rate": 4.463782683315913e-07, + "loss": 0.7883, + "num_input_tokens_seen": 141547270, + "step": 6567 + }, + { + "epoch": 0.7897553057175495, + "flos": 22606039850880.0, + "grad_norm": 2.75291756271718, + "learning_rate": 4.458878412740523e-07, + "loss": 0.7339, + "num_input_tokens_seen": 141566050, + "step": 6568 + }, + { + "epoch": 0.7898755486081885, + "flos": 14516998913280.0, + "grad_norm": 5.474740545665922, + "learning_rate": 4.453976499740919e-07, + "loss": 0.7802, + "num_input_tokens_seen": 141583695, + "step": 6569 + }, + { + "epoch": 0.7899957914988276, + "flos": 17214014948160.0, + "grad_norm": 2.79625196283755, + "learning_rate": 4.4490769450607215e-07, + "loss": 0.7834, + "num_input_tokens_seen": 141601320, + "step": 6570 + }, + { + "epoch": 0.7901160343894668, + "flos": 41219757043200.0, + "grad_norm": 3.0214094116932992, + "learning_rate": 4.4441797494431845e-07, + "loss": 0.7276, + "num_input_tokens_seen": 141623125, + "step": 6571 + }, + { + "epoch": 0.7902362772801058, + "flos": 16813113186720.0, + "grad_norm": 5.730359909178541, + "learning_rate": 4.439284913631207e-07, + "loss": 0.7802, + "num_input_tokens_seen": 141640335, + "step": 6572 + }, + { + "epoch": 0.7903565201707449, + "flos": 27087113614560.0, + "grad_norm": 2.2010018148453754, + "learning_rate": 4.434392438367347e-07, + "loss": 0.8365, + "num_input_tokens_seen": 141659760, + "step": 6573 + }, + { + "epoch": 0.790476763061384, + "flos": 30980441735040.0, + "grad_norm": 2.513437208293977, + "learning_rate": 4.4295023243937677e-07, + "loss": 0.7389, + "num_input_tokens_seen": 141679965, + "step": 6574 + }, + { + "epoch": 0.7905970059520231, + "flos": 22057476868320.0, + "grad_norm": 2.4552431134584825, + "learning_rate": 4.4246145724523123e-07, + "loss": 0.8031, + "num_input_tokens_seen": 141697710, + "step": 6575 + }, + { + "epoch": 0.7907172488426621, + "flos": 20528815525920.0, + "grad_norm": 2.9845382885425287, + "learning_rate": 4.41972918328444e-07, + "loss": 0.7741, + "num_input_tokens_seen": 141715145, + "step": 6576 + }, + { + "epoch": 0.7908374917333013, + "flos": 30038373773280.0, + "grad_norm": 4.104890995041982, + "learning_rate": 4.4148461576312646e-07, + "loss": 0.7729, + "num_input_tokens_seen": 141734810, + "step": 6577 + }, + { + "epoch": 0.7909577346239404, + "flos": 20966409574080.0, + "grad_norm": 1.7646163489966078, + "learning_rate": 4.4099654962335343e-07, + "loss": 0.7447, + "num_input_tokens_seen": 141755260, + "step": 6578 + }, + { + "epoch": 0.7910779775145794, + "flos": 26210252879040.0, + "grad_norm": 2.8384092714190525, + "learning_rate": 4.405087199831636e-07, + "loss": 0.7519, + "num_input_tokens_seen": 141775500, + "step": 6579 + }, + { + "epoch": 0.7911982204052186, + "flos": 22532172070560.0, + "grad_norm": 2.6315061684633467, + "learning_rate": 4.400211269165619e-07, + "loss": 0.6692, + "num_input_tokens_seen": 141791955, + "step": 6580 + }, + { + "epoch": 0.7913184632958576, + "flos": 23079508451040.0, + "grad_norm": 1.975587144324273, + "learning_rate": 4.3953377049751416e-07, + "loss": 0.768, + "num_input_tokens_seen": 141812380, + "step": 6581 + }, + { + "epoch": 0.7914387061864967, + "flos": 12294417892320.0, + "grad_norm": 2.8554411963765522, + "learning_rate": 4.390466507999537e-07, + "loss": 0.7845, + "num_input_tokens_seen": 141828130, + "step": 6582 + }, + { + "epoch": 0.7915589490771359, + "flos": 17578261592640.0, + "grad_norm": 9.570658475889179, + "learning_rate": 4.385597678977748e-07, + "loss": 0.7585, + "num_input_tokens_seen": 141846965, + "step": 6583 + }, + { + "epoch": 0.7916791919677749, + "flos": 25554549447360.0, + "grad_norm": 1.9931491699858563, + "learning_rate": 4.3807312186483726e-07, + "loss": 0.757, + "num_input_tokens_seen": 141867685, + "step": 6584 + }, + { + "epoch": 0.791799434858414, + "flos": 18817547654400.0, + "grad_norm": 3.1220643559823533, + "learning_rate": 4.375867127749655e-07, + "loss": 0.7834, + "num_input_tokens_seen": 141885960, + "step": 6585 + }, + { + "epoch": 0.7919196777490531, + "flos": 25775037695520.0, + "grad_norm": 2.556798873211943, + "learning_rate": 4.3710054070194744e-07, + "loss": 0.6732, + "num_input_tokens_seen": 141905655, + "step": 6586 + }, + { + "epoch": 0.7920399206396922, + "flos": 11930840303520.0, + "grad_norm": 3.15911085947048, + "learning_rate": 4.3661460571953455e-07, + "loss": 0.6664, + "num_input_tokens_seen": 141922390, + "step": 6587 + }, + { + "epoch": 0.7921601635303313, + "flos": 21549100129920.0, + "grad_norm": 2.3167734610918544, + "learning_rate": 4.36128907901443e-07, + "loss": 0.6864, + "num_input_tokens_seen": 141941985, + "step": 6588 + }, + { + "epoch": 0.7922804064209703, + "flos": 18088236630720.0, + "grad_norm": 2.6431778525396754, + "learning_rate": 4.356434473213519e-07, + "loss": 0.7275, + "num_input_tokens_seen": 141959435, + "step": 6589 + }, + { + "epoch": 0.7924006493116095, + "flos": 21621889987200.0, + "grad_norm": 1.9587323681805233, + "learning_rate": 4.351582240529068e-07, + "loss": 0.8007, + "num_input_tokens_seen": 141980265, + "step": 6590 + }, + { + "epoch": 0.7925208922022485, + "flos": 64151561369760.0, + "grad_norm": 0.6916527174113704, + "learning_rate": 4.346732381697149e-07, + "loss": 0.5809, + "num_input_tokens_seen": 142044395, + "step": 6591 + }, + { + "epoch": 0.7926411350928876, + "flos": 16917317224800.0, + "grad_norm": 5.5763080105974, + "learning_rate": 4.3418848974534825e-07, + "loss": 0.8098, + "num_input_tokens_seen": 142061335, + "step": 6592 + }, + { + "epoch": 0.7927613779835267, + "flos": 34410448599840.0, + "grad_norm": 2.2770548877760017, + "learning_rate": 4.3370397885334276e-07, + "loss": 0.6893, + "num_input_tokens_seen": 142081965, + "step": 6593 + }, + { + "epoch": 0.7928816208741658, + "flos": 18924130557120.0, + "grad_norm": 3.628377459118586, + "learning_rate": 4.3321970556719777e-07, + "loss": 0.756, + "num_input_tokens_seen": 142100260, + "step": 6594 + }, + { + "epoch": 0.7930018637648049, + "flos": 18596464690080.0, + "grad_norm": 3.102504788870731, + "learning_rate": 4.3273566996037856e-07, + "loss": 0.7183, + "num_input_tokens_seen": 142116955, + "step": 6595 + }, + { + "epoch": 0.793122106655444, + "flos": 24495379540800.0, + "grad_norm": 2.4293430488983967, + "learning_rate": 4.322518721063113e-07, + "loss": 0.8064, + "num_input_tokens_seen": 142135505, + "step": 6596 + }, + { + "epoch": 0.7932423495460831, + "flos": 34371637636800.0, + "grad_norm": 2.501740803222586, + "learning_rate": 4.3176831207838906e-07, + "loss": 0.7028, + "num_input_tokens_seen": 142158915, + "step": 6597 + }, + { + "epoch": 0.7933625924367221, + "flos": 26941682579040.0, + "grad_norm": 9.426310594336389, + "learning_rate": 4.3128498994996685e-07, + "loss": 0.745, + "num_input_tokens_seen": 142178390, + "step": 6598 + }, + { + "epoch": 0.7934828353273613, + "flos": 29526131379840.0, + "grad_norm": 7.108952972964878, + "learning_rate": 4.308019057943646e-07, + "loss": 0.7131, + "num_input_tokens_seen": 142200115, + "step": 6599 + }, + { + "epoch": 0.7936030782180004, + "flos": 28574101922400.0, + "grad_norm": 2.3164180864671766, + "learning_rate": 4.3031905968486535e-07, + "loss": 0.7497, + "num_input_tokens_seen": 142220945, + "step": 6600 + }, + { + "epoch": 0.7937233211086394, + "flos": 16369088770080.0, + "grad_norm": 5.028875134016094, + "learning_rate": 4.298364516947162e-07, + "loss": 0.6857, + "num_input_tokens_seen": 142238175, + "step": 6601 + }, + { + "epoch": 0.7938435639992786, + "flos": 21981824939520.0, + "grad_norm": 2.8743063048340534, + "learning_rate": 4.293540818971295e-07, + "loss": 0.6541, + "num_input_tokens_seen": 142255980, + "step": 6602 + }, + { + "epoch": 0.7939638068899176, + "flos": 22165212033600.0, + "grad_norm": 19.040581180453845, + "learning_rate": 4.2887195036527934e-07, + "loss": 0.7668, + "num_input_tokens_seen": 142274785, + "step": 6603 + }, + { + "epoch": 0.7940840497805567, + "flos": 17719826973120.0, + "grad_norm": 3.002781055772913, + "learning_rate": 4.28390057172306e-07, + "loss": 0.7359, + "num_input_tokens_seen": 142291290, + "step": 6604 + }, + { + "epoch": 0.7942042926711959, + "flos": 23771383792800.0, + "grad_norm": 3.442316322496167, + "learning_rate": 4.279084023913111e-07, + "loss": 0.7188, + "num_input_tokens_seen": 142309165, + "step": 6605 + }, + { + "epoch": 0.7943245355618349, + "flos": 19217557341600.0, + "grad_norm": 2.2665551742546826, + "learning_rate": 4.2742698609536096e-07, + "loss": 0.6953, + "num_input_tokens_seen": 142327475, + "step": 6606 + }, + { + "epoch": 0.794444778452474, + "flos": 24971747382240.0, + "grad_norm": 2.2311018865241805, + "learning_rate": 4.2694580835748706e-07, + "loss": 0.7883, + "num_input_tokens_seen": 142347335, + "step": 6607 + }, + { + "epoch": 0.7945650213431131, + "flos": 23188284369600.0, + "grad_norm": 3.08287479668311, + "learning_rate": 4.264648692506836e-07, + "loss": 0.7399, + "num_input_tokens_seen": 142366125, + "step": 6608 + }, + { + "epoch": 0.7946852642337522, + "flos": 26025378994560.0, + "grad_norm": 2.3550986636606486, + "learning_rate": 4.2598416884790824e-07, + "loss": 0.7234, + "num_input_tokens_seen": 142385175, + "step": 6609 + }, + { + "epoch": 0.7948055071243912, + "flos": 23735732259360.0, + "grad_norm": 2.7235239143253094, + "learning_rate": 4.255037072220828e-07, + "loss": 0.8075, + "num_input_tokens_seen": 142406545, + "step": 6610 + }, + { + "epoch": 0.7949257500150304, + "flos": 21949407175200.0, + "grad_norm": 2.017803818375484, + "learning_rate": 4.2502348444609293e-07, + "loss": 0.7188, + "num_input_tokens_seen": 142426165, + "step": 6611 + }, + { + "epoch": 0.7950459929056695, + "flos": 25737155976480.0, + "grad_norm": 4.178163967364564, + "learning_rate": 4.2454350059278844e-07, + "loss": 0.6917, + "num_input_tokens_seen": 142447225, + "step": 6612 + }, + { + "epoch": 0.7951662357963085, + "flos": 22126215221760.0, + "grad_norm": 2.3932592302218314, + "learning_rate": 4.240637557349824e-07, + "loss": 0.8447, + "num_input_tokens_seen": 142464440, + "step": 6613 + }, + { + "epoch": 0.7952864786869477, + "flos": 24606571493760.0, + "grad_norm": 2.8260444870759076, + "learning_rate": 4.235842499454516e-07, + "loss": 0.6621, + "num_input_tokens_seen": 142484355, + "step": 6614 + }, + { + "epoch": 0.7954067215775867, + "flos": 21799590108000.0, + "grad_norm": 3.9446186928624773, + "learning_rate": 4.2310498329693687e-07, + "loss": 0.8291, + "num_input_tokens_seen": 142505255, + "step": 6615 + }, + { + "epoch": 0.7955269644682258, + "flos": 24026222632800.0, + "grad_norm": 2.131757873699562, + "learning_rate": 4.2262595586214164e-07, + "loss": 0.8096, + "num_input_tokens_seen": 142526350, + "step": 6616 + }, + { + "epoch": 0.795647207358865, + "flos": 24974981151360.0, + "grad_norm": 2.8223631950857007, + "learning_rate": 4.221471677137358e-07, + "loss": 0.7717, + "num_input_tokens_seen": 142547165, + "step": 6617 + }, + { + "epoch": 0.795767450249504, + "flos": 14627670489600.0, + "grad_norm": 2.1169415433126004, + "learning_rate": 4.216686189243492e-07, + "loss": 0.7017, + "num_input_tokens_seen": 142565955, + "step": 6618 + }, + { + "epoch": 0.7958876931401431, + "flos": 18521370307680.0, + "grad_norm": 30.37478499473803, + "learning_rate": 4.211903095665785e-07, + "loss": 0.7263, + "num_input_tokens_seen": 142585340, + "step": 6619 + }, + { + "epoch": 0.7960079360307821, + "flos": 21513225577920.0, + "grad_norm": 2.818815134991791, + "learning_rate": 4.2071223971298277e-07, + "loss": 0.7516, + "num_input_tokens_seen": 142602525, + "step": 6620 + }, + { + "epoch": 0.7961281789214213, + "flos": 25445178812640.0, + "grad_norm": 2.74971527012752, + "learning_rate": 4.2023440943608433e-07, + "loss": 0.6126, + "num_input_tokens_seen": 142622490, + "step": 6621 + }, + { + "epoch": 0.7962484218120603, + "flos": 21913086586080.0, + "grad_norm": 4.0372867775940975, + "learning_rate": 4.1975681880837023e-07, + "loss": 0.7849, + "num_input_tokens_seen": 142642495, + "step": 6622 + }, + { + "epoch": 0.7963686647026994, + "flos": 18849259193280.0, + "grad_norm": 1.9519643157644302, + "learning_rate": 4.192794679022895e-07, + "loss": 0.8257, + "num_input_tokens_seen": 142660820, + "step": 6623 + }, + { + "epoch": 0.7964889075933386, + "flos": 29673941280000.0, + "grad_norm": 2.2946243158156463, + "learning_rate": 4.1880235679025743e-07, + "loss": 0.7236, + "num_input_tokens_seen": 142680915, + "step": 6624 + }, + { + "epoch": 0.7966091504839776, + "flos": 29448918321120.0, + "grad_norm": 2.5648026703894997, + "learning_rate": 4.1832548554464986e-07, + "loss": 0.6393, + "num_input_tokens_seen": 142699210, + "step": 6625 + }, + { + "epoch": 0.7967293933746167, + "flos": 67192380681120.0, + "grad_norm": 0.7505244567134168, + "learning_rate": 4.178488542378098e-07, + "loss": 0.5857, + "num_input_tokens_seen": 142756790, + "step": 6626 + }, + { + "epoch": 0.7968496362652558, + "flos": 25518005839680.0, + "grad_norm": 2.157833789275692, + "learning_rate": 4.173724629420401e-07, + "loss": 0.8896, + "num_input_tokens_seen": 142778150, + "step": 6627 + }, + { + "epoch": 0.7969698791558949, + "flos": 14480566814880.0, + "grad_norm": 3.0637349344145375, + "learning_rate": 4.168963117296087e-07, + "loss": 0.6827, + "num_input_tokens_seen": 142795485, + "step": 6628 + }, + { + "epoch": 0.797090122046534, + "flos": 22095953303520.0, + "grad_norm": 2.656841705838726, + "learning_rate": 4.1642040067274876e-07, + "loss": 0.7596, + "num_input_tokens_seen": 142815155, + "step": 6629 + }, + { + "epoch": 0.7972103649371731, + "flos": 19869172099680.0, + "grad_norm": 2.106565964541312, + "learning_rate": 4.1594472984365493e-07, + "loss": 0.7237, + "num_input_tokens_seen": 142833840, + "step": 6630 + }, + { + "epoch": 0.7973306078278122, + "flos": 36006510184320.0, + "grad_norm": 2.2735874611596714, + "learning_rate": 4.154692993144862e-07, + "loss": 0.7827, + "num_input_tokens_seen": 142853610, + "step": 6631 + }, + { + "epoch": 0.7974508507184512, + "flos": 21330098672160.0, + "grad_norm": 3.7816391421499262, + "learning_rate": 4.1499410915736476e-07, + "loss": 0.7201, + "num_input_tokens_seen": 142872540, + "step": 6632 + }, + { + "epoch": 0.7975710936090904, + "flos": 68155969933920.0, + "grad_norm": 0.7745980312422858, + "learning_rate": 4.145191594443762e-07, + "loss": 0.6417, + "num_input_tokens_seen": 142936895, + "step": 6633 + }, + { + "epoch": 0.7976913364997295, + "flos": 22460311457280.0, + "grad_norm": 2.677091981945589, + "learning_rate": 4.140444502475713e-07, + "loss": 0.7071, + "num_input_tokens_seen": 142956445, + "step": 6634 + }, + { + "epoch": 0.7978115793903685, + "flos": 15241440698400.0, + "grad_norm": 3.318969454935952, + "learning_rate": 4.1356998163896216e-07, + "loss": 0.703, + "num_input_tokens_seen": 142973495, + "step": 6635 + }, + { + "epoch": 0.7979318222810077, + "flos": 19691137451040.0, + "grad_norm": 7.023242510676214, + "learning_rate": 4.130957536905255e-07, + "loss": 0.7522, + "num_input_tokens_seen": 142991510, + "step": 6636 + }, + { + "epoch": 0.7980520651716467, + "flos": 15538249931040.0, + "grad_norm": 3.2489515227237966, + "learning_rate": 4.1262176647420134e-07, + "loss": 0.7166, + "num_input_tokens_seen": 143009675, + "step": 6637 + }, + { + "epoch": 0.7981723080622858, + "flos": 22277556249120.0, + "grad_norm": 2.18756888144899, + "learning_rate": 4.121480200618923e-07, + "loss": 0.8008, + "num_input_tokens_seen": 143029330, + "step": 6638 + }, + { + "epoch": 0.798292550952925, + "flos": 22896790412640.0, + "grad_norm": 5.068560140231306, + "learning_rate": 4.116745145254674e-07, + "loss": 0.8022, + "num_input_tokens_seen": 143048865, + "step": 6639 + }, + { + "epoch": 0.798412793843564, + "flos": 64406363040000.0, + "grad_norm": 0.7723114216201094, + "learning_rate": 4.1120124993675476e-07, + "loss": 0.5776, + "num_input_tokens_seen": 143113295, + "step": 6640 + }, + { + "epoch": 0.7985330367342031, + "flos": 13567645678560.0, + "grad_norm": 2.297728088148254, + "learning_rate": 4.107282263675498e-07, + "loss": 0.6137, + "num_input_tokens_seen": 143130965, + "step": 6641 + }, + { + "epoch": 0.7986532796248422, + "flos": 67602203184960.0, + "grad_norm": 0.7693701251051388, + "learning_rate": 4.1025544388960907e-07, + "loss": 0.5242, + "num_input_tokens_seen": 143192005, + "step": 6642 + }, + { + "epoch": 0.7987735225154813, + "flos": 22423284642720.0, + "grad_norm": 2.9535145981949182, + "learning_rate": 4.097829025746538e-07, + "loss": 0.7148, + "num_input_tokens_seen": 143213550, + "step": 6643 + }, + { + "epoch": 0.7988937654061203, + "flos": 68765725808640.0, + "grad_norm": 0.6729912467546015, + "learning_rate": 4.0931060249436757e-07, + "loss": 0.6091, + "num_input_tokens_seen": 143277390, + "step": 6644 + }, + { + "epoch": 0.7990140082967595, + "flos": 20776963809120.0, + "grad_norm": 2.8701442679587528, + "learning_rate": 4.088385437203978e-07, + "loss": 0.6976, + "num_input_tokens_seen": 143294400, + "step": 6645 + }, + { + "epoch": 0.7991342511873986, + "flos": 18958109451360.0, + "grad_norm": 7.200584377309654, + "learning_rate": 4.083667263243564e-07, + "loss": 0.7735, + "num_input_tokens_seen": 143312745, + "step": 6646 + }, + { + "epoch": 0.7992544940780376, + "flos": 20787222662880.0, + "grad_norm": 2.367351463355157, + "learning_rate": 4.0789515037781653e-07, + "loss": 0.7192, + "num_input_tokens_seen": 143333470, + "step": 6647 + }, + { + "epoch": 0.7993747369686768, + "flos": 12622975833600.0, + "grad_norm": 2.481851699861904, + "learning_rate": 4.0742381595231755e-07, + "loss": 0.8258, + "num_input_tokens_seen": 143350195, + "step": 6648 + }, + { + "epoch": 0.7994949798593158, + "flos": 20049957310560.0, + "grad_norm": 2.253324595590538, + "learning_rate": 4.06952723119359e-07, + "loss": 0.7842, + "num_input_tokens_seen": 143370070, + "step": 6649 + }, + { + "epoch": 0.7996152227499549, + "flos": 38598504446400.0, + "grad_norm": 2.732625259781203, + "learning_rate": 4.0648187195040504e-07, + "loss": 0.6716, + "num_input_tokens_seen": 143392275, + "step": 6650 + }, + { + "epoch": 0.799735465640594, + "flos": 70143194802720.0, + "grad_norm": 0.8154538088217501, + "learning_rate": 4.060112625168848e-07, + "loss": 0.6744, + "num_input_tokens_seen": 143457385, + "step": 6651 + }, + { + "epoch": 0.7998557085312331, + "flos": 24206227278720.0, + "grad_norm": 2.089235363401291, + "learning_rate": 4.055408948901886e-07, + "loss": 0.743, + "num_input_tokens_seen": 143478785, + "step": 6652 + }, + { + "epoch": 0.7999759514218722, + "flos": 27524521813920.0, + "grad_norm": 2.3777826972664937, + "learning_rate": 4.050707691416708e-07, + "loss": 0.7192, + "num_input_tokens_seen": 143500095, + "step": 6653 + }, + { + "epoch": 0.8000961943125112, + "flos": 67241487667680.0, + "grad_norm": 0.6948832140370326, + "learning_rate": 4.046008853426495e-07, + "loss": 0.5969, + "num_input_tokens_seen": 143563410, + "step": 6654 + }, + { + "epoch": 0.8002164372031504, + "flos": 28692690657600.0, + "grad_norm": 2.5889276924099485, + "learning_rate": 4.0413124356440464e-07, + "loss": 0.6288, + "num_input_tokens_seen": 143587455, + "step": 6655 + }, + { + "epoch": 0.8003366800937894, + "flos": 17614359163200.0, + "grad_norm": 2.3532004497534778, + "learning_rate": 4.0366184387818223e-07, + "loss": 0.8194, + "num_input_tokens_seen": 143605915, + "step": 6656 + }, + { + "epoch": 0.8004569229844285, + "flos": 25958090261760.0, + "grad_norm": 3.090736848448509, + "learning_rate": 4.0319268635518797e-07, + "loss": 0.8487, + "num_input_tokens_seen": 143626490, + "step": 6657 + }, + { + "epoch": 0.8005771658750677, + "flos": 20782613612640.0, + "grad_norm": 3.5679261226960124, + "learning_rate": 4.027237710665943e-07, + "loss": 0.7545, + "num_input_tokens_seen": 143644955, + "step": 6658 + }, + { + "epoch": 0.8006974087657067, + "flos": 25775669581440.0, + "grad_norm": 2.1030305747704165, + "learning_rate": 4.022550980835344e-07, + "loss": 0.6967, + "num_input_tokens_seen": 143667200, + "step": 6659 + }, + { + "epoch": 0.8008176516563458, + "flos": 17140295846880.0, + "grad_norm": 2.4880600955243084, + "learning_rate": 4.017866674771051e-07, + "loss": 0.7961, + "num_input_tokens_seen": 143684685, + "step": 6660 + }, + { + "epoch": 0.8009378945469849, + "flos": 24172694421600.0, + "grad_norm": 2.618486133473895, + "learning_rate": 4.013184793183688e-07, + "loss": 0.7461, + "num_input_tokens_seen": 143706780, + "step": 6661 + }, + { + "epoch": 0.801058137437624, + "flos": 19758946560480.0, + "grad_norm": 2.606281205317532, + "learning_rate": 4.008505336783472e-07, + "loss": 0.7283, + "num_input_tokens_seen": 143724505, + "step": 6662 + }, + { + "epoch": 0.801178380328263, + "flos": 18632785279200.0, + "grad_norm": 7.236256694569854, + "learning_rate": 4.003828306280284e-07, + "loss": 0.8109, + "num_input_tokens_seen": 143742180, + "step": 6663 + }, + { + "epoch": 0.8012986232189022, + "flos": 15684350022240.0, + "grad_norm": 2.285908997280438, + "learning_rate": 3.999153702383626e-07, + "loss": 0.7776, + "num_input_tokens_seen": 143760070, + "step": 6664 + }, + { + "epoch": 0.8014188661095413, + "flos": 28543096608960.0, + "grad_norm": 3.1929027172528794, + "learning_rate": 3.9944815258026263e-07, + "loss": 0.7375, + "num_input_tokens_seen": 143781890, + "step": 6665 + }, + { + "epoch": 0.8015391090001803, + "flos": 29268133110240.0, + "grad_norm": 3.150275590739233, + "learning_rate": 3.989811777246057e-07, + "loss": 0.8339, + "num_input_tokens_seen": 143802060, + "step": 6666 + }, + { + "epoch": 0.8016593518908195, + "flos": 70296766015680.0, + "grad_norm": 0.8588370633586256, + "learning_rate": 3.985144457422305e-07, + "loss": 0.6609, + "num_input_tokens_seen": 143856345, + "step": 6667 + }, + { + "epoch": 0.8017795947814585, + "flos": 25988649538080.0, + "grad_norm": 2.678341007312668, + "learning_rate": 3.9804795670394096e-07, + "loss": 0.7692, + "num_input_tokens_seen": 143876470, + "step": 6668 + }, + { + "epoch": 0.8018998376720976, + "flos": 22055506871040.0, + "grad_norm": 1.9093706196486553, + "learning_rate": 3.975817106805022e-07, + "loss": 0.7075, + "num_input_tokens_seen": 143895170, + "step": 6669 + }, + { + "epoch": 0.8020200805627368, + "flos": 34518815651040.0, + "grad_norm": 2.8654918177948305, + "learning_rate": 3.97115707742645e-07, + "loss": 0.6441, + "num_input_tokens_seen": 143915845, + "step": 6670 + }, + { + "epoch": 0.8021403234533758, + "flos": 20091667514880.0, + "grad_norm": 2.299362703846775, + "learning_rate": 3.966499479610599e-07, + "loss": 0.6518, + "num_input_tokens_seen": 143933940, + "step": 6671 + }, + { + "epoch": 0.8022605663440149, + "flos": 27706422117600.0, + "grad_norm": 2.201362716674326, + "learning_rate": 3.9618443140640225e-07, + "loss": 0.6538, + "num_input_tokens_seen": 143952850, + "step": 6672 + }, + { + "epoch": 0.802380809234654, + "flos": 60158867013600.0, + "grad_norm": 0.7011309641004448, + "learning_rate": 3.957191581492918e-07, + "loss": 0.5118, + "num_input_tokens_seen": 144013610, + "step": 6673 + }, + { + "epoch": 0.8025010521252931, + "flos": 15059205866880.0, + "grad_norm": 3.5575221570223676, + "learning_rate": 3.952541282603097e-07, + "loss": 0.7148, + "num_input_tokens_seen": 144028715, + "step": 6674 + }, + { + "epoch": 0.8026212950159322, + "flos": 22131865025280.0, + "grad_norm": 2.843481197857795, + "learning_rate": 3.9478934181000013e-07, + "loss": 0.8359, + "num_input_tokens_seen": 144048740, + "step": 6675 + }, + { + "epoch": 0.8027415379065713, + "flos": 17650865601120.0, + "grad_norm": 4.8070596437128685, + "learning_rate": 3.943247988688714e-07, + "loss": 0.8467, + "num_input_tokens_seen": 144067435, + "step": 6676 + }, + { + "epoch": 0.8028617807972104, + "flos": 21948143403360.0, + "grad_norm": 2.121005286519946, + "learning_rate": 3.938604995073933e-07, + "loss": 0.7238, + "num_input_tokens_seen": 144085905, + "step": 6677 + }, + { + "epoch": 0.8029820236878494, + "flos": 26390443373760.0, + "grad_norm": 2.839145495850194, + "learning_rate": 3.9339644379600157e-07, + "loss": 0.6509, + "num_input_tokens_seen": 144105735, + "step": 6678 + }, + { + "epoch": 0.8031022665784886, + "flos": 17651385977760.0, + "grad_norm": 4.45862504711834, + "learning_rate": 3.929326318050907e-07, + "loss": 0.7118, + "num_input_tokens_seen": 144123405, + "step": 6679 + }, + { + "epoch": 0.8032225094691277, + "flos": 15428396089440.0, + "grad_norm": 2.294782866468163, + "learning_rate": 3.924690636050225e-07, + "loss": 0.7901, + "num_input_tokens_seen": 144140815, + "step": 6680 + }, + { + "epoch": 0.8033427523597667, + "flos": 26141774713920.0, + "grad_norm": 3.0140767446668804, + "learning_rate": 3.9200573926611915e-07, + "loss": 0.7284, + "num_input_tokens_seen": 144162230, + "step": 6681 + }, + { + "epoch": 0.8034629952504058, + "flos": 21294484308480.0, + "grad_norm": 7.316713863327751, + "learning_rate": 3.9154265885866613e-07, + "loss": 0.7299, + "num_input_tokens_seen": 144181540, + "step": 6682 + }, + { + "epoch": 0.8035832381410449, + "flos": 21620626215360.0, + "grad_norm": 4.59337962207992, + "learning_rate": 3.9107982245291394e-07, + "loss": 0.7556, + "num_input_tokens_seen": 144199665, + "step": 6683 + }, + { + "epoch": 0.803703481031684, + "flos": 20489447016480.0, + "grad_norm": 2.9403411436353077, + "learning_rate": 3.9061723011907245e-07, + "loss": 0.7748, + "num_input_tokens_seen": 144219020, + "step": 6684 + }, + { + "epoch": 0.803823723922323, + "flos": 22822067727840.0, + "grad_norm": 2.384310392251886, + "learning_rate": 3.901548819273179e-07, + "loss": 0.7948, + "num_input_tokens_seen": 144239035, + "step": 6685 + }, + { + "epoch": 0.8039439668129622, + "flos": 21331696971840.0, + "grad_norm": 3.0908246322047224, + "learning_rate": 3.896927779477881e-07, + "loss": 0.6925, + "num_input_tokens_seen": 144258295, + "step": 6686 + }, + { + "epoch": 0.8040642097036013, + "flos": 23768224363200.0, + "grad_norm": 2.7236210825200082, + "learning_rate": 3.892309182505833e-07, + "loss": 0.6725, + "num_input_tokens_seen": 144276110, + "step": 6687 + }, + { + "epoch": 0.8041844525942403, + "flos": 25884928706880.0, + "grad_norm": 2.981909567369331, + "learning_rate": 3.887693029057675e-07, + "loss": 0.8595, + "num_input_tokens_seen": 144295620, + "step": 6688 + }, + { + "epoch": 0.8043046954848795, + "flos": 25155022967040.0, + "grad_norm": 1.9432425765780712, + "learning_rate": 3.8830793198336684e-07, + "loss": 0.811, + "num_input_tokens_seen": 144315210, + "step": 6689 + }, + { + "epoch": 0.8044249383755185, + "flos": 41659023730560.0, + "grad_norm": 1.8836852144417133, + "learning_rate": 3.878468055533721e-07, + "loss": 0.7049, + "num_input_tokens_seen": 144337750, + "step": 6690 + }, + { + "epoch": 0.8045451812661576, + "flos": 20602014250560.0, + "grad_norm": 4.407405620566278, + "learning_rate": 3.8738592368573464e-07, + "loss": 0.8469, + "num_input_tokens_seen": 144355305, + "step": 6691 + }, + { + "epoch": 0.8046654241567968, + "flos": 29668997701920.0, + "grad_norm": 2.793816796513413, + "learning_rate": 3.8692528645037137e-07, + "loss": 0.8829, + "num_input_tokens_seen": 144374485, + "step": 6692 + }, + { + "epoch": 0.8047856670474358, + "flos": 17646107871840.0, + "grad_norm": 2.7323073062504872, + "learning_rate": 3.8646489391715907e-07, + "loss": 0.7818, + "num_input_tokens_seen": 144388780, + "step": 6693 + }, + { + "epoch": 0.8049059099380749, + "flos": 17096095268640.0, + "grad_norm": 2.935910708075267, + "learning_rate": 3.8600474615593903e-07, + "loss": 0.8816, + "num_input_tokens_seen": 144403395, + "step": 6694 + }, + { + "epoch": 0.805026152828714, + "flos": 62123821462560.0, + "grad_norm": 0.7859235212474557, + "learning_rate": 3.8554484323651605e-07, + "loss": 0.595, + "num_input_tokens_seen": 144465265, + "step": 6695 + }, + { + "epoch": 0.8051463957193531, + "flos": 21657615860160.0, + "grad_norm": 2.278050726790465, + "learning_rate": 3.85085185228657e-07, + "loss": 0.7931, + "num_input_tokens_seen": 144484235, + "step": 6696 + }, + { + "epoch": 0.8052666386099921, + "flos": 32005409728800.0, + "grad_norm": 2.3293308489079685, + "learning_rate": 3.8462577220209114e-07, + "loss": 0.7322, + "num_input_tokens_seen": 144504610, + "step": 6697 + }, + { + "epoch": 0.8053868815006313, + "flos": 67062300756480.0, + "grad_norm": 0.7205389399227549, + "learning_rate": 3.8416660422651127e-07, + "loss": 0.5889, + "num_input_tokens_seen": 144574260, + "step": 6698 + }, + { + "epoch": 0.8055071243912704, + "flos": 23802723634080.0, + "grad_norm": 10.059385966982022, + "learning_rate": 3.837076813715723e-07, + "loss": 0.6849, + "num_input_tokens_seen": 144594145, + "step": 6699 + }, + { + "epoch": 0.8056273672819094, + "flos": 21294075441120.0, + "grad_norm": 2.172390444399883, + "learning_rate": 3.832490037068941e-07, + "loss": 0.753, + "num_input_tokens_seen": 144612935, + "step": 6700 + }, + { + "epoch": 0.8057476101725486, + "flos": 25731208814880.0, + "grad_norm": 2.5667875752725404, + "learning_rate": 3.827905713020554e-07, + "loss": 0.7599, + "num_input_tokens_seen": 144630580, + "step": 6701 + }, + { + "epoch": 0.8058678530631876, + "flos": 24500174439840.0, + "grad_norm": 2.413674233826378, + "learning_rate": 3.823323842266017e-07, + "loss": 0.6874, + "num_input_tokens_seen": 144649975, + "step": 6702 + }, + { + "epoch": 0.8059880959538267, + "flos": 24718283823360.0, + "grad_norm": 3.088844803359062, + "learning_rate": 3.818744425500393e-07, + "loss": 0.7347, + "num_input_tokens_seen": 144667990, + "step": 6703 + }, + { + "epoch": 0.8061083388444659, + "flos": 22164505808160.0, + "grad_norm": 2.483399205539598, + "learning_rate": 3.8141674634183675e-07, + "loss": 0.8055, + "num_input_tokens_seen": 144687020, + "step": 6704 + }, + { + "epoch": 0.8062285817351049, + "flos": 30001049600640.0, + "grad_norm": 2.6326771143846592, + "learning_rate": 3.809592956714278e-07, + "loss": 0.6625, + "num_input_tokens_seen": 144710255, + "step": 6705 + }, + { + "epoch": 0.806348824625744, + "flos": 22749984096000.0, + "grad_norm": 2.7748428437692083, + "learning_rate": 3.805020906082057e-07, + "loss": 0.7443, + "num_input_tokens_seen": 144728830, + "step": 6706 + }, + { + "epoch": 0.8064690675163831, + "flos": 23371151087040.0, + "grad_norm": 3.910495861445867, + "learning_rate": 3.8004513122152917e-07, + "loss": 0.8095, + "num_input_tokens_seen": 144747140, + "step": 6707 + }, + { + "epoch": 0.8065893104070222, + "flos": 24026482821120.0, + "grad_norm": 4.190377684256963, + "learning_rate": 3.79588417580718e-07, + "loss": 0.6733, + "num_input_tokens_seen": 144765250, + "step": 6708 + }, + { + "epoch": 0.8067095532976613, + "flos": 22273393236000.0, + "grad_norm": 2.914081645012931, + "learning_rate": 3.791319497550558e-07, + "loss": 0.7664, + "num_input_tokens_seen": 144783080, + "step": 6709 + }, + { + "epoch": 0.8068297961883004, + "flos": 17105201859840.0, + "grad_norm": 2.74945479103571, + "learning_rate": 3.78675727813788e-07, + "loss": 0.7069, + "num_input_tokens_seen": 144800645, + "step": 6710 + }, + { + "epoch": 0.8069500390789395, + "flos": 21987437573280.0, + "grad_norm": 2.160749191939011, + "learning_rate": 3.782197518261225e-07, + "loss": 0.735, + "num_input_tokens_seen": 144820085, + "step": 6711 + }, + { + "epoch": 0.8070702819695785, + "flos": 19216813946400.0, + "grad_norm": 3.5311795117930265, + "learning_rate": 3.777640218612319e-07, + "loss": 0.958, + "num_input_tokens_seen": 144838070, + "step": 6712 + }, + { + "epoch": 0.8071905248602176, + "flos": 21513597275520.0, + "grad_norm": 2.6060655816785427, + "learning_rate": 3.773085379882488e-07, + "loss": 0.72, + "num_input_tokens_seen": 144857125, + "step": 6713 + }, + { + "epoch": 0.8073107677508568, + "flos": 37214642253600.0, + "grad_norm": 2.2797622009546545, + "learning_rate": 3.768533002762715e-07, + "loss": 0.7572, + "num_input_tokens_seen": 144880660, + "step": 6714 + }, + { + "epoch": 0.8074310106414958, + "flos": 28325581941600.0, + "grad_norm": 2.0914387258317397, + "learning_rate": 3.763983087943572e-07, + "loss": 0.7657, + "num_input_tokens_seen": 144900050, + "step": 6715 + }, + { + "epoch": 0.8075512535321349, + "flos": 24246971069280.0, + "grad_norm": 1.990194602920809, + "learning_rate": 3.759435636115282e-07, + "loss": 0.8092, + "num_input_tokens_seen": 144920835, + "step": 6716 + }, + { + "epoch": 0.807671496422774, + "flos": 25993555946400.0, + "grad_norm": 2.2016534963825585, + "learning_rate": 3.7548906479676967e-07, + "loss": 0.7314, + "num_input_tokens_seen": 144940430, + "step": 6717 + }, + { + "epoch": 0.8077917393134131, + "flos": 23696735447520.0, + "grad_norm": 2.0891722554011762, + "learning_rate": 3.7503481241902855e-07, + "loss": 0.7147, + "num_input_tokens_seen": 144960405, + "step": 6718 + }, + { + "epoch": 0.8079119822040521, + "flos": 18375716253600.0, + "grad_norm": 3.0594750860440136, + "learning_rate": 3.745808065472145e-07, + "loss": 0.8061, + "num_input_tokens_seen": 144977700, + "step": 6719 + }, + { + "epoch": 0.8080322250946913, + "flos": 23587959528960.0, + "grad_norm": 2.204017079787356, + "learning_rate": 3.741270472501994e-07, + "loss": 0.7663, + "num_input_tokens_seen": 144998340, + "step": 6720 + }, + { + "epoch": 0.8081524679853304, + "flos": 22788088833600.0, + "grad_norm": 2.273558518377483, + "learning_rate": 3.736735345968183e-07, + "loss": 0.7279, + "num_input_tokens_seen": 145017950, + "step": 6721 + }, + { + "epoch": 0.8082727108759694, + "flos": 17614953879360.0, + "grad_norm": 1.8883237468591743, + "learning_rate": 3.7322026865586986e-07, + "loss": 0.7879, + "num_input_tokens_seen": 145036985, + "step": 6722 + }, + { + "epoch": 0.8083929537666086, + "flos": 25920989107680.0, + "grad_norm": 2.78115913156948, + "learning_rate": 3.7276724949611206e-07, + "loss": 0.7369, + "num_input_tokens_seen": 145057095, + "step": 6723 + }, + { + "epoch": 0.8085131966572476, + "flos": 27050346988320.0, + "grad_norm": 4.064675086292741, + "learning_rate": 3.723144771862694e-07, + "loss": 0.7508, + "num_input_tokens_seen": 145077735, + "step": 6724 + }, + { + "epoch": 0.8086334395478867, + "flos": 23954250510240.0, + "grad_norm": 2.317099867827313, + "learning_rate": 3.718619517950263e-07, + "loss": 0.7701, + "num_input_tokens_seen": 145098330, + "step": 6725 + }, + { + "epoch": 0.8087536824385259, + "flos": 20376545254560.0, + "grad_norm": 4.1621338920998925, + "learning_rate": 3.714096733910301e-07, + "loss": 0.7649, + "num_input_tokens_seen": 145115855, + "step": 6726 + }, + { + "epoch": 0.8088739253291649, + "flos": 25882438332960.0, + "grad_norm": 2.6493582686255683, + "learning_rate": 3.709576420428926e-07, + "loss": 0.7014, + "num_input_tokens_seen": 145136655, + "step": 6727 + }, + { + "epoch": 0.808994168219804, + "flos": 28361307814560.0, + "grad_norm": 3.2337238927966236, + "learning_rate": 3.7050585781918463e-07, + "loss": 0.736, + "num_input_tokens_seen": 145156955, + "step": 6728 + }, + { + "epoch": 0.8091144111104431, + "flos": 17396324119200.0, + "grad_norm": 2.7140204848582528, + "learning_rate": 3.700543207884428e-07, + "loss": 0.688, + "num_input_tokens_seen": 145173865, + "step": 6729 + }, + { + "epoch": 0.8092346540010822, + "flos": 32107346411520.0, + "grad_norm": 2.104440866362338, + "learning_rate": 3.6960303101916466e-07, + "loss": 0.7133, + "num_input_tokens_seen": 145193780, + "step": 6730 + }, + { + "epoch": 0.8093548968917212, + "flos": 57952752196320.0, + "grad_norm": 0.7650253127155068, + "learning_rate": 3.6915198857981047e-07, + "loss": 0.554, + "num_input_tokens_seen": 145258980, + "step": 6731 + }, + { + "epoch": 0.8094751397823604, + "flos": 27342361321920.0, + "grad_norm": 2.235650398807577, + "learning_rate": 3.687011935388027e-07, + "loss": 0.6788, + "num_input_tokens_seen": 145281875, + "step": 6732 + }, + { + "epoch": 0.8095953826729995, + "flos": 24026668669920.0, + "grad_norm": 3.1535285320984157, + "learning_rate": 3.6825064596452646e-07, + "loss": 0.7331, + "num_input_tokens_seen": 145302220, + "step": 6733 + }, + { + "epoch": 0.8097156255636385, + "flos": 23917149356160.0, + "grad_norm": 2.2342505787172877, + "learning_rate": 3.678003459253305e-07, + "loss": 0.7074, + "num_input_tokens_seen": 145323070, + "step": 6734 + }, + { + "epoch": 0.8098358684542777, + "flos": 21768027248160.0, + "grad_norm": 2.6622447863121255, + "learning_rate": 3.673502934895236e-07, + "loss": 0.7411, + "num_input_tokens_seen": 145342575, + "step": 6735 + }, + { + "epoch": 0.8099561113449167, + "flos": 68711892528960.0, + "grad_norm": 0.6986325113340117, + "learning_rate": 3.669004887253802e-07, + "loss": 0.5778, + "num_input_tokens_seen": 145408865, + "step": 6736 + }, + { + "epoch": 0.8100763542355558, + "flos": 23552307995520.0, + "grad_norm": 2.0759312803401224, + "learning_rate": 3.664509317011335e-07, + "loss": 0.7925, + "num_input_tokens_seen": 145429200, + "step": 6737 + }, + { + "epoch": 0.810196597126195, + "flos": 31604954004480.0, + "grad_norm": 4.801749768811902, + "learning_rate": 3.6600162248498134e-07, + "loss": 0.7404, + "num_input_tokens_seen": 145452830, + "step": 6738 + }, + { + "epoch": 0.810316840016834, + "flos": 24864829951680.0, + "grad_norm": 2.12313087388451, + "learning_rate": 3.6555256114508426e-07, + "loss": 0.7644, + "num_input_tokens_seen": 145472625, + "step": 6739 + }, + { + "epoch": 0.8104370829074731, + "flos": 27924903198720.0, + "grad_norm": 2.887370488949089, + "learning_rate": 3.651037477495642e-07, + "loss": 0.7322, + "num_input_tokens_seen": 145494075, + "step": 6740 + }, + { + "epoch": 0.8105573257981122, + "flos": 24605344891680.0, + "grad_norm": 6.235895645433174, + "learning_rate": 3.6465518236650584e-07, + "loss": 0.6785, + "num_input_tokens_seen": 145514220, + "step": 6741 + }, + { + "epoch": 0.8106775686887513, + "flos": 26321147473920.0, + "grad_norm": 2.451540976163033, + "learning_rate": 3.642068650639558e-07, + "loss": 0.7865, + "num_input_tokens_seen": 145533965, + "step": 6742 + }, + { + "epoch": 0.8107978115793903, + "flos": 27232804838400.0, + "grad_norm": 3.4767132329946313, + "learning_rate": 3.6375879590992334e-07, + "loss": 0.646, + "num_input_tokens_seen": 145554310, + "step": 6743 + }, + { + "epoch": 0.8109180544700295, + "flos": 24898548657600.0, + "grad_norm": 4.709981691685302, + "learning_rate": 3.6331097497238173e-07, + "loss": 0.8109, + "num_input_tokens_seen": 145573755, + "step": 6744 + }, + { + "epoch": 0.8110382973606686, + "flos": 21075259832160.0, + "grad_norm": 3.6034552840476404, + "learning_rate": 3.628634023192627e-07, + "loss": 0.8008, + "num_input_tokens_seen": 145594000, + "step": 6745 + }, + { + "epoch": 0.8111585402513076, + "flos": 15392261349120.0, + "grad_norm": 5.230998028847614, + "learning_rate": 3.624160780184644e-07, + "loss": 0.7539, + "num_input_tokens_seen": 145611215, + "step": 6746 + }, + { + "epoch": 0.8112787831419467, + "flos": 24061205110560.0, + "grad_norm": 2.520731795186501, + "learning_rate": 3.6196900213784496e-07, + "loss": 0.7496, + "num_input_tokens_seen": 145630530, + "step": 6747 + }, + { + "epoch": 0.8113990260325858, + "flos": 20454092841120.0, + "grad_norm": 2.483921993388335, + "learning_rate": 3.6152217474522527e-07, + "loss": 0.8668, + "num_input_tokens_seen": 145647345, + "step": 6748 + }, + { + "epoch": 0.8115192689232249, + "flos": 24865536177120.0, + "grad_norm": 2.0738874555196145, + "learning_rate": 3.6107559590838975e-07, + "loss": 0.7261, + "num_input_tokens_seen": 145666330, + "step": 6749 + }, + { + "epoch": 0.811639511813864, + "flos": 24022654335840.0, + "grad_norm": 5.082019615306628, + "learning_rate": 3.606292656950822e-07, + "loss": 0.6653, + "num_input_tokens_seen": 145684740, + "step": 6750 + }, + { + "epoch": 0.8117597547045031, + "flos": 23152000950240.0, + "grad_norm": 2.470160455627162, + "learning_rate": 3.601831841730121e-07, + "loss": 0.863, + "num_input_tokens_seen": 145702660, + "step": 6751 + }, + { + "epoch": 0.8118799975951422, + "flos": 23007350479680.0, + "grad_norm": 2.118665678883909, + "learning_rate": 3.5973735140984916e-07, + "loss": 0.7294, + "num_input_tokens_seen": 145722830, + "step": 6752 + }, + { + "epoch": 0.8120002404857812, + "flos": 24604527156960.0, + "grad_norm": 6.7055699440037, + "learning_rate": 3.5929176747322607e-07, + "loss": 0.7961, + "num_input_tokens_seen": 145741935, + "step": 6753 + }, + { + "epoch": 0.8121204833764204, + "flos": 57334335767520.0, + "grad_norm": 0.8306984053669351, + "learning_rate": 3.588464324307372e-07, + "loss": 0.5601, + "num_input_tokens_seen": 145805560, + "step": 6754 + }, + { + "epoch": 0.8122407262670595, + "flos": 19436893327200.0, + "grad_norm": 3.477946145393685, + "learning_rate": 3.584013463499391e-07, + "loss": 0.7538, + "num_input_tokens_seen": 145824180, + "step": 6755 + }, + { + "epoch": 0.8123609691576985, + "flos": 56345167986240.0, + "grad_norm": 0.7457547840830712, + "learning_rate": 3.579565092983521e-07, + "loss": 0.6432, + "num_input_tokens_seen": 145885300, + "step": 6756 + }, + { + "epoch": 0.8124812120483377, + "flos": 20602571796960.0, + "grad_norm": 2.643377460898718, + "learning_rate": 3.575119213434565e-07, + "loss": 0.8387, + "num_input_tokens_seen": 145903815, + "step": 6757 + }, + { + "epoch": 0.8126014549389767, + "flos": 22460274287520.0, + "grad_norm": 2.2147974618316764, + "learning_rate": 3.5706758255269765e-07, + "loss": 0.8181, + "num_input_tokens_seen": 145922270, + "step": 6758 + }, + { + "epoch": 0.8127216978296158, + "flos": 23254420839840.0, + "grad_norm": 3.7278738213825604, + "learning_rate": 3.566234929934795e-07, + "loss": 0.6989, + "num_input_tokens_seen": 145941470, + "step": 6759 + }, + { + "epoch": 0.812841940720255, + "flos": 25118962566240.0, + "grad_norm": 1.9853989970277213, + "learning_rate": 3.561796527331706e-07, + "loss": 0.719, + "num_input_tokens_seen": 145963145, + "step": 6760 + }, + { + "epoch": 0.812962183610894, + "flos": 26610151056960.0, + "grad_norm": 2.5494104414428818, + "learning_rate": 3.5573606183910163e-07, + "loss": 0.7734, + "num_input_tokens_seen": 145983150, + "step": 6761 + }, + { + "epoch": 0.8130824265015331, + "flos": 24931003591680.0, + "grad_norm": 2.4474353289997737, + "learning_rate": 3.5529272037856493e-07, + "loss": 0.7848, + "num_input_tokens_seen": 146001365, + "step": 6762 + }, + { + "epoch": 0.8132026693921722, + "flos": 67526328237600.0, + "grad_norm": 0.7173812105953902, + "learning_rate": 3.548496284188149e-07, + "loss": 0.5367, + "num_input_tokens_seen": 146061570, + "step": 6763 + }, + { + "epoch": 0.8133229122828113, + "flos": 19467192415200.0, + "grad_norm": 2.8554390254737085, + "learning_rate": 3.544067860270681e-07, + "loss": 0.791, + "num_input_tokens_seen": 146079145, + "step": 6764 + }, + { + "epoch": 0.8134431551734503, + "flos": 20638669367520.0, + "grad_norm": 2.313761805046216, + "learning_rate": 3.539641932705029e-07, + "loss": 0.7116, + "num_input_tokens_seen": 146097625, + "step": 6765 + }, + { + "epoch": 0.8135633980640895, + "flos": 21476644800480.0, + "grad_norm": 4.093132613124457, + "learning_rate": 3.53521850216262e-07, + "loss": 0.7734, + "num_input_tokens_seen": 146117195, + "step": 6766 + }, + { + "epoch": 0.8136836409547286, + "flos": 20524987040640.0, + "grad_norm": 2.3616100108319578, + "learning_rate": 3.530797569314461e-07, + "loss": 0.7693, + "num_input_tokens_seen": 146136220, + "step": 6767 + }, + { + "epoch": 0.8138038838453676, + "flos": 20269962351840.0, + "grad_norm": 2.7356053663267805, + "learning_rate": 3.5263791348312235e-07, + "loss": 0.78, + "num_input_tokens_seen": 146155415, + "step": 6768 + }, + { + "epoch": 0.8139241267360068, + "flos": 29747102834880.0, + "grad_norm": 2.642329100655684, + "learning_rate": 3.521963199383171e-07, + "loss": 0.7069, + "num_input_tokens_seen": 146178370, + "step": 6769 + }, + { + "epoch": 0.8140443696266458, + "flos": 19684967270880.0, + "grad_norm": 2.855003308563186, + "learning_rate": 3.517549763640197e-07, + "loss": 0.7722, + "num_input_tokens_seen": 146196010, + "step": 6770 + }, + { + "epoch": 0.8141646125172849, + "flos": 27121575715680.0, + "grad_norm": 2.5990496806041503, + "learning_rate": 3.513138828271829e-07, + "loss": 0.7167, + "num_input_tokens_seen": 146215880, + "step": 6771 + }, + { + "epoch": 0.8142848554079241, + "flos": 39615815469600.0, + "grad_norm": 2.685373929090339, + "learning_rate": 3.508730393947179e-07, + "loss": 0.702, + "num_input_tokens_seen": 146239135, + "step": 6772 + }, + { + "epoch": 0.8144050982985631, + "flos": 22205546956800.0, + "grad_norm": 2.227758574885271, + "learning_rate": 3.504324461335024e-07, + "loss": 0.721, + "num_input_tokens_seen": 146259245, + "step": 6773 + }, + { + "epoch": 0.8145253411892022, + "flos": 23005268973120.0, + "grad_norm": 2.220038563676912, + "learning_rate": 3.499921031103732e-07, + "loss": 0.8798, + "num_input_tokens_seen": 146280015, + "step": 6774 + }, + { + "epoch": 0.8146455840798413, + "flos": 24792151603680.0, + "grad_norm": 46.09619961711188, + "learning_rate": 3.4955201039212987e-07, + "loss": 0.7857, + "num_input_tokens_seen": 146300935, + "step": 6775 + }, + { + "epoch": 0.8147658269704804, + "flos": 19946050630560.0, + "grad_norm": 2.691271420051177, + "learning_rate": 3.4911216804553465e-07, + "loss": 0.6585, + "num_input_tokens_seen": 146319835, + "step": 6776 + }, + { + "epoch": 0.8148860698611194, + "flos": 21147789501120.0, + "grad_norm": 2.667712264906491, + "learning_rate": 3.4867257613731017e-07, + "loss": 0.7071, + "num_input_tokens_seen": 146339030, + "step": 6777 + }, + { + "epoch": 0.8150063127517585, + "flos": 19578867575040.0, + "grad_norm": 2.0281260015782028, + "learning_rate": 3.4823323473414343e-07, + "loss": 0.8544, + "num_input_tokens_seen": 146358550, + "step": 6778 + }, + { + "epoch": 0.8151265556423977, + "flos": 22606039850880.0, + "grad_norm": 4.124849258905276, + "learning_rate": 3.477941439026812e-07, + "loss": 0.762, + "num_input_tokens_seen": 146376770, + "step": 6779 + }, + { + "epoch": 0.8152467985330367, + "flos": 17942508237120.0, + "grad_norm": 2.230354159162389, + "learning_rate": 3.473553037095349e-07, + "loss": 0.7354, + "num_input_tokens_seen": 146395795, + "step": 6780 + }, + { + "epoch": 0.8153670414236758, + "flos": 24933122268000.0, + "grad_norm": 2.5050242482608827, + "learning_rate": 3.469167142212743e-07, + "loss": 0.8324, + "num_input_tokens_seen": 146417640, + "step": 6781 + }, + { + "epoch": 0.8154872843143149, + "flos": 31021445713920.0, + "grad_norm": 3.353982971721408, + "learning_rate": 3.4647837550443337e-07, + "loss": 0.6285, + "num_input_tokens_seen": 146436205, + "step": 6782 + }, + { + "epoch": 0.815607527204954, + "flos": 19364066300160.0, + "grad_norm": 6.075033230160852, + "learning_rate": 3.460402876255086e-07, + "loss": 0.7497, + "num_input_tokens_seen": 146453595, + "step": 6783 + }, + { + "epoch": 0.815727770095593, + "flos": 26102963750880.0, + "grad_norm": 2.478042304834762, + "learning_rate": 3.456024506509574e-07, + "loss": 0.7176, + "num_input_tokens_seen": 146474515, + "step": 6784 + }, + { + "epoch": 0.8158480129862322, + "flos": 25301234567520.0, + "grad_norm": 2.7503052455380352, + "learning_rate": 3.4516486464719873e-07, + "loss": 0.7431, + "num_input_tokens_seen": 146493905, + "step": 6785 + }, + { + "epoch": 0.8159682558768713, + "flos": 34374165180480.0, + "grad_norm": 2.51475140110807, + "learning_rate": 3.4472752968061445e-07, + "loss": 0.6188, + "num_input_tokens_seen": 146518335, + "step": 6786 + }, + { + "epoch": 0.8160884987675103, + "flos": 18626392080480.0, + "grad_norm": 4.15095078940412, + "learning_rate": 3.442904458175475e-07, + "loss": 0.7393, + "num_input_tokens_seen": 146535655, + "step": 6787 + }, + { + "epoch": 0.8162087416581495, + "flos": 31385357830560.0, + "grad_norm": 1.9109052504081927, + "learning_rate": 3.438536131243044e-07, + "loss": 0.7596, + "num_input_tokens_seen": 146556815, + "step": 6788 + }, + { + "epoch": 0.8163289845487885, + "flos": 37539669067680.0, + "grad_norm": 2.383972986010656, + "learning_rate": 3.434170316671503e-07, + "loss": 0.6242, + "num_input_tokens_seen": 146581845, + "step": 6789 + }, + { + "epoch": 0.8164492274394276, + "flos": 13934271187680.0, + "grad_norm": 3.073778365703566, + "learning_rate": 3.4298070151231583e-07, + "loss": 0.9, + "num_input_tokens_seen": 146597245, + "step": 6790 + }, + { + "epoch": 0.8165694703300668, + "flos": 28944964784160.0, + "grad_norm": 2.909755930152748, + "learning_rate": 3.425446227259916e-07, + "loss": 0.6014, + "num_input_tokens_seen": 146618210, + "step": 6791 + }, + { + "epoch": 0.8166897132207058, + "flos": 25082381788800.0, + "grad_norm": 2.283607832101423, + "learning_rate": 3.421087953743296e-07, + "loss": 0.822, + "num_input_tokens_seen": 146637975, + "step": 6792 + }, + { + "epoch": 0.8168099561113449, + "flos": 23115494512320.0, + "grad_norm": 27.345525611991693, + "learning_rate": 3.416732195234464e-07, + "loss": 0.7963, + "num_input_tokens_seen": 146658030, + "step": 6793 + }, + { + "epoch": 0.816930199001984, + "flos": 18381551905920.0, + "grad_norm": 2.0269662141698936, + "learning_rate": 3.4123789523941613e-07, + "loss": 0.7961, + "num_input_tokens_seen": 146677855, + "step": 6794 + }, + { + "epoch": 0.8170504418926231, + "flos": 21220988225760.0, + "grad_norm": 1.6945317311899601, + "learning_rate": 3.4080282258827884e-07, + "loss": 0.633, + "num_input_tokens_seen": 146696230, + "step": 6795 + }, + { + "epoch": 0.8171706847832622, + "flos": 19072051966560.0, + "grad_norm": 2.786463619239048, + "learning_rate": 3.403680016360342e-07, + "loss": 0.7214, + "num_input_tokens_seen": 146714835, + "step": 6796 + }, + { + "epoch": 0.8172909276739013, + "flos": 21439469306880.0, + "grad_norm": 2.1287333286996746, + "learning_rate": 3.3993343244864403e-07, + "loss": 0.6785, + "num_input_tokens_seen": 146734425, + "step": 6797 + }, + { + "epoch": 0.8174111705645404, + "flos": 27560210517120.0, + "grad_norm": 2.397325288932353, + "learning_rate": 3.394991150920323e-07, + "loss": 0.7298, + "num_input_tokens_seen": 146757545, + "step": 6798 + }, + { + "epoch": 0.8175314134551794, + "flos": 14044756915200.0, + "grad_norm": 2.6331311494333067, + "learning_rate": 3.3906504963208396e-07, + "loss": 0.7418, + "num_input_tokens_seen": 146774240, + "step": 6799 + }, + { + "epoch": 0.8176516563458186, + "flos": 22675521599520.0, + "grad_norm": 2.3514779414692173, + "learning_rate": 3.3863123613464774e-07, + "loss": 0.6673, + "num_input_tokens_seen": 146793140, + "step": 6800 + }, + { + "epoch": 0.8177718992364577, + "flos": 21914090169600.0, + "grad_norm": 2.3399721342856625, + "learning_rate": 3.381976746655317e-07, + "loss": 0.7528, + "num_input_tokens_seen": 146812895, + "step": 6801 + }, + { + "epoch": 0.8178921421270967, + "flos": 21985987952640.0, + "grad_norm": 2.7864476188857723, + "learning_rate": 3.3776436529050756e-07, + "loss": 0.6716, + "num_input_tokens_seen": 146832445, + "step": 6802 + }, + { + "epoch": 0.8180123850177359, + "flos": 33135436665120.0, + "grad_norm": 2.39583951424064, + "learning_rate": 3.373313080753073e-07, + "loss": 0.7254, + "num_input_tokens_seen": 146856505, + "step": 6803 + }, + { + "epoch": 0.8181326279083749, + "flos": 22059632714400.0, + "grad_norm": 1.9176992316360713, + "learning_rate": 3.3689850308562527e-07, + "loss": 0.7781, + "num_input_tokens_seen": 146876950, + "step": 6804 + }, + { + "epoch": 0.818252870799014, + "flos": 15683495117760.0, + "grad_norm": 1.9959722732040066, + "learning_rate": 3.364659503871183e-07, + "loss": 0.7767, + "num_input_tokens_seen": 146894885, + "step": 6805 + }, + { + "epoch": 0.8183731136896532, + "flos": 18743977232160.0, + "grad_norm": 2.4743652223305683, + "learning_rate": 3.3603365004540417e-07, + "loss": 0.8398, + "num_input_tokens_seen": 146913180, + "step": 6806 + }, + { + "epoch": 0.8184933565802922, + "flos": 26503754003040.0, + "grad_norm": 3.380366262421165, + "learning_rate": 3.356016021260624e-07, + "loss": 0.77, + "num_input_tokens_seen": 146933510, + "step": 6807 + }, + { + "epoch": 0.8186135994709313, + "flos": 17505211547040.0, + "grad_norm": 5.166468811348588, + "learning_rate": 3.35169806694634e-07, + "loss": 0.6554, + "num_input_tokens_seen": 146951760, + "step": 6808 + }, + { + "epoch": 0.8187338423615703, + "flos": 63390352958400.0, + "grad_norm": 0.7381435043767498, + "learning_rate": 3.3473826381662186e-07, + "loss": 0.6055, + "num_input_tokens_seen": 147022900, + "step": 6809 + }, + { + "epoch": 0.8188540852522095, + "flos": 17505062868000.0, + "grad_norm": 2.5251037954015416, + "learning_rate": 3.3430697355749216e-07, + "loss": 0.8156, + "num_input_tokens_seen": 147040590, + "step": 6810 + }, + { + "epoch": 0.8189743281428485, + "flos": 14372199763680.0, + "grad_norm": 2.770008298950984, + "learning_rate": 3.3387593598266907e-07, + "loss": 0.7561, + "num_input_tokens_seen": 147057200, + "step": 6811 + }, + { + "epoch": 0.8190945710334876, + "flos": 25044500069760.0, + "grad_norm": 2.2342703201208227, + "learning_rate": 3.3344515115754225e-07, + "loss": 0.7834, + "num_input_tokens_seen": 147076180, + "step": 6812 + }, + { + "epoch": 0.8192148139241268, + "flos": 21476458951680.0, + "grad_norm": 20.999977475108828, + "learning_rate": 3.33014619147461e-07, + "loss": 0.7999, + "num_input_tokens_seen": 147094205, + "step": 6813 + }, + { + "epoch": 0.8193350568147658, + "flos": 23917929921120.0, + "grad_norm": 2.652021295781312, + "learning_rate": 3.325843400177362e-07, + "loss": 0.7162, + "num_input_tokens_seen": 147115545, + "step": 6814 + }, + { + "epoch": 0.8194552997054049, + "flos": 20530190807040.0, + "grad_norm": 3.934643148021242, + "learning_rate": 3.32154313833642e-07, + "loss": 0.738, + "num_input_tokens_seen": 147135700, + "step": 6815 + }, + { + "epoch": 0.819575542596044, + "flos": 25994485190400.0, + "grad_norm": 2.8875412446746567, + "learning_rate": 3.3172454066041164e-07, + "loss": 0.5993, + "num_input_tokens_seen": 147155205, + "step": 6816 + }, + { + "epoch": 0.8196957854866831, + "flos": 29055227493120.0, + "grad_norm": 4.133070016492639, + "learning_rate": 3.3129502056324234e-07, + "loss": 0.758, + "num_input_tokens_seen": 147176570, + "step": 6817 + }, + { + "epoch": 0.8198160283773221, + "flos": 68934542356800.0, + "grad_norm": 0.8054459710142282, + "learning_rate": 3.3086575360729165e-07, + "loss": 0.5966, + "num_input_tokens_seen": 147234105, + "step": 6818 + }, + { + "epoch": 0.8199362712679613, + "flos": 16594483426560.0, + "grad_norm": 2.673720051801473, + "learning_rate": 3.3043673985767906e-07, + "loss": 0.7117, + "num_input_tokens_seen": 147253170, + "step": 6819 + }, + { + "epoch": 0.8200565141586004, + "flos": 21726428553120.0, + "grad_norm": 2.606497634356823, + "learning_rate": 3.3000797937948564e-07, + "loss": 0.7732, + "num_input_tokens_seen": 147271465, + "step": 6820 + }, + { + "epoch": 0.8201767570492394, + "flos": 69707161884480.0, + "grad_norm": 0.9445600115591142, + "learning_rate": 3.295794722377534e-07, + "loss": 0.6485, + "num_input_tokens_seen": 147335070, + "step": 6821 + }, + { + "epoch": 0.8202969999398786, + "flos": 23078839395360.0, + "grad_norm": 8.992239679339269, + "learning_rate": 3.291512184974876e-07, + "loss": 0.7991, + "num_input_tokens_seen": 147355460, + "step": 6822 + }, + { + "epoch": 0.8204172428305176, + "flos": 28180150906080.0, + "grad_norm": 2.3533294224188768, + "learning_rate": 3.2872321822365346e-07, + "loss": 0.6631, + "num_input_tokens_seen": 147375675, + "step": 6823 + }, + { + "epoch": 0.8205374857211567, + "flos": 20857224788160.0, + "grad_norm": 9.886998094631261, + "learning_rate": 3.282954714811783e-07, + "loss": 0.7349, + "num_input_tokens_seen": 147394580, + "step": 6824 + }, + { + "epoch": 0.8206577286117959, + "flos": 13133619927360.0, + "grad_norm": 2.8527635527143347, + "learning_rate": 3.2786797833495093e-07, + "loss": 0.7107, + "num_input_tokens_seen": 147409935, + "step": 6825 + }, + { + "epoch": 0.8207779715024349, + "flos": 25229485463520.0, + "grad_norm": 3.457557560571909, + "learning_rate": 3.274407388498213e-07, + "loss": 0.7288, + "num_input_tokens_seen": 147428065, + "step": 6826 + }, + { + "epoch": 0.820898214393074, + "flos": 19582324362720.0, + "grad_norm": 2.3014819595322686, + "learning_rate": 3.270137530906021e-07, + "loss": 0.7444, + "num_input_tokens_seen": 147447300, + "step": 6827 + }, + { + "epoch": 0.8210184572837131, + "flos": 15574533350400.0, + "grad_norm": 2.4244862907023648, + "learning_rate": 3.265870211220665e-07, + "loss": 0.8348, + "num_input_tokens_seen": 147465135, + "step": 6828 + }, + { + "epoch": 0.8211387001743522, + "flos": 20784249082080.0, + "grad_norm": 3.3659988748205145, + "learning_rate": 3.2616054300894934e-07, + "loss": 0.8177, + "num_input_tokens_seen": 147484585, + "step": 6829 + }, + { + "epoch": 0.8212589430649913, + "flos": 27664674743520.0, + "grad_norm": 2.9663347558645667, + "learning_rate": 3.2573431881594693e-07, + "loss": 0.8457, + "num_input_tokens_seen": 147504130, + "step": 6830 + }, + { + "epoch": 0.8213791859556304, + "flos": 22423544831040.0, + "grad_norm": 2.839186475584097, + "learning_rate": 3.2530834860771663e-07, + "loss": 0.6591, + "num_input_tokens_seen": 147523510, + "step": 6831 + }, + { + "epoch": 0.8214994288462695, + "flos": 16630841185440.0, + "grad_norm": 3.838890663682669, + "learning_rate": 3.248826324488794e-07, + "loss": 0.744, + "num_input_tokens_seen": 147540805, + "step": 6832 + }, + { + "epoch": 0.8216196717369085, + "flos": 25185433564320.0, + "grad_norm": 4.523749306165648, + "learning_rate": 3.244571704040138e-07, + "loss": 0.8779, + "num_input_tokens_seen": 147560965, + "step": 6833 + }, + { + "epoch": 0.8217399146275477, + "flos": 25335510819840.0, + "grad_norm": 3.7031123808955613, + "learning_rate": 3.2403196253766374e-07, + "loss": 0.7345, + "num_input_tokens_seen": 147580045, + "step": 6834 + }, + { + "epoch": 0.8218601575181868, + "flos": 25592914373280.0, + "grad_norm": 2.6591483067717805, + "learning_rate": 3.2360700891433254e-07, + "loss": 0.7932, + "num_input_tokens_seen": 147599070, + "step": 6835 + }, + { + "epoch": 0.8219804004088258, + "flos": 67563615240480.0, + "grad_norm": 0.8038445659435034, + "learning_rate": 3.231823095984847e-07, + "loss": 0.572, + "num_input_tokens_seen": 147653710, + "step": 6836 + }, + { + "epoch": 0.822100643299465, + "flos": 19436558799360.0, + "grad_norm": 2.8884349878042905, + "learning_rate": 3.2275786465454814e-07, + "loss": 0.7601, + "num_input_tokens_seen": 147670360, + "step": 6837 + }, + { + "epoch": 0.822220886190104, + "flos": 24646051512480.0, + "grad_norm": 2.4169460222863886, + "learning_rate": 3.2233367414690917e-07, + "loss": 0.7596, + "num_input_tokens_seen": 147692550, + "step": 6838 + }, + { + "epoch": 0.8223411290807431, + "flos": 27779174805120.0, + "grad_norm": 2.6153169694411753, + "learning_rate": 3.219097381399183e-07, + "loss": 0.8563, + "num_input_tokens_seen": 147709725, + "step": 6839 + }, + { + "epoch": 0.8224613719713821, + "flos": 23185236449280.0, + "grad_norm": 3.1562046765387506, + "learning_rate": 3.2148605669788584e-07, + "loss": 0.8107, + "num_input_tokens_seen": 147729615, + "step": 6840 + }, + { + "epoch": 0.8225816148620213, + "flos": 15684015494400.0, + "grad_norm": 3.2163841663757906, + "learning_rate": 3.2106262988508405e-07, + "loss": 0.7734, + "num_input_tokens_seen": 147747665, + "step": 6841 + }, + { + "epoch": 0.8227018577526604, + "flos": 18488766694560.0, + "grad_norm": 2.4390119118487585, + "learning_rate": 3.206394577657465e-07, + "loss": 0.7452, + "num_input_tokens_seen": 147765445, + "step": 6842 + }, + { + "epoch": 0.8228221006432994, + "flos": 22205026580160.0, + "grad_norm": 3.3589147917878495, + "learning_rate": 3.202165404040675e-07, + "loss": 0.7252, + "num_input_tokens_seen": 147783365, + "step": 6843 + }, + { + "epoch": 0.8229423435339386, + "flos": 24063063598560.0, + "grad_norm": 2.484260892346836, + "learning_rate": 3.1979387786420396e-07, + "loss": 0.7479, + "num_input_tokens_seen": 147803605, + "step": 6844 + }, + { + "epoch": 0.8230625864245776, + "flos": 23844396668640.0, + "grad_norm": 2.685782748874453, + "learning_rate": 3.1937147021027346e-07, + "loss": 0.8216, + "num_input_tokens_seen": 147822060, + "step": 6845 + }, + { + "epoch": 0.8231828293152167, + "flos": 16557568121280.0, + "grad_norm": 3.67977600639147, + "learning_rate": 3.189493175063547e-07, + "loss": 0.7664, + "num_input_tokens_seen": 147839295, + "step": 6846 + }, + { + "epoch": 0.8233030722058559, + "flos": 18853199187840.0, + "grad_norm": 2.237359035036656, + "learning_rate": 3.1852741981648776e-07, + "loss": 0.6726, + "num_input_tokens_seen": 147857945, + "step": 6847 + }, + { + "epoch": 0.8234233150964949, + "flos": 28398223119840.0, + "grad_norm": 2.630000755251932, + "learning_rate": 3.1810577720467404e-07, + "loss": 0.7006, + "num_input_tokens_seen": 147879675, + "step": 6848 + }, + { + "epoch": 0.823543557987134, + "flos": 33717383825760.0, + "grad_norm": 2.31026567310259, + "learning_rate": 3.176843897348769e-07, + "loss": 0.5697, + "num_input_tokens_seen": 147902870, + "step": 6849 + }, + { + "epoch": 0.8236638008777731, + "flos": 17068100705760.0, + "grad_norm": 3.0405285591599664, + "learning_rate": 3.1726325747102034e-07, + "loss": 0.7586, + "num_input_tokens_seen": 147921245, + "step": 6850 + }, + { + "epoch": 0.8237840437684122, + "flos": 61550800777920.0, + "grad_norm": 1.8983026778459011, + "learning_rate": 3.1684238047698974e-07, + "loss": 0.6422, + "num_input_tokens_seen": 147949515, + "step": 6851 + }, + { + "epoch": 0.8239042866590512, + "flos": 27269905992480.0, + "grad_norm": 3.338606827328632, + "learning_rate": 3.1642175881663155e-07, + "loss": 0.5326, + "num_input_tokens_seen": 147969245, + "step": 6852 + }, + { + "epoch": 0.8240245295496904, + "flos": 21695348900160.0, + "grad_norm": 3.0949171542664047, + "learning_rate": 3.160013925537537e-07, + "loss": 0.8377, + "num_input_tokens_seen": 147990080, + "step": 6853 + }, + { + "epoch": 0.8241447724403295, + "flos": 19980735750240.0, + "grad_norm": 2.5196539797203807, + "learning_rate": 3.155812817521266e-07, + "loss": 0.7592, + "num_input_tokens_seen": 148010455, + "step": 6854 + }, + { + "epoch": 0.8242650153309685, + "flos": 22241161320480.0, + "grad_norm": 2.776398102074631, + "learning_rate": 3.151614264754787e-07, + "loss": 0.7802, + "num_input_tokens_seen": 148028400, + "step": 6855 + }, + { + "epoch": 0.8243852582216077, + "flos": 22278113795520.0, + "grad_norm": 6.943075053514146, + "learning_rate": 3.147418267875035e-07, + "loss": 0.7986, + "num_input_tokens_seen": 148046530, + "step": 6856 + }, + { + "epoch": 0.8245055011122467, + "flos": 24610176960480.0, + "grad_norm": 2.896435809869744, + "learning_rate": 3.1432248275185315e-07, + "loss": 0.6629, + "num_input_tokens_seen": 148067040, + "step": 6857 + }, + { + "epoch": 0.8246257440028858, + "flos": 17462200401120.0, + "grad_norm": 4.209908722456583, + "learning_rate": 3.139033944321412e-07, + "loss": 0.7735, + "num_input_tokens_seen": 148084400, + "step": 6858 + }, + { + "epoch": 0.824745986893525, + "flos": 24974349265440.0, + "grad_norm": 2.6498008077360664, + "learning_rate": 3.1348456189194507e-07, + "loss": 0.7885, + "num_input_tokens_seen": 148104860, + "step": 6859 + }, + { + "epoch": 0.824866229784164, + "flos": 18746058738720.0, + "grad_norm": 2.7962013157855568, + "learning_rate": 3.1306598519479876e-07, + "loss": 0.8337, + "num_input_tokens_seen": 148124680, + "step": 6860 + }, + { + "epoch": 0.8249864726748031, + "flos": 23808076079520.0, + "grad_norm": 3.962342142998092, + "learning_rate": 3.1264766440420177e-07, + "loss": 0.7837, + "num_input_tokens_seen": 148147150, + "step": 6861 + }, + { + "epoch": 0.8251067155654422, + "flos": 20274162534720.0, + "grad_norm": 2.8465915777530624, + "learning_rate": 3.122295995836124e-07, + "loss": 0.6947, + "num_input_tokens_seen": 148167020, + "step": 6862 + }, + { + "epoch": 0.8252269584560813, + "flos": 24973903228320.0, + "grad_norm": 2.4986879720617527, + "learning_rate": 3.118117907964508e-07, + "loss": 0.7789, + "num_input_tokens_seen": 148188965, + "step": 6863 + }, + { + "epoch": 0.8253472013467203, + "flos": 17104235446080.0, + "grad_norm": 2.3454543973628312, + "learning_rate": 3.1139423810609856e-07, + "loss": 0.8059, + "num_input_tokens_seen": 148205660, + "step": 6864 + }, + { + "epoch": 0.8254674442373595, + "flos": 22382615191680.0, + "grad_norm": 3.4065106493415147, + "learning_rate": 3.1097694157589714e-07, + "loss": 0.7572, + "num_input_tokens_seen": 148225545, + "step": 6865 + }, + { + "epoch": 0.8255876871279986, + "flos": 24751147624800.0, + "grad_norm": 5.630296375340282, + "learning_rate": 3.105599012691511e-07, + "loss": 0.7622, + "num_input_tokens_seen": 148243975, + "step": 6866 + }, + { + "epoch": 0.8257079300186376, + "flos": 27416117592960.0, + "grad_norm": 1.8951890570308938, + "learning_rate": 3.101431172491249e-07, + "loss": 0.8251, + "num_input_tokens_seen": 148265520, + "step": 6867 + }, + { + "epoch": 0.8258281729092768, + "flos": 16448494844640.0, + "grad_norm": 5.317821933630145, + "learning_rate": 3.097265895790444e-07, + "loss": 0.7187, + "num_input_tokens_seen": 148283730, + "step": 6868 + }, + { + "epoch": 0.8259484157999158, + "flos": 21403223057280.0, + "grad_norm": 3.770889843821137, + "learning_rate": 3.093103183220962e-07, + "loss": 0.8396, + "num_input_tokens_seen": 148303775, + "step": 6869 + }, + { + "epoch": 0.8260686586905549, + "flos": 58239414084480.0, + "grad_norm": 0.8228399667527216, + "learning_rate": 3.0889430354142796e-07, + "loss": 0.5915, + "num_input_tokens_seen": 148365285, + "step": 6870 + }, + { + "epoch": 0.826188901581194, + "flos": 27487829527200.0, + "grad_norm": 4.944061940765011, + "learning_rate": 3.084785453001497e-07, + "loss": 0.6986, + "num_input_tokens_seen": 148386200, + "step": 6871 + }, + { + "epoch": 0.8263091444718331, + "flos": 23662161837120.0, + "grad_norm": 3.0043481469506723, + "learning_rate": 3.080630436613314e-07, + "loss": 0.8224, + "num_input_tokens_seen": 148403970, + "step": 6872 + }, + { + "epoch": 0.8264293873624722, + "flos": 17140779053760.0, + "grad_norm": 3.309077528136692, + "learning_rate": 3.076477986880039e-07, + "loss": 0.8606, + "num_input_tokens_seen": 148421765, + "step": 6873 + }, + { + "epoch": 0.8265496302531112, + "flos": 24064215861120.0, + "grad_norm": 2.951431677015715, + "learning_rate": 3.0723281044315986e-07, + "loss": 0.6974, + "num_input_tokens_seen": 148443720, + "step": 6874 + }, + { + "epoch": 0.8266698731437504, + "flos": 14079962411520.0, + "grad_norm": 2.699212037007752, + "learning_rate": 3.068180789897521e-07, + "loss": 0.7638, + "num_input_tokens_seen": 148462130, + "step": 6875 + }, + { + "epoch": 0.8267901160343895, + "flos": 30733482884160.0, + "grad_norm": 1.776317258094344, + "learning_rate": 3.064036043906966e-07, + "loss": 0.8165, + "num_input_tokens_seen": 148485570, + "step": 6876 + }, + { + "epoch": 0.8269103589250285, + "flos": 40566358136640.0, + "grad_norm": 3.1251028425460268, + "learning_rate": 3.059893867088668e-07, + "loss": 0.682, + "num_input_tokens_seen": 148509715, + "step": 6877 + }, + { + "epoch": 0.8270306018156677, + "flos": 30220199737440.0, + "grad_norm": 2.6169110052466635, + "learning_rate": 3.055754260071004e-07, + "loss": 0.6711, + "num_input_tokens_seen": 148532010, + "step": 6878 + }, + { + "epoch": 0.8271508447063067, + "flos": 25190414312160.0, + "grad_norm": 3.3785458058290345, + "learning_rate": 3.051617223481948e-07, + "loss": 0.7378, + "num_input_tokens_seen": 148553330, + "step": 6879 + }, + { + "epoch": 0.8272710875969458, + "flos": 17541643645440.0, + "grad_norm": 2.4707992354083195, + "learning_rate": 3.047482757949078e-07, + "loss": 0.7505, + "num_input_tokens_seen": 148570960, + "step": 6880 + }, + { + "epoch": 0.827391330487585, + "flos": 19727309361120.0, + "grad_norm": 5.908371317337389, + "learning_rate": 3.043350864099605e-07, + "loss": 0.8592, + "num_input_tokens_seen": 148589520, + "step": 6881 + }, + { + "epoch": 0.827511573378224, + "flos": 16811217528960.0, + "grad_norm": 4.36492391758143, + "learning_rate": 3.039221542560315e-07, + "loss": 0.8082, + "num_input_tokens_seen": 148606085, + "step": 6882 + }, + { + "epoch": 0.8276318162688631, + "flos": 18343112640480.0, + "grad_norm": 2.1169077558210403, + "learning_rate": 3.0350947939576356e-07, + "loss": 0.7354, + "num_input_tokens_seen": 148625240, + "step": 6883 + }, + { + "epoch": 0.8277520591595022, + "flos": 19325069488320.0, + "grad_norm": 2.097618855688281, + "learning_rate": 3.0309706189175876e-07, + "loss": 0.7255, + "num_input_tokens_seen": 148645075, + "step": 6884 + }, + { + "epoch": 0.8278723020501413, + "flos": 67821985207680.0, + "grad_norm": 0.7757437644115333, + "learning_rate": 3.0268490180658045e-07, + "loss": 0.5729, + "num_input_tokens_seen": 148707855, + "step": 6885 + }, + { + "epoch": 0.8279925449407803, + "flos": 18159837055680.0, + "grad_norm": 2.8422473668383117, + "learning_rate": 3.0227299920275305e-07, + "loss": 0.7918, + "num_input_tokens_seen": 148724170, + "step": 6886 + }, + { + "epoch": 0.8281127878314195, + "flos": 20602014250560.0, + "grad_norm": 2.9757255118336006, + "learning_rate": 3.018613541427613e-07, + "loss": 0.8577, + "num_input_tokens_seen": 148743690, + "step": 6887 + }, + { + "epoch": 0.8282330307220586, + "flos": 17978680147200.0, + "grad_norm": 3.436030125638015, + "learning_rate": 3.0144996668905243e-07, + "loss": 0.7369, + "num_input_tokens_seen": 148761070, + "step": 6888 + }, + { + "epoch": 0.8283532736126976, + "flos": 20055495604800.0, + "grad_norm": 2.42302100847168, + "learning_rate": 3.010388369040331e-07, + "loss": 0.8237, + "num_input_tokens_seen": 148779730, + "step": 6889 + }, + { + "epoch": 0.8284735165033368, + "flos": 31822617350880.0, + "grad_norm": 14.372202194613447, + "learning_rate": 3.0062796485007156e-07, + "loss": 0.8291, + "num_input_tokens_seen": 148800670, + "step": 6890 + }, + { + "epoch": 0.8285937593939758, + "flos": 26613124637760.0, + "grad_norm": 3.8846369856780756, + "learning_rate": 3.002173505894965e-07, + "loss": 0.6541, + "num_input_tokens_seen": 148819820, + "step": 6891 + }, + { + "epoch": 0.8287140022846149, + "flos": 20164903409280.0, + "grad_norm": 4.4607763044157585, + "learning_rate": 2.998069941845973e-07, + "loss": 0.6315, + "num_input_tokens_seen": 148838890, + "step": 6892 + }, + { + "epoch": 0.8288342451752541, + "flos": 70654656631200.0, + "grad_norm": 0.7204183620685185, + "learning_rate": 2.993968956976258e-07, + "loss": 0.5738, + "num_input_tokens_seen": 148906635, + "step": 6893 + }, + { + "epoch": 0.8289544880658931, + "flos": 24537089745120.0, + "grad_norm": 2.2614110233296305, + "learning_rate": 2.9898705519079313e-07, + "loss": 0.7057, + "num_input_tokens_seen": 148925740, + "step": 6894 + }, + { + "epoch": 0.8290747309565322, + "flos": 22241570187840.0, + "grad_norm": 2.286927878119566, + "learning_rate": 2.985774727262715e-07, + "loss": 0.7468, + "num_input_tokens_seen": 148944585, + "step": 6895 + }, + { + "epoch": 0.8291949738471713, + "flos": 23222226094080.0, + "grad_norm": 2.8341679372301387, + "learning_rate": 2.981681483661949e-07, + "loss": 0.8186, + "num_input_tokens_seen": 148964170, + "step": 6896 + }, + { + "epoch": 0.8293152167378104, + "flos": 52479840162240.0, + "grad_norm": 2.093373506772019, + "learning_rate": 2.9775908217265633e-07, + "loss": 0.7129, + "num_input_tokens_seen": 148989405, + "step": 6897 + }, + { + "epoch": 0.8294354596284494, + "flos": 63265854231360.0, + "grad_norm": 0.8374996384577185, + "learning_rate": 2.9735027420771253e-07, + "loss": 0.5016, + "num_input_tokens_seen": 149049740, + "step": 6898 + }, + { + "epoch": 0.8295557025190886, + "flos": 24791482548000.0, + "grad_norm": 1.9981862426956218, + "learning_rate": 2.969417245333774e-07, + "loss": 0.7136, + "num_input_tokens_seen": 149069120, + "step": 6899 + }, + { + "epoch": 0.8296759454097277, + "flos": 25082604807360.0, + "grad_norm": 3.4070176125830596, + "learning_rate": 2.9653343321162915e-07, + "loss": 0.7797, + "num_input_tokens_seen": 149088630, + "step": 6900 + }, + { + "epoch": 0.8297961883003667, + "flos": 24098008906560.0, + "grad_norm": 2.9556325126450833, + "learning_rate": 2.9612540030440446e-07, + "loss": 0.6449, + "num_input_tokens_seen": 149109175, + "step": 6901 + }, + { + "epoch": 0.8299164311910058, + "flos": 67350449435040.0, + "grad_norm": 0.8659267367874888, + "learning_rate": 2.9571762587360206e-07, + "loss": 0.6391, + "num_input_tokens_seen": 149165560, + "step": 6902 + }, + { + "epoch": 0.8300366740816449, + "flos": 25192570158240.0, + "grad_norm": 2.3721561471321726, + "learning_rate": 2.953101099810806e-07, + "loss": 0.7435, + "num_input_tokens_seen": 149185165, + "step": 6903 + }, + { + "epoch": 0.830156916972284, + "flos": 18015372433920.0, + "grad_norm": 2.9967592537239747, + "learning_rate": 2.9490285268865965e-07, + "loss": 0.8274, + "num_input_tokens_seen": 149202655, + "step": 6904 + }, + { + "epoch": 0.830277159862923, + "flos": 26284752545280.0, + "grad_norm": 2.648200820654909, + "learning_rate": 2.9449585405812085e-07, + "loss": 0.7999, + "num_input_tokens_seen": 149220035, + "step": 6905 + }, + { + "epoch": 0.8303974027535622, + "flos": 19910473436640.0, + "grad_norm": 2.7913795057799082, + "learning_rate": 2.940891141512043e-07, + "loss": 0.7392, + "num_input_tokens_seen": 149238055, + "step": 6906 + }, + { + "epoch": 0.8305176456442013, + "flos": 17141187921120.0, + "grad_norm": 3.99658546144668, + "learning_rate": 2.9368263302961385e-07, + "loss": 0.7264, + "num_input_tokens_seen": 149256755, + "step": 6907 + }, + { + "epoch": 0.8306378885348403, + "flos": 25590832866720.0, + "grad_norm": 2.335377603749232, + "learning_rate": 2.9327641075501075e-07, + "loss": 0.7981, + "num_input_tokens_seen": 149275745, + "step": 6908 + }, + { + "epoch": 0.8307581314254795, + "flos": 33899023941120.0, + "grad_norm": 3.00322359461105, + "learning_rate": 2.9287044738901866e-07, + "loss": 0.6609, + "num_input_tokens_seen": 149293730, + "step": 6909 + }, + { + "epoch": 0.8308783743161186, + "flos": 17537629311360.0, + "grad_norm": 2.3373193640450154, + "learning_rate": 2.9246474299322274e-07, + "loss": 0.9109, + "num_input_tokens_seen": 149309290, + "step": 6910 + }, + { + "epoch": 0.8309986172067576, + "flos": 69313768414560.0, + "grad_norm": 0.8945423775193199, + "learning_rate": 2.920592976291678e-07, + "loss": 0.6299, + "num_input_tokens_seen": 149366620, + "step": 6911 + }, + { + "epoch": 0.8311188600973968, + "flos": 22277890776960.0, + "grad_norm": 2.689184503707938, + "learning_rate": 2.916541113583595e-07, + "loss": 0.8079, + "num_input_tokens_seen": 149385120, + "step": 6912 + }, + { + "epoch": 0.8312391029880358, + "flos": 18743345346240.0, + "grad_norm": 2.676912255269513, + "learning_rate": 2.912491842422642e-07, + "loss": 0.6667, + "num_input_tokens_seen": 149402825, + "step": 6913 + }, + { + "epoch": 0.8313593458786749, + "flos": 20347621447680.0, + "grad_norm": 2.4539367887630337, + "learning_rate": 2.9084451634230857e-07, + "loss": 0.7106, + "num_input_tokens_seen": 149422125, + "step": 6914 + }, + { + "epoch": 0.831479588769314, + "flos": 32077902228000.0, + "grad_norm": 3.8193435532316684, + "learning_rate": 2.9044010771988125e-07, + "loss": 0.7146, + "num_input_tokens_seen": 149441940, + "step": 6915 + }, + { + "epoch": 0.8315998316599531, + "flos": 45120742134240.0, + "grad_norm": 2.6391537873564457, + "learning_rate": 2.900359584363303e-07, + "loss": 0.7226, + "num_input_tokens_seen": 149465045, + "step": 6916 + }, + { + "epoch": 0.8317200745505922, + "flos": 18337425667200.0, + "grad_norm": 2.7436056355446445, + "learning_rate": 2.8963206855296494e-07, + "loss": 0.8455, + "num_input_tokens_seen": 149481285, + "step": 6917 + }, + { + "epoch": 0.8318403174412313, + "flos": 24171913856640.0, + "grad_norm": 2.7051954333487784, + "learning_rate": 2.892284381310548e-07, + "loss": 0.7714, + "num_input_tokens_seen": 149502700, + "step": 6918 + }, + { + "epoch": 0.8319605603318704, + "flos": 22386741035040.0, + "grad_norm": 3.0827632755114034, + "learning_rate": 2.888250672318302e-07, + "loss": 0.7221, + "num_input_tokens_seen": 149520100, + "step": 6919 + }, + { + "epoch": 0.8320808032225094, + "flos": 37360630835520.0, + "grad_norm": 2.102793053196808, + "learning_rate": 2.884219559164831e-07, + "loss": 0.689, + "num_input_tokens_seen": 149543245, + "step": 6920 + }, + { + "epoch": 0.8322010461131486, + "flos": 12769410452640.0, + "grad_norm": 2.308722736092331, + "learning_rate": 2.880191042461635e-07, + "loss": 0.8165, + "num_input_tokens_seen": 149559185, + "step": 6921 + }, + { + "epoch": 0.8323212890037877, + "flos": 15793051601280.0, + "grad_norm": 2.2493660351573794, + "learning_rate": 2.876165122819849e-07, + "loss": 0.8029, + "num_input_tokens_seen": 149577165, + "step": 6922 + }, + { + "epoch": 0.8324415318944267, + "flos": 21688026457440.0, + "grad_norm": 1.9696702774168833, + "learning_rate": 2.872141800850201e-07, + "loss": 0.7938, + "num_input_tokens_seen": 149594340, + "step": 6923 + }, + { + "epoch": 0.8325617747850659, + "flos": 34149328070400.0, + "grad_norm": 2.2190665435646735, + "learning_rate": 2.868121077163024e-07, + "loss": 0.734, + "num_input_tokens_seen": 149613895, + "step": 6924 + }, + { + "epoch": 0.8326820176757049, + "flos": 18343335659040.0, + "grad_norm": 2.7873116776012044, + "learning_rate": 2.864102952368257e-07, + "loss": 0.7229, + "num_input_tokens_seen": 149631820, + "step": 6925 + }, + { + "epoch": 0.832802260566344, + "flos": 35939518809600.0, + "grad_norm": 1.8189973459795883, + "learning_rate": 2.860087427075444e-07, + "loss": 0.5943, + "num_input_tokens_seen": 149656070, + "step": 6926 + }, + { + "epoch": 0.8329225034569832, + "flos": 14224241184480.0, + "grad_norm": 2.9195878976537877, + "learning_rate": 2.856074501893744e-07, + "loss": 0.8631, + "num_input_tokens_seen": 149671270, + "step": 6927 + }, + { + "epoch": 0.8330427463476222, + "flos": 18051767362560.0, + "grad_norm": 2.1442986589772426, + "learning_rate": 2.8520641774319054e-07, + "loss": 0.8184, + "num_input_tokens_seen": 149689360, + "step": 6928 + }, + { + "epoch": 0.8331629892382613, + "flos": 18015186585120.0, + "grad_norm": 2.794571252738656, + "learning_rate": 2.848056454298309e-07, + "loss": 0.7601, + "num_input_tokens_seen": 149706685, + "step": 6929 + }, + { + "epoch": 0.8332832321289004, + "flos": 17432607538560.0, + "grad_norm": 2.7496484184934418, + "learning_rate": 2.844051333100905e-07, + "loss": 0.6534, + "num_input_tokens_seen": 149724230, + "step": 6930 + }, + { + "epoch": 0.8334034750195395, + "flos": 15063480389280.0, + "grad_norm": 3.2406278604698984, + "learning_rate": 2.840048814447269e-07, + "loss": 0.8384, + "num_input_tokens_seen": 149742395, + "step": 6931 + }, + { + "epoch": 0.8335237179101785, + "flos": 19399494815040.0, + "grad_norm": 2.834942940166349, + "learning_rate": 2.836048898944587e-07, + "loss": 0.7364, + "num_input_tokens_seen": 149760820, + "step": 6932 + }, + { + "epoch": 0.8336439608008177, + "flos": 21730963263840.0, + "grad_norm": 4.236746373861171, + "learning_rate": 2.832051587199642e-07, + "loss": 0.7246, + "num_input_tokens_seen": 149778075, + "step": 6933 + }, + { + "epoch": 0.8337642036914568, + "flos": 59617626473760.0, + "grad_norm": 0.8206279692991972, + "learning_rate": 2.828056879818821e-07, + "loss": 0.5751, + "num_input_tokens_seen": 149837150, + "step": 6934 + }, + { + "epoch": 0.8338844465820958, + "flos": 27123248354880.0, + "grad_norm": 2.3677321207239244, + "learning_rate": 2.824064777408117e-07, + "loss": 0.8346, + "num_input_tokens_seen": 149856940, + "step": 6935 + }, + { + "epoch": 0.8340046894727349, + "flos": 30437454216480.0, + "grad_norm": 3.190883112675443, + "learning_rate": 2.8200752805731263e-07, + "loss": 0.759, + "num_input_tokens_seen": 149879930, + "step": 6936 + }, + { + "epoch": 0.834124932363374, + "flos": 27087299463360.0, + "grad_norm": 2.3666635065550143, + "learning_rate": 2.8160883899190625e-07, + "loss": 0.8081, + "num_input_tokens_seen": 149903200, + "step": 6937 + }, + { + "epoch": 0.8342451752540131, + "flos": 24534636540960.0, + "grad_norm": 4.416027718636691, + "learning_rate": 2.8121041060507234e-07, + "loss": 0.739, + "num_input_tokens_seen": 149922230, + "step": 6938 + }, + { + "epoch": 0.8343654181446521, + "flos": 26576952727680.0, + "grad_norm": 2.1171025896113216, + "learning_rate": 2.808122429572528e-07, + "loss": 0.7115, + "num_input_tokens_seen": 149942435, + "step": 6939 + }, + { + "epoch": 0.8344856610352913, + "flos": 20747631134880.0, + "grad_norm": 3.3245371790422173, + "learning_rate": 2.804143361088489e-07, + "loss": 0.7596, + "num_input_tokens_seen": 149961615, + "step": 6940 + }, + { + "epoch": 0.8346059039259304, + "flos": 26058279965760.0, + "grad_norm": 3.1755313939227654, + "learning_rate": 2.8001669012022277e-07, + "loss": 0.7809, + "num_input_tokens_seen": 149979585, + "step": 6941 + }, + { + "epoch": 0.8347261468165694, + "flos": 28986228951360.0, + "grad_norm": 1.8671379561614796, + "learning_rate": 2.7961930505169795e-07, + "loss": 0.6918, + "num_input_tokens_seen": 150003830, + "step": 6942 + }, + { + "epoch": 0.8348463897072086, + "flos": 26358657495360.0, + "grad_norm": 2.6507408006338107, + "learning_rate": 2.792221809635558e-07, + "loss": 0.7672, + "num_input_tokens_seen": 150024460, + "step": 6943 + }, + { + "epoch": 0.8349666325978476, + "flos": 23334310121280.0, + "grad_norm": 2.887533152996821, + "learning_rate": 2.788253179160411e-07, + "loss": 0.7517, + "num_input_tokens_seen": 150045620, + "step": 6944 + }, + { + "epoch": 0.8350868754884867, + "flos": 12878372220000.0, + "grad_norm": 2.57242838026765, + "learning_rate": 2.7842871596935725e-07, + "loss": 0.6507, + "num_input_tokens_seen": 150064135, + "step": 6945 + }, + { + "epoch": 0.8352071183791259, + "flos": 26467544923200.0, + "grad_norm": 1.8949496406969732, + "learning_rate": 2.780323751836682e-07, + "loss": 0.6908, + "num_input_tokens_seen": 150085350, + "step": 6946 + }, + { + "epoch": 0.8353273612697649, + "flos": 20638595028000.0, + "grad_norm": 2.0308358577211156, + "learning_rate": 2.7763629561909876e-07, + "loss": 0.7889, + "num_input_tokens_seen": 150106180, + "step": 6947 + }, + { + "epoch": 0.835447604160404, + "flos": 19725934080000.0, + "grad_norm": 2.703146168179399, + "learning_rate": 2.772404773357335e-07, + "loss": 0.7687, + "num_input_tokens_seen": 150125585, + "step": 6948 + }, + { + "epoch": 0.8355678470510431, + "flos": 23402156400480.0, + "grad_norm": 6.467374056373931, + "learning_rate": 2.7684492039361853e-07, + "loss": 0.785, + "num_input_tokens_seen": 150144810, + "step": 6949 + }, + { + "epoch": 0.8356880899416822, + "flos": 21184295939040.0, + "grad_norm": 1.9364291277306362, + "learning_rate": 2.764496248527586e-07, + "loss": 0.8399, + "num_input_tokens_seen": 150163785, + "step": 6950 + }, + { + "epoch": 0.8358083328323213, + "flos": 27997321358400.0, + "grad_norm": 2.3449687586413397, + "learning_rate": 2.760545907731211e-07, + "loss": 0.7896, + "num_input_tokens_seen": 150184360, + "step": 6951 + }, + { + "epoch": 0.8359285757229604, + "flos": 27743857799520.0, + "grad_norm": 5.898263850403321, + "learning_rate": 2.75659818214631e-07, + "loss": 0.681, + "num_input_tokens_seen": 150205465, + "step": 6952 + }, + { + "epoch": 0.8360488186135995, + "flos": 21404561168640.0, + "grad_norm": 2.427376092327581, + "learning_rate": 2.752653072371749e-07, + "loss": 0.7857, + "num_input_tokens_seen": 150224900, + "step": 6953 + }, + { + "epoch": 0.8361690615042385, + "flos": 27592553941920.0, + "grad_norm": 3.181542457672559, + "learning_rate": 2.7487105790060105e-07, + "loss": 0.7503, + "num_input_tokens_seen": 150244310, + "step": 6954 + }, + { + "epoch": 0.8362893043948777, + "flos": 39146063845440.0, + "grad_norm": 3.0769379936370203, + "learning_rate": 2.7447707026471587e-07, + "loss": 0.6891, + "num_input_tokens_seen": 150267285, + "step": 6955 + }, + { + "epoch": 0.8364095472855168, + "flos": 24749660834400.0, + "grad_norm": 2.993979930609111, + "learning_rate": 2.740833443892874e-07, + "loss": 0.7993, + "num_input_tokens_seen": 150285455, + "step": 6956 + }, + { + "epoch": 0.8365297901761558, + "flos": 22710764265600.0, + "grad_norm": 2.494794862834075, + "learning_rate": 2.7368988033404327e-07, + "loss": 0.8019, + "num_input_tokens_seen": 150302970, + "step": 6957 + }, + { + "epoch": 0.836650033066795, + "flos": 28355026125120.0, + "grad_norm": 2.357114707068592, + "learning_rate": 2.732966781586712e-07, + "loss": 0.8473, + "num_input_tokens_seen": 150322715, + "step": 6958 + }, + { + "epoch": 0.836770275957434, + "flos": 22205100919680.0, + "grad_norm": 2.090679593364171, + "learning_rate": 2.729037379228205e-07, + "loss": 0.6701, + "num_input_tokens_seen": 150342900, + "step": 6959 + }, + { + "epoch": 0.8368905188480731, + "flos": 22459791080640.0, + "grad_norm": 1.8200465978169618, + "learning_rate": 2.725110596860998e-07, + "loss": 0.8059, + "num_input_tokens_seen": 150363580, + "step": 6960 + }, + { + "epoch": 0.8370107617387123, + "flos": 13351543462080.0, + "grad_norm": 2.278344904329078, + "learning_rate": 2.7211864350807776e-07, + "loss": 0.7032, + "num_input_tokens_seen": 150381780, + "step": 6961 + }, + { + "epoch": 0.8371310046293513, + "flos": 25224987922560.0, + "grad_norm": 1.91133858065666, + "learning_rate": 2.717264894482836e-07, + "loss": 0.7381, + "num_input_tokens_seen": 150402120, + "step": 6962 + }, + { + "epoch": 0.8372512475199904, + "flos": 19764522024480.0, + "grad_norm": 2.384436582317507, + "learning_rate": 2.7133459756620646e-07, + "loss": 0.8138, + "num_input_tokens_seen": 150419315, + "step": 6963 + }, + { + "epoch": 0.8373714904106295, + "flos": 19363620263040.0, + "grad_norm": 2.321551265198988, + "learning_rate": 2.7094296792129733e-07, + "loss": 0.7354, + "num_input_tokens_seen": 150438915, + "step": 6964 + }, + { + "epoch": 0.8374917333012686, + "flos": 14954741640480.0, + "grad_norm": 2.29824134513481, + "learning_rate": 2.7055160057296424e-07, + "loss": 0.7547, + "num_input_tokens_seen": 150456155, + "step": 6965 + }, + { + "epoch": 0.8376119761919076, + "flos": 30288120356160.0, + "grad_norm": 1.9707289543765378, + "learning_rate": 2.7016049558057896e-07, + "loss": 0.7257, + "num_input_tokens_seen": 150478115, + "step": 6966 + }, + { + "epoch": 0.8377322190825467, + "flos": 29381072041920.0, + "grad_norm": 2.311799536115605, + "learning_rate": 2.6976965300347074e-07, + "loss": 0.7077, + "num_input_tokens_seen": 150500725, + "step": 6967 + }, + { + "epoch": 0.8378524619731859, + "flos": 26648813340960.0, + "grad_norm": 3.2115173558789647, + "learning_rate": 2.693790729009309e-07, + "loss": 0.6938, + "num_input_tokens_seen": 150522335, + "step": 6968 + }, + { + "epoch": 0.8379727048638249, + "flos": 20674320900960.0, + "grad_norm": 3.9779398093677996, + "learning_rate": 2.6898875533220946e-07, + "loss": 0.8835, + "num_input_tokens_seen": 150541640, + "step": 6969 + }, + { + "epoch": 0.838092947754464, + "flos": 20055458435040.0, + "grad_norm": 2.966980095354613, + "learning_rate": 2.685987003565171e-07, + "loss": 0.8229, + "num_input_tokens_seen": 150559680, + "step": 6970 + }, + { + "epoch": 0.8382131906451031, + "flos": 18087493235520.0, + "grad_norm": 4.56042077749923, + "learning_rate": 2.6820890803302566e-07, + "loss": 0.75, + "num_input_tokens_seen": 150575205, + "step": 6971 + }, + { + "epoch": 0.8383334335357422, + "flos": 17068398063840.0, + "grad_norm": 2.7423413877426475, + "learning_rate": 2.6781937842086557e-07, + "loss": 0.8186, + "num_input_tokens_seen": 150593995, + "step": 6972 + }, + { + "epoch": 0.8384536764263812, + "flos": 20675212975200.0, + "grad_norm": 2.112387872011215, + "learning_rate": 2.6743011157912933e-07, + "loss": 0.6741, + "num_input_tokens_seen": 150613715, + "step": 6973 + }, + { + "epoch": 0.8385739193170204, + "flos": 28945262142240.0, + "grad_norm": 2.3370760691337114, + "learning_rate": 2.6704110756686725e-07, + "loss": 0.6546, + "num_input_tokens_seen": 150634540, + "step": 6974 + }, + { + "epoch": 0.8386941622076595, + "flos": 23404460925600.0, + "grad_norm": 2.6216719283891643, + "learning_rate": 2.6665236644309085e-07, + "loss": 0.8414, + "num_input_tokens_seen": 150654850, + "step": 6975 + }, + { + "epoch": 0.8388144050982985, + "flos": 22969840458240.0, + "grad_norm": 2.8173741808251656, + "learning_rate": 2.662638882667727e-07, + "loss": 0.8014, + "num_input_tokens_seen": 150673790, + "step": 6976 + }, + { + "epoch": 0.8389346479889377, + "flos": 24245632957920.0, + "grad_norm": 7.168770178290041, + "learning_rate": 2.658756730968443e-07, + "loss": 0.7286, + "num_input_tokens_seen": 150692255, + "step": 6977 + }, + { + "epoch": 0.8390548908795767, + "flos": 21184741976160.0, + "grad_norm": 2.5075833700208214, + "learning_rate": 2.654877209921975e-07, + "loss": 0.8867, + "num_input_tokens_seen": 150709790, + "step": 6978 + }, + { + "epoch": 0.8391751337702158, + "flos": 35576052730080.0, + "grad_norm": 2.9608144608753375, + "learning_rate": 2.651000320116843e-07, + "loss": 0.6269, + "num_input_tokens_seen": 150730675, + "step": 6979 + }, + { + "epoch": 0.839295376660855, + "flos": 21294818836320.0, + "grad_norm": 2.266261254780802, + "learning_rate": 2.647126062141163e-07, + "loss": 0.7596, + "num_input_tokens_seen": 150749750, + "step": 6980 + }, + { + "epoch": 0.839415619551494, + "flos": 18416311365120.0, + "grad_norm": 2.436465312746078, + "learning_rate": 2.643254436582669e-07, + "loss": 0.8404, + "num_input_tokens_seen": 150769240, + "step": 6981 + }, + { + "epoch": 0.8395358624421331, + "flos": 23188767576480.0, + "grad_norm": 2.2764221734631556, + "learning_rate": 2.6393854440286743e-07, + "loss": 0.8213, + "num_input_tokens_seen": 150788410, + "step": 6982 + }, + { + "epoch": 0.8396561053327722, + "flos": 24346491717600.0, + "grad_norm": 2.5061390840335473, + "learning_rate": 2.6355190850661045e-07, + "loss": 0.704, + "num_input_tokens_seen": 150805245, + "step": 6983 + }, + { + "epoch": 0.8397763482234113, + "flos": 22206030163680.0, + "grad_norm": 3.66844192069171, + "learning_rate": 2.631655360281486e-07, + "loss": 0.8691, + "num_input_tokens_seen": 150824920, + "step": 6984 + }, + { + "epoch": 0.8398965911140504, + "flos": 22132088043840.0, + "grad_norm": 2.680626618448642, + "learning_rate": 2.6277942702609323e-07, + "loss": 0.6589, + "num_input_tokens_seen": 150844400, + "step": 6985 + }, + { + "epoch": 0.8400168340046895, + "flos": 21511998975840.0, + "grad_norm": 2.713136336402435, + "learning_rate": 2.623935815590186e-07, + "loss": 0.8744, + "num_input_tokens_seen": 150862780, + "step": 6986 + }, + { + "epoch": 0.8401370768953286, + "flos": 22459828250400.0, + "grad_norm": 2.33322281802808, + "learning_rate": 2.6200799968545516e-07, + "loss": 0.812, + "num_input_tokens_seen": 150883075, + "step": 6987 + }, + { + "epoch": 0.8402573197859676, + "flos": 59154788424960.0, + "grad_norm": 2.5713874401102963, + "learning_rate": 2.616226814638969e-07, + "loss": 0.5636, + "num_input_tokens_seen": 150948180, + "step": 6988 + }, + { + "epoch": 0.8403775626766068, + "flos": 22642211760960.0, + "grad_norm": 2.35384493854639, + "learning_rate": 2.612376269527954e-07, + "loss": 0.773, + "num_input_tokens_seen": 150967885, + "step": 6989 + }, + { + "epoch": 0.8404978055672458, + "flos": 19581766816320.0, + "grad_norm": 4.464398593020484, + "learning_rate": 2.608528362105635e-07, + "loss": 0.6756, + "num_input_tokens_seen": 150987255, + "step": 6990 + }, + { + "epoch": 0.8406180484578849, + "flos": 27487457829600.0, + "grad_norm": 2.5914919112511856, + "learning_rate": 2.6046830929557374e-07, + "loss": 0.736, + "num_input_tokens_seen": 151006905, + "step": 6991 + }, + { + "epoch": 0.8407382913485241, + "flos": 22095878964000.0, + "grad_norm": 2.5225390083616257, + "learning_rate": 2.6008404626615776e-07, + "loss": 0.8497, + "num_input_tokens_seen": 151025405, + "step": 6992 + }, + { + "epoch": 0.8408585342391631, + "flos": 13898842672800.0, + "grad_norm": 3.5915146219475544, + "learning_rate": 2.597000471806092e-07, + "loss": 0.7321, + "num_input_tokens_seen": 151041970, + "step": 6993 + }, + { + "epoch": 0.8409787771298022, + "flos": 20158881908160.0, + "grad_norm": 2.4890022610584746, + "learning_rate": 2.593163120971793e-07, + "loss": 0.7355, + "num_input_tokens_seen": 151060835, + "step": 6994 + }, + { + "epoch": 0.8410990200204413, + "flos": 23109212822880.0, + "grad_norm": 2.244458474009429, + "learning_rate": 2.5893284107408165e-07, + "loss": 0.6901, + "num_input_tokens_seen": 151078205, + "step": 6995 + }, + { + "epoch": 0.8412192629110804, + "flos": 23992949964000.0, + "grad_norm": 2.987449141193172, + "learning_rate": 2.5854963416948726e-07, + "loss": 0.7787, + "num_input_tokens_seen": 151100470, + "step": 6996 + }, + { + "epoch": 0.8413395058017195, + "flos": 25554995484480.0, + "grad_norm": 2.2965771748012553, + "learning_rate": 2.5816669144152816e-07, + "loss": 0.6954, + "num_input_tokens_seen": 151121650, + "step": 6997 + }, + { + "epoch": 0.8414597486923585, + "flos": 63544227263040.0, + "grad_norm": 0.8627431236596601, + "learning_rate": 2.5778401294829777e-07, + "loss": 0.6622, + "num_input_tokens_seen": 151180390, + "step": 6998 + }, + { + "epoch": 0.8415799915829977, + "flos": 19071531589920.0, + "grad_norm": 2.33613386572744, + "learning_rate": 2.574015987478473e-07, + "loss": 0.6521, + "num_input_tokens_seen": 151198520, + "step": 6999 + }, + { + "epoch": 0.8417002344736367, + "flos": 19793111303520.0, + "grad_norm": 3.193191682571384, + "learning_rate": 2.570194488981887e-07, + "loss": 0.8739, + "num_input_tokens_seen": 151215065, + "step": 7000 + }, + { + "epoch": 0.8418204773642758, + "flos": 62073296291520.0, + "grad_norm": 0.847401277367928, + "learning_rate": 2.566375634572939e-07, + "loss": 0.6018, + "num_input_tokens_seen": 151275705, + "step": 7001 + }, + { + "epoch": 0.841940720254915, + "flos": 17068658252160.0, + "grad_norm": 2.3225144749448012, + "learning_rate": 2.562559424830943e-07, + "loss": 0.7641, + "num_input_tokens_seen": 151293665, + "step": 7002 + }, + { + "epoch": 0.842060963145554, + "flos": 16260275681760.0, + "grad_norm": 2.948104546876098, + "learning_rate": 2.5587458603348256e-07, + "loss": 0.7049, + "num_input_tokens_seen": 151310955, + "step": 7003 + }, + { + "epoch": 0.8421812060361931, + "flos": 21876951845760.0, + "grad_norm": 2.378940325212646, + "learning_rate": 2.554934941663085e-07, + "loss": 0.8391, + "num_input_tokens_seen": 151328490, + "step": 7004 + }, + { + "epoch": 0.8423014489268322, + "flos": 27737687619360.0, + "grad_norm": 2.6675332469351978, + "learning_rate": 2.5511266693938484e-07, + "loss": 0.7346, + "num_input_tokens_seen": 151346950, + "step": 7005 + }, + { + "epoch": 0.8424216918174713, + "flos": 25081601223840.0, + "grad_norm": 1.9237507915928969, + "learning_rate": 2.547321044104822e-07, + "loss": 0.7802, + "num_input_tokens_seen": 151368445, + "step": 7006 + }, + { + "epoch": 0.8425419347081103, + "flos": 24712819868640.0, + "grad_norm": 2.999863462519954, + "learning_rate": 2.5435180663733113e-07, + "loss": 0.7697, + "num_input_tokens_seen": 151388745, + "step": 7007 + }, + { + "epoch": 0.8426621775987495, + "flos": 24785163688800.0, + "grad_norm": 8.097535179841428, + "learning_rate": 2.539717736776241e-07, + "loss": 0.7146, + "num_input_tokens_seen": 151404970, + "step": 7008 + }, + { + "epoch": 0.8427824204893886, + "flos": 23516396273760.0, + "grad_norm": 1.8715091634409338, + "learning_rate": 2.535920055890097e-07, + "loss": 0.7619, + "num_input_tokens_seen": 151426265, + "step": 7009 + }, + { + "epoch": 0.8429026633800276, + "flos": 16120977656640.0, + "grad_norm": 2.6729016924761004, + "learning_rate": 2.5321250242910006e-07, + "loss": 0.6488, + "num_input_tokens_seen": 151444180, + "step": 7010 + }, + { + "epoch": 0.8430229062706668, + "flos": 22166773163520.0, + "grad_norm": 2.1745259963301935, + "learning_rate": 2.5283326425546493e-07, + "loss": 0.8651, + "num_input_tokens_seen": 151463290, + "step": 7011 + }, + { + "epoch": 0.8431431491613058, + "flos": 35279317836960.0, + "grad_norm": 4.646994794829121, + "learning_rate": 2.5245429112563443e-07, + "loss": 0.6991, + "num_input_tokens_seen": 151483965, + "step": 7012 + }, + { + "epoch": 0.8432633920519449, + "flos": 25775409393120.0, + "grad_norm": 2.8449895483807155, + "learning_rate": 2.5207558309709865e-07, + "loss": 0.8235, + "num_input_tokens_seen": 151503700, + "step": 7013 + }, + { + "epoch": 0.8433836349425841, + "flos": 64866822224160.0, + "grad_norm": 0.6736574039632497, + "learning_rate": 2.516971402273065e-07, + "loss": 0.5611, + "num_input_tokens_seen": 151569765, + "step": 7014 + }, + { + "epoch": 0.8435038778332231, + "flos": 20201149658880.0, + "grad_norm": 2.2956809748537936, + "learning_rate": 2.513189625736687e-07, + "loss": 0.6765, + "num_input_tokens_seen": 151586660, + "step": 7015 + }, + { + "epoch": 0.8436241207238622, + "flos": 20962135051680.0, + "grad_norm": 2.908791036213911, + "learning_rate": 2.509410501935534e-07, + "loss": 0.718, + "num_input_tokens_seen": 151602295, + "step": 7016 + }, + { + "epoch": 0.8437443636145013, + "flos": 14661426365280.0, + "grad_norm": 3.7848614557224773, + "learning_rate": 2.5056340314429116e-07, + "loss": 0.7547, + "num_input_tokens_seen": 151619760, + "step": 7017 + }, + { + "epoch": 0.8438646065051404, + "flos": 21577689408960.0, + "grad_norm": 3.2606259279869474, + "learning_rate": 2.5018602148316904e-07, + "loss": 0.8022, + "num_input_tokens_seen": 151635795, + "step": 7018 + }, + { + "epoch": 0.8439848493957794, + "flos": 23256465176640.0, + "grad_norm": 2.168771845902671, + "learning_rate": 2.498089052674359e-07, + "loss": 0.8033, + "num_input_tokens_seen": 151653770, + "step": 7019 + }, + { + "epoch": 0.8441050922864186, + "flos": 19691248960320.0, + "grad_norm": 9.244463114026772, + "learning_rate": 2.494320545543007e-07, + "loss": 0.7521, + "num_input_tokens_seen": 151673340, + "step": 7020 + }, + { + "epoch": 0.8442253351770577, + "flos": 21804087648960.0, + "grad_norm": 2.332734735090837, + "learning_rate": 2.490554694009308e-07, + "loss": 0.6681, + "num_input_tokens_seen": 151694205, + "step": 7021 + }, + { + "epoch": 0.8443455780676967, + "flos": 34296840612480.0, + "grad_norm": 18.914120585412668, + "learning_rate": 2.4867914986445426e-07, + "loss": 0.7864, + "num_input_tokens_seen": 151716595, + "step": 7022 + }, + { + "epoch": 0.8444658209583359, + "flos": 48145015168800.0, + "grad_norm": 2.3585019548968877, + "learning_rate": 2.483030960019581e-07, + "loss": 0.71, + "num_input_tokens_seen": 151740525, + "step": 7023 + }, + { + "epoch": 0.8445860638489749, + "flos": 68386425411360.0, + "grad_norm": 0.7519067923403294, + "learning_rate": 2.479273078704891e-07, + "loss": 0.5537, + "num_input_tokens_seen": 151793890, + "step": 7024 + }, + { + "epoch": 0.844706306739614, + "flos": 62743352984160.0, + "grad_norm": 0.7868099731773208, + "learning_rate": 2.475517855270552e-07, + "loss": 0.646, + "num_input_tokens_seen": 151853970, + "step": 7025 + }, + { + "epoch": 0.8448265496302532, + "flos": 14954927489280.0, + "grad_norm": 2.454709070063567, + "learning_rate": 2.4717652902862143e-07, + "loss": 0.7289, + "num_input_tokens_seen": 151872945, + "step": 7026 + }, + { + "epoch": 0.8449467925208922, + "flos": 23407769034240.0, + "grad_norm": 2.466779767699736, + "learning_rate": 2.4680153843211495e-07, + "loss": 0.8164, + "num_input_tokens_seen": 151892875, + "step": 7027 + }, + { + "epoch": 0.8450670354115313, + "flos": 22715559164640.0, + "grad_norm": 2.2531860893528464, + "learning_rate": 2.464268137944212e-07, + "loss": 0.725, + "num_input_tokens_seen": 151914170, + "step": 7028 + }, + { + "epoch": 0.8451872783021703, + "flos": 29782754368320.0, + "grad_norm": 2.1654228038993257, + "learning_rate": 2.46052355172385e-07, + "loss": 0.783, + "num_input_tokens_seen": 151932210, + "step": 7029 + }, + { + "epoch": 0.8453075211928095, + "flos": 21838735598880.0, + "grad_norm": 4.589926029847612, + "learning_rate": 2.456781626228128e-07, + "loss": 0.7472, + "num_input_tokens_seen": 151951715, + "step": 7030 + }, + { + "epoch": 0.8454277640834486, + "flos": 58668013050720.0, + "grad_norm": 0.9202297529704773, + "learning_rate": 2.453042362024675e-07, + "loss": 0.6617, + "num_input_tokens_seen": 152004960, + "step": 7031 + }, + { + "epoch": 0.8455480069740876, + "flos": 27051276232320.0, + "grad_norm": 2.1085349211484243, + "learning_rate": 2.449305759680751e-07, + "loss": 0.7331, + "num_input_tokens_seen": 152026285, + "step": 7032 + }, + { + "epoch": 0.8456682498647268, + "flos": 27159680453280.0, + "grad_norm": 7.550777208353807, + "learning_rate": 2.445571819763188e-07, + "loss": 0.7555, + "num_input_tokens_seen": 152048585, + "step": 7033 + }, + { + "epoch": 0.8457884927553658, + "flos": 20602497457440.0, + "grad_norm": 3.957861192887892, + "learning_rate": 2.4418405428384227e-07, + "loss": 0.5868, + "num_input_tokens_seen": 152068795, + "step": 7034 + }, + { + "epoch": 0.8459087356460049, + "flos": 15277166571360.0, + "grad_norm": 2.2065850541080856, + "learning_rate": 2.4381119294724864e-07, + "loss": 0.7159, + "num_input_tokens_seen": 152086240, + "step": 7035 + }, + { + "epoch": 0.846028978536644, + "flos": 18816321052320.0, + "grad_norm": 3.4539792484351444, + "learning_rate": 2.434385980231004e-07, + "loss": 0.5447, + "num_input_tokens_seen": 152105080, + "step": 7036 + }, + { + "epoch": 0.8461492214272831, + "flos": 52585530990720.0, + "grad_norm": 2.107162445476085, + "learning_rate": 2.4306626956792043e-07, + "loss": 0.6549, + "num_input_tokens_seen": 152130025, + "step": 7037 + }, + { + "epoch": 0.8462694643179222, + "flos": 18561965419200.0, + "grad_norm": 2.0019119184021377, + "learning_rate": 2.4269420763819017e-07, + "loss": 0.7561, + "num_input_tokens_seen": 152148070, + "step": 7038 + }, + { + "epoch": 0.8463897072085613, + "flos": 24352104351360.0, + "grad_norm": 5.932592994420835, + "learning_rate": 2.4232241229035223e-07, + "loss": 0.839, + "num_input_tokens_seen": 152165825, + "step": 7039 + }, + { + "epoch": 0.8465099500992004, + "flos": 68603419702080.0, + "grad_norm": 0.750895680609673, + "learning_rate": 2.419508835808064e-07, + "loss": 0.5659, + "num_input_tokens_seen": 152222380, + "step": 7040 + }, + { + "epoch": 0.8466301929898394, + "flos": 13716347652960.0, + "grad_norm": 4.094164412394521, + "learning_rate": 2.415796215659134e-07, + "loss": 0.6277, + "num_input_tokens_seen": 152239085, + "step": 7041 + }, + { + "epoch": 0.8467504358804786, + "flos": 19213728856320.0, + "grad_norm": 2.928586055971626, + "learning_rate": 2.412086263019939e-07, + "loss": 0.7699, + "num_input_tokens_seen": 152257110, + "step": 7042 + }, + { + "epoch": 0.8468706787711177, + "flos": 21293666573760.0, + "grad_norm": 2.494258582089799, + "learning_rate": 2.408378978453276e-07, + "loss": 0.8012, + "num_input_tokens_seen": 152277230, + "step": 7043 + }, + { + "epoch": 0.8469909216617567, + "flos": 64784516908320.0, + "grad_norm": 0.8465214881805878, + "learning_rate": 2.404674362521533e-07, + "loss": 0.6386, + "num_input_tokens_seen": 152335725, + "step": 7044 + }, + { + "epoch": 0.8471111645523959, + "flos": 19253097365760.0, + "grad_norm": 2.6040158546011973, + "learning_rate": 2.4009724157866997e-07, + "loss": 0.7449, + "num_input_tokens_seen": 152352785, + "step": 7045 + }, + { + "epoch": 0.8472314074430349, + "flos": 21984315313440.0, + "grad_norm": 3.086085680530546, + "learning_rate": 2.3972731388103564e-07, + "loss": 0.7666, + "num_input_tokens_seen": 152371265, + "step": 7046 + }, + { + "epoch": 0.847351650333674, + "flos": 57800221736640.0, + "grad_norm": 0.8098371592832532, + "learning_rate": 2.393576532153687e-07, + "loss": 0.6228, + "num_input_tokens_seen": 152435050, + "step": 7047 + }, + { + "epoch": 0.8474718932243132, + "flos": 41226144508320.0, + "grad_norm": 0.9368430697017671, + "learning_rate": 2.389882596377453e-07, + "loss": 0.577, + "num_input_tokens_seen": 152489315, + "step": 7048 + }, + { + "epoch": 0.8475921361149522, + "flos": 38125927920480.0, + "grad_norm": 2.0545423333120536, + "learning_rate": 2.386191332042031e-07, + "loss": 0.7657, + "num_input_tokens_seen": 152511210, + "step": 7049 + }, + { + "epoch": 0.8477123790055913, + "flos": 25338744588960.0, + "grad_norm": 11.415988725641961, + "learning_rate": 2.3825027397073794e-07, + "loss": 0.7303, + "num_input_tokens_seen": 152531685, + "step": 7050 + }, + { + "epoch": 0.8478326218962304, + "flos": 30181723302240.0, + "grad_norm": 2.4131799955110815, + "learning_rate": 2.3788168199330515e-07, + "loss": 0.6691, + "num_input_tokens_seen": 152553245, + "step": 7051 + }, + { + "epoch": 0.8479528647868695, + "flos": 38158643042880.0, + "grad_norm": 2.0904542405622673, + "learning_rate": 2.3751335732782074e-07, + "loss": 0.727, + "num_input_tokens_seen": 152574505, + "step": 7052 + }, + { + "epoch": 0.8480731076775085, + "flos": 20927821629600.0, + "grad_norm": 3.2529899341932937, + "learning_rate": 2.371453000301582e-07, + "loss": 0.7984, + "num_input_tokens_seen": 152593190, + "step": 7053 + }, + { + "epoch": 0.8481933505681477, + "flos": 32549586679680.0, + "grad_norm": 2.2602781144908235, + "learning_rate": 2.3677751015615222e-07, + "loss": 0.7437, + "num_input_tokens_seen": 152615265, + "step": 7054 + }, + { + "epoch": 0.8483135934587868, + "flos": 20711607903840.0, + "grad_norm": 2.7329695222999755, + "learning_rate": 2.3640998776159593e-07, + "loss": 0.858, + "num_input_tokens_seen": 152632770, + "step": 7055 + }, + { + "epoch": 0.8484338363494258, + "flos": 21622150175520.0, + "grad_norm": 2.2299267432663505, + "learning_rate": 2.3604273290224253e-07, + "loss": 0.8135, + "num_input_tokens_seen": 152653485, + "step": 7056 + }, + { + "epoch": 0.848554079240065, + "flos": 14992772038560.0, + "grad_norm": 2.820168362048889, + "learning_rate": 2.356757456338039e-07, + "loss": 0.7451, + "num_input_tokens_seen": 152670970, + "step": 7057 + }, + { + "epoch": 0.848674322130704, + "flos": 67963216060320.0, + "grad_norm": 0.7576877451978634, + "learning_rate": 2.3530902601195147e-07, + "loss": 0.5889, + "num_input_tokens_seen": 152739460, + "step": 7058 + }, + { + "epoch": 0.8487945650213431, + "flos": 18449398185120.0, + "grad_norm": 3.00605607705973, + "learning_rate": 2.34942574092317e-07, + "loss": 0.7887, + "num_input_tokens_seen": 152754710, + "step": 7059 + }, + { + "epoch": 0.8489148079119821, + "flos": 23439480573120.0, + "grad_norm": 2.875868624619236, + "learning_rate": 2.3457638993049045e-07, + "loss": 0.7666, + "num_input_tokens_seen": 152772700, + "step": 7060 + }, + { + "epoch": 0.8490350508026213, + "flos": 19909172495040.0, + "grad_norm": 2.6175932156962216, + "learning_rate": 2.3421047358202252e-07, + "loss": 0.6455, + "num_input_tokens_seen": 152791550, + "step": 7061 + }, + { + "epoch": 0.8491552936932604, + "flos": 24245224090560.0, + "grad_norm": 3.0691651974364746, + "learning_rate": 2.3384482510242144e-07, + "loss": 0.8326, + "num_input_tokens_seen": 152809410, + "step": 7062 + }, + { + "epoch": 0.8492755365838994, + "flos": 22490127338400.0, + "grad_norm": 2.4075860612011035, + "learning_rate": 2.3347944454715575e-07, + "loss": 0.7762, + "num_input_tokens_seen": 152825800, + "step": 7063 + }, + { + "epoch": 0.8493957794745386, + "flos": 26941794088320.0, + "grad_norm": 2.7227094034837966, + "learning_rate": 2.331143319716542e-07, + "loss": 0.6734, + "num_input_tokens_seen": 152845330, + "step": 7064 + }, + { + "epoch": 0.8495160223651776, + "flos": 29819112127200.0, + "grad_norm": 2.4601454656127753, + "learning_rate": 2.3274948743130363e-07, + "loss": 0.6626, + "num_input_tokens_seen": 152865165, + "step": 7065 + }, + { + "epoch": 0.8496362652558167, + "flos": 23042444466720.0, + "grad_norm": 1.9024597383266695, + "learning_rate": 2.3238491098145085e-07, + "loss": 0.7953, + "num_input_tokens_seen": 152884695, + "step": 7066 + }, + { + "epoch": 0.8497565081464559, + "flos": 14588376319680.0, + "grad_norm": 2.7541788602050237, + "learning_rate": 2.3202060267740141e-07, + "loss": 0.7329, + "num_input_tokens_seen": 152902220, + "step": 7067 + }, + { + "epoch": 0.8498767510370949, + "flos": 21106079296800.0, + "grad_norm": 3.1875213921520977, + "learning_rate": 2.3165656257442044e-07, + "loss": 0.7711, + "num_input_tokens_seen": 152920740, + "step": 7068 + }, + { + "epoch": 0.849996993927734, + "flos": 23620525972320.0, + "grad_norm": 2.6202006289417956, + "learning_rate": 2.31292790727734e-07, + "loss": 0.9037, + "num_input_tokens_seen": 152938305, + "step": 7069 + }, + { + "epoch": 0.8501172368183731, + "flos": 20528741186400.0, + "grad_norm": 2.7594553779136444, + "learning_rate": 2.3092928719252392e-07, + "loss": 0.8013, + "num_input_tokens_seen": 152956705, + "step": 7070 + }, + { + "epoch": 0.8502374797090122, + "flos": 22240789622880.0, + "grad_norm": 2.5917095753183133, + "learning_rate": 2.3056605202393475e-07, + "loss": 0.7835, + "num_input_tokens_seen": 152974265, + "step": 7071 + }, + { + "epoch": 0.8503577225996513, + "flos": 23625989927040.0, + "grad_norm": 2.7724058999688572, + "learning_rate": 2.3020308527706888e-07, + "loss": 0.6677, + "num_input_tokens_seen": 152993680, + "step": 7072 + }, + { + "epoch": 0.8504779654902904, + "flos": 26720079238080.0, + "grad_norm": 2.0941503341327343, + "learning_rate": 2.2984038700698715e-07, + "loss": 0.8938, + "num_input_tokens_seen": 153012990, + "step": 7073 + }, + { + "epoch": 0.8505982083809295, + "flos": 26430295090080.0, + "grad_norm": 3.111627577495461, + "learning_rate": 2.2947795726871222e-07, + "loss": 0.7918, + "num_input_tokens_seen": 153034015, + "step": 7074 + }, + { + "epoch": 0.8507184512715685, + "flos": 20274459892800.0, + "grad_norm": 4.506373025217423, + "learning_rate": 2.2911579611722253e-07, + "loss": 0.858, + "num_input_tokens_seen": 153053160, + "step": 7075 + }, + { + "epoch": 0.8508386941622077, + "flos": 18998072676960.0, + "grad_norm": 2.3541902073946304, + "learning_rate": 2.2875390360745905e-07, + "loss": 0.8735, + "num_input_tokens_seen": 153072565, + "step": 7076 + }, + { + "epoch": 0.8509589370528468, + "flos": 16410018409440.0, + "grad_norm": 3.142891289166327, + "learning_rate": 2.2839227979432008e-07, + "loss": 0.7758, + "num_input_tokens_seen": 153090215, + "step": 7077 + }, + { + "epoch": 0.8510791799434858, + "flos": 18233742005760.0, + "grad_norm": 2.1199065481137596, + "learning_rate": 2.2803092473266373e-07, + "loss": 0.8518, + "num_input_tokens_seen": 153107740, + "step": 7078 + }, + { + "epoch": 0.851199422834125, + "flos": 23407880543520.0, + "grad_norm": 4.607391033902183, + "learning_rate": 2.2766983847730724e-07, + "loss": 0.8694, + "num_input_tokens_seen": 153127410, + "step": 7079 + }, + { + "epoch": 0.851319665724764, + "flos": 16266297182880.0, + "grad_norm": 2.2777731541325648, + "learning_rate": 2.2730902108302663e-07, + "loss": 0.6674, + "num_input_tokens_seen": 153144325, + "step": 7080 + }, + { + "epoch": 0.8514399086154031, + "flos": 18962309634240.0, + "grad_norm": 2.3434242530791205, + "learning_rate": 2.269484726045583e-07, + "loss": 0.6899, + "num_input_tokens_seen": 153163240, + "step": 7081 + }, + { + "epoch": 0.8515601515060423, + "flos": 24536011822080.0, + "grad_norm": 2.1390826383859025, + "learning_rate": 2.2658819309659672e-07, + "loss": 0.7953, + "num_input_tokens_seen": 153183440, + "step": 7082 + }, + { + "epoch": 0.8516803943966813, + "flos": 19501357158240.0, + "grad_norm": 2.1551163499051316, + "learning_rate": 2.2622818261379706e-07, + "loss": 0.8473, + "num_input_tokens_seen": 153200460, + "step": 7083 + }, + { + "epoch": 0.8518006372873204, + "flos": 20236875531840.0, + "grad_norm": 2.3155633380211595, + "learning_rate": 2.2586844121077142e-07, + "loss": 0.7513, + "num_input_tokens_seen": 153218970, + "step": 7084 + }, + { + "epoch": 0.8519208801779595, + "flos": 24098677962240.0, + "grad_norm": 1.9966915303031785, + "learning_rate": 2.2550896894209215e-07, + "loss": 0.7203, + "num_input_tokens_seen": 153238485, + "step": 7085 + }, + { + "epoch": 0.8520411230685986, + "flos": 62945659486080.0, + "grad_norm": 0.6891911470900138, + "learning_rate": 2.2514976586229184e-07, + "loss": 0.5656, + "num_input_tokens_seen": 153306430, + "step": 7086 + }, + { + "epoch": 0.8521613659592376, + "flos": 65742939564480.0, + "grad_norm": 0.7550428450345249, + "learning_rate": 2.247908320258609e-07, + "loss": 0.5451, + "num_input_tokens_seen": 153366520, + "step": 7087 + }, + { + "epoch": 0.8522816088498768, + "flos": 23079285432480.0, + "grad_norm": 2.3840350907450603, + "learning_rate": 2.2443216748724914e-07, + "loss": 0.7965, + "num_input_tokens_seen": 153387230, + "step": 7088 + }, + { + "epoch": 0.8524018517405159, + "flos": 31712466151200.0, + "grad_norm": 17.61074663995814, + "learning_rate": 2.2407377230086588e-07, + "loss": 0.7428, + "num_input_tokens_seen": 153409585, + "step": 7089 + }, + { + "epoch": 0.8525220946311549, + "flos": 18663456064800.0, + "grad_norm": 2.514524989384916, + "learning_rate": 2.23715646521079e-07, + "loss": 0.8319, + "num_input_tokens_seen": 153427105, + "step": 7090 + }, + { + "epoch": 0.852642337521794, + "flos": 21761894237760.0, + "grad_norm": 2.4256451210734538, + "learning_rate": 2.2335779020221724e-07, + "loss": 0.8456, + "num_input_tokens_seen": 153443725, + "step": 7091 + }, + { + "epoch": 0.8527625804124331, + "flos": 68941270083360.0, + "grad_norm": 0.8070936431843339, + "learning_rate": 2.2300020339856497e-07, + "loss": 0.5632, + "num_input_tokens_seen": 153506410, + "step": 7092 + }, + { + "epoch": 0.8528828233030722, + "flos": 26940716165280.0, + "grad_norm": 2.756793993077788, + "learning_rate": 2.2264288616436966e-07, + "loss": 0.7755, + "num_input_tokens_seen": 153526665, + "step": 7093 + }, + { + "epoch": 0.8530030661937112, + "flos": 17462534928960.0, + "grad_norm": 2.5697181269516007, + "learning_rate": 2.222858385538351e-07, + "loss": 0.7236, + "num_input_tokens_seen": 153543465, + "step": 7094 + }, + { + "epoch": 0.8531233090843504, + "flos": 22129040123520.0, + "grad_norm": 2.5125074320061684, + "learning_rate": 2.2192906062112527e-07, + "loss": 0.6826, + "num_input_tokens_seen": 153563340, + "step": 7095 + }, + { + "epoch": 0.8532435519749895, + "flos": 37581267762720.0, + "grad_norm": 2.102028947422176, + "learning_rate": 2.2157255242036377e-07, + "loss": 0.7087, + "num_input_tokens_seen": 153587005, + "step": 7096 + }, + { + "epoch": 0.8533637948656285, + "flos": 21367645863360.0, + "grad_norm": 2.1792855934771405, + "learning_rate": 2.2121631400563135e-07, + "loss": 0.7489, + "num_input_tokens_seen": 153606835, + "step": 7097 + }, + { + "epoch": 0.8534840377562677, + "flos": 53268888723840.0, + "grad_norm": 0.7702555853998072, + "learning_rate": 2.208603454309701e-07, + "loss": 0.5287, + "num_input_tokens_seen": 153664925, + "step": 7098 + }, + { + "epoch": 0.8536042806469067, + "flos": 20784472100640.0, + "grad_norm": 2.274463540010353, + "learning_rate": 2.2050464675037994e-07, + "loss": 0.7116, + "num_input_tokens_seen": 153683550, + "step": 7099 + }, + { + "epoch": 0.8537245235375458, + "flos": 24646051512480.0, + "grad_norm": 2.368675242832132, + "learning_rate": 2.2014921801782016e-07, + "loss": 0.7311, + "num_input_tokens_seen": 153703040, + "step": 7100 + }, + { + "epoch": 0.853844766428185, + "flos": 24349762656480.0, + "grad_norm": 2.391709723622223, + "learning_rate": 2.1979405928720872e-07, + "loss": 0.7398, + "num_input_tokens_seen": 153722695, + "step": 7101 + }, + { + "epoch": 0.853965009318824, + "flos": 20925665783520.0, + "grad_norm": 1.974770257359371, + "learning_rate": 2.1943917061242257e-07, + "loss": 0.7946, + "num_input_tokens_seen": 153742550, + "step": 7102 + }, + { + "epoch": 0.8540852522094631, + "flos": 24167007448320.0, + "grad_norm": 2.952941081460755, + "learning_rate": 2.1908455204729903e-07, + "loss": 0.6648, + "num_input_tokens_seen": 153761700, + "step": 7103 + }, + { + "epoch": 0.8542054951001022, + "flos": 25047027613440.0, + "grad_norm": 2.686842993336884, + "learning_rate": 2.1873020364563265e-07, + "loss": 0.7872, + "num_input_tokens_seen": 153780765, + "step": 7104 + }, + { + "epoch": 0.8543257379907413, + "flos": 24281767698240.0, + "grad_norm": 3.7936284506115068, + "learning_rate": 2.183761254611789e-07, + "loss": 0.7648, + "num_input_tokens_seen": 153801090, + "step": 7105 + }, + { + "epoch": 0.8544459808813804, + "flos": 55206523399200.0, + "grad_norm": 3.499279765546544, + "learning_rate": 2.1802231754764987e-07, + "loss": 0.7042, + "num_input_tokens_seen": 153826530, + "step": 7106 + }, + { + "epoch": 0.8545662237720195, + "flos": 25739348992320.0, + "grad_norm": 2.372916261194567, + "learning_rate": 2.17668779958718e-07, + "loss": 0.7636, + "num_input_tokens_seen": 153849110, + "step": 7107 + }, + { + "epoch": 0.8546864666626586, + "flos": 11092753361280.0, + "grad_norm": 3.4612477596188915, + "learning_rate": 2.1731551274801553e-07, + "loss": 0.8069, + "num_input_tokens_seen": 153865550, + "step": 7108 + }, + { + "epoch": 0.8548067095532976, + "flos": 25483618078080.0, + "grad_norm": 12.508130900460182, + "learning_rate": 2.169625159691324e-07, + "loss": 0.6148, + "num_input_tokens_seen": 153885425, + "step": 7109 + }, + { + "epoch": 0.8549269524439368, + "flos": 24682483610880.0, + "grad_norm": 3.870729985549588, + "learning_rate": 2.1660978967561784e-07, + "loss": 0.7448, + "num_input_tokens_seen": 153904760, + "step": 7110 + }, + { + "epoch": 0.8550471953345758, + "flos": 19797237146880.0, + "grad_norm": 2.6871586072681373, + "learning_rate": 2.1625733392098035e-07, + "loss": 0.7934, + "num_input_tokens_seen": 153920370, + "step": 7111 + }, + { + "epoch": 0.8551674382252149, + "flos": 22787828645280.0, + "grad_norm": 3.134760148943815, + "learning_rate": 2.159051487586867e-07, + "loss": 0.7959, + "num_input_tokens_seen": 153940500, + "step": 7112 + }, + { + "epoch": 0.8552876811158541, + "flos": 20601828401760.0, + "grad_norm": 5.658003224484876, + "learning_rate": 2.155532342421642e-07, + "loss": 0.728, + "num_input_tokens_seen": 153958500, + "step": 7113 + }, + { + "epoch": 0.8554079240064931, + "flos": 23079322602240.0, + "grad_norm": 2.277149853050315, + "learning_rate": 2.1520159042479636e-07, + "loss": 0.7843, + "num_input_tokens_seen": 153976790, + "step": 7114 + }, + { + "epoch": 0.8555281668971322, + "flos": 22096473680160.0, + "grad_norm": 13.213922868261317, + "learning_rate": 2.148502173599287e-07, + "loss": 0.7139, + "num_input_tokens_seen": 153994930, + "step": 7115 + }, + { + "epoch": 0.8556484097877713, + "flos": 31094309910720.0, + "grad_norm": 2.0083087946024483, + "learning_rate": 2.1449911510086372e-07, + "loss": 0.66, + "num_input_tokens_seen": 154014400, + "step": 7116 + }, + { + "epoch": 0.8557686526784104, + "flos": 24281767698240.0, + "grad_norm": 2.4372284566796005, + "learning_rate": 2.141482837008628e-07, + "loss": 0.7714, + "num_input_tokens_seen": 154034940, + "step": 7117 + }, + { + "epoch": 0.8558888955690495, + "flos": 17687557887840.0, + "grad_norm": 2.785033663369891, + "learning_rate": 2.1379772321314826e-07, + "loss": 0.7224, + "num_input_tokens_seen": 154052985, + "step": 7118 + }, + { + "epoch": 0.8560091384596886, + "flos": 19144395786720.0, + "grad_norm": 3.2686804390210686, + "learning_rate": 2.1344743369089802e-07, + "loss": 0.8164, + "num_input_tokens_seen": 154068765, + "step": 7119 + }, + { + "epoch": 0.8561293813503277, + "flos": 23880642918240.0, + "grad_norm": 2.148448486300035, + "learning_rate": 2.130974151872522e-07, + "loss": 0.8222, + "num_input_tokens_seen": 154089100, + "step": 7120 + }, + { + "epoch": 0.8562496242409667, + "flos": 22497040913760.0, + "grad_norm": 2.1556582537281868, + "learning_rate": 2.1274766775530773e-07, + "loss": 0.7883, + "num_input_tokens_seen": 154108965, + "step": 7121 + }, + { + "epoch": 0.8563698671316058, + "flos": 14691651113760.0, + "grad_norm": 3.8510362989150755, + "learning_rate": 2.1239819144812077e-07, + "loss": 0.8002, + "num_input_tokens_seen": 154123425, + "step": 7122 + }, + { + "epoch": 0.856490110022245, + "flos": 39111415895520.0, + "grad_norm": 2.267332604232143, + "learning_rate": 2.1204898631870716e-07, + "loss": 0.7022, + "num_input_tokens_seen": 154148315, + "step": 7123 + }, + { + "epoch": 0.856610352912884, + "flos": 29017643132160.0, + "grad_norm": 3.8129544847094388, + "learning_rate": 2.1170005242004006e-07, + "loss": 0.762, + "num_input_tokens_seen": 154169665, + "step": 7124 + }, + { + "epoch": 0.8567305958035231, + "flos": 23844173650080.0, + "grad_norm": 2.467009304924385, + "learning_rate": 2.1135138980505384e-07, + "loss": 0.7801, + "num_input_tokens_seen": 154190335, + "step": 7125 + }, + { + "epoch": 0.8568508386941622, + "flos": 22168408632960.0, + "grad_norm": 2.027647962145385, + "learning_rate": 2.110029985266395e-07, + "loss": 0.7205, + "num_input_tokens_seen": 154210040, + "step": 7126 + }, + { + "epoch": 0.8569710815848013, + "flos": 17282790471360.0, + "grad_norm": 4.254942768577174, + "learning_rate": 2.1065487863764787e-07, + "loss": 0.7404, + "num_input_tokens_seen": 154228385, + "step": 7127 + }, + { + "epoch": 0.8570913244754403, + "flos": 23698333747200.0, + "grad_norm": 3.2514357696576113, + "learning_rate": 2.1030703019088846e-07, + "loss": 0.8585, + "num_input_tokens_seen": 154249015, + "step": 7128 + }, + { + "epoch": 0.8572115673660795, + "flos": 20020215768960.0, + "grad_norm": 2.458097532784729, + "learning_rate": 2.099594532391291e-07, + "loss": 0.7117, + "num_input_tokens_seen": 154267700, + "step": 7129 + }, + { + "epoch": 0.8573318102567186, + "flos": 27121389866880.0, + "grad_norm": 2.3191160463197193, + "learning_rate": 2.0961214783509806e-07, + "loss": 0.7869, + "num_input_tokens_seen": 154290580, + "step": 7130 + }, + { + "epoch": 0.8574520531473576, + "flos": 24900890352480.0, + "grad_norm": 2.289366405740655, + "learning_rate": 2.0926511403148051e-07, + "loss": 0.7504, + "num_input_tokens_seen": 154309545, + "step": 7131 + }, + { + "epoch": 0.8575722960379968, + "flos": 18744460439040.0, + "grad_norm": 2.4379950778651756, + "learning_rate": 2.0891835188092143e-07, + "loss": 0.7609, + "num_input_tokens_seen": 154326545, + "step": 7132 + }, + { + "epoch": 0.8576925389286358, + "flos": 22168594481760.0, + "grad_norm": 4.941539994793283, + "learning_rate": 2.0857186143602434e-07, + "loss": 0.8154, + "num_input_tokens_seen": 154345190, + "step": 7133 + }, + { + "epoch": 0.8578127818192749, + "flos": 22861659255840.0, + "grad_norm": 2.4714679484140274, + "learning_rate": 2.0822564274935094e-07, + "loss": 0.6748, + "num_input_tokens_seen": 154364615, + "step": 7134 + }, + { + "epoch": 0.8579330247099141, + "flos": 34774769583840.0, + "grad_norm": 7.636918056062899, + "learning_rate": 2.078796958734239e-07, + "loss": 0.6667, + "num_input_tokens_seen": 154387335, + "step": 7135 + }, + { + "epoch": 0.8580532676005531, + "flos": 19727532379680.0, + "grad_norm": 2.604654210546681, + "learning_rate": 2.0753402086072124e-07, + "loss": 0.7518, + "num_input_tokens_seen": 154405965, + "step": 7136 + }, + { + "epoch": 0.8581735104911922, + "flos": 22706526912960.0, + "grad_norm": 4.241595766165688, + "learning_rate": 2.071886177636828e-07, + "loss": 0.7576, + "num_input_tokens_seen": 154422750, + "step": 7137 + }, + { + "epoch": 0.8582937533818313, + "flos": 23116052058720.0, + "grad_norm": 2.264341951319116, + "learning_rate": 2.0684348663470575e-07, + "loss": 0.8331, + "num_input_tokens_seen": 154442930, + "step": 7138 + }, + { + "epoch": 0.8584139962724704, + "flos": 19470835051680.0, + "grad_norm": 2.7295480029791794, + "learning_rate": 2.0649862752614555e-07, + "loss": 0.6178, + "num_input_tokens_seen": 154461640, + "step": 7139 + }, + { + "epoch": 0.8585342391631094, + "flos": 71174853353280.0, + "grad_norm": 0.7579685325834568, + "learning_rate": 2.0615404049031838e-07, + "loss": 0.5694, + "num_input_tokens_seen": 154518610, + "step": 7140 + }, + { + "epoch": 0.8586544820537486, + "flos": 10800255820800.0, + "grad_norm": 3.5714582278518976, + "learning_rate": 2.0580972557949616e-07, + "loss": 0.7806, + "num_input_tokens_seen": 154533700, + "step": 7141 + }, + { + "epoch": 0.8587747249443877, + "flos": 64718603456640.0, + "grad_norm": 0.8042568795262997, + "learning_rate": 2.054656828459125e-07, + "loss": 0.5414, + "num_input_tokens_seen": 154597810, + "step": 7142 + }, + { + "epoch": 0.8588949678350267, + "flos": 26796177204000.0, + "grad_norm": 2.1220943556651064, + "learning_rate": 2.051219123417578e-07, + "loss": 0.7731, + "num_input_tokens_seen": 154617870, + "step": 7143 + }, + { + "epoch": 0.8590152107256659, + "flos": 26066866180320.0, + "grad_norm": 2.6598728381860335, + "learning_rate": 2.0477841411918196e-07, + "loss": 0.602, + "num_input_tokens_seen": 154637395, + "step": 7144 + }, + { + "epoch": 0.859135453616305, + "flos": 26103446957760.0, + "grad_norm": 3.0418829296691383, + "learning_rate": 2.0443518823029326e-07, + "loss": 0.7444, + "num_input_tokens_seen": 154657935, + "step": 7145 + }, + { + "epoch": 0.859255696506944, + "flos": 12951050568000.0, + "grad_norm": 4.48831715913888, + "learning_rate": 2.0409223472715854e-07, + "loss": 0.7678, + "num_input_tokens_seen": 154675080, + "step": 7146 + }, + { + "epoch": 0.8593759393975832, + "flos": 18449361015360.0, + "grad_norm": 2.188021897319731, + "learning_rate": 2.0374955366180434e-07, + "loss": 0.7466, + "num_input_tokens_seen": 154692720, + "step": 7147 + }, + { + "epoch": 0.8594961822882222, + "flos": 22168445802720.0, + "grad_norm": 2.3362159180336537, + "learning_rate": 2.034071450862147e-07, + "loss": 0.7257, + "num_input_tokens_seen": 154710190, + "step": 7148 + }, + { + "epoch": 0.8596164251788613, + "flos": 23261445924480.0, + "grad_norm": 2.0248416461958674, + "learning_rate": 2.030650090523327e-07, + "loss": 0.7694, + "num_input_tokens_seen": 154729380, + "step": 7149 + }, + { + "epoch": 0.8597366680695004, + "flos": 31604433627840.0, + "grad_norm": 2.470484064214065, + "learning_rate": 2.0272314561205995e-07, + "loss": 0.5952, + "num_input_tokens_seen": 154749775, + "step": 7150 + }, + { + "epoch": 0.8598569109601395, + "flos": 21257345984640.0, + "grad_norm": 2.8023381076029583, + "learning_rate": 2.023815548172567e-07, + "loss": 0.7324, + "num_input_tokens_seen": 154769845, + "step": 7151 + }, + { + "epoch": 0.8599771538507786, + "flos": 25411385767200.0, + "grad_norm": 2.0925332704585786, + "learning_rate": 2.0204023671974267e-07, + "loss": 0.6608, + "num_input_tokens_seen": 154789230, + "step": 7152 + }, + { + "epoch": 0.8600973967414177, + "flos": 16700099915520.0, + "grad_norm": 2.886027045479437, + "learning_rate": 2.0169919137129532e-07, + "loss": 0.813, + "num_input_tokens_seen": 154807670, + "step": 7153 + }, + { + "epoch": 0.8602176396320568, + "flos": 25191789593280.0, + "grad_norm": 3.1508735716990524, + "learning_rate": 2.013584188236508e-07, + "loss": 0.7066, + "num_input_tokens_seen": 154825575, + "step": 7154 + }, + { + "epoch": 0.8603378825226958, + "flos": 20383124302080.0, + "grad_norm": 3.156497303555946, + "learning_rate": 2.0101791912850396e-07, + "loss": 0.8007, + "num_input_tokens_seen": 154845115, + "step": 7155 + }, + { + "epoch": 0.8604581254133349, + "flos": 34879828526400.0, + "grad_norm": 8.202606940486493, + "learning_rate": 2.006776923375082e-07, + "loss": 0.6394, + "num_input_tokens_seen": 154866770, + "step": 7156 + }, + { + "epoch": 0.860578368303974, + "flos": 22564069458240.0, + "grad_norm": 2.72966749143555, + "learning_rate": 2.003377385022764e-07, + "loss": 0.7134, + "num_input_tokens_seen": 154885595, + "step": 7157 + }, + { + "epoch": 0.8606986111946131, + "flos": 21293889592320.0, + "grad_norm": 2.4585411393434105, + "learning_rate": 1.9999805767437826e-07, + "loss": 0.7733, + "num_input_tokens_seen": 154904485, + "step": 7158 + }, + { + "epoch": 0.8608188540852522, + "flos": 28835779998240.0, + "grad_norm": 2.017537237530395, + "learning_rate": 1.9965864990534386e-07, + "loss": 0.718, + "num_input_tokens_seen": 154925010, + "step": 7159 + }, + { + "epoch": 0.8609390969758913, + "flos": 29673532412640.0, + "grad_norm": 2.2299597805826807, + "learning_rate": 1.9931951524666092e-07, + "loss": 0.7774, + "num_input_tokens_seen": 154946100, + "step": 7160 + }, + { + "epoch": 0.8610593398665304, + "flos": 21218943888960.0, + "grad_norm": 2.0507765996971083, + "learning_rate": 1.9898065374977534e-07, + "loss": 0.8119, + "num_input_tokens_seen": 154966305, + "step": 7161 + }, + { + "epoch": 0.8611795827571694, + "flos": 14809124756160.0, + "grad_norm": 2.4847985632323013, + "learning_rate": 1.9864206546609342e-07, + "loss": 0.7308, + "num_input_tokens_seen": 154985145, + "step": 7162 + }, + { + "epoch": 0.8612998256478086, + "flos": 24208568973600.0, + "grad_norm": 2.667703417071294, + "learning_rate": 1.983037504469771e-07, + "loss": 0.8406, + "num_input_tokens_seen": 155003855, + "step": 7163 + }, + { + "epoch": 0.8614200685384477, + "flos": 21222363506880.0, + "grad_norm": 2.8497469204081924, + "learning_rate": 1.9796570874374984e-07, + "loss": 0.6697, + "num_input_tokens_seen": 155023960, + "step": 7164 + }, + { + "epoch": 0.8615403114290867, + "flos": 19979100280800.0, + "grad_norm": 2.9004138059364055, + "learning_rate": 1.976279404076917e-07, + "loss": 0.7787, + "num_input_tokens_seen": 155043360, + "step": 7165 + }, + { + "epoch": 0.8616605543197259, + "flos": 29746619628000.0, + "grad_norm": 3.412536699146613, + "learning_rate": 1.9729044549004193e-07, + "loss": 0.7581, + "num_input_tokens_seen": 155064745, + "step": 7166 + }, + { + "epoch": 0.8617807972103649, + "flos": 28870056250560.0, + "grad_norm": 2.316964145012538, + "learning_rate": 1.9695322404199822e-07, + "loss": 0.7057, + "num_input_tokens_seen": 155086790, + "step": 7167 + }, + { + "epoch": 0.861901040101004, + "flos": 27634152636960.0, + "grad_norm": 2.8265554382674623, + "learning_rate": 1.9661627611471654e-07, + "loss": 0.8248, + "num_input_tokens_seen": 155106585, + "step": 7168 + }, + { + "epoch": 0.8620212829916432, + "flos": 49676278394400.0, + "grad_norm": 2.613272491577251, + "learning_rate": 1.9627960175931246e-07, + "loss": 0.703, + "num_input_tokens_seen": 155131285, + "step": 7169 + }, + { + "epoch": 0.8621415258822822, + "flos": 21108383821920.0, + "grad_norm": 2.411394045627162, + "learning_rate": 1.9594320102685847e-07, + "loss": 0.745, + "num_input_tokens_seen": 155150555, + "step": 7170 + }, + { + "epoch": 0.8622617687729213, + "flos": 21658061897280.0, + "grad_norm": 2.8661685141132134, + "learning_rate": 1.956070739683864e-07, + "loss": 0.6411, + "num_input_tokens_seen": 155169890, + "step": 7171 + }, + { + "epoch": 0.8623820116635604, + "flos": 26213003441280.0, + "grad_norm": 16.843846931249388, + "learning_rate": 1.9527122063488678e-07, + "loss": 0.7445, + "num_input_tokens_seen": 155191005, + "step": 7172 + }, + { + "epoch": 0.8625022545541995, + "flos": 19619611365600.0, + "grad_norm": 1.8474510030482378, + "learning_rate": 1.9493564107730755e-07, + "loss": 0.8048, + "num_input_tokens_seen": 155211005, + "step": 7173 + }, + { + "epoch": 0.8626224974448385, + "flos": 21877100524800.0, + "grad_norm": 4.251362301768509, + "learning_rate": 1.9460033534655684e-07, + "loss": 0.6094, + "num_input_tokens_seen": 155230365, + "step": 7174 + }, + { + "epoch": 0.8627427403354777, + "flos": 23297840853120.0, + "grad_norm": 2.429051846581824, + "learning_rate": 1.9426530349349978e-07, + "loss": 0.8432, + "num_input_tokens_seen": 155250815, + "step": 7175 + }, + { + "epoch": 0.8628629832261168, + "flos": 16339681756320.0, + "grad_norm": 2.4574857209606704, + "learning_rate": 1.9393054556896038e-07, + "loss": 0.6481, + "num_input_tokens_seen": 155268515, + "step": 7176 + }, + { + "epoch": 0.8629832261167558, + "flos": 28063383489120.0, + "grad_norm": 5.264922596071756, + "learning_rate": 1.9359606162372133e-07, + "loss": 0.6911, + "num_input_tokens_seen": 155290120, + "step": 7177 + }, + { + "epoch": 0.863103469007395, + "flos": 20201223998400.0, + "grad_norm": 1.982202015979532, + "learning_rate": 1.9326185170852293e-07, + "loss": 0.7083, + "num_input_tokens_seen": 155309085, + "step": 7178 + }, + { + "epoch": 0.863223711898034, + "flos": 24463482153120.0, + "grad_norm": 2.316841380531681, + "learning_rate": 1.9292791587406598e-07, + "loss": 0.7244, + "num_input_tokens_seen": 155327945, + "step": 7179 + }, + { + "epoch": 0.8633439547886731, + "flos": 17650233715200.0, + "grad_norm": 9.318416891917963, + "learning_rate": 1.9259425417100661e-07, + "loss": 0.8712, + "num_input_tokens_seen": 155344060, + "step": 7180 + }, + { + "epoch": 0.8634641976793123, + "flos": 12876848259840.0, + "grad_norm": 2.981279455395302, + "learning_rate": 1.9226086664996234e-07, + "loss": 0.7601, + "num_input_tokens_seen": 155362305, + "step": 7181 + }, + { + "epoch": 0.8635844405699513, + "flos": 23844842705760.0, + "grad_norm": 2.7397637156774888, + "learning_rate": 1.9192775336150712e-07, + "loss": 0.7431, + "num_input_tokens_seen": 155382605, + "step": 7182 + }, + { + "epoch": 0.8637046834605904, + "flos": 60366568864320.0, + "grad_norm": 0.7881027276615239, + "learning_rate": 1.915949143561739e-07, + "loss": 0.5617, + "num_input_tokens_seen": 155437280, + "step": 7183 + }, + { + "epoch": 0.8638249263512295, + "flos": 20529447411840.0, + "grad_norm": 2.1049404295332312, + "learning_rate": 1.9126234968445498e-07, + "loss": 0.7807, + "num_input_tokens_seen": 155456970, + "step": 7184 + }, + { + "epoch": 0.8639451692418686, + "flos": 26577658953120.0, + "grad_norm": 1.8605164942882928, + "learning_rate": 1.9093005939679884e-07, + "loss": 0.6799, + "num_input_tokens_seen": 155478925, + "step": 7185 + }, + { + "epoch": 0.8640654121325076, + "flos": 15101064750240.0, + "grad_norm": 2.2104508905760465, + "learning_rate": 1.9059804354361452e-07, + "loss": 0.7661, + "num_input_tokens_seen": 155496700, + "step": 7186 + }, + { + "epoch": 0.8641856550231467, + "flos": 31823286406560.0, + "grad_norm": 2.0640439121469174, + "learning_rate": 1.902663021752684e-07, + "loss": 0.7058, + "num_input_tokens_seen": 155518130, + "step": 7187 + }, + { + "epoch": 0.8643058979137859, + "flos": 14955373526400.0, + "grad_norm": 2.5972798239655024, + "learning_rate": 1.8993483534208556e-07, + "loss": 0.8264, + "num_input_tokens_seen": 155537040, + "step": 7188 + }, + { + "epoch": 0.8644261408044249, + "flos": 13097262168480.0, + "grad_norm": 2.505816559637636, + "learning_rate": 1.8960364309434884e-07, + "loss": 0.7506, + "num_input_tokens_seen": 155554535, + "step": 7189 + }, + { + "epoch": 0.864546383695064, + "flos": 20821201557120.0, + "grad_norm": 2.1681294149715087, + "learning_rate": 1.8927272548229967e-07, + "loss": 0.784, + "num_input_tokens_seen": 155574225, + "step": 7190 + }, + { + "epoch": 0.8646666265857031, + "flos": 21294149780640.0, + "grad_norm": 1.851705618047208, + "learning_rate": 1.8894208255613876e-07, + "loss": 0.8334, + "num_input_tokens_seen": 155593540, + "step": 7191 + }, + { + "epoch": 0.8647868694763422, + "flos": 19722514462080.0, + "grad_norm": 3.291531539911909, + "learning_rate": 1.8861171436602397e-07, + "loss": 0.7833, + "num_input_tokens_seen": 155610655, + "step": 7192 + }, + { + "epoch": 0.8649071123669813, + "flos": 26139247170240.0, + "grad_norm": 2.9119525051016497, + "learning_rate": 1.882816209620719e-07, + "loss": 0.8062, + "num_input_tokens_seen": 155627365, + "step": 7193 + }, + { + "epoch": 0.8650273552576204, + "flos": 20273939516160.0, + "grad_norm": 2.1085998241086905, + "learning_rate": 1.8795180239435738e-07, + "loss": 0.7706, + "num_input_tokens_seen": 155646220, + "step": 7194 + }, + { + "epoch": 0.8651475981482595, + "flos": 23916814828320.0, + "grad_norm": 3.7607273526469935, + "learning_rate": 1.8762225871291348e-07, + "loss": 0.7584, + "num_input_tokens_seen": 155662095, + "step": 7195 + }, + { + "epoch": 0.8652678410388985, + "flos": 21653490016800.0, + "grad_norm": 2.407436227082554, + "learning_rate": 1.8729298996773201e-07, + "loss": 0.8094, + "num_input_tokens_seen": 155680845, + "step": 7196 + }, + { + "epoch": 0.8653880839295377, + "flos": 65131919917920.0, + "grad_norm": 0.848535951413592, + "learning_rate": 1.8696399620876301e-07, + "loss": 0.608, + "num_input_tokens_seen": 155737875, + "step": 7197 + }, + { + "epoch": 0.8655083268201768, + "flos": 17723952816480.0, + "grad_norm": 3.859937372045469, + "learning_rate": 1.866352774859141e-07, + "loss": 0.7922, + "num_input_tokens_seen": 155753100, + "step": 7198 + }, + { + "epoch": 0.8656285697108158, + "flos": 20674432410240.0, + "grad_norm": 3.0271030650291038, + "learning_rate": 1.8630683384905188e-07, + "loss": 0.6946, + "num_input_tokens_seen": 155772430, + "step": 7199 + }, + { + "epoch": 0.865748812601455, + "flos": 18627023966400.0, + "grad_norm": 2.5472919146276967, + "learning_rate": 1.8597866534800045e-07, + "loss": 0.8884, + "num_input_tokens_seen": 155787545, + "step": 7200 + }, + { + "epoch": 0.865869055492094, + "flos": 70550272477920.0, + "grad_norm": 2.3685974353787027, + "learning_rate": 1.8565077203254398e-07, + "loss": 0.7429, + "num_input_tokens_seen": 155813005, + "step": 7201 + }, + { + "epoch": 0.8659892983827331, + "flos": 17359260134880.0, + "grad_norm": 3.6325233262344243, + "learning_rate": 1.8532315395242203e-07, + "loss": 0.727, + "num_input_tokens_seen": 155829455, + "step": 7202 + }, + { + "epoch": 0.8661095412733723, + "flos": 17869792719360.0, + "grad_norm": 5.714633815244332, + "learning_rate": 1.849958111573353e-07, + "loss": 0.7245, + "num_input_tokens_seen": 155848060, + "step": 7203 + }, + { + "epoch": 0.8662297841640113, + "flos": 18198276321120.0, + "grad_norm": 2.6695701071590823, + "learning_rate": 1.8466874369694074e-07, + "loss": 0.6447, + "num_input_tokens_seen": 155865755, + "step": 7204 + }, + { + "epoch": 0.8663500270546504, + "flos": 16338938361120.0, + "grad_norm": 3.082027846263171, + "learning_rate": 1.843419516208542e-07, + "loss": 0.6974, + "num_input_tokens_seen": 155882680, + "step": 7205 + }, + { + "epoch": 0.8664702699452895, + "flos": 17868974984640.0, + "grad_norm": 3.183426891236955, + "learning_rate": 1.8401543497865047e-07, + "loss": 0.7963, + "num_input_tokens_seen": 155900680, + "step": 7206 + }, + { + "epoch": 0.8665905128359286, + "flos": 30692181547200.0, + "grad_norm": 2.556485631104438, + "learning_rate": 1.836891938198608e-07, + "loss": 0.6388, + "num_input_tokens_seen": 155921105, + "step": 7207 + }, + { + "epoch": 0.8667107557265676, + "flos": 18629700189120.0, + "grad_norm": 3.481604188608357, + "learning_rate": 1.8336322819397677e-07, + "loss": 0.713, + "num_input_tokens_seen": 155938640, + "step": 7208 + }, + { + "epoch": 0.8668309986172068, + "flos": 20055086737440.0, + "grad_norm": 2.758773506593311, + "learning_rate": 1.8303753815044654e-07, + "loss": 0.6289, + "num_input_tokens_seen": 155957945, + "step": 7209 + }, + { + "epoch": 0.8669512415078459, + "flos": 21584194116960.0, + "grad_norm": 5.6828423177017084, + "learning_rate": 1.827121237386773e-07, + "loss": 0.7082, + "num_input_tokens_seen": 155975390, + "step": 7210 + }, + { + "epoch": 0.8670714843984849, + "flos": 17678116768800.0, + "grad_norm": 3.5720180052317354, + "learning_rate": 1.8238698500803374e-07, + "loss": 0.7523, + "num_input_tokens_seen": 155988155, + "step": 7211 + }, + { + "epoch": 0.8671917272891241, + "flos": 60619506312960.0, + "grad_norm": 0.729901833269812, + "learning_rate": 1.820621220078391e-07, + "loss": 0.5616, + "num_input_tokens_seen": 156052065, + "step": 7212 + }, + { + "epoch": 0.8673119701797631, + "flos": 20422827339360.0, + "grad_norm": 2.1397127309843427, + "learning_rate": 1.8173753478737553e-07, + "loss": 0.6795, + "num_input_tokens_seen": 156072500, + "step": 7213 + }, + { + "epoch": 0.8674322130704022, + "flos": 19619462686560.0, + "grad_norm": 2.302752183322785, + "learning_rate": 1.8141322339588205e-07, + "loss": 0.7975, + "num_input_tokens_seen": 156089840, + "step": 7214 + }, + { + "epoch": 0.8675524559610414, + "flos": 25987497275520.0, + "grad_norm": 4.211798924818879, + "learning_rate": 1.810891878825569e-07, + "loss": 0.7029, + "num_input_tokens_seen": 156109815, + "step": 7215 + }, + { + "epoch": 0.8676726988516804, + "flos": 15028163383680.0, + "grad_norm": 5.202308751137622, + "learning_rate": 1.8076542829655561e-07, + "loss": 0.7178, + "num_input_tokens_seen": 156128620, + "step": 7216 + }, + { + "epoch": 0.8677929417423195, + "flos": 16264959071520.0, + "grad_norm": 2.55515231415688, + "learning_rate": 1.8044194468699203e-07, + "loss": 0.7936, + "num_input_tokens_seen": 156145930, + "step": 7217 + }, + { + "epoch": 0.8679131846329585, + "flos": 18817361805600.0, + "grad_norm": 3.6519297319234854, + "learning_rate": 1.8011873710293912e-07, + "loss": 0.761, + "num_input_tokens_seen": 156164435, + "step": 7218 + }, + { + "epoch": 0.8680334275235977, + "flos": 33572398827360.0, + "grad_norm": 4.699000986940581, + "learning_rate": 1.7979580559342677e-07, + "loss": 0.6975, + "num_input_tokens_seen": 156185165, + "step": 7219 + }, + { + "epoch": 0.8681536704142367, + "flos": 24645828493920.0, + "grad_norm": 7.023716395933, + "learning_rate": 1.7947315020744358e-07, + "loss": 0.667, + "num_input_tokens_seen": 156206260, + "step": 7220 + }, + { + "epoch": 0.8682739133048758, + "flos": 19982519898720.0, + "grad_norm": 2.3780213261988306, + "learning_rate": 1.7915077099393594e-07, + "loss": 0.802, + "num_input_tokens_seen": 156222860, + "step": 7221 + }, + { + "epoch": 0.868394156195515, + "flos": 16631138543520.0, + "grad_norm": 3.839061575902285, + "learning_rate": 1.788286680018083e-07, + "loss": 0.7348, + "num_input_tokens_seen": 156241570, + "step": 7222 + }, + { + "epoch": 0.868514399086154, + "flos": 27961223787840.0, + "grad_norm": 2.0265144901279957, + "learning_rate": 1.7850684127992443e-07, + "loss": 0.7247, + "num_input_tokens_seen": 156261315, + "step": 7223 + }, + { + "epoch": 0.8686346419767931, + "flos": 20055978811680.0, + "grad_norm": 3.310879292860815, + "learning_rate": 1.7818529087710378e-07, + "loss": 0.7054, + "num_input_tokens_seen": 156281670, + "step": 7224 + }, + { + "epoch": 0.8687548848674322, + "flos": 18197607265440.0, + "grad_norm": 2.4447356204768482, + "learning_rate": 1.7786401684212637e-07, + "loss": 0.8412, + "num_input_tokens_seen": 156300500, + "step": 7225 + }, + { + "epoch": 0.8688751277580713, + "flos": 70356769475520.0, + "grad_norm": 0.7427142231740231, + "learning_rate": 1.7754301922372883e-07, + "loss": 0.5579, + "num_input_tokens_seen": 156350145, + "step": 7226 + }, + { + "epoch": 0.8689953706487104, + "flos": 26868223666080.0, + "grad_norm": 2.35813568104727, + "learning_rate": 1.7722229807060617e-07, + "loss": 0.812, + "num_input_tokens_seen": 156368725, + "step": 7227 + }, + { + "epoch": 0.8691156135393495, + "flos": 34587405325440.0, + "grad_norm": 2.9782154369732687, + "learning_rate": 1.7690185343141172e-07, + "loss": 0.8199, + "num_input_tokens_seen": 156388640, + "step": 7228 + }, + { + "epoch": 0.8692358564299886, + "flos": 18962792841120.0, + "grad_norm": 2.7090458055061686, + "learning_rate": 1.7658168535475615e-07, + "loss": 0.7015, + "num_input_tokens_seen": 156406425, + "step": 7229 + }, + { + "epoch": 0.8693560993206276, + "flos": 30327154337760.0, + "grad_norm": 4.277575844516473, + "learning_rate": 1.7626179388920948e-07, + "loss": 0.6462, + "num_input_tokens_seen": 156427705, + "step": 7230 + }, + { + "epoch": 0.8694763422112668, + "flos": 27161501771520.0, + "grad_norm": 2.1816077302452364, + "learning_rate": 1.7594217908329866e-07, + "loss": 0.804, + "num_input_tokens_seen": 156449890, + "step": 7231 + }, + { + "epoch": 0.8695965851019059, + "flos": 26101662809280.0, + "grad_norm": 3.2541953664792977, + "learning_rate": 1.7562284098550895e-07, + "loss": 0.7391, + "num_input_tokens_seen": 156469600, + "step": 7232 + }, + { + "epoch": 0.8697168279925449, + "flos": 67236506919840.0, + "grad_norm": 0.8447796931674543, + "learning_rate": 1.753037796442838e-07, + "loss": 0.6207, + "num_input_tokens_seen": 156529040, + "step": 7233 + }, + { + "epoch": 0.8698370708831841, + "flos": 19690579904640.0, + "grad_norm": 4.841812829302577, + "learning_rate": 1.74984995108024e-07, + "loss": 0.7532, + "num_input_tokens_seen": 156546970, + "step": 7234 + }, + { + "epoch": 0.8699573137738231, + "flos": 12841865782080.0, + "grad_norm": 2.581229265697785, + "learning_rate": 1.7466648742508981e-07, + "loss": 0.8328, + "num_input_tokens_seen": 156563155, + "step": 7235 + }, + { + "epoch": 0.8700775566644622, + "flos": 17395989591360.0, + "grad_norm": 2.586489154715507, + "learning_rate": 1.7434825664379837e-07, + "loss": 0.846, + "num_input_tokens_seen": 156581660, + "step": 7236 + }, + { + "epoch": 0.8701977995551013, + "flos": 13715827276320.0, + "grad_norm": 3.2322143119253917, + "learning_rate": 1.740303028124246e-07, + "loss": 0.8585, + "num_input_tokens_seen": 156597720, + "step": 7237 + }, + { + "epoch": 0.8703180424457404, + "flos": 30511842373440.0, + "grad_norm": 2.610049954224977, + "learning_rate": 1.7371262597920212e-07, + "loss": 0.7576, + "num_input_tokens_seen": 156619780, + "step": 7238 + }, + { + "epoch": 0.8704382853363795, + "flos": 19581023421120.0, + "grad_norm": 1.6509432437413394, + "learning_rate": 1.7339522619232195e-07, + "loss": 0.7641, + "num_input_tokens_seen": 156638160, + "step": 7239 + }, + { + "epoch": 0.8705585282270186, + "flos": 26575726125600.0, + "grad_norm": 2.9388606390008953, + "learning_rate": 1.730781034999338e-07, + "loss": 0.7581, + "num_input_tokens_seen": 156659740, + "step": 7240 + }, + { + "epoch": 0.8706787711176577, + "flos": 34041295547040.0, + "grad_norm": 2.766725482851694, + "learning_rate": 1.7276125795014497e-07, + "loss": 0.7344, + "num_input_tokens_seen": 156678190, + "step": 7241 + }, + { + "epoch": 0.8707990140082967, + "flos": 14590866693600.0, + "grad_norm": 2.3955592675947983, + "learning_rate": 1.7244468959102054e-07, + "loss": 0.6759, + "num_input_tokens_seen": 156695245, + "step": 7242 + }, + { + "epoch": 0.8709192568989359, + "flos": 20055532774560.0, + "grad_norm": 2.96960308466784, + "learning_rate": 1.7212839847058348e-07, + "loss": 0.8514, + "num_input_tokens_seen": 156710375, + "step": 7243 + }, + { + "epoch": 0.871039499789575, + "flos": 16703333684640.0, + "grad_norm": 6.905430809906806, + "learning_rate": 1.718123846368147e-07, + "loss": 0.7369, + "num_input_tokens_seen": 156729420, + "step": 7244 + }, + { + "epoch": 0.871159742680214, + "flos": 21038716224480.0, + "grad_norm": 2.349392012818427, + "learning_rate": 1.714966481376543e-07, + "loss": 0.7185, + "num_input_tokens_seen": 156746100, + "step": 7245 + }, + { + "epoch": 0.8712799855708532, + "flos": 28215988288320.0, + "grad_norm": 3.3697743403103804, + "learning_rate": 1.7118118902099797e-07, + "loss": 0.831, + "num_input_tokens_seen": 156764185, + "step": 7246 + }, + { + "epoch": 0.8714002284614922, + "flos": 22204915070880.0, + "grad_norm": 2.1354958000097146, + "learning_rate": 1.7086600733470146e-07, + "loss": 0.8105, + "num_input_tokens_seen": 156784855, + "step": 7247 + }, + { + "epoch": 0.8715204713521313, + "flos": 21401253060000.0, + "grad_norm": 2.478500426285649, + "learning_rate": 1.7055110312657738e-07, + "loss": 0.7749, + "num_input_tokens_seen": 156804350, + "step": 7248 + }, + { + "epoch": 0.8716407142427703, + "flos": 23406914129760.0, + "grad_norm": 3.3014991476152318, + "learning_rate": 1.702364764443962e-07, + "loss": 0.7429, + "num_input_tokens_seen": 156823425, + "step": 7249 + }, + { + "epoch": 0.8717609571334095, + "flos": 27918361320960.0, + "grad_norm": 2.2930156436524616, + "learning_rate": 1.6992212733588685e-07, + "loss": 0.7269, + "num_input_tokens_seen": 156843090, + "step": 7250 + }, + { + "epoch": 0.8718812000240486, + "flos": 25442837117760.0, + "grad_norm": 2.3461953185939026, + "learning_rate": 1.6960805584873538e-07, + "loss": 0.748, + "num_input_tokens_seen": 156861090, + "step": 7251 + }, + { + "epoch": 0.8720014429146876, + "flos": 23369552787360.0, + "grad_norm": 2.4982028223532553, + "learning_rate": 1.6929426203058684e-07, + "loss": 0.785, + "num_input_tokens_seen": 156881515, + "step": 7252 + }, + { + "epoch": 0.8721216858053268, + "flos": 24317976778080.0, + "grad_norm": 2.5232668603416752, + "learning_rate": 1.689807459290431e-07, + "loss": 0.8023, + "num_input_tokens_seen": 156900170, + "step": 7253 + }, + { + "epoch": 0.8722419286959658, + "flos": 33820770129120.0, + "grad_norm": 2.5939586509520773, + "learning_rate": 1.6866750759166437e-07, + "loss": 0.7088, + "num_input_tokens_seen": 156920150, + "step": 7254 + }, + { + "epoch": 0.8723621715866049, + "flos": 18343261319520.0, + "grad_norm": 3.5496545059359272, + "learning_rate": 1.6835454706596865e-07, + "loss": 0.7709, + "num_input_tokens_seen": 156937980, + "step": 7255 + }, + { + "epoch": 0.8724824144772441, + "flos": 21981750600000.0, + "grad_norm": 2.717836248375873, + "learning_rate": 1.680418643994317e-07, + "loss": 0.7362, + "num_input_tokens_seen": 156956550, + "step": 7256 + }, + { + "epoch": 0.8726026573678831, + "flos": 66603036738240.0, + "grad_norm": 1.756704701923326, + "learning_rate": 1.6772945963948738e-07, + "loss": 0.6442, + "num_input_tokens_seen": 157014715, + "step": 7257 + }, + { + "epoch": 0.8727229002585222, + "flos": 13352063838720.0, + "grad_norm": 6.254530218521174, + "learning_rate": 1.6741733283352733e-07, + "loss": 0.7748, + "num_input_tokens_seen": 157032320, + "step": 7258 + }, + { + "epoch": 0.8728431431491613, + "flos": 21765797062560.0, + "grad_norm": 1.8239078624400895, + "learning_rate": 1.6710548402890102e-07, + "loss": 0.84, + "num_input_tokens_seen": 157052845, + "step": 7259 + }, + { + "epoch": 0.8729633860398004, + "flos": 36121753641120.0, + "grad_norm": 2.7706276857934036, + "learning_rate": 1.6679391327291527e-07, + "loss": 0.6689, + "num_input_tokens_seen": 157074050, + "step": 7260 + }, + { + "epoch": 0.8730836289304394, + "flos": 16336373647680.0, + "grad_norm": 3.2129401948113774, + "learning_rate": 1.6648262061283492e-07, + "loss": 0.6841, + "num_input_tokens_seen": 157089350, + "step": 7261 + }, + { + "epoch": 0.8732038718210786, + "flos": 21184853485440.0, + "grad_norm": 2.7136681076337745, + "learning_rate": 1.6617160609588353e-07, + "loss": 0.7352, + "num_input_tokens_seen": 157108525, + "step": 7262 + }, + { + "epoch": 0.8733241147117177, + "flos": 16586863625760.0, + "grad_norm": 3.2951337365200724, + "learning_rate": 1.6586086976924163e-07, + "loss": 0.7243, + "num_input_tokens_seen": 157124025, + "step": 7263 + }, + { + "epoch": 0.8734443576023567, + "flos": 20165163597600.0, + "grad_norm": 3.914182131951062, + "learning_rate": 1.6555041168004747e-07, + "loss": 0.7832, + "num_input_tokens_seen": 157143345, + "step": 7264 + }, + { + "epoch": 0.8735646004929959, + "flos": 18015521112960.0, + "grad_norm": 2.281816595077654, + "learning_rate": 1.6524023187539715e-07, + "loss": 0.6925, + "num_input_tokens_seen": 157161630, + "step": 7265 + }, + { + "epoch": 0.873684843383635, + "flos": 20233307234880.0, + "grad_norm": 3.601017511470879, + "learning_rate": 1.649303304023446e-07, + "loss": 0.7511, + "num_input_tokens_seen": 157181385, + "step": 7266 + }, + { + "epoch": 0.873805086274274, + "flos": 16923264386400.0, + "grad_norm": 2.107013878558002, + "learning_rate": 1.6462070730790246e-07, + "loss": 0.7922, + "num_input_tokens_seen": 157200545, + "step": 7267 + }, + { + "epoch": 0.8739253291649132, + "flos": 18015632622240.0, + "grad_norm": 3.4417354499895607, + "learning_rate": 1.6431136263903912e-07, + "loss": 0.7866, + "num_input_tokens_seen": 157219545, + "step": 7268 + }, + { + "epoch": 0.8740455720555522, + "flos": 21294670157280.0, + "grad_norm": 2.408136495270211, + "learning_rate": 1.6400229644268282e-07, + "loss": 0.7327, + "num_input_tokens_seen": 157237900, + "step": 7269 + }, + { + "epoch": 0.8741658149461913, + "flos": 15137236660320.0, + "grad_norm": 2.663216912546311, + "learning_rate": 1.6369350876571852e-07, + "loss": 0.8165, + "num_input_tokens_seen": 157256055, + "step": 7270 + }, + { + "epoch": 0.8742860578368304, + "flos": 23006941612320.0, + "grad_norm": 2.7524115652348353, + "learning_rate": 1.6338499965498874e-07, + "loss": 0.8226, + "num_input_tokens_seen": 157274785, + "step": 7271 + }, + { + "epoch": 0.8744063007274695, + "flos": 28104907844640.0, + "grad_norm": 4.173045019320588, + "learning_rate": 1.630767691572943e-07, + "loss": 0.7741, + "num_input_tokens_seen": 157294805, + "step": 7272 + }, + { + "epoch": 0.8745265436181086, + "flos": 63942229783200.0, + "grad_norm": 0.7554679862711852, + "learning_rate": 1.6276881731939306e-07, + "loss": 0.5338, + "num_input_tokens_seen": 157356695, + "step": 7273 + }, + { + "epoch": 0.8746467865087477, + "flos": 28617521935680.0, + "grad_norm": 2.1427616120796995, + "learning_rate": 1.6246114418800193e-07, + "loss": 0.7507, + "num_input_tokens_seen": 157376975, + "step": 7274 + }, + { + "epoch": 0.8747670293993868, + "flos": 23951462778240.0, + "grad_norm": 2.045370932517355, + "learning_rate": 1.6215374980979423e-07, + "loss": 0.765, + "num_input_tokens_seen": 157397385, + "step": 7275 + }, + { + "epoch": 0.8748872722900258, + "flos": 45156356497920.0, + "grad_norm": 4.395864164174995, + "learning_rate": 1.6184663423140133e-07, + "loss": 0.6888, + "num_input_tokens_seen": 157417685, + "step": 7276 + }, + { + "epoch": 0.875007515180665, + "flos": 19726082759040.0, + "grad_norm": 3.0122951643701183, + "learning_rate": 1.615397974994126e-07, + "loss": 0.643, + "num_input_tokens_seen": 157435490, + "step": 7277 + }, + { + "epoch": 0.875127758071304, + "flos": 22678272161760.0, + "grad_norm": 1.8992924903855737, + "learning_rate": 1.6123323966037438e-07, + "loss": 0.8065, + "num_input_tokens_seen": 157454010, + "step": 7278 + }, + { + "epoch": 0.8752480009619431, + "flos": 23370370522080.0, + "grad_norm": 2.743803151513047, + "learning_rate": 1.6092696076079216e-07, + "loss": 0.7906, + "num_input_tokens_seen": 157472385, + "step": 7279 + }, + { + "epoch": 0.8753682438525822, + "flos": 26175530589600.0, + "grad_norm": 2.446466143322565, + "learning_rate": 1.6062096084712785e-07, + "loss": 0.7418, + "num_input_tokens_seen": 157493735, + "step": 7280 + }, + { + "epoch": 0.8754884867432213, + "flos": 23293231802880.0, + "grad_norm": 2.4064481579324344, + "learning_rate": 1.6031523996580098e-07, + "loss": 0.7107, + "num_input_tokens_seen": 157511295, + "step": 7281 + }, + { + "epoch": 0.8756087296338604, + "flos": 12477656307360.0, + "grad_norm": 4.157032598944411, + "learning_rate": 1.6000979816318981e-07, + "loss": 0.6661, + "num_input_tokens_seen": 157529760, + "step": 7282 + }, + { + "epoch": 0.8757289725244994, + "flos": 18925914705600.0, + "grad_norm": 3.3804696584038676, + "learning_rate": 1.5970463548562886e-07, + "loss": 0.7503, + "num_input_tokens_seen": 157548915, + "step": 7283 + }, + { + "epoch": 0.8758492154151386, + "flos": 25229819991360.0, + "grad_norm": 2.026306474994103, + "learning_rate": 1.5939975197941192e-07, + "loss": 0.7117, + "num_input_tokens_seen": 157570570, + "step": 7284 + }, + { + "epoch": 0.8759694583057777, + "flos": 65477767531200.0, + "grad_norm": 0.8222403986814787, + "learning_rate": 1.5909514769078892e-07, + "loss": 0.5321, + "num_input_tokens_seen": 157635675, + "step": 7285 + }, + { + "epoch": 0.8760897011964167, + "flos": 25409713128000.0, + "grad_norm": 2.0157340559917283, + "learning_rate": 1.5879082266596867e-07, + "loss": 0.7797, + "num_input_tokens_seen": 157657015, + "step": 7286 + }, + { + "epoch": 0.8762099440870559, + "flos": 28943180635680.0, + "grad_norm": 2.1595153409673977, + "learning_rate": 1.5848677695111645e-07, + "loss": 0.7224, + "num_input_tokens_seen": 157678615, + "step": 7287 + }, + { + "epoch": 0.8763301869776949, + "flos": 21578358464640.0, + "grad_norm": 3.160659278471022, + "learning_rate": 1.5818301059235562e-07, + "loss": 0.6987, + "num_input_tokens_seen": 157693790, + "step": 7288 + }, + { + "epoch": 0.876450429868334, + "flos": 24609136207200.0, + "grad_norm": 2.271392951627777, + "learning_rate": 1.578795236357684e-07, + "loss": 0.8156, + "num_input_tokens_seen": 157715255, + "step": 7289 + }, + { + "epoch": 0.8765706727589732, + "flos": 20230928370240.0, + "grad_norm": 2.4353855449775397, + "learning_rate": 1.5757631612739218e-07, + "loss": 0.8572, + "num_input_tokens_seen": 157728800, + "step": 7290 + }, + { + "epoch": 0.8766909156496122, + "flos": 71268761367840.0, + "grad_norm": 0.7786240122787769, + "learning_rate": 1.572733881132242e-07, + "loss": 0.6127, + "num_input_tokens_seen": 157789445, + "step": 7291 + }, + { + "epoch": 0.8768111585402513, + "flos": 69424254142080.0, + "grad_norm": 0.7883411547277535, + "learning_rate": 1.5697073963921814e-07, + "loss": 0.5844, + "num_input_tokens_seen": 157848685, + "step": 7292 + }, + { + "epoch": 0.8769314014308904, + "flos": 18811563323040.0, + "grad_norm": 2.974090751376682, + "learning_rate": 1.566683707512857e-07, + "loss": 0.8522, + "num_input_tokens_seen": 157866360, + "step": 7293 + }, + { + "epoch": 0.8770516443215295, + "flos": 14955224847360.0, + "grad_norm": 3.6945170217729304, + "learning_rate": 1.5636628149529553e-07, + "loss": 0.7942, + "num_input_tokens_seen": 157884150, + "step": 7294 + }, + { + "epoch": 0.8771718872121685, + "flos": 31604173439520.0, + "grad_norm": 2.669642447058313, + "learning_rate": 1.560644719170743e-07, + "loss": 0.7938, + "num_input_tokens_seen": 157905020, + "step": 7295 + }, + { + "epoch": 0.8772921301028077, + "flos": 36043313980320.0, + "grad_norm": 2.5948362718502995, + "learning_rate": 1.5576294206240692e-07, + "loss": 0.7231, + "num_input_tokens_seen": 157924545, + "step": 7296 + }, + { + "epoch": 0.8774123729934468, + "flos": 57034410008640.0, + "grad_norm": 2.030062519106723, + "learning_rate": 1.5546169197703507e-07, + "loss": 0.6764, + "num_input_tokens_seen": 157950820, + "step": 7297 + }, + { + "epoch": 0.8775326158840858, + "flos": 23880271220640.0, + "grad_norm": 3.4743810088669593, + "learning_rate": 1.5516072170665774e-07, + "loss": 0.7811, + "num_input_tokens_seen": 157968790, + "step": 7298 + }, + { + "epoch": 0.877652858774725, + "flos": 17098771491360.0, + "grad_norm": 2.464613140460163, + "learning_rate": 1.5486003129693214e-07, + "loss": 0.8649, + "num_input_tokens_seen": 157986155, + "step": 7299 + }, + { + "epoch": 0.877773101665364, + "flos": 16485558828960.0, + "grad_norm": 3.034759849084235, + "learning_rate": 1.545596207934725e-07, + "loss": 0.7798, + "num_input_tokens_seen": 158004640, + "step": 7300 + }, + { + "epoch": 0.8778933445560031, + "flos": 22021676655840.0, + "grad_norm": 2.206937696786558, + "learning_rate": 1.5425949024185147e-07, + "loss": 0.7777, + "num_input_tokens_seen": 158024665, + "step": 7301 + }, + { + "epoch": 0.8780135874466423, + "flos": 22532506598400.0, + "grad_norm": 2.648225578899169, + "learning_rate": 1.5395963968759818e-07, + "loss": 0.6758, + "num_input_tokens_seen": 158044940, + "step": 7302 + }, + { + "epoch": 0.8781338303372813, + "flos": 61443028442880.0, + "grad_norm": 1.964841687180515, + "learning_rate": 1.536600691761998e-07, + "loss": 0.6487, + "num_input_tokens_seen": 158070770, + "step": 7303 + }, + { + "epoch": 0.8782540732279204, + "flos": 22642360440000.0, + "grad_norm": 2.4097401462246215, + "learning_rate": 1.5336077875310084e-07, + "loss": 0.7181, + "num_input_tokens_seen": 158089995, + "step": 7304 + }, + { + "epoch": 0.8783743161185595, + "flos": 16047555913440.0, + "grad_norm": 2.590391627569022, + "learning_rate": 1.5306176846370321e-07, + "loss": 0.743, + "num_input_tokens_seen": 158108140, + "step": 7305 + }, + { + "epoch": 0.8784945590091986, + "flos": 26030545591200.0, + "grad_norm": 2.6870836334614805, + "learning_rate": 1.5276303835336712e-07, + "loss": 0.7459, + "num_input_tokens_seen": 158128680, + "step": 7306 + }, + { + "epoch": 0.8786148018998376, + "flos": 62630711410560.0, + "grad_norm": 0.7756958560101643, + "learning_rate": 1.524645884674094e-07, + "loss": 0.5343, + "num_input_tokens_seen": 158185610, + "step": 7307 + }, + { + "epoch": 0.8787350447904768, + "flos": 21621815647680.0, + "grad_norm": 2.7186153860857476, + "learning_rate": 1.521664188511047e-07, + "loss": 0.7916, + "num_input_tokens_seen": 158205635, + "step": 7308 + }, + { + "epoch": 0.8788552876811159, + "flos": 25441796364480.0, + "grad_norm": 2.2825627142656266, + "learning_rate": 1.518685295496851e-07, + "loss": 0.8044, + "num_input_tokens_seen": 158223495, + "step": 7309 + }, + { + "epoch": 0.8789755305717549, + "flos": 22278373983840.0, + "grad_norm": 2.193362201203099, + "learning_rate": 1.5157092060833975e-07, + "loss": 0.8573, + "num_input_tokens_seen": 158243145, + "step": 7310 + }, + { + "epoch": 0.879095773462394, + "flos": 29268764996160.0, + "grad_norm": 2.1981847036159037, + "learning_rate": 1.5127359207221658e-07, + "loss": 0.6612, + "num_input_tokens_seen": 158262625, + "step": 7311 + }, + { + "epoch": 0.8792160163530331, + "flos": 16666306870080.0, + "grad_norm": 2.7926933526264928, + "learning_rate": 1.5097654398641923e-07, + "loss": 0.7341, + "num_input_tokens_seen": 158280530, + "step": 7312 + }, + { + "epoch": 0.8793362592436722, + "flos": 24464262718080.0, + "grad_norm": 1.599061380683786, + "learning_rate": 1.5067977639601014e-07, + "loss": 0.7306, + "num_input_tokens_seen": 158303290, + "step": 7313 + }, + { + "epoch": 0.8794565021343113, + "flos": 14517853817760.0, + "grad_norm": 3.2412122615525316, + "learning_rate": 1.5038328934600864e-07, + "loss": 0.7111, + "num_input_tokens_seen": 158319925, + "step": 7314 + }, + { + "epoch": 0.8795767450249504, + "flos": 39473320845120.0, + "grad_norm": 2.3949372536995024, + "learning_rate": 1.5008708288139161e-07, + "loss": 0.7011, + "num_input_tokens_seen": 158342285, + "step": 7315 + }, + { + "epoch": 0.8796969879155895, + "flos": 22927647047040.0, + "grad_norm": 2.3864958763638398, + "learning_rate": 1.497911570470931e-07, + "loss": 0.7332, + "num_input_tokens_seen": 158362290, + "step": 7316 + }, + { + "epoch": 0.8798172308062285, + "flos": 28321604777280.0, + "grad_norm": 2.0994850299298773, + "learning_rate": 1.494955118880048e-07, + "loss": 0.8581, + "num_input_tokens_seen": 158383275, + "step": 7317 + }, + { + "epoch": 0.8799374736968677, + "flos": 23954696547360.0, + "grad_norm": 2.1983981456365633, + "learning_rate": 1.4920014744897634e-07, + "loss": 0.7289, + "num_input_tokens_seen": 158402690, + "step": 7318 + }, + { + "epoch": 0.8800577165875068, + "flos": 25593583428960.0, + "grad_norm": 2.3178382294219273, + "learning_rate": 1.4890506377481392e-07, + "loss": 0.8641, + "num_input_tokens_seen": 158421780, + "step": 7319 + }, + { + "epoch": 0.8801779594781458, + "flos": 23406802620480.0, + "grad_norm": 2.0393823833139026, + "learning_rate": 1.486102609102815e-07, + "loss": 0.6425, + "num_input_tokens_seen": 158442330, + "step": 7320 + }, + { + "epoch": 0.880298202368785, + "flos": 11492391350880.0, + "grad_norm": 2.996718837579513, + "learning_rate": 1.483157389001004e-07, + "loss": 0.857, + "num_input_tokens_seen": 158459890, + "step": 7321 + }, + { + "epoch": 0.880418445259424, + "flos": 22638940822080.0, + "grad_norm": 6.930277320850623, + "learning_rate": 1.4802149778894933e-07, + "loss": 0.7912, + "num_input_tokens_seen": 158478680, + "step": 7322 + }, + { + "epoch": 0.8805386881500631, + "flos": 20492903804160.0, + "grad_norm": 3.103371196096548, + "learning_rate": 1.4772753762146484e-07, + "loss": 0.8746, + "num_input_tokens_seen": 158497935, + "step": 7323 + }, + { + "epoch": 0.8806589310407023, + "flos": 36486297643680.0, + "grad_norm": 2.020131123507717, + "learning_rate": 1.474338584422401e-07, + "loss": 0.7089, + "num_input_tokens_seen": 158523145, + "step": 7324 + }, + { + "epoch": 0.8807791739313413, + "flos": 23406653941440.0, + "grad_norm": 2.289892781675291, + "learning_rate": 1.4714046029582595e-07, + "loss": 0.7603, + "num_input_tokens_seen": 158542210, + "step": 7325 + }, + { + "epoch": 0.8808994168219804, + "flos": 25919613826560.0, + "grad_norm": 2.4440258407038327, + "learning_rate": 1.46847343226731e-07, + "loss": 0.759, + "num_input_tokens_seen": 158563250, + "step": 7326 + }, + { + "epoch": 0.8810196597126195, + "flos": 17067803347680.0, + "grad_norm": 2.6791882286173125, + "learning_rate": 1.465545072794203e-07, + "loss": 0.6978, + "num_input_tokens_seen": 158581125, + "step": 7327 + }, + { + "epoch": 0.8811399026032586, + "flos": 22969951967520.0, + "grad_norm": 2.3049183875342223, + "learning_rate": 1.4626195249831774e-07, + "loss": 0.7576, + "num_input_tokens_seen": 158602025, + "step": 7328 + }, + { + "epoch": 0.8812601454938976, + "flos": 14443316981760.0, + "grad_norm": 2.3835510110039824, + "learning_rate": 1.4596967892780244e-07, + "loss": 0.7214, + "num_input_tokens_seen": 158618355, + "step": 7329 + }, + { + "epoch": 0.8813803883845368, + "flos": 22461129192000.0, + "grad_norm": 3.1760432474281095, + "learning_rate": 1.4567768661221314e-07, + "loss": 0.7478, + "num_input_tokens_seen": 158638925, + "step": 7330 + }, + { + "epoch": 0.8815006312751759, + "flos": 21476198763360.0, + "grad_norm": 2.6437418222864526, + "learning_rate": 1.4538597559584442e-07, + "loss": 0.7446, + "num_input_tokens_seen": 158656715, + "step": 7331 + }, + { + "epoch": 0.8816208741658149, + "flos": 22752028432800.0, + "grad_norm": 2.231124968458273, + "learning_rate": 1.4509454592294823e-07, + "loss": 0.7932, + "num_input_tokens_seen": 158677200, + "step": 7332 + }, + { + "epoch": 0.8817411170564541, + "flos": 17753917376640.0, + "grad_norm": 3.195851026450413, + "learning_rate": 1.448033976377354e-07, + "loss": 0.7912, + "num_input_tokens_seen": 158692185, + "step": 7333 + }, + { + "epoch": 0.8818613599470931, + "flos": 18525421811520.0, + "grad_norm": 2.680030378197972, + "learning_rate": 1.445125307843713e-07, + "loss": 0.7462, + "num_input_tokens_seen": 158710410, + "step": 7334 + }, + { + "epoch": 0.8819816028377322, + "flos": 27560210517120.0, + "grad_norm": 3.021972459012455, + "learning_rate": 1.442219454069813e-07, + "loss": 0.7551, + "num_input_tokens_seen": 158731435, + "step": 7335 + }, + { + "epoch": 0.8821018457283714, + "flos": 23371151087040.0, + "grad_norm": 2.9322724598627645, + "learning_rate": 1.4393164154964676e-07, + "loss": 0.666, + "num_input_tokens_seen": 158750965, + "step": 7336 + }, + { + "epoch": 0.8822220886190104, + "flos": 29090656008000.0, + "grad_norm": 2.7830897900444143, + "learning_rate": 1.4364161925640649e-07, + "loss": 0.9457, + "num_input_tokens_seen": 158772365, + "step": 7337 + }, + { + "epoch": 0.8823423315096495, + "flos": 20456211517440.0, + "grad_norm": 2.5933183977371996, + "learning_rate": 1.4335187857125663e-07, + "loss": 0.8519, + "num_input_tokens_seen": 158791065, + "step": 7338 + }, + { + "epoch": 0.8824625744002886, + "flos": 24171728007840.0, + "grad_norm": 2.115614303043431, + "learning_rate": 1.4306241953815023e-07, + "loss": 0.7548, + "num_input_tokens_seen": 158812125, + "step": 7339 + }, + { + "epoch": 0.8825828172909277, + "flos": 24644118684960.0, + "grad_norm": 2.3140764791103763, + "learning_rate": 1.4277324220099862e-07, + "loss": 0.7131, + "num_input_tokens_seen": 158834035, + "step": 7340 + }, + { + "epoch": 0.8827030601815667, + "flos": 22424102377440.0, + "grad_norm": 2.39646537294448, + "learning_rate": 1.4248434660366938e-07, + "loss": 0.7429, + "num_input_tokens_seen": 158853510, + "step": 7341 + }, + { + "epoch": 0.8828233030722058, + "flos": 19837349051520.0, + "grad_norm": 2.2199535060859605, + "learning_rate": 1.4219573278998808e-07, + "loss": 0.7042, + "num_input_tokens_seen": 158871970, + "step": 7342 + }, + { + "epoch": 0.882943545962845, + "flos": 39290491297440.0, + "grad_norm": 3.281210181585036, + "learning_rate": 1.4190740080373685e-07, + "loss": 0.6503, + "num_input_tokens_seen": 158892250, + "step": 7343 + }, + { + "epoch": 0.883063788853484, + "flos": 19027107993120.0, + "grad_norm": 3.232599333072559, + "learning_rate": 1.4161935068865538e-07, + "loss": 0.8464, + "num_input_tokens_seen": 158907395, + "step": 7344 + }, + { + "epoch": 0.8831840317441231, + "flos": 18706504380480.0, + "grad_norm": 4.16640614012985, + "learning_rate": 1.4133158248844113e-07, + "loss": 0.7614, + "num_input_tokens_seen": 158926300, + "step": 7345 + }, + { + "epoch": 0.8833042746347622, + "flos": 26789040610080.0, + "grad_norm": 2.3686941658626175, + "learning_rate": 1.4104409624674785e-07, + "loss": 0.7368, + "num_input_tokens_seen": 158946085, + "step": 7346 + }, + { + "epoch": 0.8834245175254013, + "flos": 26066680331520.0, + "grad_norm": 2.388780268582656, + "learning_rate": 1.407568920071873e-07, + "loss": 0.7841, + "num_input_tokens_seen": 158964950, + "step": 7347 + }, + { + "epoch": 0.8835447604160404, + "flos": 30585078267840.0, + "grad_norm": 2.7462280587876897, + "learning_rate": 1.4046996981332782e-07, + "loss": 0.6803, + "num_input_tokens_seen": 158984835, + "step": 7348 + }, + { + "epoch": 0.8836650033066795, + "flos": 24683003987520.0, + "grad_norm": 2.4043643052587402, + "learning_rate": 1.4018332970869516e-07, + "loss": 0.7841, + "num_input_tokens_seen": 159002775, + "step": 7349 + }, + { + "epoch": 0.8837852461973186, + "flos": 25377072345120.0, + "grad_norm": 2.219277332663149, + "learning_rate": 1.398969717367733e-07, + "loss": 0.8491, + "num_input_tokens_seen": 159024100, + "step": 7350 + }, + { + "epoch": 0.8839054890879576, + "flos": 17797263050400.0, + "grad_norm": 2.0494713078258897, + "learning_rate": 1.396108959410014e-07, + "loss": 0.7663, + "num_input_tokens_seen": 159043315, + "step": 7351 + }, + { + "epoch": 0.8840257319785968, + "flos": 23735026033920.0, + "grad_norm": 2.3703832425931375, + "learning_rate": 1.3932510236477745e-07, + "loss": 0.8142, + "num_input_tokens_seen": 159063985, + "step": 7352 + }, + { + "epoch": 0.8841459748692359, + "flos": 29018089169280.0, + "grad_norm": 2.322529462580232, + "learning_rate": 1.3903959105145636e-07, + "loss": 0.5624, + "num_input_tokens_seen": 159084475, + "step": 7353 + }, + { + "epoch": 0.8842662177598749, + "flos": 24276563931840.0, + "grad_norm": 2.6771468415996194, + "learning_rate": 1.387543620443492e-07, + "loss": 0.8326, + "num_input_tokens_seen": 159101320, + "step": 7354 + }, + { + "epoch": 0.8843864606505141, + "flos": 24972007570560.0, + "grad_norm": 2.015396696660404, + "learning_rate": 1.3846941538672606e-07, + "loss": 0.8406, + "num_input_tokens_seen": 159120025, + "step": 7355 + }, + { + "epoch": 0.8845067035411531, + "flos": 28143495789120.0, + "grad_norm": 2.698989703870754, + "learning_rate": 1.3818475112181193e-07, + "loss": 0.8116, + "num_input_tokens_seen": 159138820, + "step": 7356 + }, + { + "epoch": 0.8846269464317922, + "flos": 12834877867200.0, + "grad_norm": 3.182497025193999, + "learning_rate": 1.3790036929279091e-07, + "loss": 0.795, + "num_input_tokens_seen": 159155975, + "step": 7357 + }, + { + "epoch": 0.8847471893224313, + "flos": 18597839971200.0, + "grad_norm": 3.3722370282147858, + "learning_rate": 1.3761626994280363e-07, + "loss": 0.5911, + "num_input_tokens_seen": 159173445, + "step": 7358 + }, + { + "epoch": 0.8848674322130704, + "flos": 35717841129120.0, + "grad_norm": 2.5011229853906256, + "learning_rate": 1.3733245311494735e-07, + "loss": 0.7387, + "num_input_tokens_seen": 159196100, + "step": 7359 + }, + { + "epoch": 0.8849876751037095, + "flos": 24209089350240.0, + "grad_norm": 3.0769111834809717, + "learning_rate": 1.3704891885227676e-07, + "loss": 0.7051, + "num_input_tokens_seen": 159216145, + "step": 7360 + }, + { + "epoch": 0.8851079179943486, + "flos": 21469656885600.0, + "grad_norm": 2.413167744465197, + "learning_rate": 1.367656671978037e-07, + "loss": 0.7792, + "num_input_tokens_seen": 159233610, + "step": 7361 + }, + { + "epoch": 0.8852281608849877, + "flos": 15278913550080.0, + "grad_norm": 3.094511304491074, + "learning_rate": 1.36482698194498e-07, + "loss": 0.737, + "num_input_tokens_seen": 159250155, + "step": 7362 + }, + { + "epoch": 0.8853484037756267, + "flos": 23262449508000.0, + "grad_norm": 2.652601361580641, + "learning_rate": 1.3620001188528506e-07, + "loss": 0.7175, + "num_input_tokens_seen": 159270875, + "step": 7363 + }, + { + "epoch": 0.8854686466662659, + "flos": 25078664812800.0, + "grad_norm": 3.3557921710038454, + "learning_rate": 1.3591760831304865e-07, + "loss": 0.7383, + "num_input_tokens_seen": 159288565, + "step": 7364 + }, + { + "epoch": 0.885588889556905, + "flos": 21360100402080.0, + "grad_norm": 2.4364078716017517, + "learning_rate": 1.356354875206287e-07, + "loss": 0.7935, + "num_input_tokens_seen": 159306400, + "step": 7365 + }, + { + "epoch": 0.885709132447544, + "flos": 26868186496320.0, + "grad_norm": 2.672812670981211, + "learning_rate": 1.3535364955082296e-07, + "loss": 0.7029, + "num_input_tokens_seen": 159326250, + "step": 7366 + }, + { + "epoch": 0.8858293753381832, + "flos": 26065974106080.0, + "grad_norm": 1.977582645642419, + "learning_rate": 1.3507209444638613e-07, + "loss": 0.6441, + "num_input_tokens_seen": 159348250, + "step": 7367 + }, + { + "epoch": 0.8859496182288222, + "flos": 23259438757440.0, + "grad_norm": 3.002149765334997, + "learning_rate": 1.347908222500298e-07, + "loss": 0.7396, + "num_input_tokens_seen": 159368325, + "step": 7368 + }, + { + "epoch": 0.8860698611194613, + "flos": 16848950568960.0, + "grad_norm": 2.385979820530654, + "learning_rate": 1.3450983300442276e-07, + "loss": 0.7028, + "num_input_tokens_seen": 159387555, + "step": 7369 + }, + { + "epoch": 0.8861901040101005, + "flos": 24646088682240.0, + "grad_norm": 2.580701338344985, + "learning_rate": 1.3422912675219068e-07, + "loss": 0.7359, + "num_input_tokens_seen": 159407310, + "step": 7370 + }, + { + "epoch": 0.8863103469007395, + "flos": 24387978903360.0, + "grad_norm": 2.1832115644127006, + "learning_rate": 1.339487035359166e-07, + "loss": 0.7915, + "num_input_tokens_seen": 159427680, + "step": 7371 + }, + { + "epoch": 0.8864305897913786, + "flos": 22022457220800.0, + "grad_norm": 1.9306438099595322, + "learning_rate": 1.336685633981409e-07, + "loss": 0.8478, + "num_input_tokens_seen": 159448765, + "step": 7372 + }, + { + "epoch": 0.8865508326820177, + "flos": 19072274985120.0, + "grad_norm": 2.3563091240302665, + "learning_rate": 1.333887063813597e-07, + "loss": 0.7497, + "num_input_tokens_seen": 159466870, + "step": 7373 + }, + { + "epoch": 0.8866710755726568, + "flos": 15392372858400.0, + "grad_norm": 3.9109320085390182, + "learning_rate": 1.331091325280278e-07, + "loss": 0.6658, + "num_input_tokens_seen": 159485840, + "step": 7374 + }, + { + "epoch": 0.8867913184632958, + "flos": 20055049567680.0, + "grad_norm": 2.062190408301409, + "learning_rate": 1.3282984188055625e-07, + "loss": 0.7868, + "num_input_tokens_seen": 159505630, + "step": 7375 + }, + { + "epoch": 0.8869115613539349, + "flos": 23332377293760.0, + "grad_norm": 2.6040147559045255, + "learning_rate": 1.3255083448131288e-07, + "loss": 0.7984, + "num_input_tokens_seen": 159524675, + "step": 7376 + }, + { + "epoch": 0.8870318042445741, + "flos": 21256416740640.0, + "grad_norm": 2.7538402792962615, + "learning_rate": 1.3227211037262365e-07, + "loss": 0.7931, + "num_input_tokens_seen": 159541915, + "step": 7377 + }, + { + "epoch": 0.8871520471352131, + "flos": 19982334049920.0, + "grad_norm": 2.98701352424229, + "learning_rate": 1.319936695967696e-07, + "loss": 0.8539, + "num_input_tokens_seen": 159559740, + "step": 7378 + }, + { + "epoch": 0.8872722900258522, + "flos": 22569124545600.0, + "grad_norm": 6.427971463741254, + "learning_rate": 1.3171551219599097e-07, + "loss": 0.8226, + "num_input_tokens_seen": 159578265, + "step": 7379 + }, + { + "epoch": 0.8873925329164913, + "flos": 22131604836960.0, + "grad_norm": 2.4132408546235737, + "learning_rate": 1.3143763821248377e-07, + "loss": 0.7818, + "num_input_tokens_seen": 159595350, + "step": 7380 + }, + { + "epoch": 0.8875127758071304, + "flos": 19180865054880.0, + "grad_norm": 2.0913951783711893, + "learning_rate": 1.3116004768840118e-07, + "loss": 0.7214, + "num_input_tokens_seen": 159613725, + "step": 7381 + }, + { + "epoch": 0.8876330186977694, + "flos": 18085040031360.0, + "grad_norm": 2.0212451738006214, + "learning_rate": 1.3088274066585348e-07, + "loss": 0.7381, + "num_input_tokens_seen": 159631335, + "step": 7382 + }, + { + "epoch": 0.8877532615884086, + "flos": 21977699096160.0, + "grad_norm": 2.502948643333165, + "learning_rate": 1.3060571718690749e-07, + "loss": 0.9042, + "num_input_tokens_seen": 159648830, + "step": 7383 + }, + { + "epoch": 0.8878735044790477, + "flos": 72033389397120.0, + "grad_norm": 0.7545163151978556, + "learning_rate": 1.3032897729358805e-07, + "loss": 0.5686, + "num_input_tokens_seen": 159709785, + "step": 7384 + }, + { + "epoch": 0.8879937473696867, + "flos": 27486863113440.0, + "grad_norm": 2.48644606916591, + "learning_rate": 1.3005252102787645e-07, + "loss": 0.7992, + "num_input_tokens_seen": 159728725, + "step": 7385 + }, + { + "epoch": 0.8881139902603259, + "flos": 22202238848160.0, + "grad_norm": 1.7901428159289747, + "learning_rate": 1.297763484317105e-07, + "loss": 0.7351, + "num_input_tokens_seen": 159747020, + "step": 7386 + }, + { + "epoch": 0.888234233150965, + "flos": 20270482728480.0, + "grad_norm": 3.4298301441186387, + "learning_rate": 1.2950045954698551e-07, + "loss": 0.7018, + "num_input_tokens_seen": 159764855, + "step": 7387 + }, + { + "epoch": 0.888354476041604, + "flos": 18122104015680.0, + "grad_norm": 2.2901741947736696, + "learning_rate": 1.2922485441555343e-07, + "loss": 0.7564, + "num_input_tokens_seen": 159783935, + "step": 7388 + }, + { + "epoch": 0.8884747189322432, + "flos": 21982717013760.0, + "grad_norm": 2.0642340335420193, + "learning_rate": 1.2894953307922363e-07, + "loss": 0.8185, + "num_input_tokens_seen": 159802895, + "step": 7389 + }, + { + "epoch": 0.8885949618228822, + "flos": 19758389014080.0, + "grad_norm": 2.141452156139878, + "learning_rate": 1.2867449557976208e-07, + "loss": 0.8394, + "num_input_tokens_seen": 159820995, + "step": 7390 + }, + { + "epoch": 0.8887152047135213, + "flos": 20019026336640.0, + "grad_norm": 2.695618277633567, + "learning_rate": 1.283997419588916e-07, + "loss": 0.7594, + "num_input_tokens_seen": 159840465, + "step": 7391 + }, + { + "epoch": 0.8888354476041604, + "flos": 18561668061120.0, + "grad_norm": 2.809092003950336, + "learning_rate": 1.2812527225829216e-07, + "loss": 0.6233, + "num_input_tokens_seen": 159858690, + "step": 7392 + }, + { + "epoch": 0.8889556904947995, + "flos": 21658842462240.0, + "grad_norm": 2.3480861987354498, + "learning_rate": 1.2785108651960052e-07, + "loss": 0.7623, + "num_input_tokens_seen": 159880325, + "step": 7393 + }, + { + "epoch": 0.8890759333854386, + "flos": 27341952454560.0, + "grad_norm": 2.830722690681319, + "learning_rate": 1.2757718478441094e-07, + "loss": 0.8093, + "num_input_tokens_seen": 159901070, + "step": 7394 + }, + { + "epoch": 0.8891961762760777, + "flos": 24463779511200.0, + "grad_norm": 2.505592100023066, + "learning_rate": 1.2730356709427302e-07, + "loss": 0.7767, + "num_input_tokens_seen": 159919750, + "step": 7395 + }, + { + "epoch": 0.8893164191667168, + "flos": 41439910763520.0, + "grad_norm": 1.946481192791179, + "learning_rate": 1.2703023349069542e-07, + "loss": 0.6005, + "num_input_tokens_seen": 159944600, + "step": 7396 + }, + { + "epoch": 0.8894366620573558, + "flos": 33535334843040.0, + "grad_norm": 2.3173769481792106, + "learning_rate": 1.2675718401514223e-07, + "loss": 0.6185, + "num_input_tokens_seen": 159968780, + "step": 7397 + }, + { + "epoch": 0.889556904947995, + "flos": 16885828704480.0, + "grad_norm": 4.174877900491078, + "learning_rate": 1.264844187090346e-07, + "loss": 0.7479, + "num_input_tokens_seen": 159985905, + "step": 7398 + }, + { + "epoch": 0.889677147838634, + "flos": 25993555946400.0, + "grad_norm": 2.403762720860918, + "learning_rate": 1.262119376137516e-07, + "loss": 0.753, + "num_input_tokens_seen": 160006965, + "step": 7399 + }, + { + "epoch": 0.8897973907292731, + "flos": 26430332259840.0, + "grad_norm": 2.1269498742729405, + "learning_rate": 1.2593974077062707e-07, + "loss": 0.8513, + "num_input_tokens_seen": 160028585, + "step": 7400 + }, + { + "epoch": 0.8899176336199123, + "flos": 26212074197280.0, + "grad_norm": 2.601413848570743, + "learning_rate": 1.2566782822095423e-07, + "loss": 0.639, + "num_input_tokens_seen": 160048840, + "step": 7401 + }, + { + "epoch": 0.8900378765105513, + "flos": 20781981726720.0, + "grad_norm": 5.763404480623878, + "learning_rate": 1.2539620000598162e-07, + "loss": 0.7169, + "num_input_tokens_seen": 160068175, + "step": 7402 + }, + { + "epoch": 0.8901581194011904, + "flos": 16448978051520.0, + "grad_norm": 2.301287506311748, + "learning_rate": 1.2512485616691492e-07, + "loss": 0.797, + "num_input_tokens_seen": 160085230, + "step": 7403 + }, + { + "epoch": 0.8902783622918296, + "flos": 35104777145760.0, + "grad_norm": 1.9171002630083054, + "learning_rate": 1.2485379674491681e-07, + "loss": 0.8094, + "num_input_tokens_seen": 160111425, + "step": 7404 + }, + { + "epoch": 0.8903986051824686, + "flos": 17176616436000.0, + "grad_norm": 3.094458296642274, + "learning_rate": 1.2458302178110657e-07, + "loss": 0.7902, + "num_input_tokens_seen": 160129790, + "step": 7405 + }, + { + "epoch": 0.8905188480731077, + "flos": 25447371828480.0, + "grad_norm": 2.4307715643915193, + "learning_rate": 1.2431253131656118e-07, + "loss": 0.8237, + "num_input_tokens_seen": 160149265, + "step": 7406 + }, + { + "epoch": 0.8906390909637467, + "flos": 23332414463520.0, + "grad_norm": 2.380857772764022, + "learning_rate": 1.240423253923133e-07, + "loss": 0.7694, + "num_input_tokens_seen": 160168870, + "step": 7407 + }, + { + "epoch": 0.8907593338543859, + "flos": 21038716224480.0, + "grad_norm": 2.313928935892658, + "learning_rate": 1.237724040493533e-07, + "loss": 0.6956, + "num_input_tokens_seen": 160187495, + "step": 7408 + }, + { + "epoch": 0.8908795767450249, + "flos": 21838252392000.0, + "grad_norm": 3.3688024909578025, + "learning_rate": 1.2350276732862773e-07, + "loss": 0.7334, + "num_input_tokens_seen": 160208520, + "step": 7409 + }, + { + "epoch": 0.890999819635664, + "flos": 66213137225760.0, + "grad_norm": 0.8504951157023162, + "learning_rate": 1.2323341527103993e-07, + "loss": 0.5658, + "num_input_tokens_seen": 160263720, + "step": 7410 + }, + { + "epoch": 0.8911200625263032, + "flos": 26831382700320.0, + "grad_norm": 2.723214353610536, + "learning_rate": 1.2296434791745135e-07, + "loss": 0.8538, + "num_input_tokens_seen": 160282170, + "step": 7411 + }, + { + "epoch": 0.8912403054169422, + "flos": 20855998186080.0, + "grad_norm": 2.2696715950044233, + "learning_rate": 1.2269556530867875e-07, + "loss": 0.7697, + "num_input_tokens_seen": 160300725, + "step": 7412 + }, + { + "epoch": 0.8913605483075813, + "flos": 26977854489120.0, + "grad_norm": 2.694670050528149, + "learning_rate": 1.2242706748549614e-07, + "loss": 0.8199, + "num_input_tokens_seen": 160318720, + "step": 7413 + }, + { + "epoch": 0.8914807911982204, + "flos": 23587662170880.0, + "grad_norm": 2.4267527037169567, + "learning_rate": 1.2215885448863473e-07, + "loss": 0.8214, + "num_input_tokens_seen": 160339595, + "step": 7414 + }, + { + "epoch": 0.8916010340888595, + "flos": 24427273073280.0, + "grad_norm": 2.240441895562154, + "learning_rate": 1.2189092635878152e-07, + "loss": 0.8055, + "num_input_tokens_seen": 160361045, + "step": 7415 + }, + { + "epoch": 0.8917212769794985, + "flos": 21185522541120.0, + "grad_norm": 2.697670079650693, + "learning_rate": 1.216232831365822e-07, + "loss": 0.777, + "num_input_tokens_seen": 160379990, + "step": 7416 + }, + { + "epoch": 0.8918415198701377, + "flos": 25477224879360.0, + "grad_norm": 1.99004551044737, + "learning_rate": 1.2135592486263678e-07, + "loss": 0.8128, + "num_input_tokens_seen": 160399240, + "step": 7417 + }, + { + "epoch": 0.8919617627607768, + "flos": 37798522241760.0, + "grad_norm": 5.3544057156922005, + "learning_rate": 1.2108885157750415e-07, + "loss": 0.6133, + "num_input_tokens_seen": 160421600, + "step": 7418 + }, + { + "epoch": 0.8920820056514158, + "flos": 26176125305760.0, + "grad_norm": 4.062918304035127, + "learning_rate": 1.2082206332169897e-07, + "loss": 0.8034, + "num_input_tokens_seen": 160441695, + "step": 7419 + }, + { + "epoch": 0.892202248542055, + "flos": 17358405230400.0, + "grad_norm": 2.90598090782116, + "learning_rate": 1.2055556013569225e-07, + "loss": 0.7356, + "num_input_tokens_seen": 160457205, + "step": 7420 + }, + { + "epoch": 0.892322491432694, + "flos": 21293778083040.0, + "grad_norm": 2.0765553907010075, + "learning_rate": 1.2028934205991315e-07, + "loss": 0.8219, + "num_input_tokens_seen": 160476525, + "step": 7421 + }, + { + "epoch": 0.8924427343233331, + "flos": 23995180149600.0, + "grad_norm": 1.6840452929587648, + "learning_rate": 1.2002340913474607e-07, + "loss": 0.7661, + "num_input_tokens_seen": 160500160, + "step": 7422 + }, + { + "epoch": 0.8925629772139723, + "flos": 29964989199840.0, + "grad_norm": 6.193099445625607, + "learning_rate": 1.1975776140053317e-07, + "loss": 0.7347, + "num_input_tokens_seen": 160520130, + "step": 7423 + }, + { + "epoch": 0.8926832201046113, + "flos": 22569421903680.0, + "grad_norm": 6.663559889091402, + "learning_rate": 1.194923988975729e-07, + "loss": 0.7351, + "num_input_tokens_seen": 160539730, + "step": 7424 + }, + { + "epoch": 0.8928034629952504, + "flos": 13278976623360.0, + "grad_norm": 7.949527788833101, + "learning_rate": 1.192273216661206e-07, + "loss": 0.7351, + "num_input_tokens_seen": 160557820, + "step": 7425 + }, + { + "epoch": 0.8929237058858895, + "flos": 54776617757760.0, + "grad_norm": 0.7881927985026184, + "learning_rate": 1.189625297463881e-07, + "loss": 0.5746, + "num_input_tokens_seen": 160619510, + "step": 7426 + }, + { + "epoch": 0.8930439487765286, + "flos": 28842656403840.0, + "grad_norm": 2.0780691268408167, + "learning_rate": 1.1869802317854394e-07, + "loss": 0.7953, + "num_input_tokens_seen": 160643295, + "step": 7427 + }, + { + "epoch": 0.8931641916671677, + "flos": 22387261411680.0, + "grad_norm": 2.675494324830862, + "learning_rate": 1.1843380200271425e-07, + "loss": 0.7245, + "num_input_tokens_seen": 160662495, + "step": 7428 + }, + { + "epoch": 0.8932844345578068, + "flos": 25805931499680.0, + "grad_norm": 2.434845873206255, + "learning_rate": 1.181698662589805e-07, + "loss": 0.806, + "num_input_tokens_seen": 160682080, + "step": 7429 + }, + { + "epoch": 0.8934046774484459, + "flos": 22892887587840.0, + "grad_norm": 2.2215974830258776, + "learning_rate": 1.1790621598738249e-07, + "loss": 0.7607, + "num_input_tokens_seen": 160700395, + "step": 7430 + }, + { + "epoch": 0.8935249203390849, + "flos": 24427087224480.0, + "grad_norm": 3.9764702866810073, + "learning_rate": 1.1764285122791461e-07, + "loss": 0.7501, + "num_input_tokens_seen": 160721115, + "step": 7431 + }, + { + "epoch": 0.8936451632297241, + "flos": 15720447592800.0, + "grad_norm": 3.454724523931667, + "learning_rate": 1.173797720205294e-07, + "loss": 0.7732, + "num_input_tokens_seen": 160739150, + "step": 7432 + }, + { + "epoch": 0.8937654061203631, + "flos": 35064962599200.0, + "grad_norm": 3.0511840085555217, + "learning_rate": 1.1711697840513602e-07, + "loss": 0.7196, + "num_input_tokens_seen": 160758585, + "step": 7433 + }, + { + "epoch": 0.8938856490110022, + "flos": 16084619897760.0, + "grad_norm": 2.8941099861614643, + "learning_rate": 1.1685447042160012e-07, + "loss": 0.71, + "num_input_tokens_seen": 160776620, + "step": 7434 + }, + { + "epoch": 0.8940058919016414, + "flos": 20674655428800.0, + "grad_norm": 2.3181637662976895, + "learning_rate": 1.1659224810974367e-07, + "loss": 0.7173, + "num_input_tokens_seen": 160796580, + "step": 7435 + }, + { + "epoch": 0.8941261347922804, + "flos": 25192941855840.0, + "grad_norm": 1.9111825880957276, + "learning_rate": 1.1633031150934591e-07, + "loss": 0.6856, + "num_input_tokens_seen": 160819610, + "step": 7436 + }, + { + "epoch": 0.8942463776829195, + "flos": 19509422996160.0, + "grad_norm": 2.9694437922861705, + "learning_rate": 1.1606866066014176e-07, + "loss": 0.7949, + "num_input_tokens_seen": 160838660, + "step": 7437 + }, + { + "epoch": 0.8943666205735585, + "flos": 22269416071680.0, + "grad_norm": 3.213779352054812, + "learning_rate": 1.1580729560182434e-07, + "loss": 0.7537, + "num_input_tokens_seen": 160854515, + "step": 7438 + }, + { + "epoch": 0.8944868634641977, + "flos": 18885839970720.0, + "grad_norm": 2.612796938142308, + "learning_rate": 1.1554621637404171e-07, + "loss": 0.7089, + "num_input_tokens_seen": 160872605, + "step": 7439 + }, + { + "epoch": 0.8946071063548368, + "flos": 14440306231200.0, + "grad_norm": 2.3366145042804494, + "learning_rate": 1.1528542301639999e-07, + "loss": 0.6142, + "num_input_tokens_seen": 160889395, + "step": 7440 + }, + { + "epoch": 0.8947273492454758, + "flos": 20055458435040.0, + "grad_norm": 3.110222169109973, + "learning_rate": 1.1502491556846105e-07, + "loss": 0.8247, + "num_input_tokens_seen": 160907890, + "step": 7441 + }, + { + "epoch": 0.894847592136115, + "flos": 18524120869920.0, + "grad_norm": 3.131160919079675, + "learning_rate": 1.1476469406974331e-07, + "loss": 0.8182, + "num_input_tokens_seen": 160923490, + "step": 7442 + }, + { + "epoch": 0.894967835026754, + "flos": 23444944527840.0, + "grad_norm": 2.140034837884851, + "learning_rate": 1.1450475855972341e-07, + "loss": 0.7716, + "num_input_tokens_seen": 160944280, + "step": 7443 + }, + { + "epoch": 0.8950880779173931, + "flos": 15166569334560.0, + "grad_norm": 2.736261207693541, + "learning_rate": 1.1424510907783158e-07, + "loss": 0.706, + "num_input_tokens_seen": 160961560, + "step": 7444 + }, + { + "epoch": 0.8952083208080323, + "flos": 22060450449120.0, + "grad_norm": 5.352296831746355, + "learning_rate": 1.1398574566345787e-07, + "loss": 0.8256, + "num_input_tokens_seen": 160982195, + "step": 7445 + }, + { + "epoch": 0.8953285636986713, + "flos": 23220627794400.0, + "grad_norm": 2.7281230843483755, + "learning_rate": 1.1372666835594702e-07, + "loss": 0.8293, + "num_input_tokens_seen": 161000520, + "step": 7446 + }, + { + "epoch": 0.8954488065893104, + "flos": 16339458737760.0, + "grad_norm": 2.8813545654917374, + "learning_rate": 1.1346787719460071e-07, + "loss": 0.7183, + "num_input_tokens_seen": 161019405, + "step": 7447 + }, + { + "epoch": 0.8955690494799495, + "flos": 18231846348000.0, + "grad_norm": 2.3688459472549437, + "learning_rate": 1.1320937221867732e-07, + "loss": 0.7256, + "num_input_tokens_seen": 161036985, + "step": 7448 + }, + { + "epoch": 0.8956892923705886, + "flos": 25411051239360.0, + "grad_norm": 2.07759408694143, + "learning_rate": 1.1295115346739192e-07, + "loss": 0.7966, + "num_input_tokens_seen": 161056985, + "step": 7449 + }, + { + "epoch": 0.8958095352612276, + "flos": 52585679669760.0, + "grad_norm": 2.7271322185272027, + "learning_rate": 1.1269322097991629e-07, + "loss": 0.7266, + "num_input_tokens_seen": 161080270, + "step": 7450 + }, + { + "epoch": 0.8959297781518668, + "flos": 23153227552320.0, + "grad_norm": 2.9248995706646275, + "learning_rate": 1.1243557479537846e-07, + "loss": 0.6784, + "num_input_tokens_seen": 161100160, + "step": 7451 + }, + { + "epoch": 0.8960500210425059, + "flos": 20305130678400.0, + "grad_norm": 2.3739666949653975, + "learning_rate": 1.121782149528634e-07, + "loss": 0.6908, + "num_input_tokens_seen": 161117260, + "step": 7452 + }, + { + "epoch": 0.8961702639331449, + "flos": 19873335112800.0, + "grad_norm": 2.5635288313123623, + "learning_rate": 1.1192114149141208e-07, + "loss": 0.7866, + "num_input_tokens_seen": 161136895, + "step": 7453 + }, + { + "epoch": 0.8962905068237841, + "flos": 12878186371200.0, + "grad_norm": 3.060363958816795, + "learning_rate": 1.1166435445002197e-07, + "loss": 0.6476, + "num_input_tokens_seen": 161154565, + "step": 7454 + }, + { + "epoch": 0.8964107497144231, + "flos": 23406802620480.0, + "grad_norm": 3.1558127525295796, + "learning_rate": 1.1140785386764818e-07, + "loss": 0.6865, + "num_input_tokens_seen": 161173265, + "step": 7455 + }, + { + "epoch": 0.8965309926050622, + "flos": 19472396181600.0, + "grad_norm": 2.6793942304899465, + "learning_rate": 1.1115163978320153e-07, + "loss": 0.6993, + "num_input_tokens_seen": 161191995, + "step": 7456 + }, + { + "epoch": 0.8966512354957014, + "flos": 28616890049760.0, + "grad_norm": 3.4183871116729105, + "learning_rate": 1.1089571223554917e-07, + "loss": 0.8264, + "num_input_tokens_seen": 161212880, + "step": 7457 + }, + { + "epoch": 0.8967714783863404, + "flos": 23337618229920.0, + "grad_norm": 2.589177064542229, + "learning_rate": 1.1064007126351537e-07, + "loss": 0.8562, + "num_input_tokens_seen": 161233595, + "step": 7458 + }, + { + "epoch": 0.8968917212769795, + "flos": 24500062930560.0, + "grad_norm": 2.778969933801885, + "learning_rate": 1.1038471690588003e-07, + "loss": 0.7621, + "num_input_tokens_seen": 161252740, + "step": 7459 + }, + { + "epoch": 0.8970119641676186, + "flos": 23441785098240.0, + "grad_norm": 3.5374382957226156, + "learning_rate": 1.1012964920138145e-07, + "loss": 0.8016, + "num_input_tokens_seen": 161272325, + "step": 7460 + }, + { + "epoch": 0.8971322070582577, + "flos": 24503185190400.0, + "grad_norm": 2.259731546341189, + "learning_rate": 1.0987486818871205e-07, + "loss": 0.761, + "num_input_tokens_seen": 161295915, + "step": 7461 + }, + { + "epoch": 0.8972524499488967, + "flos": 21766280269440.0, + "grad_norm": 3.102702433385567, + "learning_rate": 1.0962037390652245e-07, + "loss": 0.7322, + "num_input_tokens_seen": 161314935, + "step": 7462 + }, + { + "epoch": 0.8973726928395359, + "flos": 21695237390880.0, + "grad_norm": 2.3014204241686302, + "learning_rate": 1.0936616639341911e-07, + "loss": 0.717, + "num_input_tokens_seen": 161335225, + "step": 7463 + }, + { + "epoch": 0.897492935730175, + "flos": 53760793524960.0, + "grad_norm": 0.7508313022406821, + "learning_rate": 1.0911224568796473e-07, + "loss": 0.5464, + "num_input_tokens_seen": 161393420, + "step": 7464 + }, + { + "epoch": 0.897613178620814, + "flos": 18264115433280.0, + "grad_norm": 2.8570034844601233, + "learning_rate": 1.0885861182867984e-07, + "loss": 0.7103, + "num_input_tokens_seen": 161411525, + "step": 7465 + }, + { + "epoch": 0.8977334215114532, + "flos": 32946065239680.0, + "grad_norm": 2.22011534898094, + "learning_rate": 1.0860526485403942e-07, + "loss": 0.7076, + "num_input_tokens_seen": 161432800, + "step": 7466 + }, + { + "epoch": 0.8978536644020922, + "flos": 15174151965600.0, + "grad_norm": 1.8670790892217008, + "learning_rate": 1.0835220480247675e-07, + "loss": 0.7744, + "num_input_tokens_seen": 161451295, + "step": 7467 + }, + { + "epoch": 0.8979739072927313, + "flos": 17978791656480.0, + "grad_norm": 4.150825574635702, + "learning_rate": 1.0809943171238067e-07, + "loss": 0.8363, + "num_input_tokens_seen": 161469220, + "step": 7468 + }, + { + "epoch": 0.8980941501833704, + "flos": 22239674530080.0, + "grad_norm": 2.7995143775795577, + "learning_rate": 1.078469456220965e-07, + "loss": 0.6292, + "num_input_tokens_seen": 161489375, + "step": 7469 + }, + { + "epoch": 0.8982143930740095, + "flos": 37506656587200.0, + "grad_norm": 5.255497733226625, + "learning_rate": 1.0759474656992606e-07, + "loss": 0.6979, + "num_input_tokens_seen": 161512145, + "step": 7470 + }, + { + "epoch": 0.8983346359646486, + "flos": 18051172646400.0, + "grad_norm": 2.8854870472023877, + "learning_rate": 1.0734283459412785e-07, + "loss": 0.7803, + "num_input_tokens_seen": 161528995, + "step": 7471 + }, + { + "epoch": 0.8984548788552876, + "flos": 20529596090880.0, + "grad_norm": 4.984206815650486, + "learning_rate": 1.0709120973291707e-07, + "loss": 0.8081, + "num_input_tokens_seen": 161548775, + "step": 7472 + }, + { + "epoch": 0.8985751217459268, + "flos": 17760161896320.0, + "grad_norm": 2.651228163079345, + "learning_rate": 1.0683987202446475e-07, + "loss": 0.7787, + "num_input_tokens_seen": 161566960, + "step": 7473 + }, + { + "epoch": 0.8986953646365659, + "flos": 21586164114240.0, + "grad_norm": 2.353007129488162, + "learning_rate": 1.0658882150689862e-07, + "loss": 0.6971, + "num_input_tokens_seen": 161585820, + "step": 7474 + }, + { + "epoch": 0.8988156075272049, + "flos": 14007544251840.0, + "grad_norm": 3.5750419294125915, + "learning_rate": 1.0633805821830288e-07, + "loss": 0.7915, + "num_input_tokens_seen": 161602505, + "step": 7475 + }, + { + "epoch": 0.8989358504178441, + "flos": 29017494453120.0, + "grad_norm": 2.827556236780404, + "learning_rate": 1.0608758219671753e-07, + "loss": 0.8301, + "num_input_tokens_seen": 161621545, + "step": 7476 + }, + { + "epoch": 0.8990560933084831, + "flos": 20201595696000.0, + "grad_norm": 2.002588028133796, + "learning_rate": 1.0583739348014065e-07, + "loss": 0.7034, + "num_input_tokens_seen": 161642630, + "step": 7477 + }, + { + "epoch": 0.8991763361991222, + "flos": 25484138454720.0, + "grad_norm": 2.493994270159052, + "learning_rate": 1.0558749210652518e-07, + "loss": 0.8457, + "num_input_tokens_seen": 161662560, + "step": 7478 + }, + { + "epoch": 0.8992965790897613, + "flos": 25083236693280.0, + "grad_norm": 2.4478236945062006, + "learning_rate": 1.053378781137808e-07, + "loss": 0.8577, + "num_input_tokens_seen": 161683480, + "step": 7479 + }, + { + "epoch": 0.8994168219804004, + "flos": 16048039120320.0, + "grad_norm": 2.106052862614263, + "learning_rate": 1.0508855153977392e-07, + "loss": 0.7773, + "num_input_tokens_seen": 161699945, + "step": 7480 + }, + { + "epoch": 0.8995370648710395, + "flos": 24790181606400.0, + "grad_norm": 21.69957983779222, + "learning_rate": 1.0483951242232669e-07, + "loss": 0.6723, + "num_input_tokens_seen": 161720420, + "step": 7481 + }, + { + "epoch": 0.8996573077616786, + "flos": 63026849709120.0, + "grad_norm": 0.9796563540531665, + "learning_rate": 1.0459076079921936e-07, + "loss": 0.5762, + "num_input_tokens_seen": 161773080, + "step": 7482 + }, + { + "epoch": 0.8997775506523177, + "flos": 18193630101120.0, + "grad_norm": 6.129244520919099, + "learning_rate": 1.0434229670818618e-07, + "loss": 0.8499, + "num_input_tokens_seen": 161789755, + "step": 7483 + }, + { + "epoch": 0.8998977935429567, + "flos": 24132062140320.0, + "grad_norm": 1.8744624320941392, + "learning_rate": 1.0409412018691944e-07, + "loss": 0.7997, + "num_input_tokens_seen": 161810220, + "step": 7484 + }, + { + "epoch": 0.9000180364335959, + "flos": 20743245103200.0, + "grad_norm": 2.2233309668662287, + "learning_rate": 1.0384623127306724e-07, + "loss": 0.75, + "num_input_tokens_seen": 161827565, + "step": 7485 + }, + { + "epoch": 0.900138279324235, + "flos": 19178300341440.0, + "grad_norm": 2.0735466796830755, + "learning_rate": 1.0359863000423397e-07, + "loss": 0.7946, + "num_input_tokens_seen": 161845700, + "step": 7486 + }, + { + "epoch": 0.900258522214874, + "flos": 28689940095360.0, + "grad_norm": 1.9345863953970868, + "learning_rate": 1.0335131641798112e-07, + "loss": 0.7211, + "num_input_tokens_seen": 161867660, + "step": 7487 + }, + { + "epoch": 0.9003787651055132, + "flos": 58197629540640.0, + "grad_norm": 0.8373952615065856, + "learning_rate": 1.0310429055182512e-07, + "loss": 0.5555, + "num_input_tokens_seen": 161921980, + "step": 7488 + }, + { + "epoch": 0.9004990079961522, + "flos": 25519864327680.0, + "grad_norm": 2.0089687476606373, + "learning_rate": 1.0285755244324024e-07, + "loss": 0.7426, + "num_input_tokens_seen": 161942725, + "step": 7489 + }, + { + "epoch": 0.9006192508867913, + "flos": 23301966696480.0, + "grad_norm": 3.2820359514906965, + "learning_rate": 1.0261110212965629e-07, + "loss": 0.6881, + "num_input_tokens_seen": 161964520, + "step": 7490 + }, + { + "epoch": 0.9007394937774305, + "flos": 18014629038720.0, + "grad_norm": 2.340224195320353, + "learning_rate": 1.023649396484596e-07, + "loss": 0.7943, + "num_input_tokens_seen": 161983575, + "step": 7491 + }, + { + "epoch": 0.9008597366680695, + "flos": 43006305145920.0, + "grad_norm": 4.937329253129529, + "learning_rate": 1.0211906503699275e-07, + "loss": 0.6768, + "num_input_tokens_seen": 162006550, + "step": 7492 + }, + { + "epoch": 0.9009799795587086, + "flos": 14917900674720.0, + "grad_norm": 3.2173832334065318, + "learning_rate": 1.0187347833255455e-07, + "loss": 0.8188, + "num_input_tokens_seen": 162022455, + "step": 7493 + }, + { + "epoch": 0.9011002224493477, + "flos": 21548765602080.0, + "grad_norm": 5.007477128676063, + "learning_rate": 1.0162817957240056e-07, + "loss": 0.7925, + "num_input_tokens_seen": 162042350, + "step": 7494 + }, + { + "epoch": 0.9012204653399868, + "flos": 71064776493120.0, + "grad_norm": 0.8820736924697071, + "learning_rate": 1.0138316879374253e-07, + "loss": 0.6282, + "num_input_tokens_seen": 162111640, + "step": 7495 + }, + { + "epoch": 0.9013407082306258, + "flos": 15572823541440.0, + "grad_norm": 2.8604125761672905, + "learning_rate": 1.0113844603374833e-07, + "loss": 0.7404, + "num_input_tokens_seen": 162128165, + "step": 7496 + }, + { + "epoch": 0.901460951121265, + "flos": 15027866025600.0, + "grad_norm": 4.834144568052758, + "learning_rate": 1.0089401132954178e-07, + "loss": 0.7194, + "num_input_tokens_seen": 162146145, + "step": 7497 + }, + { + "epoch": 0.9015811940119041, + "flos": 22205100919680.0, + "grad_norm": 1.991453746230744, + "learning_rate": 1.006498647182037e-07, + "loss": 0.725, + "num_input_tokens_seen": 162164800, + "step": 7498 + }, + { + "epoch": 0.9017014369025431, + "flos": 24938103015840.0, + "grad_norm": 7.531158066817651, + "learning_rate": 1.004060062367713e-07, + "loss": 0.7175, + "num_input_tokens_seen": 162185895, + "step": 7499 + }, + { + "epoch": 0.9018216797931822, + "flos": 18088422479520.0, + "grad_norm": 2.236908869736071, + "learning_rate": 1.0016243592223728e-07, + "loss": 0.6959, + "num_input_tokens_seen": 162206295, + "step": 7500 + }, + { + "epoch": 0.9019419226838213, + "flos": 37214790932640.0, + "grad_norm": 2.1350445927199817, + "learning_rate": 9.991915381155114e-08, + "loss": 0.6596, + "num_input_tokens_seen": 162229275, + "step": 7501 + }, + { + "epoch": 0.9020621655744604, + "flos": 23407620355200.0, + "grad_norm": 2.810823407065625, + "learning_rate": 9.967615994161871e-08, + "loss": 0.7452, + "num_input_tokens_seen": 162248935, + "step": 7502 + }, + { + "epoch": 0.9021824084650995, + "flos": 22824818290080.0, + "grad_norm": 2.234194128011674, + "learning_rate": 9.943345434930161e-08, + "loss": 0.7841, + "num_input_tokens_seen": 162269185, + "step": 7503 + }, + { + "epoch": 0.9023026513557386, + "flos": 22095990473280.0, + "grad_norm": 2.450265177175026, + "learning_rate": 9.919103707141885e-08, + "loss": 0.6898, + "num_input_tokens_seen": 162288065, + "step": 7504 + }, + { + "epoch": 0.9024228942463777, + "flos": 24164182546560.0, + "grad_norm": 5.301765277468727, + "learning_rate": 9.89489081447441e-08, + "loss": 0.7657, + "num_input_tokens_seen": 162305265, + "step": 7505 + }, + { + "epoch": 0.9025431371370167, + "flos": 24972416437920.0, + "grad_norm": 3.308951438475999, + "learning_rate": 9.870706760600844e-08, + "loss": 0.8327, + "num_input_tokens_seen": 162325215, + "step": 7506 + }, + { + "epoch": 0.9026633800276559, + "flos": 18925877535840.0, + "grad_norm": 2.1598767725797443, + "learning_rate": 9.846551549189918e-08, + "loss": 0.7283, + "num_input_tokens_seen": 162344820, + "step": 7507 + }, + { + "epoch": 0.902783622918295, + "flos": 32369470524480.0, + "grad_norm": 2.5202798367090637, + "learning_rate": 9.822425183905902e-08, + "loss": 0.6853, + "num_input_tokens_seen": 162365345, + "step": 7508 + }, + { + "epoch": 0.902903865808934, + "flos": 63626495409120.0, + "grad_norm": 0.9148057051777213, + "learning_rate": 9.798327668408823e-08, + "loss": 0.7498, + "num_input_tokens_seen": 162426980, + "step": 7509 + }, + { + "epoch": 0.9030241086995732, + "flos": 23769934172160.0, + "grad_norm": 87.09499180582227, + "learning_rate": 9.774259006354158e-08, + "loss": 0.6892, + "num_input_tokens_seen": 162447050, + "step": 7510 + }, + { + "epoch": 0.9031443515902122, + "flos": 26358062779200.0, + "grad_norm": 3.2133821337570216, + "learning_rate": 9.750219201393184e-08, + "loss": 0.76, + "num_input_tokens_seen": 162467015, + "step": 7511 + }, + { + "epoch": 0.9032645944808513, + "flos": 24903752424000.0, + "grad_norm": 2.4489536659091318, + "learning_rate": 9.726208257172697e-08, + "loss": 0.7781, + "num_input_tokens_seen": 162488710, + "step": 7512 + }, + { + "epoch": 0.9033848373714904, + "flos": 21148607235840.0, + "grad_norm": 5.909603075359696, + "learning_rate": 9.702226177335115e-08, + "loss": 0.7493, + "num_input_tokens_seen": 162507205, + "step": 7513 + }, + { + "epoch": 0.9035050802621295, + "flos": 26248692144480.0, + "grad_norm": 2.0574437209969023, + "learning_rate": 9.67827296551853e-08, + "loss": 0.7262, + "num_input_tokens_seen": 162528490, + "step": 7514 + }, + { + "epoch": 0.9036253231527686, + "flos": 24169832350080.0, + "grad_norm": 6.529815370248324, + "learning_rate": 9.65434862535659e-08, + "loss": 0.6868, + "num_input_tokens_seen": 162547860, + "step": 7515 + }, + { + "epoch": 0.9037455660434077, + "flos": 18046786614720.0, + "grad_norm": 8.168431097925874, + "learning_rate": 9.630453160478635e-08, + "loss": 0.6507, + "num_input_tokens_seen": 162563215, + "step": 7516 + }, + { + "epoch": 0.9038658089340468, + "flos": 24026371311840.0, + "grad_norm": 2.0895804266055253, + "learning_rate": 9.60658657450959e-08, + "loss": 0.8222, + "num_input_tokens_seen": 162583515, + "step": 7517 + }, + { + "epoch": 0.9039860518246858, + "flos": 21803084065440.0, + "grad_norm": 2.2709407139392273, + "learning_rate": 9.582748871069979e-08, + "loss": 0.7981, + "num_input_tokens_seen": 162602955, + "step": 7518 + }, + { + "epoch": 0.904106294715325, + "flos": 26576729709120.0, + "grad_norm": 2.8995331289757376, + "learning_rate": 9.558940053775954e-08, + "loss": 0.8382, + "num_input_tokens_seen": 162621595, + "step": 7519 + }, + { + "epoch": 0.904226537605964, + "flos": 17760422084640.0, + "grad_norm": 2.4653502607361033, + "learning_rate": 9.535160126239294e-08, + "loss": 0.6788, + "num_input_tokens_seen": 162638220, + "step": 7520 + }, + { + "epoch": 0.9043467804966031, + "flos": 24754938940320.0, + "grad_norm": 1.8355538474642792, + "learning_rate": 9.511409092067424e-08, + "loss": 0.706, + "num_input_tokens_seen": 162658575, + "step": 7521 + }, + { + "epoch": 0.9044670233872423, + "flos": 22599572312640.0, + "grad_norm": 2.212668452342708, + "learning_rate": 9.487686954863327e-08, + "loss": 0.6765, + "num_input_tokens_seen": 162678295, + "step": 7522 + }, + { + "epoch": 0.9045872662778813, + "flos": 23737404898560.0, + "grad_norm": 2.795938345637289, + "learning_rate": 9.46399371822566e-08, + "loss": 0.7703, + "num_input_tokens_seen": 162698070, + "step": 7523 + }, + { + "epoch": 0.9047075091685204, + "flos": 15169394236320.0, + "grad_norm": 2.669623513841972, + "learning_rate": 9.440329385748657e-08, + "loss": 0.7234, + "num_input_tokens_seen": 162715140, + "step": 7524 + }, + { + "epoch": 0.9048277520591596, + "flos": 17978085431040.0, + "grad_norm": 2.791947013860461, + "learning_rate": 9.416693961022137e-08, + "loss": 0.7083, + "num_input_tokens_seen": 162733390, + "step": 7525 + }, + { + "epoch": 0.9049479949497986, + "flos": 21841263142560.0, + "grad_norm": 2.1406172035242372, + "learning_rate": 9.393087447631654e-08, + "loss": 0.7715, + "num_input_tokens_seen": 162751670, + "step": 7526 + }, + { + "epoch": 0.9050682378404377, + "flos": 20744174347200.0, + "grad_norm": 5.158427570490837, + "learning_rate": 9.36950984915823e-08, + "loss": 0.731, + "num_input_tokens_seen": 162770025, + "step": 7527 + }, + { + "epoch": 0.9051884807310768, + "flos": 21549211639200.0, + "grad_norm": 2.1572508769566814, + "learning_rate": 9.345961169178607e-08, + "loss": 0.6923, + "num_input_tokens_seen": 162789710, + "step": 7528 + }, + { + "epoch": 0.9053087236217159, + "flos": 21876765996960.0, + "grad_norm": 1.6383973029147987, + "learning_rate": 9.322441411265081e-08, + "loss": 0.7282, + "num_input_tokens_seen": 162810695, + "step": 7529 + }, + { + "epoch": 0.9054289665123549, + "flos": 17031594267840.0, + "grad_norm": 3.2073399050236975, + "learning_rate": 9.298950578985554e-08, + "loss": 0.733, + "num_input_tokens_seen": 162827770, + "step": 7530 + }, + { + "epoch": 0.905549209402994, + "flos": 20747631134880.0, + "grad_norm": 2.243408827491513, + "learning_rate": 9.275488675903665e-08, + "loss": 0.7133, + "num_input_tokens_seen": 162846715, + "step": 7531 + }, + { + "epoch": 0.9056694522936332, + "flos": 21655757372160.0, + "grad_norm": 2.412494236440148, + "learning_rate": 9.252055705578454e-08, + "loss": 0.7368, + "num_input_tokens_seen": 162866215, + "step": 7532 + }, + { + "epoch": 0.9057896951842722, + "flos": 29526651756480.0, + "grad_norm": 2.3385990761646354, + "learning_rate": 9.228651671564747e-08, + "loss": 0.7226, + "num_input_tokens_seen": 162888245, + "step": 7533 + }, + { + "epoch": 0.9059099380749113, + "flos": 27852262020480.0, + "grad_norm": 1.7731710897646822, + "learning_rate": 9.205276577412901e-08, + "loss": 0.779, + "num_input_tokens_seen": 162911025, + "step": 7534 + }, + { + "epoch": 0.9060301809655504, + "flos": 17723766967680.0, + "grad_norm": 4.332582017372073, + "learning_rate": 9.181930426668905e-08, + "loss": 0.7658, + "num_input_tokens_seen": 162927820, + "step": 7535 + }, + { + "epoch": 0.9061504238561895, + "flos": 31713804262560.0, + "grad_norm": 1.839662320207863, + "learning_rate": 9.158613222874346e-08, + "loss": 0.6784, + "num_input_tokens_seen": 162949445, + "step": 7536 + }, + { + "epoch": 0.9062706667468285, + "flos": 20019509543520.0, + "grad_norm": 2.06955068715214, + "learning_rate": 9.135324969566394e-08, + "loss": 0.8197, + "num_input_tokens_seen": 162968945, + "step": 7537 + }, + { + "epoch": 0.9063909096374677, + "flos": 18410884580160.0, + "grad_norm": 2.4302449951371488, + "learning_rate": 9.112065670277913e-08, + "loss": 0.7559, + "num_input_tokens_seen": 162985310, + "step": 7538 + }, + { + "epoch": 0.9065111525281068, + "flos": 33499125763200.0, + "grad_norm": 2.268819411716088, + "learning_rate": 9.088835328537303e-08, + "loss": 0.7378, + "num_input_tokens_seen": 163006900, + "step": 7539 + }, + { + "epoch": 0.9066313954187458, + "flos": 23334458800320.0, + "grad_norm": 2.7912348133294143, + "learning_rate": 9.065633947868568e-08, + "loss": 0.7129, + "num_input_tokens_seen": 163026505, + "step": 7540 + }, + { + "epoch": 0.906751638309385, + "flos": 26211999857760.0, + "grad_norm": 2.750790395724199, + "learning_rate": 9.042461531791379e-08, + "loss": 0.8004, + "num_input_tokens_seen": 163044925, + "step": 7541 + }, + { + "epoch": 0.906871881200024, + "flos": 16520987343840.0, + "grad_norm": 2.35884896786898, + "learning_rate": 9.019318083820903e-08, + "loss": 0.7802, + "num_input_tokens_seen": 163063505, + "step": 7542 + }, + { + "epoch": 0.9069921240906631, + "flos": 24569804867520.0, + "grad_norm": 2.5592336538779734, + "learning_rate": 8.996203607468045e-08, + "loss": 0.8513, + "num_input_tokens_seen": 163082505, + "step": 7543 + }, + { + "epoch": 0.9071123669813023, + "flos": 25338410061120.0, + "grad_norm": 2.571751525065601, + "learning_rate": 8.973118106239241e-08, + "loss": 0.7559, + "num_input_tokens_seen": 163105110, + "step": 7544 + }, + { + "epoch": 0.9072326098719413, + "flos": 26687549964480.0, + "grad_norm": 3.4385700294382655, + "learning_rate": 8.95006158363656e-08, + "loss": 0.9461, + "num_input_tokens_seen": 163125295, + "step": 7545 + }, + { + "epoch": 0.9073528527625804, + "flos": 23844656856960.0, + "grad_norm": 2.1267950666564466, + "learning_rate": 8.9270340431576e-08, + "loss": 0.7727, + "num_input_tokens_seen": 163144385, + "step": 7546 + }, + { + "epoch": 0.9074730956532195, + "flos": 37797630167520.0, + "grad_norm": 2.28450146284618, + "learning_rate": 8.904035488295658e-08, + "loss": 0.7406, + "num_input_tokens_seen": 163163885, + "step": 7547 + }, + { + "epoch": 0.9075933385438586, + "flos": 65080099538880.0, + "grad_norm": 0.6693886506000376, + "learning_rate": 8.881065922539632e-08, + "loss": 0.5319, + "num_input_tokens_seen": 163224325, + "step": 7548 + }, + { + "epoch": 0.9077135814344977, + "flos": 19902927975360.0, + "grad_norm": 5.75851407169982, + "learning_rate": 8.85812534937389e-08, + "loss": 0.7361, + "num_input_tokens_seen": 163242425, + "step": 7549 + }, + { + "epoch": 0.9078338243251368, + "flos": 17504728340160.0, + "grad_norm": 3.1316687572310533, + "learning_rate": 8.835213772278583e-08, + "loss": 0.6787, + "num_input_tokens_seen": 163259280, + "step": 7550 + }, + { + "epoch": 0.9079540672157759, + "flos": 28762060896960.0, + "grad_norm": 2.1160752869536927, + "learning_rate": 8.812331194729373e-08, + "loss": 0.793, + "num_input_tokens_seen": 163278925, + "step": 7551 + }, + { + "epoch": 0.9080743101064149, + "flos": 23480224363680.0, + "grad_norm": 2.361343169184138, + "learning_rate": 8.789477620197461e-08, + "loss": 0.7227, + "num_input_tokens_seen": 163298450, + "step": 7552 + }, + { + "epoch": 0.9081945529970541, + "flos": 22745337876000.0, + "grad_norm": 4.553299288271015, + "learning_rate": 8.766653052149831e-08, + "loss": 0.7926, + "num_input_tokens_seen": 163313635, + "step": 7553 + }, + { + "epoch": 0.9083147958876931, + "flos": 18846620140320.0, + "grad_norm": 2.4847748632548647, + "learning_rate": 8.743857494048823e-08, + "loss": 0.7472, + "num_input_tokens_seen": 163330450, + "step": 7554 + }, + { + "epoch": 0.9084350387783322, + "flos": 18882197334240.0, + "grad_norm": 3.780760299877446, + "learning_rate": 8.721090949352605e-08, + "loss": 0.6292, + "num_input_tokens_seen": 163346690, + "step": 7555 + }, + { + "epoch": 0.9085552816689714, + "flos": 20566139698560.0, + "grad_norm": 2.4775659110870327, + "learning_rate": 8.698353421514793e-08, + "loss": 0.7303, + "num_input_tokens_seen": 163365065, + "step": 7556 + }, + { + "epoch": 0.9086755245596104, + "flos": 18087381726240.0, + "grad_norm": 4.5761862545788805, + "learning_rate": 8.67564491398467e-08, + "loss": 0.8046, + "num_input_tokens_seen": 163382070, + "step": 7557 + }, + { + "epoch": 0.9087957674502495, + "flos": 19102351054560.0, + "grad_norm": 2.1861158078746543, + "learning_rate": 8.652965430207104e-08, + "loss": 0.739, + "num_input_tokens_seen": 163399975, + "step": 7558 + }, + { + "epoch": 0.9089160103408886, + "flos": 18084036447840.0, + "grad_norm": 2.7072267668165098, + "learning_rate": 8.630314973622521e-08, + "loss": 0.6589, + "num_input_tokens_seen": 163417520, + "step": 7559 + }, + { + "epoch": 0.9090362532315277, + "flos": 33316890931680.0, + "grad_norm": 3.130810485572358, + "learning_rate": 8.607693547666995e-08, + "loss": 0.7113, + "num_input_tokens_seen": 163440060, + "step": 7560 + }, + { + "epoch": 0.9091564961221668, + "flos": 71378020493280.0, + "grad_norm": 0.8925715807251201, + "learning_rate": 8.585101155772201e-08, + "loss": 0.5785, + "num_input_tokens_seen": 163502180, + "step": 7561 + }, + { + "epoch": 0.9092767390128058, + "flos": 24677019656160.0, + "grad_norm": 2.7716892803811866, + "learning_rate": 8.562537801365377e-08, + "loss": 0.6856, + "num_input_tokens_seen": 163521915, + "step": 7562 + }, + { + "epoch": 0.909396981903445, + "flos": 23552754032640.0, + "grad_norm": 2.5931276241131993, + "learning_rate": 8.540003487869362e-08, + "loss": 0.7011, + "num_input_tokens_seen": 163543585, + "step": 7563 + }, + { + "epoch": 0.909517224794084, + "flos": 23370705049920.0, + "grad_norm": 2.476101805379627, + "learning_rate": 8.517498218702557e-08, + "loss": 0.8014, + "num_input_tokens_seen": 163561515, + "step": 7564 + }, + { + "epoch": 0.9096374676847231, + "flos": 19181273922240.0, + "grad_norm": 3.264398225763393, + "learning_rate": 8.49502199727905e-08, + "loss": 0.696, + "num_input_tokens_seen": 163579540, + "step": 7565 + }, + { + "epoch": 0.9097577105753623, + "flos": 33244658620800.0, + "grad_norm": 3.4020561508397824, + "learning_rate": 8.472574827008428e-08, + "loss": 0.6594, + "num_input_tokens_seen": 163600425, + "step": 7566 + }, + { + "epoch": 0.9098779534660013, + "flos": 21875093357760.0, + "grad_norm": 2.029247764308892, + "learning_rate": 8.450156711295942e-08, + "loss": 0.8388, + "num_input_tokens_seen": 163620595, + "step": 7567 + }, + { + "epoch": 0.9099981963566404, + "flos": 25693178416800.0, + "grad_norm": 3.50709278225216, + "learning_rate": 8.427767653542383e-08, + "loss": 0.8679, + "num_input_tokens_seen": 163635765, + "step": 7568 + }, + { + "epoch": 0.9101184392472795, + "flos": 21039385280160.0, + "grad_norm": 3.224396724609256, + "learning_rate": 8.405407657144125e-08, + "loss": 0.7061, + "num_input_tokens_seen": 163654925, + "step": 7569 + }, + { + "epoch": 0.9102386821379186, + "flos": 24717131560800.0, + "grad_norm": 2.3739298366986064, + "learning_rate": 8.383076725493232e-08, + "loss": 0.7254, + "num_input_tokens_seen": 163672245, + "step": 7570 + }, + { + "epoch": 0.9103589250285576, + "flos": 22530499431360.0, + "grad_norm": 2.3603520675488183, + "learning_rate": 8.360774861977216e-08, + "loss": 0.6787, + "num_input_tokens_seen": 163691365, + "step": 7571 + }, + { + "epoch": 0.9104791679191968, + "flos": 25337109119520.0, + "grad_norm": 2.1645638336194275, + "learning_rate": 8.338502069979281e-08, + "loss": 0.7441, + "num_input_tokens_seen": 163711675, + "step": 7572 + }, + { + "epoch": 0.9105994108098359, + "flos": 14406215827680.0, + "grad_norm": 3.5812916964621753, + "learning_rate": 8.316258352878214e-08, + "loss": 0.8007, + "num_input_tokens_seen": 163728095, + "step": 7573 + }, + { + "epoch": 0.9107196537004749, + "flos": 26679781484640.0, + "grad_norm": 2.223720409760389, + "learning_rate": 8.294043714048338e-08, + "loss": 0.7149, + "num_input_tokens_seen": 163747175, + "step": 7574 + }, + { + "epoch": 0.9108398965911141, + "flos": 66437788487040.0, + "grad_norm": 0.7642953576356857, + "learning_rate": 8.271858156859624e-08, + "loss": 0.6038, + "num_input_tokens_seen": 163812545, + "step": 7575 + }, + { + "epoch": 0.9109601394817531, + "flos": 25374693480480.0, + "grad_norm": 1.9476375549863638, + "learning_rate": 8.249701684677557e-08, + "loss": 0.7386, + "num_input_tokens_seen": 163833870, + "step": 7576 + }, + { + "epoch": 0.9110803823723922, + "flos": 22715298976320.0, + "grad_norm": 2.256366834495021, + "learning_rate": 8.227574300863294e-08, + "loss": 0.8139, + "num_input_tokens_seen": 163854040, + "step": 7577 + }, + { + "epoch": 0.9112006252630314, + "flos": 48399965518080.0, + "grad_norm": 2.1138673769836145, + "learning_rate": 8.205476008773548e-08, + "loss": 0.6939, + "num_input_tokens_seen": 163878040, + "step": 7578 + }, + { + "epoch": 0.9113208681536704, + "flos": 29966624669280.0, + "grad_norm": 13.51605343883796, + "learning_rate": 8.183406811760596e-08, + "loss": 0.8299, + "num_input_tokens_seen": 163897770, + "step": 7579 + }, + { + "epoch": 0.9114411110443095, + "flos": 25558377932640.0, + "grad_norm": 2.0245984844994576, + "learning_rate": 8.161366713172313e-08, + "loss": 0.7401, + "num_input_tokens_seen": 163920160, + "step": 7580 + }, + { + "epoch": 0.9115613539349486, + "flos": 18373337388960.0, + "grad_norm": 5.744785764200655, + "learning_rate": 8.139355716352137e-08, + "loss": 0.8428, + "num_input_tokens_seen": 163935000, + "step": 7581 + }, + { + "epoch": 0.9116815968255877, + "flos": 21695163051360.0, + "grad_norm": 3.1373936365343775, + "learning_rate": 8.117373824639196e-08, + "loss": 0.7032, + "num_input_tokens_seen": 163955265, + "step": 7582 + }, + { + "epoch": 0.9118018397162267, + "flos": 65270325868800.0, + "grad_norm": 0.73287933305126, + "learning_rate": 8.095421041368067e-08, + "loss": 0.592, + "num_input_tokens_seen": 164014680, + "step": 7583 + }, + { + "epoch": 0.9119220826068659, + "flos": 20893025000640.0, + "grad_norm": 8.484895187926464, + "learning_rate": 8.073497369868999e-08, + "loss": 0.7079, + "num_input_tokens_seen": 164033140, + "step": 7584 + }, + { + "epoch": 0.912042325497505, + "flos": 28325767790400.0, + "grad_norm": 2.1186464357524506, + "learning_rate": 8.051602813467772e-08, + "loss": 0.7582, + "num_input_tokens_seen": 164054995, + "step": 7585 + }, + { + "epoch": 0.912162568388144, + "flos": 17541829494240.0, + "grad_norm": 1.9514280958599954, + "learning_rate": 8.029737375485756e-08, + "loss": 0.7138, + "num_input_tokens_seen": 164074225, + "step": 7586 + }, + { + "epoch": 0.9122828112787832, + "flos": 19799616011520.0, + "grad_norm": 2.3596652022995466, + "learning_rate": 8.007901059239986e-08, + "loss": 0.7317, + "num_input_tokens_seen": 164093505, + "step": 7587 + }, + { + "epoch": 0.9124030541694222, + "flos": 20784100403040.0, + "grad_norm": 1.9970279306703964, + "learning_rate": 7.986093868042964e-08, + "loss": 0.8035, + "num_input_tokens_seen": 164113750, + "step": 7588 + }, + { + "epoch": 0.9125232970600613, + "flos": 25156249569120.0, + "grad_norm": 2.329787204995929, + "learning_rate": 7.964315805202826e-08, + "loss": 0.6769, + "num_input_tokens_seen": 164134330, + "step": 7589 + }, + { + "epoch": 0.9126435399507005, + "flos": 19691509148640.0, + "grad_norm": 2.0407693668723685, + "learning_rate": 7.942566874023304e-08, + "loss": 0.7355, + "num_input_tokens_seen": 164153515, + "step": 7590 + }, + { + "epoch": 0.9127637828413395, + "flos": 19545743585280.0, + "grad_norm": 2.6312164993146934, + "learning_rate": 7.920847077803649e-08, + "loss": 0.6989, + "num_input_tokens_seen": 164171305, + "step": 7591 + }, + { + "epoch": 0.9128840257319786, + "flos": 20201818714560.0, + "grad_norm": 2.8237457934614634, + "learning_rate": 7.899156419838826e-08, + "loss": 0.8225, + "num_input_tokens_seen": 164190275, + "step": 7592 + }, + { + "epoch": 0.9130042686226177, + "flos": 24822896728800.0, + "grad_norm": 2.4388639106523793, + "learning_rate": 7.87749490341918e-08, + "loss": 0.6536, + "num_input_tokens_seen": 164210550, + "step": 7593 + }, + { + "epoch": 0.9131245115132568, + "flos": 23547847624320.0, + "grad_norm": 3.0680398465071894, + "learning_rate": 7.855862531830836e-08, + "loss": 0.8348, + "num_input_tokens_seen": 164226660, + "step": 7594 + }, + { + "epoch": 0.9132447544038959, + "flos": 19902965145120.0, + "grad_norm": 2.1671486342946116, + "learning_rate": 7.834259308355373e-08, + "loss": 0.7272, + "num_input_tokens_seen": 164245425, + "step": 7595 + }, + { + "epoch": 0.9133649972945349, + "flos": 21950373588960.0, + "grad_norm": 3.006471012738931, + "learning_rate": 7.812685236269989e-08, + "loss": 0.753, + "num_input_tokens_seen": 164264085, + "step": 7596 + }, + { + "epoch": 0.9134852401851741, + "flos": 71138867292000.0, + "grad_norm": 0.7992015982210944, + "learning_rate": 7.791140318847445e-08, + "loss": 0.5857, + "num_input_tokens_seen": 164322220, + "step": 7597 + }, + { + "epoch": 0.9136054830758131, + "flos": 23593274804640.0, + "grad_norm": 1.7484403199419423, + "learning_rate": 7.769624559356081e-08, + "loss": 0.8057, + "num_input_tokens_seen": 164345615, + "step": 7598 + }, + { + "epoch": 0.9137257259664522, + "flos": 23405278660320.0, + "grad_norm": 3.780501614747234, + "learning_rate": 7.748137961059842e-08, + "loss": 0.7577, + "num_input_tokens_seen": 164359595, + "step": 7599 + }, + { + "epoch": 0.9138459688570914, + "flos": 19100343887520.0, + "grad_norm": 3.097779395408457, + "learning_rate": 7.726680527218211e-08, + "loss": 0.6608, + "num_input_tokens_seen": 164376635, + "step": 7600 + }, + { + "epoch": 0.9139662117477304, + "flos": 46214485651200.0, + "grad_norm": 2.3109479283632077, + "learning_rate": 7.70525226108627e-08, + "loss": 0.7598, + "num_input_tokens_seen": 164400305, + "step": 7601 + }, + { + "epoch": 0.9140864546383695, + "flos": 22241384339040.0, + "grad_norm": 2.8596388064567795, + "learning_rate": 7.683853165914666e-08, + "loss": 0.8002, + "num_input_tokens_seen": 164418075, + "step": 7602 + }, + { + "epoch": 0.9142066975290086, + "flos": 17505100037760.0, + "grad_norm": 2.244547807822944, + "learning_rate": 7.662483244949602e-08, + "loss": 0.7731, + "num_input_tokens_seen": 164435335, + "step": 7603 + }, + { + "epoch": 0.9143269404196477, + "flos": 17686963171680.0, + "grad_norm": 2.880899430335943, + "learning_rate": 7.641142501432951e-08, + "loss": 0.8048, + "num_input_tokens_seen": 164451530, + "step": 7604 + }, + { + "epoch": 0.9144471833102867, + "flos": 33275329406400.0, + "grad_norm": 5.203551312185258, + "learning_rate": 7.619830938602013e-08, + "loss": 0.7363, + "num_input_tokens_seen": 164472755, + "step": 7605 + }, + { + "epoch": 0.9145674262009259, + "flos": 21038902073280.0, + "grad_norm": 2.51979997961264, + "learning_rate": 7.598548559689777e-08, + "loss": 0.8242, + "num_input_tokens_seen": 164491155, + "step": 7606 + }, + { + "epoch": 0.914687669091565, + "flos": 16776420900000.0, + "grad_norm": 2.699927053525771, + "learning_rate": 7.577295367924751e-08, + "loss": 0.8112, + "num_input_tokens_seen": 164507555, + "step": 7607 + }, + { + "epoch": 0.914807911982204, + "flos": 25736635599840.0, + "grad_norm": 2.2764314947614386, + "learning_rate": 7.556071366531002e-08, + "loss": 0.8215, + "num_input_tokens_seen": 164528525, + "step": 7608 + }, + { + "epoch": 0.9149281548728432, + "flos": 19180679206080.0, + "grad_norm": 2.576181314840078, + "learning_rate": 7.53487655872822e-08, + "loss": 0.7938, + "num_input_tokens_seen": 164547695, + "step": 7609 + }, + { + "epoch": 0.9150483977634822, + "flos": 26832051756000.0, + "grad_norm": 11.889246054856871, + "learning_rate": 7.513710947731656e-08, + "loss": 0.7405, + "num_input_tokens_seen": 164568905, + "step": 7610 + }, + { + "epoch": 0.9151686406541213, + "flos": 21877695240960.0, + "grad_norm": 2.3526740504547385, + "learning_rate": 7.492574536752095e-08, + "loss": 0.8528, + "num_input_tokens_seen": 164588895, + "step": 7611 + }, + { + "epoch": 0.9152888835447605, + "flos": 27269422785600.0, + "grad_norm": 2.38439228656547, + "learning_rate": 7.471467328995907e-08, + "loss": 0.784, + "num_input_tokens_seen": 164607705, + "step": 7612 + }, + { + "epoch": 0.9154091264353995, + "flos": 13351877989920.0, + "grad_norm": 5.088031577941981, + "learning_rate": 7.450389327665018e-08, + "loss": 0.608, + "num_input_tokens_seen": 164625540, + "step": 7613 + }, + { + "epoch": 0.9155293693260386, + "flos": 20164234353600.0, + "grad_norm": 3.194987758559295, + "learning_rate": 7.429340535957029e-08, + "loss": 0.6824, + "num_input_tokens_seen": 164640735, + "step": 7614 + }, + { + "epoch": 0.9156496122166777, + "flos": 19327448352960.0, + "grad_norm": 3.5198883654491957, + "learning_rate": 7.40832095706494e-08, + "loss": 0.7123, + "num_input_tokens_seen": 164657430, + "step": 7615 + }, + { + "epoch": 0.9157698551073168, + "flos": 21077229829440.0, + "grad_norm": 2.6162343581270173, + "learning_rate": 7.387330594177443e-08, + "loss": 0.8045, + "num_input_tokens_seen": 164679505, + "step": 7616 + }, + { + "epoch": 0.9158900979979558, + "flos": 25157141643360.0, + "grad_norm": 2.4174784469827184, + "learning_rate": 7.366369450478749e-08, + "loss": 0.7945, + "num_input_tokens_seen": 164700615, + "step": 7617 + }, + { + "epoch": 0.916010340888595, + "flos": 30103097792640.0, + "grad_norm": 2.2338751387165705, + "learning_rate": 7.345437529148646e-08, + "loss": 0.6646, + "num_input_tokens_seen": 164719625, + "step": 7618 + }, + { + "epoch": 0.9161305837792341, + "flos": 17067543159360.0, + "grad_norm": 4.819075019057769, + "learning_rate": 7.324534833362483e-08, + "loss": 0.7285, + "num_input_tokens_seen": 164737200, + "step": 7619 + }, + { + "epoch": 0.9162508266698731, + "flos": 22860990200160.0, + "grad_norm": 2.315532448755628, + "learning_rate": 7.303661366291192e-08, + "loss": 0.6855, + "num_input_tokens_seen": 164757870, + "step": 7620 + }, + { + "epoch": 0.9163710695605123, + "flos": 19946347988640.0, + "grad_norm": 2.1628683493130043, + "learning_rate": 7.28281713110126e-08, + "loss": 0.8176, + "num_input_tokens_seen": 164775945, + "step": 7621 + }, + { + "epoch": 0.9164913124511513, + "flos": 22751247867840.0, + "grad_norm": 5.530785492611624, + "learning_rate": 7.262002130954759e-08, + "loss": 0.7785, + "num_input_tokens_seen": 164794400, + "step": 7622 + }, + { + "epoch": 0.9166115553417904, + "flos": 24865238819040.0, + "grad_norm": 2.0463845633675417, + "learning_rate": 7.241216369009296e-08, + "loss": 0.7927, + "num_input_tokens_seen": 164814585, + "step": 7623 + }, + { + "epoch": 0.9167317982324296, + "flos": 25666113097920.0, + "grad_norm": 2.24662930945213, + "learning_rate": 7.220459848418037e-08, + "loss": 0.664, + "num_input_tokens_seen": 164834660, + "step": 7624 + }, + { + "epoch": 0.9168520411230686, + "flos": 15610668090720.0, + "grad_norm": 2.10192448217772, + "learning_rate": 7.199732572329708e-08, + "loss": 0.7978, + "num_input_tokens_seen": 164852560, + "step": 7625 + }, + { + "epoch": 0.9169722840137077, + "flos": 30214512764160.0, + "grad_norm": 2.9619815163457432, + "learning_rate": 7.179034543888684e-08, + "loss": 0.7566, + "num_input_tokens_seen": 164872065, + "step": 7626 + }, + { + "epoch": 0.9170925269043467, + "flos": 22459493722560.0, + "grad_norm": 3.119546480473636, + "learning_rate": 7.158365766234808e-08, + "loss": 0.7792, + "num_input_tokens_seen": 164890915, + "step": 7627 + }, + { + "epoch": 0.9172127697949859, + "flos": 22861213218720.0, + "grad_norm": 2.887693496099784, + "learning_rate": 7.137726242503527e-08, + "loss": 0.7255, + "num_input_tokens_seen": 164909835, + "step": 7628 + }, + { + "epoch": 0.917333012685625, + "flos": 17426883395520.0, + "grad_norm": 3.2020033942829467, + "learning_rate": 7.11711597582585e-08, + "loss": 0.7812, + "num_input_tokens_seen": 164927195, + "step": 7629 + }, + { + "epoch": 0.917453255576264, + "flos": 14298703680960.0, + "grad_norm": 2.131113794175042, + "learning_rate": 7.096534969328271e-08, + "loss": 0.8002, + "num_input_tokens_seen": 164944640, + "step": 7630 + }, + { + "epoch": 0.9175734984669032, + "flos": 20711607903840.0, + "grad_norm": 2.3320164256687006, + "learning_rate": 7.075983226132987e-08, + "loss": 0.8459, + "num_input_tokens_seen": 164963570, + "step": 7631 + }, + { + "epoch": 0.9176937413575422, + "flos": 14809533623520.0, + "grad_norm": 3.068828816382604, + "learning_rate": 7.055460749357656e-08, + "loss": 0.7928, + "num_input_tokens_seen": 164979850, + "step": 7632 + }, + { + "epoch": 0.9178139842481813, + "flos": 18448171583040.0, + "grad_norm": 5.84576903362444, + "learning_rate": 7.034967542115521e-08, + "loss": 0.7024, + "num_input_tokens_seen": 164998115, + "step": 7633 + }, + { + "epoch": 0.9179342271388204, + "flos": 20018877657600.0, + "grad_norm": 2.9372266276968095, + "learning_rate": 7.014503607515388e-08, + "loss": 0.7542, + "num_input_tokens_seen": 165017420, + "step": 7634 + }, + { + "epoch": 0.9180544700294595, + "flos": 24640736236800.0, + "grad_norm": 2.7577581211164515, + "learning_rate": 6.994068948661592e-08, + "loss": 0.6872, + "num_input_tokens_seen": 165035575, + "step": 7635 + }, + { + "epoch": 0.9181747129200986, + "flos": 16703742552000.0, + "grad_norm": 3.1660794834707318, + "learning_rate": 6.973663568654142e-08, + "loss": 0.7689, + "num_input_tokens_seen": 165053280, + "step": 7636 + }, + { + "epoch": 0.9182949558107377, + "flos": 24236600706240.0, + "grad_norm": 2.526986381346126, + "learning_rate": 6.953287470588386e-08, + "loss": 0.66, + "num_input_tokens_seen": 165071235, + "step": 7637 + }, + { + "epoch": 0.9184151987013768, + "flos": 22050265934880.0, + "grad_norm": 4.881280814450893, + "learning_rate": 6.932940657555452e-08, + "loss": 0.8614, + "num_input_tokens_seen": 165087365, + "step": 7638 + }, + { + "epoch": 0.9185354415920158, + "flos": 32120727525120.0, + "grad_norm": 10.606428363583849, + "learning_rate": 6.912623132641938e-08, + "loss": 0.7659, + "num_input_tokens_seen": 165112455, + "step": 7639 + }, + { + "epoch": 0.918655684482655, + "flos": 20967933534240.0, + "grad_norm": 3.5553313290571107, + "learning_rate": 6.892334898929952e-08, + "loss": 0.7679, + "num_input_tokens_seen": 165132700, + "step": 7640 + }, + { + "epoch": 0.918775927373294, + "flos": 15538695968160.0, + "grad_norm": 4.1102855529342675, + "learning_rate": 6.872075959497236e-08, + "loss": 0.8475, + "num_input_tokens_seen": 165151475, + "step": 7641 + }, + { + "epoch": 0.9188961702639331, + "flos": 29891047080000.0, + "grad_norm": 2.2442471816725242, + "learning_rate": 6.85184631741702e-08, + "loss": 0.8298, + "num_input_tokens_seen": 165172040, + "step": 7642 + }, + { + "epoch": 0.9190164131545723, + "flos": 20671607508480.0, + "grad_norm": 3.22795462883471, + "learning_rate": 6.831645975758161e-08, + "loss": 0.7752, + "num_input_tokens_seen": 165189980, + "step": 7643 + }, + { + "epoch": 0.9191366560452113, + "flos": 25593211731360.0, + "grad_norm": 2.3882526269956372, + "learning_rate": 6.811474937585026e-08, + "loss": 0.6744, + "num_input_tokens_seen": 165209770, + "step": 7644 + }, + { + "epoch": 0.9192568989358504, + "flos": 21403743433920.0, + "grad_norm": 2.6177154464007724, + "learning_rate": 6.79133320595755e-08, + "loss": 0.7933, + "num_input_tokens_seen": 165229690, + "step": 7645 + }, + { + "epoch": 0.9193771418264896, + "flos": 23151926610720.0, + "grad_norm": 2.2583647055986447, + "learning_rate": 6.771220783931198e-08, + "loss": 0.7533, + "num_input_tokens_seen": 165249850, + "step": 7646 + }, + { + "epoch": 0.9194973847171286, + "flos": 70482234616320.0, + "grad_norm": 0.8345654002481845, + "learning_rate": 6.751137674556994e-08, + "loss": 0.6445, + "num_input_tokens_seen": 165310235, + "step": 7647 + }, + { + "epoch": 0.9196176276077677, + "flos": 14699865630720.0, + "grad_norm": 2.7644473224712134, + "learning_rate": 6.731083880881572e-08, + "loss": 0.7814, + "num_input_tokens_seen": 165326455, + "step": 7648 + }, + { + "epoch": 0.9197378704984068, + "flos": 23261594603520.0, + "grad_norm": 2.648225578899169, + "learning_rate": 6.711059405947072e-08, + "loss": 0.8112, + "num_input_tokens_seen": 165344645, + "step": 7649 + }, + { + "epoch": 0.9198581133890459, + "flos": 20273604988320.0, + "grad_norm": 2.235902810777325, + "learning_rate": 6.691064252791156e-08, + "loss": 0.77, + "num_input_tokens_seen": 165364120, + "step": 7650 + }, + { + "epoch": 0.9199783562796849, + "flos": 17650531073280.0, + "grad_norm": 2.4627276010428916, + "learning_rate": 6.67109842444713e-08, + "loss": 0.7819, + "num_input_tokens_seen": 165383050, + "step": 7651 + }, + { + "epoch": 0.9200985991703241, + "flos": 17650902770880.0, + "grad_norm": 2.5107271838695038, + "learning_rate": 6.651161923943704e-08, + "loss": 0.7658, + "num_input_tokens_seen": 165400955, + "step": 7652 + }, + { + "epoch": 0.9202188420609632, + "flos": 20966632592640.0, + "grad_norm": 2.6246018789019403, + "learning_rate": 6.631254754305326e-08, + "loss": 0.7719, + "num_input_tokens_seen": 165420415, + "step": 7653 + }, + { + "epoch": 0.9203390849516022, + "flos": 13898582484480.0, + "grad_norm": 3.575687961140734, + "learning_rate": 6.611376918551848e-08, + "loss": 0.7801, + "num_input_tokens_seen": 165439200, + "step": 7654 + }, + { + "epoch": 0.9204593278422414, + "flos": 21145596485280.0, + "grad_norm": 2.757434332749607, + "learning_rate": 6.591528419698744e-08, + "loss": 0.7947, + "num_input_tokens_seen": 165458985, + "step": 7655 + }, + { + "epoch": 0.9205795707328804, + "flos": 14480864172960.0, + "grad_norm": 2.867291773425341, + "learning_rate": 6.571709260756986e-08, + "loss": 0.8348, + "num_input_tokens_seen": 165475630, + "step": 7656 + }, + { + "epoch": 0.9206998136235195, + "flos": 22385477263200.0, + "grad_norm": 3.071003688053561, + "learning_rate": 6.551919444733122e-08, + "loss": 0.7659, + "num_input_tokens_seen": 165493445, + "step": 7657 + }, + { + "epoch": 0.9208200565141585, + "flos": 53281309157280.0, + "grad_norm": 2.958653678921694, + "learning_rate": 6.53215897462931e-08, + "loss": 0.6573, + "num_input_tokens_seen": 165517200, + "step": 7658 + }, + { + "epoch": 0.9209402994047977, + "flos": 30545115042240.0, + "grad_norm": 2.9340693142952277, + "learning_rate": 6.512427853443103e-08, + "loss": 0.7539, + "num_input_tokens_seen": 165539280, + "step": 7659 + }, + { + "epoch": 0.9210605422954368, + "flos": 29090916196320.0, + "grad_norm": 1.8017113656360877, + "learning_rate": 6.492726084167799e-08, + "loss": 0.7583, + "num_input_tokens_seen": 165561395, + "step": 7660 + }, + { + "epoch": 0.9211807851860758, + "flos": 54775874362560.0, + "grad_norm": 0.787642494812559, + "learning_rate": 6.473053669792072e-08, + "loss": 0.5744, + "num_input_tokens_seen": 165616085, + "step": 7661 + }, + { + "epoch": 0.921301028076715, + "flos": 19174397516640.0, + "grad_norm": 3.055927776052427, + "learning_rate": 6.453410613300248e-08, + "loss": 0.7313, + "num_input_tokens_seen": 165634015, + "step": 7662 + }, + { + "epoch": 0.921421270967354, + "flos": 27487420659840.0, + "grad_norm": 2.5437913258637774, + "learning_rate": 6.43379691767214e-08, + "loss": 0.5834, + "num_input_tokens_seen": 165653220, + "step": 7663 + }, + { + "epoch": 0.9215415138579931, + "flos": 70108807041120.0, + "grad_norm": 0.7447267161683555, + "learning_rate": 6.414212585883105e-08, + "loss": 0.5499, + "num_input_tokens_seen": 165715850, + "step": 7664 + }, + { + "epoch": 0.9216617567486323, + "flos": 35501924761440.0, + "grad_norm": 1.792292315636582, + "learning_rate": 6.394657620904143e-08, + "loss": 0.696, + "num_input_tokens_seen": 165738960, + "step": 7665 + }, + { + "epoch": 0.9217819996392713, + "flos": 29490925883520.0, + "grad_norm": 4.293441010619562, + "learning_rate": 6.375132025701657e-08, + "loss": 0.7191, + "num_input_tokens_seen": 165761260, + "step": 7666 + }, + { + "epoch": 0.9219022425299104, + "flos": 14554062897600.0, + "grad_norm": 3.8217709493626235, + "learning_rate": 6.355635803237724e-08, + "loss": 0.6931, + "num_input_tokens_seen": 165776960, + "step": 7667 + }, + { + "epoch": 0.9220224854205495, + "flos": 18051804532320.0, + "grad_norm": 3.0000945712124247, + "learning_rate": 6.336168956469867e-08, + "loss": 0.799, + "num_input_tokens_seen": 165795435, + "step": 7668 + }, + { + "epoch": 0.9221427283111886, + "flos": 24754976110080.0, + "grad_norm": 2.3274889595597568, + "learning_rate": 6.316731488351168e-08, + "loss": 0.7185, + "num_input_tokens_seen": 165816625, + "step": 7669 + }, + { + "epoch": 0.9222629712018277, + "flos": 13825495269120.0, + "grad_norm": 2.5044346577655183, + "learning_rate": 6.297323401830334e-08, + "loss": 0.634, + "num_input_tokens_seen": 165835410, + "step": 7670 + }, + { + "epoch": 0.9223832140924668, + "flos": 21585309209760.0, + "grad_norm": 5.823468104293072, + "learning_rate": 6.277944699851523e-08, + "loss": 0.6918, + "num_input_tokens_seen": 165854930, + "step": 7671 + }, + { + "epoch": 0.9225034569831059, + "flos": 21111914949120.0, + "grad_norm": 2.65401822747509, + "learning_rate": 6.25859538535447e-08, + "loss": 0.7346, + "num_input_tokens_seen": 165875725, + "step": 7672 + }, + { + "epoch": 0.9226236998737449, + "flos": 12477470458560.0, + "grad_norm": 3.5340829473928217, + "learning_rate": 6.239275461274474e-08, + "loss": 0.7784, + "num_input_tokens_seen": 165892100, + "step": 7673 + }, + { + "epoch": 0.9227439427643841, + "flos": 26176534173120.0, + "grad_norm": 2.0914933298740093, + "learning_rate": 6.219984930542299e-08, + "loss": 0.8622, + "num_input_tokens_seen": 165912840, + "step": 7674 + }, + { + "epoch": 0.9228641856550232, + "flos": 17942025030240.0, + "grad_norm": 2.4081924326861173, + "learning_rate": 6.200723796084383e-08, + "loss": 0.7636, + "num_input_tokens_seen": 165930935, + "step": 7675 + }, + { + "epoch": 0.9229844285456622, + "flos": 70319519642400.0, + "grad_norm": 0.7754478145235283, + "learning_rate": 6.181492060822546e-08, + "loss": 0.6293, + "num_input_tokens_seen": 165991110, + "step": 7676 + }, + { + "epoch": 0.9231046714363014, + "flos": 17942285218560.0, + "grad_norm": 2.565135716794523, + "learning_rate": 6.162289727674274e-08, + "loss": 0.8224, + "num_input_tokens_seen": 166008790, + "step": 7677 + }, + { + "epoch": 0.9232249143269404, + "flos": 17832840244320.0, + "grad_norm": 2.8064802385453747, + "learning_rate": 6.143116799552527e-08, + "loss": 0.8794, + "num_input_tokens_seen": 166025035, + "step": 7678 + }, + { + "epoch": 0.9233451572175795, + "flos": 23370928068480.0, + "grad_norm": 2.8798692353020066, + "learning_rate": 6.123973279365802e-08, + "loss": 0.5573, + "num_input_tokens_seen": 166044010, + "step": 7679 + }, + { + "epoch": 0.9234654001082186, + "flos": 17973550720320.0, + "grad_norm": 2.32199477532429, + "learning_rate": 6.10485917001824e-08, + "loss": 0.7831, + "num_input_tokens_seen": 166061865, + "step": 7680 + }, + { + "epoch": 0.9235856429988577, + "flos": 24715198733280.0, + "grad_norm": 1.9075149341563595, + "learning_rate": 6.085774474409322e-08, + "loss": 0.8108, + "num_input_tokens_seen": 166082425, + "step": 7681 + }, + { + "epoch": 0.9237058858894968, + "flos": 14079962411520.0, + "grad_norm": 3.265521034155456, + "learning_rate": 6.066719195434267e-08, + "loss": 0.7026, + "num_input_tokens_seen": 166100225, + "step": 7682 + }, + { + "epoch": 0.9238261287801359, + "flos": 28653359317920.0, + "grad_norm": 5.252220229296952, + "learning_rate": 6.047693335983717e-08, + "loss": 0.6676, + "num_input_tokens_seen": 166122570, + "step": 7683 + }, + { + "epoch": 0.923946371670775, + "flos": 23078802225600.0, + "grad_norm": 3.1310850789083173, + "learning_rate": 6.028696898943853e-08, + "loss": 0.8266, + "num_input_tokens_seen": 166141630, + "step": 7684 + }, + { + "epoch": 0.924066614561414, + "flos": 21835724848320.0, + "grad_norm": 27.466536813570983, + "learning_rate": 6.00972988719648e-08, + "loss": 0.7061, + "num_input_tokens_seen": 166159865, + "step": 7685 + }, + { + "epoch": 0.9241868574520532, + "flos": 28470975807360.0, + "grad_norm": 5.9155980967510935, + "learning_rate": 5.990792303618807e-08, + "loss": 0.7014, + "num_input_tokens_seen": 166179445, + "step": 7686 + }, + { + "epoch": 0.9243071003426923, + "flos": 30474481031040.0, + "grad_norm": 2.386643828419505, + "learning_rate": 5.971884151083695e-08, + "loss": 0.6971, + "num_input_tokens_seen": 166201855, + "step": 7687 + }, + { + "epoch": 0.9244273432333313, + "flos": 28617224577600.0, + "grad_norm": 2.5963621659125353, + "learning_rate": 5.9530054324595124e-08, + "loss": 0.7465, + "num_input_tokens_seen": 166221970, + "step": 7688 + }, + { + "epoch": 0.9245475861239704, + "flos": 66135812657760.0, + "grad_norm": 0.7318346848692201, + "learning_rate": 5.934156150610103e-08, + "loss": 0.5748, + "num_input_tokens_seen": 166279335, + "step": 7689 + }, + { + "epoch": 0.9246678290146095, + "flos": 24205112185920.0, + "grad_norm": 23.68373774541827, + "learning_rate": 5.915336308394914e-08, + "loss": 0.7887, + "num_input_tokens_seen": 166298170, + "step": 7690 + }, + { + "epoch": 0.9247880719052486, + "flos": 18961752087840.0, + "grad_norm": 1.8373250060541015, + "learning_rate": 5.89654590866886e-08, + "loss": 0.7721, + "num_input_tokens_seen": 166317670, + "step": 7691 + }, + { + "epoch": 0.9249083147958876, + "flos": 23989716194880.0, + "grad_norm": 2.5254703045966336, + "learning_rate": 5.877784954282483e-08, + "loss": 0.8864, + "num_input_tokens_seen": 166337010, + "step": 7692 + }, + { + "epoch": 0.9250285576865268, + "flos": 30729765908160.0, + "grad_norm": 2.4172570285771866, + "learning_rate": 5.8590534480817963e-08, + "loss": 0.7254, + "num_input_tokens_seen": 166358735, + "step": 7693 + }, + { + "epoch": 0.9251488005771659, + "flos": 10618690044960.0, + "grad_norm": 2.7284597010195863, + "learning_rate": 5.840351392908349e-08, + "loss": 0.7295, + "num_input_tokens_seen": 166374455, + "step": 7694 + }, + { + "epoch": 0.9252690434678049, + "flos": 23552679693120.0, + "grad_norm": 5.1596021448569935, + "learning_rate": 5.821678791599205e-08, + "loss": 0.7076, + "num_input_tokens_seen": 166393900, + "step": 7695 + }, + { + "epoch": 0.9253892863584441, + "flos": 21438465723360.0, + "grad_norm": 2.9516686779369334, + "learning_rate": 5.803035646986965e-08, + "loss": 0.8103, + "num_input_tokens_seen": 166413235, + "step": 7696 + }, + { + "epoch": 0.9255095292490831, + "flos": 17431864143360.0, + "grad_norm": 2.585174724643113, + "learning_rate": 5.7844219618998766e-08, + "loss": 0.6774, + "num_input_tokens_seen": 166427560, + "step": 7697 + }, + { + "epoch": 0.9256297721397222, + "flos": 24715124393760.0, + "grad_norm": 2.2000547618986177, + "learning_rate": 5.765837739161505e-08, + "loss": 0.7173, + "num_input_tokens_seen": 166446680, + "step": 7698 + }, + { + "epoch": 0.9257500150303614, + "flos": 23079173923200.0, + "grad_norm": 1.9211970428336023, + "learning_rate": 5.7472829815911504e-08, + "loss": 0.7466, + "num_input_tokens_seen": 166465505, + "step": 7699 + }, + { + "epoch": 0.9258702579210004, + "flos": 22532432258880.0, + "grad_norm": 2.749254125533894, + "learning_rate": 5.7287576920035164e-08, + "loss": 0.8159, + "num_input_tokens_seen": 166484520, + "step": 7700 + }, + { + "epoch": 0.9259905008116395, + "flos": 29961086375040.0, + "grad_norm": 3.2995506385262114, + "learning_rate": 5.7102618732088435e-08, + "loss": 0.7678, + "num_input_tokens_seen": 166503850, + "step": 7701 + }, + { + "epoch": 0.9261107437022786, + "flos": 24536978235840.0, + "grad_norm": 2.6230457842747175, + "learning_rate": 5.6917955280130216e-08, + "loss": 0.7445, + "num_input_tokens_seen": 166525330, + "step": 7702 + }, + { + "epoch": 0.9262309865929177, + "flos": 21987028705920.0, + "grad_norm": 3.009613686798059, + "learning_rate": 5.6733586592172755e-08, + "loss": 0.7191, + "num_input_tokens_seen": 166544825, + "step": 7703 + }, + { + "epoch": 0.9263512294835567, + "flos": 20310483123840.0, + "grad_norm": 2.714865705003401, + "learning_rate": 5.6549512696185244e-08, + "loss": 0.8031, + "num_input_tokens_seen": 166563325, + "step": 7704 + }, + { + "epoch": 0.9264714723741959, + "flos": 21185113673760.0, + "grad_norm": 2.632419636107428, + "learning_rate": 5.636573362009156e-08, + "loss": 0.684, + "num_input_tokens_seen": 166583385, + "step": 7705 + }, + { + "epoch": 0.926591715264835, + "flos": 17979386372640.0, + "grad_norm": 5.080760283760212, + "learning_rate": 5.618224939177074e-08, + "loss": 0.7707, + "num_input_tokens_seen": 166601290, + "step": 7706 + }, + { + "epoch": 0.926711958155474, + "flos": 36115694970240.0, + "grad_norm": 2.913683010069301, + "learning_rate": 5.599906003905719e-08, + "loss": 0.706, + "num_input_tokens_seen": 166623835, + "step": 7707 + }, + { + "epoch": 0.9268322010461132, + "flos": 21002730163200.0, + "grad_norm": 2.8270165394534326, + "learning_rate": 5.581616558974023e-08, + "loss": 0.8162, + "num_input_tokens_seen": 166642400, + "step": 7708 + }, + { + "epoch": 0.9269524439367522, + "flos": 22931958739200.0, + "grad_norm": 2.378643253114249, + "learning_rate": 5.5633566071565444e-08, + "loss": 0.7909, + "num_input_tokens_seen": 166661640, + "step": 7709 + }, + { + "epoch": 0.9270726868273913, + "flos": 41915052002880.0, + "grad_norm": 2.6910818316925487, + "learning_rate": 5.5451261512232896e-08, + "loss": 0.7078, + "num_input_tokens_seen": 166684320, + "step": 7710 + }, + { + "epoch": 0.9271929297180305, + "flos": 19763555610720.0, + "grad_norm": 2.412769057534081, + "learning_rate": 5.5269251939397576e-08, + "loss": 0.6285, + "num_input_tokens_seen": 166703835, + "step": 7711 + }, + { + "epoch": 0.9273131726086695, + "flos": 19939657431840.0, + "grad_norm": 3.2574395348184204, + "learning_rate": 5.508753738067073e-08, + "loss": 0.7647, + "num_input_tokens_seen": 166723375, + "step": 7712 + }, + { + "epoch": 0.9274334154993086, + "flos": 23225348353920.0, + "grad_norm": 3.553630831099403, + "learning_rate": 5.4906117863617875e-08, + "loss": 0.7905, + "num_input_tokens_seen": 166742760, + "step": 7713 + }, + { + "epoch": 0.9275536583899477, + "flos": 31749641644800.0, + "grad_norm": 3.202140694085636, + "learning_rate": 5.4724993415760533e-08, + "loss": 0.7837, + "num_input_tokens_seen": 166761115, + "step": 7714 + }, + { + "epoch": 0.9276739012805868, + "flos": 18670369640160.0, + "grad_norm": 4.087875694251281, + "learning_rate": 5.454416406457496e-08, + "loss": 0.7487, + "num_input_tokens_seen": 166780210, + "step": 7715 + }, + { + "epoch": 0.9277941441712259, + "flos": 13859399823840.0, + "grad_norm": 3.065008731848272, + "learning_rate": 5.436362983749299e-08, + "loss": 0.7377, + "num_input_tokens_seen": 166795380, + "step": 7716 + }, + { + "epoch": 0.927914387061865, + "flos": 23225422693440.0, + "grad_norm": 11.214551899938145, + "learning_rate": 5.418339076190137e-08, + "loss": 0.6452, + "num_input_tokens_seen": 166814890, + "step": 7717 + }, + { + "epoch": 0.9280346299525041, + "flos": 18047901707520.0, + "grad_norm": 2.1830499343787837, + "learning_rate": 5.400344686514202e-08, + "loss": 0.8859, + "num_input_tokens_seen": 166832475, + "step": 7718 + }, + { + "epoch": 0.9281548728431431, + "flos": 22310866087680.0, + "grad_norm": 2.601669171869936, + "learning_rate": 5.38237981745131e-08, + "loss": 0.6669, + "num_input_tokens_seen": 166850590, + "step": 7719 + }, + { + "epoch": 0.9282751157337822, + "flos": 18816952938240.0, + "grad_norm": 2.548669101241287, + "learning_rate": 5.364444471726592e-08, + "loss": 0.8137, + "num_input_tokens_seen": 166869795, + "step": 7720 + }, + { + "epoch": 0.9283953586244214, + "flos": 25520236025280.0, + "grad_norm": 2.8283530296159882, + "learning_rate": 5.346538652060939e-08, + "loss": 0.8007, + "num_input_tokens_seen": 166891150, + "step": 7721 + }, + { + "epoch": 0.9285156015150604, + "flos": 18197570095680.0, + "grad_norm": 2.630038285502557, + "learning_rate": 5.3286623611705994e-08, + "loss": 0.7017, + "num_input_tokens_seen": 166909195, + "step": 7722 + }, + { + "epoch": 0.9286358444056995, + "flos": 66305850074400.0, + "grad_norm": 0.8314278433866314, + "learning_rate": 5.3108156017673824e-08, + "loss": 0.605, + "num_input_tokens_seen": 166970955, + "step": 7723 + }, + { + "epoch": 0.9287560872963386, + "flos": 22313690989440.0, + "grad_norm": 2.390764219930073, + "learning_rate": 5.2929983765586775e-08, + "loss": 0.7192, + "num_input_tokens_seen": 166989735, + "step": 7724 + }, + { + "epoch": 0.9288763301869777, + "flos": 25663325365920.0, + "grad_norm": 2.1813225851579174, + "learning_rate": 5.275210688247278e-08, + "loss": 0.6235, + "num_input_tokens_seen": 167010060, + "step": 7725 + }, + { + "epoch": 0.9289965730776167, + "flos": 12294715250400.0, + "grad_norm": 2.5194634007034376, + "learning_rate": 5.257452539531604e-08, + "loss": 0.8513, + "num_input_tokens_seen": 167027130, + "step": 7726 + }, + { + "epoch": 0.9291168159682559, + "flos": 26647661078400.0, + "grad_norm": 2.0487595829439904, + "learning_rate": 5.2397239331055445e-08, + "loss": 0.6853, + "num_input_tokens_seen": 167049130, + "step": 7727 + }, + { + "epoch": 0.929237058858895, + "flos": 14517705138720.0, + "grad_norm": 5.924281132221339, + "learning_rate": 5.2220248716585036e-08, + "loss": 0.8136, + "num_input_tokens_seen": 167066810, + "step": 7728 + }, + { + "epoch": 0.929357301749534, + "flos": 23801125334400.0, + "grad_norm": 3.494130799315419, + "learning_rate": 5.204355357875445e-08, + "loss": 0.7546, + "num_input_tokens_seen": 167085155, + "step": 7729 + }, + { + "epoch": 0.9294775446401732, + "flos": 12951050568000.0, + "grad_norm": 2.754050479656215, + "learning_rate": 5.1867153944367584e-08, + "loss": 0.7089, + "num_input_tokens_seen": 167101545, + "step": 7730 + }, + { + "epoch": 0.9295977875308122, + "flos": 26174192478240.0, + "grad_norm": 2.2502606558817093, + "learning_rate": 5.16910498401848e-08, + "loss": 0.7341, + "num_input_tokens_seen": 167120995, + "step": 7731 + }, + { + "epoch": 0.9297180304214513, + "flos": 16448940881760.0, + "grad_norm": 2.612120320596833, + "learning_rate": 5.151524129292073e-08, + "loss": 0.8305, + "num_input_tokens_seen": 167138890, + "step": 7732 + }, + { + "epoch": 0.9298382733120905, + "flos": 24025999614240.0, + "grad_norm": 2.2218076027736235, + "learning_rate": 5.1339728329245155e-08, + "loss": 0.667, + "num_input_tokens_seen": 167159455, + "step": 7733 + }, + { + "epoch": 0.9299585162027295, + "flos": 22096064812800.0, + "grad_norm": 6.013213548728931, + "learning_rate": 5.116451097578367e-08, + "loss": 0.7972, + "num_input_tokens_seen": 167178495, + "step": 7734 + }, + { + "epoch": 0.9300787590933686, + "flos": 21441179115840.0, + "grad_norm": 2.0639169622913336, + "learning_rate": 5.0989589259115895e-08, + "loss": 0.7431, + "num_input_tokens_seen": 167199380, + "step": 7735 + }, + { + "epoch": 0.9301990019840077, + "flos": 17754103225440.0, + "grad_norm": 2.3711949033072366, + "learning_rate": 5.081496320577816e-08, + "loss": 0.7189, + "num_input_tokens_seen": 167216500, + "step": 7736 + }, + { + "epoch": 0.9303192448746468, + "flos": 58811994465600.0, + "grad_norm": 0.9921267933934447, + "learning_rate": 5.0640632842260835e-08, + "loss": 0.6109, + "num_input_tokens_seen": 167276760, + "step": 7737 + }, + { + "epoch": 0.9304394877652858, + "flos": 57580148089440.0, + "grad_norm": 1.9956429825428652, + "learning_rate": 5.0466598195009426e-08, + "loss": 0.727, + "num_input_tokens_seen": 167303630, + "step": 7738 + }, + { + "epoch": 0.930559730655925, + "flos": 20966186555520.0, + "grad_norm": 2.56040357105785, + "learning_rate": 5.0292859290425036e-08, + "loss": 0.7045, + "num_input_tokens_seen": 167321500, + "step": 7739 + }, + { + "epoch": 0.9306799735465641, + "flos": 23225013826080.0, + "grad_norm": 2.3252712911843645, + "learning_rate": 5.011941615486348e-08, + "loss": 0.7831, + "num_input_tokens_seen": 167340615, + "step": 7740 + }, + { + "epoch": 0.9308002164372031, + "flos": 15209989347840.0, + "grad_norm": 5.791574527730253, + "learning_rate": 4.994626881463659e-08, + "loss": 0.8487, + "num_input_tokens_seen": 167356870, + "step": 7741 + }, + { + "epoch": 0.9309204593278423, + "flos": 30803187651360.0, + "grad_norm": 2.0933858070761984, + "learning_rate": 4.9773417296009814e-08, + "loss": 0.7085, + "num_input_tokens_seen": 167378390, + "step": 7742 + }, + { + "epoch": 0.9310407022184813, + "flos": 23004265389600.0, + "grad_norm": 2.2025694451008677, + "learning_rate": 4.960086162520527e-08, + "loss": 0.6544, + "num_input_tokens_seen": 167398480, + "step": 7743 + }, + { + "epoch": 0.9311609451091204, + "flos": 22096399340640.0, + "grad_norm": 3.129330038452798, + "learning_rate": 4.942860182839936e-08, + "loss": 0.8258, + "num_input_tokens_seen": 167416825, + "step": 7744 + }, + { + "epoch": 0.9312811879997596, + "flos": 21068866633440.0, + "grad_norm": 2.7100294719655946, + "learning_rate": 4.925663793172341e-08, + "loss": 0.7977, + "num_input_tokens_seen": 167434785, + "step": 7745 + }, + { + "epoch": 0.9314014308903986, + "flos": 67052264921280.0, + "grad_norm": 0.7907833935841027, + "learning_rate": 4.908496996126477e-08, + "loss": 0.5637, + "num_input_tokens_seen": 167498245, + "step": 7746 + }, + { + "epoch": 0.9315216737810377, + "flos": 22533361502880.0, + "grad_norm": 2.0232382422175914, + "learning_rate": 4.89135979430646e-08, + "loss": 0.767, + "num_input_tokens_seen": 167518290, + "step": 7747 + }, + { + "epoch": 0.9316419166716768, + "flos": 23949604290240.0, + "grad_norm": 2.332196458239361, + "learning_rate": 4.874252190312078e-08, + "loss": 0.8562, + "num_input_tokens_seen": 167538675, + "step": 7748 + }, + { + "epoch": 0.9317621595623159, + "flos": 30187187256960.0, + "grad_norm": 2.058755654730845, + "learning_rate": 4.857174186738477e-08, + "loss": 0.6483, + "num_input_tokens_seen": 167562025, + "step": 7749 + }, + { + "epoch": 0.931882402452955, + "flos": 15720447592800.0, + "grad_norm": 2.9804234732492327, + "learning_rate": 4.840125786176408e-08, + "loss": 0.7299, + "num_input_tokens_seen": 167578230, + "step": 7750 + }, + { + "epoch": 0.932002645343594, + "flos": 28325842129920.0, + "grad_norm": 2.204401282339566, + "learning_rate": 4.823106991212067e-08, + "loss": 0.7714, + "num_input_tokens_seen": 167597470, + "step": 7751 + }, + { + "epoch": 0.9321228882342332, + "flos": 15319434322080.0, + "grad_norm": 2.770372357181706, + "learning_rate": 4.806117804427212e-08, + "loss": 0.8393, + "num_input_tokens_seen": 167615405, + "step": 7752 + }, + { + "epoch": 0.9322431311248722, + "flos": 17869941398400.0, + "grad_norm": 2.348813297054269, + "learning_rate": 4.7891582283990926e-08, + "loss": 0.6476, + "num_input_tokens_seen": 167634130, + "step": 7753 + }, + { + "epoch": 0.9323633740155113, + "flos": 24135110060640.0, + "grad_norm": 2.054770464364418, + "learning_rate": 4.772228265700473e-08, + "loss": 0.7312, + "num_input_tokens_seen": 167654990, + "step": 7754 + }, + { + "epoch": 0.9324836169061504, + "flos": 15021993203520.0, + "grad_norm": 2.3541082751929205, + "learning_rate": 4.75532791889961e-08, + "loss": 0.7598, + "num_input_tokens_seen": 167671690, + "step": 7755 + }, + { + "epoch": 0.9326038597967895, + "flos": 18598397517600.0, + "grad_norm": 2.99906843344827, + "learning_rate": 4.738457190560252e-08, + "loss": 0.6579, + "num_input_tokens_seen": 167690800, + "step": 7756 + }, + { + "epoch": 0.9327241026874286, + "flos": 18925654517280.0, + "grad_norm": 3.4459118656925347, + "learning_rate": 4.721616083241664e-08, + "loss": 0.7931, + "num_input_tokens_seen": 167709055, + "step": 7757 + }, + { + "epoch": 0.9328443455780677, + "flos": 29527841188800.0, + "grad_norm": 2.1832012990772243, + "learning_rate": 4.7048045994986684e-08, + "loss": 0.7755, + "num_input_tokens_seen": 167729915, + "step": 7758 + }, + { + "epoch": 0.9329645884687068, + "flos": 30038559622080.0, + "grad_norm": 2.635318095860465, + "learning_rate": 4.688022741881559e-08, + "loss": 0.9155, + "num_input_tokens_seen": 167750440, + "step": 7759 + }, + { + "epoch": 0.9330848313593458, + "flos": 21836616922560.0, + "grad_norm": 2.069706666222076, + "learning_rate": 4.671270512936076e-08, + "loss": 0.7538, + "num_input_tokens_seen": 167769600, + "step": 7760 + }, + { + "epoch": 0.933205074249985, + "flos": 22095618775680.0, + "grad_norm": 3.8230696949833645, + "learning_rate": 4.6545479152035884e-08, + "loss": 0.8302, + "num_input_tokens_seen": 167788770, + "step": 7761 + }, + { + "epoch": 0.9333253171406241, + "flos": 15319545831360.0, + "grad_norm": 2.442950780965305, + "learning_rate": 4.637854951220821e-08, + "loss": 0.7618, + "num_input_tokens_seen": 167805555, + "step": 7762 + }, + { + "epoch": 0.9334455600312631, + "flos": 15683903985120.0, + "grad_norm": 2.698160275884131, + "learning_rate": 4.621191623520171e-08, + "loss": 0.7481, + "num_input_tokens_seen": 167823985, + "step": 7763 + }, + { + "epoch": 0.9335658029219023, + "flos": 22131902195040.0, + "grad_norm": 2.9772668040294086, + "learning_rate": 4.604557934629372e-08, + "loss": 0.8483, + "num_input_tokens_seen": 167843060, + "step": 7764 + }, + { + "epoch": 0.9336860458125413, + "flos": 20237953454880.0, + "grad_norm": 6.522525783033137, + "learning_rate": 4.587953887071805e-08, + "loss": 0.8088, + "num_input_tokens_seen": 167862880, + "step": 7765 + }, + { + "epoch": 0.9338062887031804, + "flos": 20889716892000.0, + "grad_norm": 9.711082451276392, + "learning_rate": 4.5713794833662554e-08, + "loss": 0.8582, + "num_input_tokens_seen": 167882095, + "step": 7766 + }, + { + "epoch": 0.9339265315938196, + "flos": 23188433048640.0, + "grad_norm": 2.3520799238677497, + "learning_rate": 4.5548347260270236e-08, + "loss": 0.634, + "num_input_tokens_seen": 167901385, + "step": 7767 + }, + { + "epoch": 0.9340467744844586, + "flos": 22787642796480.0, + "grad_norm": 2.205139321490835, + "learning_rate": 4.538319617564012e-08, + "loss": 0.6968, + "num_input_tokens_seen": 167920405, + "step": 7768 + }, + { + "epoch": 0.9341670173750977, + "flos": 23626621812960.0, + "grad_norm": 3.0150912904598974, + "learning_rate": 4.521834160482485e-08, + "loss": 0.7471, + "num_input_tokens_seen": 167939145, + "step": 7769 + }, + { + "epoch": 0.9342872602657368, + "flos": 24788620476480.0, + "grad_norm": 3.6045694438516946, + "learning_rate": 4.5053783572832846e-08, + "loss": 0.8241, + "num_input_tokens_seen": 167959325, + "step": 7770 + }, + { + "epoch": 0.9344075031563759, + "flos": 25734777111840.0, + "grad_norm": 2.7161625794247723, + "learning_rate": 4.488952210462771e-08, + "loss": 0.762, + "num_input_tokens_seen": 167979530, + "step": 7771 + }, + { + "epoch": 0.9345277460470149, + "flos": 25514511882240.0, + "grad_norm": 2.389284942630087, + "learning_rate": 4.4725557225127495e-08, + "loss": 0.8608, + "num_input_tokens_seen": 167997870, + "step": 7772 + }, + { + "epoch": 0.9346479889376541, + "flos": 34263753792480.0, + "grad_norm": 2.7468527210360647, + "learning_rate": 4.456188895920565e-08, + "loss": 0.7948, + "num_input_tokens_seen": 168019625, + "step": 7773 + }, + { + "epoch": 0.9347682318282932, + "flos": 19066364993280.0, + "grad_norm": 3.2072267646991595, + "learning_rate": 4.439851733169031e-08, + "loss": 0.857, + "num_input_tokens_seen": 168037415, + "step": 7774 + }, + { + "epoch": 0.9348884747189322, + "flos": 26211962688000.0, + "grad_norm": 3.7874454154434196, + "learning_rate": 4.4235442367365204e-08, + "loss": 0.689, + "num_input_tokens_seen": 168057725, + "step": 7775 + }, + { + "epoch": 0.9350087176095714, + "flos": 18307015069920.0, + "grad_norm": 2.990455385059971, + "learning_rate": 4.4072664090968545e-08, + "loss": 0.7982, + "num_input_tokens_seen": 168076110, + "step": 7776 + }, + { + "epoch": 0.9351289605002104, + "flos": 19290756066240.0, + "grad_norm": 2.340546924220914, + "learning_rate": 4.391018252719347e-08, + "loss": 0.8501, + "num_input_tokens_seen": 168095905, + "step": 7777 + }, + { + "epoch": 0.9352492033908495, + "flos": 18772975378560.0, + "grad_norm": 2.916578091684256, + "learning_rate": 4.374799770068849e-08, + "loss": 0.6958, + "num_input_tokens_seen": 168112580, + "step": 7778 + }, + { + "epoch": 0.9353694462814887, + "flos": 29487617774880.0, + "grad_norm": 2.6069972917713167, + "learning_rate": 4.358610963605658e-08, + "loss": 0.7495, + "num_input_tokens_seen": 168133030, + "step": 7779 + }, + { + "epoch": 0.9354896891721277, + "flos": 30621436026720.0, + "grad_norm": 2.7631706079749865, + "learning_rate": 4.342451835785677e-08, + "loss": 0.6849, + "num_input_tokens_seen": 168153940, + "step": 7780 + }, + { + "epoch": 0.9356099320627668, + "flos": 19436187101760.0, + "grad_norm": 1.975998568094175, + "learning_rate": 4.3263223890601665e-08, + "loss": 0.751, + "num_input_tokens_seen": 168172650, + "step": 7781 + }, + { + "epoch": 0.9357301749534058, + "flos": 19070936873760.0, + "grad_norm": 2.1131330004464184, + "learning_rate": 4.31022262587597e-08, + "loss": 0.7933, + "num_input_tokens_seen": 168191325, + "step": 7782 + }, + { + "epoch": 0.935850417844045, + "flos": 23516693631840.0, + "grad_norm": 2.6655922354398482, + "learning_rate": 4.2941525486754225e-08, + "loss": 0.6554, + "num_input_tokens_seen": 168211645, + "step": 7783 + }, + { + "epoch": 0.935970660734684, + "flos": 18561891079680.0, + "grad_norm": 2.3040023561174214, + "learning_rate": 4.278112159896286e-08, + "loss": 0.8001, + "num_input_tokens_seen": 168229485, + "step": 7784 + }, + { + "epoch": 0.9360909036253231, + "flos": 20601828401760.0, + "grad_norm": 2.601825872613101, + "learning_rate": 4.2621014619719896e-08, + "loss": 0.6747, + "num_input_tokens_seen": 168247250, + "step": 7785 + }, + { + "epoch": 0.9362111465159623, + "flos": 61703548522560.0, + "grad_norm": 0.7279700724271843, + "learning_rate": 4.246120457331215e-08, + "loss": 0.5851, + "num_input_tokens_seen": 168309425, + "step": 7786 + }, + { + "epoch": 0.9363313894066013, + "flos": 24135816286080.0, + "grad_norm": 2.5377291420891233, + "learning_rate": 4.2301691483983325e-08, + "loss": 0.7231, + "num_input_tokens_seen": 168329545, + "step": 7787 + }, + { + "epoch": 0.9364516322972404, + "flos": 20091667514880.0, + "grad_norm": 3.632626239554693, + "learning_rate": 4.214247537593163e-08, + "loss": 0.7579, + "num_input_tokens_seen": 168348795, + "step": 7788 + }, + { + "epoch": 0.9365718751878795, + "flos": 20674209391680.0, + "grad_norm": 2.5505838361585838, + "learning_rate": 4.1983556273309293e-08, + "loss": 0.8052, + "num_input_tokens_seen": 168367545, + "step": 7789 + }, + { + "epoch": 0.9366921180785186, + "flos": 18628510756800.0, + "grad_norm": 3.0971710094405105, + "learning_rate": 4.182493420022526e-08, + "loss": 0.6893, + "num_input_tokens_seen": 168383215, + "step": 7790 + }, + { + "epoch": 0.9368123609691577, + "flos": 25738010880960.0, + "grad_norm": 2.223838987580235, + "learning_rate": 4.166660918074139e-08, + "loss": 0.7876, + "num_input_tokens_seen": 168403120, + "step": 7791 + }, + { + "epoch": 0.9369326038597968, + "flos": 25516667728320.0, + "grad_norm": 1.7143677765641576, + "learning_rate": 4.15085812388758e-08, + "loss": 0.7345, + "num_input_tokens_seen": 168425340, + "step": 7792 + }, + { + "epoch": 0.9370528467504359, + "flos": 23186760409440.0, + "grad_norm": 1.9752361687563544, + "learning_rate": 4.135085039860153e-08, + "loss": 0.7854, + "num_input_tokens_seen": 168444740, + "step": 7793 + }, + { + "epoch": 0.9371730896410749, + "flos": 24931523968320.0, + "grad_norm": 5.014021097109172, + "learning_rate": 4.1193416683845906e-08, + "loss": 0.7853, + "num_input_tokens_seen": 168463670, + "step": 7794 + }, + { + "epoch": 0.9372933325317141, + "flos": 15530815979040.0, + "grad_norm": 2.704456067775566, + "learning_rate": 4.103628011849136e-08, + "loss": 0.8329, + "num_input_tokens_seen": 168479030, + "step": 7795 + }, + { + "epoch": 0.9374135754223532, + "flos": 21840631256640.0, + "grad_norm": 2.3995579391895983, + "learning_rate": 4.0879440726375506e-08, + "loss": 0.7596, + "num_input_tokens_seen": 168496815, + "step": 7796 + }, + { + "epoch": 0.9375338183129922, + "flos": 22598457219840.0, + "grad_norm": 2.9372597454109703, + "learning_rate": 4.0722898531291074e-08, + "loss": 0.5626, + "num_input_tokens_seen": 168514055, + "step": 7797 + }, + { + "epoch": 0.9376540612036314, + "flos": 26066903350080.0, + "grad_norm": 2.271489203227349, + "learning_rate": 4.0566653556985295e-08, + "loss": 0.7678, + "num_input_tokens_seen": 168534600, + "step": 7798 + }, + { + "epoch": 0.9377743040942704, + "flos": 19689427642080.0, + "grad_norm": 3.147425032034185, + "learning_rate": 4.0410705827159886e-08, + "loss": 0.8201, + "num_input_tokens_seen": 168551895, + "step": 7799 + }, + { + "epoch": 0.9378945469849095, + "flos": 15245901069600.0, + "grad_norm": 2.676221736097687, + "learning_rate": 4.0255055365472356e-08, + "loss": 0.7182, + "num_input_tokens_seen": 168569990, + "step": 7800 + }, + { + "epoch": 0.9380147898755486, + "flos": 20561902345920.0, + "grad_norm": 2.6776085621599175, + "learning_rate": 4.009970219553471e-08, + "loss": 0.7496, + "num_input_tokens_seen": 168586940, + "step": 7801 + }, + { + "epoch": 0.9381350327661877, + "flos": 26243265359520.0, + "grad_norm": 4.797759629977414, + "learning_rate": 3.99446463409141e-08, + "loss": 0.7692, + "num_input_tokens_seen": 168604795, + "step": 7802 + }, + { + "epoch": 0.9382552756568268, + "flos": 23552605353600.0, + "grad_norm": 3.7438655268655086, + "learning_rate": 3.978988782513215e-08, + "loss": 0.6928, + "num_input_tokens_seen": 168622290, + "step": 7803 + }, + { + "epoch": 0.9383755185474659, + "flos": 28398000101280.0, + "grad_norm": 2.3444380703860026, + "learning_rate": 3.963542667166586e-08, + "loss": 0.7666, + "num_input_tokens_seen": 168642395, + "step": 7804 + }, + { + "epoch": 0.938495761438105, + "flos": 20420448474720.0, + "grad_norm": 2.053342661939626, + "learning_rate": 3.9481262903946486e-08, + "loss": 0.6841, + "num_input_tokens_seen": 168661840, + "step": 7805 + }, + { + "epoch": 0.938616004328744, + "flos": 69203728724160.0, + "grad_norm": 0.7833627362177744, + "learning_rate": 3.932739654536066e-08, + "loss": 0.5438, + "num_input_tokens_seen": 168724540, + "step": 7806 + }, + { + "epoch": 0.9387362472193832, + "flos": 18884539029120.0, + "grad_norm": 2.8482166466231242, + "learning_rate": 3.917382761925014e-08, + "loss": 0.7396, + "num_input_tokens_seen": 168740375, + "step": 7807 + }, + { + "epoch": 0.9388564901100223, + "flos": 26463865116960.0, + "grad_norm": 2.1343151570053114, + "learning_rate": 3.9020556148910754e-08, + "loss": 0.7912, + "num_input_tokens_seen": 168759730, + "step": 7808 + }, + { + "epoch": 0.9389767330006613, + "flos": 58857198627360.0, + "grad_norm": 0.8324488277001105, + "learning_rate": 3.8867582157593895e-08, + "loss": 0.5658, + "num_input_tokens_seen": 168813935, + "step": 7809 + }, + { + "epoch": 0.9390969758913005, + "flos": 31057877812320.0, + "grad_norm": 1.9912035138869708, + "learning_rate": 3.871490566850544e-08, + "loss": 0.7656, + "num_input_tokens_seen": 168838145, + "step": 7810 + }, + { + "epoch": 0.9392172187819395, + "flos": 22387372920960.0, + "grad_norm": 2.1176365450055785, + "learning_rate": 3.856252670480642e-08, + "loss": 0.7119, + "num_input_tokens_seen": 168856805, + "step": 7811 + }, + { + "epoch": 0.9393374616725786, + "flos": 19691397639360.0, + "grad_norm": 2.168613757160685, + "learning_rate": 3.841044528961279e-08, + "loss": 0.8167, + "num_input_tokens_seen": 168874600, + "step": 7812 + }, + { + "epoch": 0.9394577045632178, + "flos": 24135630437280.0, + "grad_norm": 2.577235767869528, + "learning_rate": 3.825866144599477e-08, + "loss": 0.7893, + "num_input_tokens_seen": 168893655, + "step": 7813 + }, + { + "epoch": 0.9395779474538568, + "flos": 18996771735360.0, + "grad_norm": 2.687056393939097, + "learning_rate": 3.8107175196978145e-08, + "loss": 0.7568, + "num_input_tokens_seen": 168909960, + "step": 7814 + }, + { + "epoch": 0.9396981903444959, + "flos": 14298889529760.0, + "grad_norm": 2.3532777531502056, + "learning_rate": 3.7955986565542996e-08, + "loss": 0.7709, + "num_input_tokens_seen": 168928040, + "step": 7815 + }, + { + "epoch": 0.9398184332351349, + "flos": 34737519750720.0, + "grad_norm": 2.575365751641934, + "learning_rate": 3.780509557462497e-08, + "loss": 0.6848, + "num_input_tokens_seen": 168948240, + "step": 7816 + }, + { + "epoch": 0.9399386761257741, + "flos": 25335845347680.0, + "grad_norm": 1.7840703754314704, + "learning_rate": 3.765450224711375e-08, + "loss": 0.7535, + "num_input_tokens_seen": 168968745, + "step": 7817 + }, + { + "epoch": 0.9400589190164131, + "flos": 27265705809600.0, + "grad_norm": 2.686181831897255, + "learning_rate": 3.750420660585396e-08, + "loss": 0.7961, + "num_input_tokens_seen": 168990685, + "step": 7818 + }, + { + "epoch": 0.9401791619070522, + "flos": 23366244678720.0, + "grad_norm": 2.508107010610266, + "learning_rate": 3.735420867364603e-08, + "loss": 0.801, + "num_input_tokens_seen": 169011665, + "step": 7819 + }, + { + "epoch": 0.9402994047976914, + "flos": 35829739307520.0, + "grad_norm": 1.990701996708012, + "learning_rate": 3.7204508473244186e-08, + "loss": 0.6193, + "num_input_tokens_seen": 169035290, + "step": 7820 + }, + { + "epoch": 0.9404196476883304, + "flos": 22205063749920.0, + "grad_norm": 1.939846217952392, + "learning_rate": 3.7055106027357395e-08, + "loss": 0.6933, + "num_input_tokens_seen": 169055155, + "step": 7821 + }, + { + "epoch": 0.9405398905789695, + "flos": 18888850721280.0, + "grad_norm": 2.920479997349236, + "learning_rate": 3.690600135865063e-08, + "loss": 0.7222, + "num_input_tokens_seen": 169072080, + "step": 7822 + }, + { + "epoch": 0.9406601334696086, + "flos": 70173233702400.0, + "grad_norm": 0.7998776804275062, + "learning_rate": 3.675719448974246e-08, + "loss": 0.5794, + "num_input_tokens_seen": 169137170, + "step": 7823 + }, + { + "epoch": 0.9407803763602477, + "flos": 22133463324960.0, + "grad_norm": 2.525388359192719, + "learning_rate": 3.6608685443207054e-08, + "loss": 0.595, + "num_input_tokens_seen": 169156320, + "step": 7824 + }, + { + "epoch": 0.9409006192508867, + "flos": 18852604471680.0, + "grad_norm": 2.691634081317511, + "learning_rate": 3.646047424157306e-08, + "loss": 0.6686, + "num_input_tokens_seen": 169173295, + "step": 7825 + }, + { + "epoch": 0.9410208621415259, + "flos": 23334867667680.0, + "grad_norm": 3.788958298341678, + "learning_rate": 3.631256090732382e-08, + "loss": 0.691, + "num_input_tokens_seen": 169193755, + "step": 7826 + }, + { + "epoch": 0.941141105032165, + "flos": 22709537663520.0, + "grad_norm": 1.9088632150585623, + "learning_rate": 3.6164945462897833e-08, + "loss": 0.8283, + "num_input_tokens_seen": 169213045, + "step": 7827 + }, + { + "epoch": 0.941261347922804, + "flos": 20675398824000.0, + "grad_norm": 2.422731900133466, + "learning_rate": 3.6017627930687856e-08, + "loss": 0.7587, + "num_input_tokens_seen": 169232870, + "step": 7828 + }, + { + "epoch": 0.9413815908134432, + "flos": 19394142369600.0, + "grad_norm": 6.907356829434876, + "learning_rate": 3.587060833304267e-08, + "loss": 0.7677, + "num_input_tokens_seen": 169251060, + "step": 7829 + }, + { + "epoch": 0.9415018337040822, + "flos": 17468593599840.0, + "grad_norm": 2.3380211560260507, + "learning_rate": 3.5723886692264225e-08, + "loss": 0.6412, + "num_input_tokens_seen": 169268600, + "step": 7830 + }, + { + "epoch": 0.9416220765947213, + "flos": 31785850724640.0, + "grad_norm": 2.610909838618386, + "learning_rate": 3.557746303061071e-08, + "loss": 0.6242, + "num_input_tokens_seen": 169290355, + "step": 7831 + }, + { + "epoch": 0.9417423194853605, + "flos": 23478291536160.0, + "grad_norm": 4.374126020056013, + "learning_rate": 3.543133737029391e-08, + "loss": 0.7249, + "num_input_tokens_seen": 169310865, + "step": 7832 + }, + { + "epoch": 0.9418625623759995, + "flos": 23880903106560.0, + "grad_norm": 2.3329360714560905, + "learning_rate": 3.5285509733481214e-08, + "loss": 0.6899, + "num_input_tokens_seen": 169330590, + "step": 7833 + }, + { + "epoch": 0.9419828052666386, + "flos": 18051246985920.0, + "grad_norm": 8.381576916309001, + "learning_rate": 3.513998014229469e-08, + "loss": 0.7671, + "num_input_tokens_seen": 169349540, + "step": 7834 + }, + { + "epoch": 0.9421030481572777, + "flos": 17687149020480.0, + "grad_norm": 11.030731967076898, + "learning_rate": 3.499474861881069e-08, + "loss": 0.8652, + "num_input_tokens_seen": 169366765, + "step": 7835 + }, + { + "epoch": 0.9422232910479168, + "flos": 20165349446400.0, + "grad_norm": 2.653053155112758, + "learning_rate": 3.4849815185061136e-08, + "loss": 0.6772, + "num_input_tokens_seen": 169386655, + "step": 7836 + }, + { + "epoch": 0.9423435339385559, + "flos": 18416348534880.0, + "grad_norm": 9.501016662807887, + "learning_rate": 3.470517986303223e-08, + "loss": 0.7616, + "num_input_tokens_seen": 169405640, + "step": 7837 + }, + { + "epoch": 0.942463776829195, + "flos": 20051890138080.0, + "grad_norm": 2.9231965855186495, + "learning_rate": 3.4560842674664856e-08, + "loss": 0.7951, + "num_input_tokens_seen": 169424155, + "step": 7838 + }, + { + "epoch": 0.9425840197198341, + "flos": 22603846835040.0, + "grad_norm": 2.742967282791636, + "learning_rate": 3.441680364185506e-08, + "loss": 0.7522, + "num_input_tokens_seen": 169444025, + "step": 7839 + }, + { + "epoch": 0.9427042626104731, + "flos": 19909469853120.0, + "grad_norm": 4.398117243489035, + "learning_rate": 3.427306278645314e-08, + "loss": 0.7522, + "num_input_tokens_seen": 169462480, + "step": 7840 + }, + { + "epoch": 0.9428245055011123, + "flos": 22824335083200.0, + "grad_norm": 3.1559665665976917, + "learning_rate": 3.4129620130264767e-08, + "loss": 0.7313, + "num_input_tokens_seen": 169480840, + "step": 7841 + }, + { + "epoch": 0.9429447483917514, + "flos": 20921577109920.0, + "grad_norm": 10.785499332418938, + "learning_rate": 3.398647569505009e-08, + "loss": 0.7774, + "num_input_tokens_seen": 169498265, + "step": 7842 + }, + { + "epoch": 0.9430649912823904, + "flos": 18816841428960.0, + "grad_norm": 3.094030194165407, + "learning_rate": 3.384362950252373e-08, + "loss": 0.743, + "num_input_tokens_seen": 169515235, + "step": 7843 + }, + { + "epoch": 0.9431852341730296, + "flos": 32509028737920.0, + "grad_norm": 3.355104508797545, + "learning_rate": 3.3701081574355473e-08, + "loss": 0.5674, + "num_input_tokens_seen": 169537195, + "step": 7844 + }, + { + "epoch": 0.9433054770636686, + "flos": 66809357574240.0, + "grad_norm": 0.6450406691201374, + "learning_rate": 3.3558831932169796e-08, + "loss": 0.5155, + "num_input_tokens_seen": 169605865, + "step": 7845 + }, + { + "epoch": 0.9434257199543077, + "flos": 26102889411360.0, + "grad_norm": 3.225627773012456, + "learning_rate": 3.341688059754588e-08, + "loss": 0.89, + "num_input_tokens_seen": 169621520, + "step": 7846 + }, + { + "epoch": 0.9435459628449467, + "flos": 24967472859840.0, + "grad_norm": 2.926935230639782, + "learning_rate": 3.327522759201762e-08, + "loss": 0.7789, + "num_input_tokens_seen": 169641390, + "step": 7847 + }, + { + "epoch": 0.9436662057355859, + "flos": 22131567667200.0, + "grad_norm": 6.751133576328474, + "learning_rate": 3.313387293707359e-08, + "loss": 0.6676, + "num_input_tokens_seen": 169660095, + "step": 7848 + }, + { + "epoch": 0.943786448626225, + "flos": 20089771857120.0, + "grad_norm": 6.817934440980194, + "learning_rate": 3.29928166541571e-08, + "loss": 0.6853, + "num_input_tokens_seen": 169679050, + "step": 7849 + }, + { + "epoch": 0.943906691516864, + "flos": 22059261016800.0, + "grad_norm": 2.4753036428359776, + "learning_rate": 3.2852058764666346e-08, + "loss": 0.8032, + "num_input_tokens_seen": 169698150, + "step": 7850 + }, + { + "epoch": 0.9440269344075032, + "flos": 35247717807360.0, + "grad_norm": 2.3684073852122403, + "learning_rate": 3.2711599289954264e-08, + "loss": 0.69, + "num_input_tokens_seen": 169722185, + "step": 7851 + }, + { + "epoch": 0.9441471772981422, + "flos": 19210495087200.0, + "grad_norm": 3.0955589477466923, + "learning_rate": 3.257143825132847e-08, + "loss": 0.7812, + "num_input_tokens_seen": 169740355, + "step": 7852 + }, + { + "epoch": 0.9442674201887813, + "flos": 25702954063680.0, + "grad_norm": 2.1401937008477616, + "learning_rate": 3.243157567005106e-08, + "loss": 0.7598, + "num_input_tokens_seen": 169758535, + "step": 7853 + }, + { + "epoch": 0.9443876630794205, + "flos": 15501334625760.0, + "grad_norm": 2.960614108741446, + "learning_rate": 3.2292011567339296e-08, + "loss": 0.6434, + "num_input_tokens_seen": 169776340, + "step": 7854 + }, + { + "epoch": 0.9445079059700595, + "flos": 13381693871040.0, + "grad_norm": 4.117125653665587, + "learning_rate": 3.21527459643649e-08, + "loss": 0.5594, + "num_input_tokens_seen": 169793225, + "step": 7855 + }, + { + "epoch": 0.9446281488606986, + "flos": 23625395210880.0, + "grad_norm": 2.78550497603156, + "learning_rate": 3.2013778882254536e-08, + "loss": 0.7411, + "num_input_tokens_seen": 169812020, + "step": 7856 + }, + { + "epoch": 0.9447483917513377, + "flos": 25520756401920.0, + "grad_norm": 3.2496582731907506, + "learning_rate": 3.1875110342088676e-08, + "loss": 0.7587, + "num_input_tokens_seen": 169833470, + "step": 7857 + }, + { + "epoch": 0.9448686346419768, + "flos": 24500360288640.0, + "grad_norm": 2.2465782638083764, + "learning_rate": 3.1736740364904035e-08, + "loss": 0.6556, + "num_input_tokens_seen": 169854000, + "step": 7858 + }, + { + "epoch": 0.9449888775326158, + "flos": 14700460346880.0, + "grad_norm": 2.9007261550976637, + "learning_rate": 3.159866897169094e-08, + "loss": 0.7693, + "num_input_tokens_seen": 169872200, + "step": 7859 + }, + { + "epoch": 0.945109120423255, + "flos": 15425905715520.0, + "grad_norm": 2.2159162881374184, + "learning_rate": 3.146089618339487e-08, + "loss": 0.7592, + "num_input_tokens_seen": 169889055, + "step": 7860 + }, + { + "epoch": 0.9452293633138941, + "flos": 25411683125280.0, + "grad_norm": 3.798226324038388, + "learning_rate": 3.132342202091554e-08, + "loss": 0.6842, + "num_input_tokens_seen": 169909280, + "step": 7861 + }, + { + "epoch": 0.9453496062045331, + "flos": 21185076504000.0, + "grad_norm": 3.2058088270506246, + "learning_rate": 3.1186246505107595e-08, + "loss": 0.6851, + "num_input_tokens_seen": 169928420, + "step": 7862 + }, + { + "epoch": 0.9454698490951723, + "flos": 19982371219680.0, + "grad_norm": 2.450877040441748, + "learning_rate": 3.104936965678084e-08, + "loss": 0.8368, + "num_input_tokens_seen": 169946750, + "step": 7863 + }, + { + "epoch": 0.9455900919858113, + "flos": 21039608298720.0, + "grad_norm": 2.306839691280495, + "learning_rate": 3.091279149669956e-08, + "loss": 0.8201, + "num_input_tokens_seen": 169965540, + "step": 7864 + }, + { + "epoch": 0.9457103348764504, + "flos": 20711273376000.0, + "grad_norm": 2.1368547636883073, + "learning_rate": 3.0776512045581624e-08, + "loss": 0.7364, + "num_input_tokens_seen": 169984330, + "step": 7865 + }, + { + "epoch": 0.9458305777670896, + "flos": 21397573253760.0, + "grad_norm": 2.5846232506043694, + "learning_rate": 3.0640531324101384e-08, + "loss": 0.7795, + "num_input_tokens_seen": 170002095, + "step": 7866 + }, + { + "epoch": 0.9459508206577286, + "flos": 19982668577760.0, + "grad_norm": 4.265932847536678, + "learning_rate": 3.0504849352886554e-08, + "loss": 0.7612, + "num_input_tokens_seen": 170020240, + "step": 7867 + }, + { + "epoch": 0.9460710635483677, + "flos": 12149321384640.0, + "grad_norm": 3.1704153827275765, + "learning_rate": 3.036946615252023e-08, + "loss": 0.7153, + "num_input_tokens_seen": 170035770, + "step": 7868 + }, + { + "epoch": 0.9461913064390068, + "flos": 34226875656960.0, + "grad_norm": 3.276181766362721, + "learning_rate": 3.0234381743539984e-08, + "loss": 0.674, + "num_input_tokens_seen": 170053385, + "step": 7869 + }, + { + "epoch": 0.9463115493296459, + "flos": 19436038422720.0, + "grad_norm": 3.1386111051210936, + "learning_rate": 3.0099596146437863e-08, + "loss": 0.802, + "num_input_tokens_seen": 170070490, + "step": 7870 + }, + { + "epoch": 0.946431792220285, + "flos": 70469671237440.0, + "grad_norm": 0.7806146326896098, + "learning_rate": 2.996510938166086e-08, + "loss": 0.5991, + "num_input_tokens_seen": 170133465, + "step": 7871 + }, + { + "epoch": 0.9465520351109241, + "flos": 18920748108960.0, + "grad_norm": 2.2558292535529594, + "learning_rate": 2.983092146960997e-08, + "loss": 0.7388, + "num_input_tokens_seen": 170150810, + "step": 7872 + }, + { + "epoch": 0.9466722780015632, + "flos": 19108409725440.0, + "grad_norm": 3.1287450094027007, + "learning_rate": 2.9697032430642256e-08, + "loss": 0.804, + "num_input_tokens_seen": 170169025, + "step": 7873 + }, + { + "epoch": 0.9467925208922022, + "flos": 17213420232000.0, + "grad_norm": 3.157629032835996, + "learning_rate": 2.9563442285067906e-08, + "loss": 0.7359, + "num_input_tokens_seen": 170186070, + "step": 7874 + }, + { + "epoch": 0.9469127637828414, + "flos": 29127162445920.0, + "grad_norm": 3.437725545246228, + "learning_rate": 2.943015105315294e-08, + "loss": 0.796, + "num_input_tokens_seen": 170206265, + "step": 7875 + }, + { + "epoch": 0.9470330066734804, + "flos": 25993147079040.0, + "grad_norm": 2.625113348557246, + "learning_rate": 2.929715875511718e-08, + "loss": 0.6652, + "num_input_tokens_seen": 170225300, + "step": 7876 + }, + { + "epoch": 0.9471532495641195, + "flos": 23407397336640.0, + "grad_norm": 2.6331744298996957, + "learning_rate": 2.9164465411135375e-08, + "loss": 0.7009, + "num_input_tokens_seen": 170245070, + "step": 7877 + }, + { + "epoch": 0.9472734924547586, + "flos": 15793460468640.0, + "grad_norm": 2.823330519253328, + "learning_rate": 2.9032071041337426e-08, + "loss": 0.809, + "num_input_tokens_seen": 170263305, + "step": 7878 + }, + { + "epoch": 0.9473937353453977, + "flos": 11166137934720.0, + "grad_norm": 2.1189220122333436, + "learning_rate": 2.889997566580704e-08, + "loss": 0.73, + "num_input_tokens_seen": 170281460, + "step": 7879 + }, + { + "epoch": 0.9475139782360368, + "flos": 25738605597120.0, + "grad_norm": 2.6249716620959247, + "learning_rate": 2.8768179304583086e-08, + "loss": 0.7032, + "num_input_tokens_seen": 170303515, + "step": 7880 + }, + { + "epoch": 0.9476342211266758, + "flos": 22787419777920.0, + "grad_norm": 2.552100221746709, + "learning_rate": 2.8636681977659117e-08, + "loss": 0.7348, + "num_input_tokens_seen": 170323165, + "step": 7881 + }, + { + "epoch": 0.947754464017315, + "flos": 20164829069760.0, + "grad_norm": 2.6428861248238866, + "learning_rate": 2.850548370498318e-08, + "loss": 0.7821, + "num_input_tokens_seen": 170341005, + "step": 7882 + }, + { + "epoch": 0.9478747069079541, + "flos": 24682557950400.0, + "grad_norm": 1.9125804535143511, + "learning_rate": 2.8374584506457798e-08, + "loss": 0.7141, + "num_input_tokens_seen": 170362280, + "step": 7883 + }, + { + "epoch": 0.9479949497985931, + "flos": 21330879237120.0, + "grad_norm": 3.140357618194751, + "learning_rate": 2.824398440193998e-08, + "loss": 0.6708, + "num_input_tokens_seen": 170381080, + "step": 7884 + }, + { + "epoch": 0.9481151926892323, + "flos": 18122735901600.0, + "grad_norm": 2.8105640105996916, + "learning_rate": 2.811368341124232e-08, + "loss": 0.7122, + "num_input_tokens_seen": 170400150, + "step": 7885 + }, + { + "epoch": 0.9482354355798713, + "flos": 22096436510400.0, + "grad_norm": 4.169693851713409, + "learning_rate": 2.7983681554131222e-08, + "loss": 0.6844, + "num_input_tokens_seen": 170420410, + "step": 7886 + }, + { + "epoch": 0.9483556784705104, + "flos": 19035954396000.0, + "grad_norm": 2.8889131993306068, + "learning_rate": 2.7853978850327365e-08, + "loss": 0.7051, + "num_input_tokens_seen": 170439580, + "step": 7887 + }, + { + "epoch": 0.9484759213611496, + "flos": 25740241066560.0, + "grad_norm": 2.133759899507519, + "learning_rate": 2.7724575319507225e-08, + "loss": 0.8726, + "num_input_tokens_seen": 170459290, + "step": 7888 + }, + { + "epoch": 0.9485961642517886, + "flos": 20638260500160.0, + "grad_norm": 2.4069003922978625, + "learning_rate": 2.759547098130044e-08, + "loss": 0.7756, + "num_input_tokens_seen": 170478020, + "step": 7889 + }, + { + "epoch": 0.9487164071424277, + "flos": 22641654214560.0, + "grad_norm": 3.862774050968481, + "learning_rate": 2.746666585529267e-08, + "loss": 0.7676, + "num_input_tokens_seen": 170498295, + "step": 7890 + }, + { + "epoch": 0.9488366500330668, + "flos": 38670625248000.0, + "grad_norm": 2.871496470745933, + "learning_rate": 2.73381599610234e-08, + "loss": 0.7426, + "num_input_tokens_seen": 170518695, + "step": 7891 + }, + { + "epoch": 0.9489568929237059, + "flos": 27850886739360.0, + "grad_norm": 2.0615862353873826, + "learning_rate": 2.7209953317987033e-08, + "loss": 0.7162, + "num_input_tokens_seen": 170539735, + "step": 7892 + }, + { + "epoch": 0.9490771358143449, + "flos": 33532249752960.0, + "grad_norm": 2.6257584702225207, + "learning_rate": 2.7082045945631793e-08, + "loss": 0.7787, + "num_input_tokens_seen": 170561950, + "step": 7893 + }, + { + "epoch": 0.9491973787049841, + "flos": 14772878506560.0, + "grad_norm": 3.0112586158512, + "learning_rate": 2.6954437863361712e-08, + "loss": 0.6941, + "num_input_tokens_seen": 170579865, + "step": 7894 + }, + { + "epoch": 0.9493176215956232, + "flos": 25296067970880.0, + "grad_norm": 2.916997418496314, + "learning_rate": 2.6827129090534862e-08, + "loss": 0.7113, + "num_input_tokens_seen": 170600165, + "step": 7895 + }, + { + "epoch": 0.9494378644862622, + "flos": 20999868091680.0, + "grad_norm": 7.180214282480293, + "learning_rate": 2.670011964646335e-08, + "loss": 0.7802, + "num_input_tokens_seen": 170618845, + "step": 7896 + }, + { + "epoch": 0.9495581073769014, + "flos": 15173631588960.0, + "grad_norm": 2.4771169529187556, + "learning_rate": 2.657340955041487e-08, + "loss": 0.6849, + "num_input_tokens_seen": 170637620, + "step": 7897 + }, + { + "epoch": 0.9496783502675404, + "flos": 28575328524480.0, + "grad_norm": 2.861321625685878, + "learning_rate": 2.6446998821611167e-08, + "loss": 0.7169, + "num_input_tokens_seen": 170657815, + "step": 7898 + }, + { + "epoch": 0.9497985931581795, + "flos": 14845854212640.0, + "grad_norm": 3.5107656846831894, + "learning_rate": 2.6320887479228228e-08, + "loss": 0.7113, + "num_input_tokens_seen": 170674415, + "step": 7899 + }, + { + "epoch": 0.9499188360488187, + "flos": 27154142159040.0, + "grad_norm": 6.916852037021667, + "learning_rate": 2.619507554239786e-08, + "loss": 0.7311, + "num_input_tokens_seen": 170692975, + "step": 7900 + }, + { + "epoch": 0.9500390789394577, + "flos": 24281693358720.0, + "grad_norm": 2.2418443370686396, + "learning_rate": 2.606956303020502e-08, + "loss": 0.7, + "num_input_tokens_seen": 170713780, + "step": 7901 + }, + { + "epoch": 0.9501593218300968, + "flos": 14334429553920.0, + "grad_norm": 2.1272140637113988, + "learning_rate": 2.5944349961690036e-08, + "loss": 0.8439, + "num_input_tokens_seen": 170730800, + "step": 7902 + }, + { + "epoch": 0.9502795647207359, + "flos": 38672335056960.0, + "grad_norm": 2.8168125573772986, + "learning_rate": 2.581943635584749e-08, + "loss": 0.7336, + "num_input_tokens_seen": 170753630, + "step": 7903 + }, + { + "epoch": 0.950399807611375, + "flos": 40749224854080.0, + "grad_norm": 1.9122081259612373, + "learning_rate": 2.569482223162689e-08, + "loss": 0.655, + "num_input_tokens_seen": 170777605, + "step": 7904 + }, + { + "epoch": 0.950520050502014, + "flos": 23406876960000.0, + "grad_norm": 2.8652774778945216, + "learning_rate": 2.5570507607932e-08, + "loss": 0.7298, + "num_input_tokens_seen": 170797520, + "step": 7905 + }, + { + "epoch": 0.9506402933926532, + "flos": 17758377747840.0, + "grad_norm": 7.919247765315657, + "learning_rate": 2.54464925036213e-08, + "loss": 0.6374, + "num_input_tokens_seen": 170814810, + "step": 7906 + }, + { + "epoch": 0.9507605362832923, + "flos": 32515124578560.0, + "grad_norm": 2.6793187724630037, + "learning_rate": 2.532277693750773e-08, + "loss": 0.6083, + "num_input_tokens_seen": 170835735, + "step": 7907 + }, + { + "epoch": 0.9508807791739313, + "flos": 19573849657440.0, + "grad_norm": 12.250184972495378, + "learning_rate": 2.5199360928358948e-08, + "loss": 0.7576, + "num_input_tokens_seen": 170852970, + "step": 7908 + }, + { + "epoch": 0.9510010220645704, + "flos": 21441253455360.0, + "grad_norm": 2.244917745803798, + "learning_rate": 2.507624449489665e-08, + "loss": 0.8701, + "num_input_tokens_seen": 170871600, + "step": 7909 + }, + { + "epoch": 0.9511212649552095, + "flos": 18853124848320.0, + "grad_norm": 2.226016543194082, + "learning_rate": 2.495342765579811e-08, + "loss": 0.6495, + "num_input_tokens_seen": 170890260, + "step": 7910 + }, + { + "epoch": 0.9512415078458486, + "flos": 20780717954880.0, + "grad_norm": 2.6209887783446337, + "learning_rate": 2.4830910429693984e-08, + "loss": 0.711, + "num_input_tokens_seen": 170910220, + "step": 7911 + }, + { + "epoch": 0.9513617507364877, + "flos": 18343447168320.0, + "grad_norm": 2.5621337628908667, + "learning_rate": 2.470869283517052e-08, + "loss": 0.7957, + "num_input_tokens_seen": 170928255, + "step": 7912 + }, + { + "epoch": 0.9514819936271268, + "flos": 24974386435200.0, + "grad_norm": 2.0091366687351075, + "learning_rate": 2.458677489076777e-08, + "loss": 0.7747, + "num_input_tokens_seen": 170949265, + "step": 7913 + }, + { + "epoch": 0.9516022365177659, + "flos": 18635610180960.0, + "grad_norm": 2.3190632035367837, + "learning_rate": 2.446515661498072e-08, + "loss": 0.8305, + "num_input_tokens_seen": 170968595, + "step": 7914 + }, + { + "epoch": 0.9517224794084049, + "flos": 25335696668640.0, + "grad_norm": 3.0526681454747884, + "learning_rate": 2.434383802625861e-08, + "loss": 0.741, + "num_input_tokens_seen": 170987550, + "step": 7915 + }, + { + "epoch": 0.9518427222990441, + "flos": 21440287041600.0, + "grad_norm": 2.2642408681127075, + "learning_rate": 2.4222819143005168e-08, + "loss": 0.738, + "num_input_tokens_seen": 171007145, + "step": 7916 + }, + { + "epoch": 0.9519629651896832, + "flos": 21003659407200.0, + "grad_norm": 2.360968954506663, + "learning_rate": 2.4102099983579706e-08, + "loss": 0.8085, + "num_input_tokens_seen": 171026295, + "step": 7917 + }, + { + "epoch": 0.9520832080803222, + "flos": 21658247746080.0, + "grad_norm": 2.861038641233008, + "learning_rate": 2.3981680566294236e-08, + "loss": 0.7725, + "num_input_tokens_seen": 171045895, + "step": 7918 + }, + { + "epoch": 0.9522034509709614, + "flos": 23112074894400.0, + "grad_norm": 2.3463065890203216, + "learning_rate": 2.3861560909416822e-08, + "loss": 0.7354, + "num_input_tokens_seen": 171065445, + "step": 7919 + }, + { + "epoch": 0.9523236938616004, + "flos": 24644044345440.0, + "grad_norm": 3.761529365121699, + "learning_rate": 2.3741741031169325e-08, + "loss": 0.8265, + "num_input_tokens_seen": 171085015, + "step": 7920 + }, + { + "epoch": 0.9524439367522395, + "flos": 22639832896320.0, + "grad_norm": 2.5657391428676384, + "learning_rate": 2.3622220949728544e-08, + "loss": 0.7116, + "num_input_tokens_seen": 171103900, + "step": 7921 + }, + { + "epoch": 0.9525641796428787, + "flos": 34006461748320.0, + "grad_norm": 3.31426177767862, + "learning_rate": 2.3503000683225526e-08, + "loss": 0.6141, + "num_input_tokens_seen": 171121615, + "step": 7922 + }, + { + "epoch": 0.9526844225335177, + "flos": 16703333684640.0, + "grad_norm": 2.624207513256495, + "learning_rate": 2.3384080249745585e-08, + "loss": 0.8409, + "num_input_tokens_seen": 171139505, + "step": 7923 + }, + { + "epoch": 0.9528046654241568, + "flos": 36886158651840.0, + "grad_norm": 3.7468365359459717, + "learning_rate": 2.3265459667329178e-08, + "loss": 0.8306, + "num_input_tokens_seen": 171158995, + "step": 7924 + }, + { + "epoch": 0.9529249083147959, + "flos": 18228872767200.0, + "grad_norm": 3.0240799568747003, + "learning_rate": 2.31471389539708e-08, + "loss": 0.8603, + "num_input_tokens_seen": 171176120, + "step": 7925 + }, + { + "epoch": 0.953045151205435, + "flos": 28617336086880.0, + "grad_norm": 2.6541134488106786, + "learning_rate": 2.3029118127619872e-08, + "loss": 0.7294, + "num_input_tokens_seen": 171195835, + "step": 7926 + }, + { + "epoch": 0.953165394096074, + "flos": 21804050479200.0, + "grad_norm": 2.3990799729889947, + "learning_rate": 2.2911397206179628e-08, + "loss": 0.8681, + "num_input_tokens_seen": 171212585, + "step": 7927 + }, + { + "epoch": 0.9532856369867132, + "flos": 19946608176960.0, + "grad_norm": 2.836740931679232, + "learning_rate": 2.279397620750845e-08, + "loss": 0.6301, + "num_input_tokens_seen": 171231630, + "step": 7928 + }, + { + "epoch": 0.9534058798773523, + "flos": 15028312062720.0, + "grad_norm": 4.795667266400923, + "learning_rate": 2.2676855149419195e-08, + "loss": 0.7887, + "num_input_tokens_seen": 171248800, + "step": 7929 + }, + { + "epoch": 0.9535261227679913, + "flos": 17577778385760.0, + "grad_norm": 3.2941317603065383, + "learning_rate": 2.2560034049678988e-08, + "loss": 0.7593, + "num_input_tokens_seen": 171263150, + "step": 7930 + }, + { + "epoch": 0.9536463656586305, + "flos": 23109658860000.0, + "grad_norm": 3.0529535157569234, + "learning_rate": 2.2443512926008988e-08, + "loss": 0.7556, + "num_input_tokens_seen": 171282480, + "step": 7931 + }, + { + "epoch": 0.9537666085492695, + "flos": 18598546196640.0, + "grad_norm": 3.1321640294799953, + "learning_rate": 2.2327291796085946e-08, + "loss": 0.6952, + "num_input_tokens_seen": 171300840, + "step": 7932 + }, + { + "epoch": 0.9538868514399086, + "flos": 18962309634240.0, + "grad_norm": 3.7296728433513597, + "learning_rate": 2.2211370677540197e-08, + "loss": 0.7786, + "num_input_tokens_seen": 171317365, + "step": 7933 + }, + { + "epoch": 0.9540070943305478, + "flos": 16776718258080.0, + "grad_norm": 3.7077471380505007, + "learning_rate": 2.2095749587957012e-08, + "loss": 0.7896, + "num_input_tokens_seen": 171335270, + "step": 7934 + }, + { + "epoch": 0.9541273372211868, + "flos": 20128173952800.0, + "grad_norm": 8.014876362304495, + "learning_rate": 2.1980428544876138e-08, + "loss": 0.6947, + "num_input_tokens_seen": 171353910, + "step": 7935 + }, + { + "epoch": 0.9542475801118259, + "flos": 26687252606400.0, + "grad_norm": 2.094360490303015, + "learning_rate": 2.1865407565791584e-08, + "loss": 0.7389, + "num_input_tokens_seen": 171375480, + "step": 7936 + }, + { + "epoch": 0.954367823002465, + "flos": 23297543495040.0, + "grad_norm": 8.686959009684431, + "learning_rate": 2.175068666815183e-08, + "loss": 0.7752, + "num_input_tokens_seen": 171396320, + "step": 7937 + }, + { + "epoch": 0.9544880658931041, + "flos": 14881617255360.0, + "grad_norm": 5.929795898733342, + "learning_rate": 2.163626586935985e-08, + "loss": 0.7947, + "num_input_tokens_seen": 171412860, + "step": 7938 + }, + { + "epoch": 0.9546083087837431, + "flos": 29054149570080.0, + "grad_norm": 3.363259274436409, + "learning_rate": 2.1522145186773755e-08, + "loss": 0.633, + "num_input_tokens_seen": 171431780, + "step": 7939 + }, + { + "epoch": 0.9547285516743822, + "flos": 21112175137440.0, + "grad_norm": 2.8033555462290183, + "learning_rate": 2.140832463770481e-08, + "loss": 0.8588, + "num_input_tokens_seen": 171450975, + "step": 7940 + }, + { + "epoch": 0.9548487945650214, + "flos": 27450728373120.0, + "grad_norm": 2.555902687858966, + "learning_rate": 2.129480423941987e-08, + "loss": 0.7641, + "num_input_tokens_seen": 171467645, + "step": 7941 + }, + { + "epoch": 0.9549690374556604, + "flos": 22241384339040.0, + "grad_norm": 2.90321815431605, + "learning_rate": 2.1181584009140052e-08, + "loss": 0.8023, + "num_input_tokens_seen": 171487185, + "step": 7942 + }, + { + "epoch": 0.9550892803462995, + "flos": 17570009905920.0, + "grad_norm": 3.0024797998871646, + "learning_rate": 2.10686639640405e-08, + "loss": 0.8408, + "num_input_tokens_seen": 171501275, + "step": 7943 + }, + { + "epoch": 0.9552095232369386, + "flos": 24318311305920.0, + "grad_norm": 2.4130126250371893, + "learning_rate": 2.0956044121251294e-08, + "loss": 0.8158, + "num_input_tokens_seen": 171520060, + "step": 7944 + }, + { + "epoch": 0.9553297661275777, + "flos": 22714332562560.0, + "grad_norm": 2.7199143295260124, + "learning_rate": 2.084372449785654e-08, + "loss": 0.8111, + "num_input_tokens_seen": 171539895, + "step": 7945 + }, + { + "epoch": 0.9554500090182168, + "flos": 15392038330560.0, + "grad_norm": 4.043370913346376, + "learning_rate": 2.0731705110895282e-08, + "loss": 0.6862, + "num_input_tokens_seen": 171557385, + "step": 7946 + }, + { + "epoch": 0.9555702519088559, + "flos": 23479889835840.0, + "grad_norm": 2.2126279373679596, + "learning_rate": 2.0619985977360587e-08, + "loss": 0.8687, + "num_input_tokens_seen": 171575705, + "step": 7947 + }, + { + "epoch": 0.955690494799495, + "flos": 22929096667680.0, + "grad_norm": 2.004306448372917, + "learning_rate": 2.0508567114200237e-08, + "loss": 0.7698, + "num_input_tokens_seen": 171595620, + "step": 7948 + }, + { + "epoch": 0.955810737690134, + "flos": 25993964813760.0, + "grad_norm": 8.199786802172007, + "learning_rate": 2.0397448538316485e-08, + "loss": 0.7872, + "num_input_tokens_seen": 171616660, + "step": 7949 + }, + { + "epoch": 0.9559309805807732, + "flos": 20820197973600.0, + "grad_norm": 3.152615560278583, + "learning_rate": 2.028663026656563e-08, + "loss": 0.6662, + "num_input_tokens_seen": 171635515, + "step": 7950 + }, + { + "epoch": 0.9560512234714122, + "flos": 21547353151200.0, + "grad_norm": 2.4766970817508986, + "learning_rate": 2.0176112315758885e-08, + "loss": 0.7189, + "num_input_tokens_seen": 171653305, + "step": 7951 + }, + { + "epoch": 0.9561714663620513, + "flos": 17425099247040.0, + "grad_norm": 3.19741371664707, + "learning_rate": 2.0065894702661957e-08, + "loss": 0.696, + "num_input_tokens_seen": 171669980, + "step": 7952 + }, + { + "epoch": 0.9562917092526905, + "flos": 26060919018720.0, + "grad_norm": 2.863865062944636, + "learning_rate": 1.9955977443994577e-08, + "loss": 0.7796, + "num_input_tokens_seen": 171689970, + "step": 7953 + }, + { + "epoch": 0.9564119521433295, + "flos": 24062245863840.0, + "grad_norm": 3.113659844059996, + "learning_rate": 1.9846360556430965e-08, + "loss": 0.6292, + "num_input_tokens_seen": 171708270, + "step": 7954 + }, + { + "epoch": 0.9565321950339686, + "flos": 31962695940960.0, + "grad_norm": 2.5826566691850914, + "learning_rate": 1.973704405660004e-08, + "loss": 0.6155, + "num_input_tokens_seen": 171729385, + "step": 7955 + }, + { + "epoch": 0.9566524379246077, + "flos": 23554984218240.0, + "grad_norm": 3.448829153658953, + "learning_rate": 1.9628027961085203e-08, + "loss": 0.7811, + "num_input_tokens_seen": 171752615, + "step": 7956 + }, + { + "epoch": 0.9567726808152468, + "flos": 38012765970240.0, + "grad_norm": 2.4942510307090413, + "learning_rate": 1.9519312286423894e-08, + "loss": 0.8432, + "num_input_tokens_seen": 171775810, + "step": 7957 + }, + { + "epoch": 0.9568929237058859, + "flos": 22711433321280.0, + "grad_norm": 2.329695613445274, + "learning_rate": 1.9410897049108255e-08, + "loss": 0.7774, + "num_input_tokens_seen": 171795090, + "step": 7958 + }, + { + "epoch": 0.957013166596525, + "flos": 23807518533120.0, + "grad_norm": 2.257714292508935, + "learning_rate": 1.9302782265584905e-08, + "loss": 0.9132, + "num_input_tokens_seen": 171815755, + "step": 7959 + }, + { + "epoch": 0.9571334094871641, + "flos": 17614210484160.0, + "grad_norm": 2.8454216036965776, + "learning_rate": 1.9194967952254282e-08, + "loss": 0.8718, + "num_input_tokens_seen": 171834330, + "step": 7960 + }, + { + "epoch": 0.9572536523778031, + "flos": 15347577564000.0, + "grad_norm": 3.1568113526563106, + "learning_rate": 1.9087454125472635e-08, + "loss": 0.8087, + "num_input_tokens_seen": 171848805, + "step": 7961 + }, + { + "epoch": 0.9573738952684423, + "flos": 24933977172480.0, + "grad_norm": 2.449204148244452, + "learning_rate": 1.8980240801548696e-08, + "loss": 0.7883, + "num_input_tokens_seen": 171867995, + "step": 7962 + }, + { + "epoch": 0.9574941381590814, + "flos": 25732323907680.0, + "grad_norm": 2.8571111336718014, + "learning_rate": 1.8873327996747458e-08, + "loss": 0.7435, + "num_input_tokens_seen": 171886495, + "step": 7963 + }, + { + "epoch": 0.9576143810497204, + "flos": 32260768945440.0, + "grad_norm": 2.3566443196828426, + "learning_rate": 1.8766715727287053e-08, + "loss": 0.6636, + "num_input_tokens_seen": 171908200, + "step": 7964 + }, + { + "epoch": 0.9577346239403596, + "flos": 27705938910720.0, + "grad_norm": 2.3248464554158765, + "learning_rate": 1.8660404009340546e-08, + "loss": 0.795, + "num_input_tokens_seen": 171928650, + "step": 7965 + }, + { + "epoch": 0.9578548668309986, + "flos": 57386087540640.0, + "grad_norm": 0.8874439853130612, + "learning_rate": 1.8554392859035485e-08, + "loss": 0.5944, + "num_input_tokens_seen": 171986400, + "step": 7966 + }, + { + "epoch": 0.9579751097216377, + "flos": 19727904077280.0, + "grad_norm": 2.3958307570291573, + "learning_rate": 1.8448682292453444e-08, + "loss": 0.7909, + "num_input_tokens_seen": 172005475, + "step": 7967 + }, + { + "epoch": 0.9580953526122769, + "flos": 18015632622240.0, + "grad_norm": 2.1135707240945383, + "learning_rate": 1.8343272325631154e-08, + "loss": 0.6649, + "num_input_tokens_seen": 172024420, + "step": 7968 + }, + { + "epoch": 0.9582155955029159, + "flos": 24235076746080.0, + "grad_norm": 2.790674080582619, + "learning_rate": 1.8238162974558492e-08, + "loss": 0.7832, + "num_input_tokens_seen": 172038350, + "step": 7969 + }, + { + "epoch": 0.958335838393555, + "flos": 22750727491200.0, + "grad_norm": 3.1708322699633174, + "learning_rate": 1.8133354255181144e-08, + "loss": 0.7505, + "num_input_tokens_seen": 172058665, + "step": 7970 + }, + { + "epoch": 0.958456081284194, + "flos": 16887538513440.0, + "grad_norm": 1.985744574395204, + "learning_rate": 1.802884618339795e-08, + "loss": 0.751, + "num_input_tokens_seen": 172077470, + "step": 7971 + }, + { + "epoch": 0.9585763241748332, + "flos": 19946496667680.0, + "grad_norm": 2.3682238636135047, + "learning_rate": 1.7924638775062894e-08, + "loss": 0.809, + "num_input_tokens_seen": 172096590, + "step": 7972 + }, + { + "epoch": 0.9586965670654722, + "flos": 21365043980160.0, + "grad_norm": 2.4984733688709344, + "learning_rate": 1.7820732045984444e-08, + "loss": 0.8149, + "num_input_tokens_seen": 172116735, + "step": 7973 + }, + { + "epoch": 0.9588168099561113, + "flos": 21404561168640.0, + "grad_norm": 2.553776663231635, + "learning_rate": 1.7717126011924655e-08, + "loss": 0.7398, + "num_input_tokens_seen": 172138320, + "step": 7974 + }, + { + "epoch": 0.9589370528467505, + "flos": 11747193021120.0, + "grad_norm": 3.3144652097231075, + "learning_rate": 1.7613820688600957e-08, + "loss": 0.7713, + "num_input_tokens_seen": 172154225, + "step": 7975 + }, + { + "epoch": 0.9590572957373895, + "flos": 23408029222560.0, + "grad_norm": 2.319861165858298, + "learning_rate": 1.7510816091684588e-08, + "loss": 0.7873, + "num_input_tokens_seen": 172174150, + "step": 7976 + }, + { + "epoch": 0.9591775386280286, + "flos": 22496074500000.0, + "grad_norm": 3.3163158103374357, + "learning_rate": 1.740811223680083e-08, + "loss": 0.7868, + "num_input_tokens_seen": 172191005, + "step": 7977 + }, + { + "epoch": 0.9592977815186677, + "flos": 18161100827520.0, + "grad_norm": 3.1929343777778927, + "learning_rate": 1.7305709139530334e-08, + "loss": 0.7433, + "num_input_tokens_seen": 172209785, + "step": 7978 + }, + { + "epoch": 0.9594180244093068, + "flos": 16514073768480.0, + "grad_norm": 2.9771418771043967, + "learning_rate": 1.7203606815407334e-08, + "loss": 0.7438, + "num_input_tokens_seen": 172224380, + "step": 7979 + }, + { + "epoch": 0.9595382672999458, + "flos": 20524875531360.0, + "grad_norm": 2.6234644985570528, + "learning_rate": 1.7101805279920557e-08, + "loss": 0.7963, + "num_input_tokens_seen": 172242540, + "step": 7980 + }, + { + "epoch": 0.959658510190585, + "flos": 22605928341600.0, + "grad_norm": 2.483367331696659, + "learning_rate": 1.7000304548513643e-08, + "loss": 0.8118, + "num_input_tokens_seen": 172262645, + "step": 7981 + }, + { + "epoch": 0.9597787530812241, + "flos": 19108000858080.0, + "grad_norm": 3.729014487686719, + "learning_rate": 1.6899104636583394e-08, + "loss": 0.8248, + "num_input_tokens_seen": 172280695, + "step": 7982 + }, + { + "epoch": 0.9598989959718631, + "flos": 60012172206240.0, + "grad_norm": 0.7506329329402593, + "learning_rate": 1.6798205559482638e-08, + "loss": 0.6179, + "num_input_tokens_seen": 172343075, + "step": 7983 + }, + { + "epoch": 0.9600192388625023, + "flos": 20857001769600.0, + "grad_norm": 2.7294329641942614, + "learning_rate": 1.669760733251713e-08, + "loss": 0.7669, + "num_input_tokens_seen": 172361950, + "step": 7984 + }, + { + "epoch": 0.9601394817531413, + "flos": 20416471310400.0, + "grad_norm": 3.318041359053287, + "learning_rate": 1.659730997094755e-08, + "loss": 0.8243, + "num_input_tokens_seen": 172380440, + "step": 7985 + }, + { + "epoch": 0.9602597246437804, + "flos": 21476458951680.0, + "grad_norm": 2.8181785900902927, + "learning_rate": 1.6497313489989283e-08, + "loss": 0.6221, + "num_input_tokens_seen": 172400265, + "step": 7986 + }, + { + "epoch": 0.9603799675344196, + "flos": 29892125003040.0, + "grad_norm": 3.2798970657572304, + "learning_rate": 1.639761790481131e-08, + "loss": 0.6994, + "num_input_tokens_seen": 172421145, + "step": 7987 + }, + { + "epoch": 0.9605002104250586, + "flos": 27961372466880.0, + "grad_norm": 2.225915861762304, + "learning_rate": 1.6298223230537754e-08, + "loss": 0.7886, + "num_input_tokens_seen": 172438945, + "step": 7988 + }, + { + "epoch": 0.9606204533156977, + "flos": 35539286103840.0, + "grad_norm": 3.0107439296641894, + "learning_rate": 1.619912948224611e-08, + "loss": 0.6934, + "num_input_tokens_seen": 172460150, + "step": 7989 + }, + { + "epoch": 0.9607406962063368, + "flos": 26536394785920.0, + "grad_norm": 2.5557990499493353, + "learning_rate": 1.6100336674969682e-08, + "loss": 0.6081, + "num_input_tokens_seen": 172478990, + "step": 7990 + }, + { + "epoch": 0.9608609390969759, + "flos": 25295287405920.0, + "grad_norm": 2.128557929145235, + "learning_rate": 1.600184482369449e-08, + "loss": 0.7707, + "num_input_tokens_seen": 172498905, + "step": 7991 + }, + { + "epoch": 0.960981181987615, + "flos": 21039050752320.0, + "grad_norm": 3.3110947417239323, + "learning_rate": 1.5903653943362126e-08, + "loss": 0.8928, + "num_input_tokens_seen": 172517900, + "step": 7992 + }, + { + "epoch": 0.9611014248782541, + "flos": 17797969275840.0, + "grad_norm": 2.4213726876389448, + "learning_rate": 1.580576404886802e-08, + "loss": 0.7716, + "num_input_tokens_seen": 172536430, + "step": 7993 + }, + { + "epoch": 0.9612216677688932, + "flos": 19327076655360.0, + "grad_norm": 3.298868308943383, + "learning_rate": 1.570817515506162e-08, + "loss": 0.7993, + "num_input_tokens_seen": 172555120, + "step": 7994 + }, + { + "epoch": 0.9613419106595322, + "flos": 15786100856160.0, + "grad_norm": 3.8757927606646625, + "learning_rate": 1.561088727674753e-08, + "loss": 0.8137, + "num_input_tokens_seen": 172569330, + "step": 7995 + }, + { + "epoch": 0.9614621535501714, + "flos": 25665667060800.0, + "grad_norm": 4.7101847741558345, + "learning_rate": 1.551390042868417e-08, + "loss": 0.714, + "num_input_tokens_seen": 172591100, + "step": 7996 + }, + { + "epoch": 0.9615823964408104, + "flos": 17793620413920.0, + "grad_norm": 4.060806860589259, + "learning_rate": 1.5417214625584207e-08, + "loss": 0.7077, + "num_input_tokens_seen": 172608755, + "step": 7997 + }, + { + "epoch": 0.9617026393314495, + "flos": 20161112093760.0, + "grad_norm": 1.9424612452348957, + "learning_rate": 1.5320829882114806e-08, + "loss": 0.8532, + "num_input_tokens_seen": 172626830, + "step": 7998 + }, + { + "epoch": 0.9618228822220887, + "flos": 20238139303680.0, + "grad_norm": 2.5743059945264064, + "learning_rate": 1.5224746212897378e-08, + "loss": 0.7899, + "num_input_tokens_seen": 172646125, + "step": 7999 + }, + { + "epoch": 0.9619431251127277, + "flos": 21001243372800.0, + "grad_norm": 2.164023278497237, + "learning_rate": 1.512896363250804e-08, + "loss": 0.7752, + "num_input_tokens_seen": 172666235, + "step": 8000 + }, + { + "epoch": 0.9620633680033668, + "flos": 22350680634240.0, + "grad_norm": 2.57221789015242, + "learning_rate": 1.503348215547673e-08, + "loss": 0.7565, + "num_input_tokens_seen": 172687325, + "step": 8001 + }, + { + "epoch": 0.962183610894006, + "flos": 18445198002240.0, + "grad_norm": 2.2315696765125597, + "learning_rate": 1.4938301796288078e-08, + "loss": 0.8126, + "num_input_tokens_seen": 172703895, + "step": 8002 + }, + { + "epoch": 0.962303853784645, + "flos": 18408765903840.0, + "grad_norm": 8.332542636234631, + "learning_rate": 1.4843422569380537e-08, + "loss": 0.814, + "num_input_tokens_seen": 172720650, + "step": 8003 + }, + { + "epoch": 0.9624240966752841, + "flos": 26353082031360.0, + "grad_norm": 2.6789426964088072, + "learning_rate": 1.4748844489147483e-08, + "loss": 0.8313, + "num_input_tokens_seen": 172737640, + "step": 8004 + }, + { + "epoch": 0.9625443395659231, + "flos": 14626964264160.0, + "grad_norm": 4.16286361068461, + "learning_rate": 1.4654567569936326e-08, + "loss": 0.7066, + "num_input_tokens_seen": 172755215, + "step": 8005 + }, + { + "epoch": 0.9626645824565623, + "flos": 18341440001280.0, + "grad_norm": 2.9991579463486073, + "learning_rate": 1.456059182604874e-08, + "loss": 0.8335, + "num_input_tokens_seen": 172774020, + "step": 8006 + }, + { + "epoch": 0.9627848253472013, + "flos": 16556861895840.0, + "grad_norm": 2.7451947451301493, + "learning_rate": 1.4466917271740653e-08, + "loss": 0.7671, + "num_input_tokens_seen": 172792220, + "step": 8007 + }, + { + "epoch": 0.9629050682378404, + "flos": 20857076109120.0, + "grad_norm": 3.094379707753532, + "learning_rate": 1.4373543921222697e-08, + "loss": 0.6727, + "num_input_tokens_seen": 172811805, + "step": 8008 + }, + { + "epoch": 0.9630253111284796, + "flos": 16995087829920.0, + "grad_norm": 2.359534352965323, + "learning_rate": 1.428047178865932e-08, + "loss": 0.7828, + "num_input_tokens_seen": 172828595, + "step": 8009 + }, + { + "epoch": 0.9631455540191186, + "flos": 20309925577440.0, + "grad_norm": 2.4320500408344903, + "learning_rate": 1.4187700888169451e-08, + "loss": 0.7506, + "num_input_tokens_seen": 172847770, + "step": 8010 + }, + { + "epoch": 0.9632657969097577, + "flos": 65862977920320.0, + "grad_norm": 0.7618945579115152, + "learning_rate": 1.40952312338265e-08, + "loss": 0.569, + "num_input_tokens_seen": 172912415, + "step": 8011 + }, + { + "epoch": 0.9633860398003968, + "flos": 44355445049280.0, + "grad_norm": 3.1240648014708436, + "learning_rate": 1.4003062839657909e-08, + "loss": 0.6864, + "num_input_tokens_seen": 172934895, + "step": 8012 + }, + { + "epoch": 0.9635062826910359, + "flos": 24792188773440.0, + "grad_norm": 2.1970259200847337, + "learning_rate": 1.391119571964583e-08, + "loss": 0.8002, + "num_input_tokens_seen": 172955835, + "step": 8013 + }, + { + "epoch": 0.9636265255816749, + "flos": 15938742825120.0, + "grad_norm": 2.3072586459056654, + "learning_rate": 1.3819629887726225e-08, + "loss": 0.7317, + "num_input_tokens_seen": 172973925, + "step": 8014 + }, + { + "epoch": 0.9637467684723141, + "flos": 22569496243200.0, + "grad_norm": 4.042750786697078, + "learning_rate": 1.3728365357789317e-08, + "loss": 0.7639, + "num_input_tokens_seen": 172993290, + "step": 8015 + }, + { + "epoch": 0.9638670113629532, + "flos": 17540825910720.0, + "grad_norm": 3.214826692742188, + "learning_rate": 1.3637402143680254e-08, + "loss": 0.7684, + "num_input_tokens_seen": 173008190, + "step": 8016 + }, + { + "epoch": 0.9639872542535922, + "flos": 54993835067040.0, + "grad_norm": 0.7363011686617068, + "learning_rate": 1.3546740259197998e-08, + "loss": 0.5496, + "num_input_tokens_seen": 173061000, + "step": 8017 + }, + { + "epoch": 0.9641074971442314, + "flos": 24099978903840.0, + "grad_norm": 2.4132793847617426, + "learning_rate": 1.3456379718095989e-08, + "loss": 0.7049, + "num_input_tokens_seen": 173081415, + "step": 8018 + }, + { + "epoch": 0.9642277400348704, + "flos": 66652106555040.0, + "grad_norm": 0.8469490581942981, + "learning_rate": 1.3366320534081487e-08, + "loss": 0.6189, + "num_input_tokens_seen": 173144095, + "step": 8019 + }, + { + "epoch": 0.9643479829255095, + "flos": 30875865999360.0, + "grad_norm": 3.3778770334596, + "learning_rate": 1.3276562720816675e-08, + "loss": 0.7605, + "num_input_tokens_seen": 173165605, + "step": 8020 + }, + { + "epoch": 0.9644682258161487, + "flos": 20019695392320.0, + "grad_norm": 2.727680665631457, + "learning_rate": 1.3187106291917549e-08, + "loss": 0.827, + "num_input_tokens_seen": 173182595, + "step": 8021 + }, + { + "epoch": 0.9645884687067877, + "flos": 21148495726560.0, + "grad_norm": 2.166881391693538, + "learning_rate": 1.309795126095503e-08, + "loss": 0.7057, + "num_input_tokens_seen": 173200895, + "step": 8022 + }, + { + "epoch": 0.9647087115974268, + "flos": 18918740941920.0, + "grad_norm": 3.431511396168769, + "learning_rate": 1.3009097641453192e-08, + "loss": 0.81, + "num_input_tokens_seen": 173218745, + "step": 8023 + }, + { + "epoch": 0.9648289544880659, + "flos": 16521619229760.0, + "grad_norm": 2.212130814262665, + "learning_rate": 1.2920545446891474e-08, + "loss": 0.759, + "num_input_tokens_seen": 173235465, + "step": 8024 + }, + { + "epoch": 0.964949197378705, + "flos": 24022988863680.0, + "grad_norm": 3.480092654474594, + "learning_rate": 1.2832294690703127e-08, + "loss": 0.7073, + "num_input_tokens_seen": 173254440, + "step": 8025 + }, + { + "epoch": 0.965069440269344, + "flos": 23332079935680.0, + "grad_norm": 3.0680908241008185, + "learning_rate": 1.2744345386275668e-08, + "loss": 0.7738, + "num_input_tokens_seen": 173273980, + "step": 8026 + }, + { + "epoch": 0.9651896831599832, + "flos": 25338558740160.0, + "grad_norm": 2.0766401220414115, + "learning_rate": 1.265669754695109e-08, + "loss": 0.7892, + "num_input_tokens_seen": 173293550, + "step": 8027 + }, + { + "epoch": 0.9653099260506223, + "flos": 22241049811200.0, + "grad_norm": 2.4117433339735936, + "learning_rate": 1.2569351186025201e-08, + "loss": 0.8225, + "num_input_tokens_seen": 173312005, + "step": 8028 + }, + { + "epoch": 0.9654301689412613, + "flos": 26722495272480.0, + "grad_norm": 1.6637154517251687, + "learning_rate": 1.2482306316748737e-08, + "loss": 0.7544, + "num_input_tokens_seen": 173332450, + "step": 8029 + }, + { + "epoch": 0.9655504118319005, + "flos": 17387663565120.0, + "grad_norm": 2.4656257658402447, + "learning_rate": 1.2395562952326021e-08, + "loss": 0.7865, + "num_input_tokens_seen": 173349610, + "step": 8030 + }, + { + "epoch": 0.9656706547225395, + "flos": 22095358587360.0, + "grad_norm": 2.9879464554080855, + "learning_rate": 1.2309121105916309e-08, + "loss": 0.8146, + "num_input_tokens_seen": 173367900, + "step": 8031 + }, + { + "epoch": 0.9657908976131786, + "flos": 36995938153920.0, + "grad_norm": 2.4614501419178616, + "learning_rate": 1.222298079063222e-08, + "loss": 0.694, + "num_input_tokens_seen": 173389040, + "step": 8032 + }, + { + "epoch": 0.9659111405038178, + "flos": 24354929253120.0, + "grad_norm": 2.6407768612143467, + "learning_rate": 1.2137142019541524e-08, + "loss": 0.7286, + "num_input_tokens_seen": 173407595, + "step": 8033 + }, + { + "epoch": 0.9660313833944568, + "flos": 24973605870240.0, + "grad_norm": 2.507564449223688, + "learning_rate": 1.2051604805666027e-08, + "loss": 0.7366, + "num_input_tokens_seen": 173424720, + "step": 8034 + }, + { + "epoch": 0.9661516262850959, + "flos": 11785000400640.0, + "grad_norm": 3.105151197404119, + "learning_rate": 1.196636916198135e-08, + "loss": 0.7839, + "num_input_tokens_seen": 173442260, + "step": 8035 + }, + { + "epoch": 0.9662718691757349, + "flos": 20019212185440.0, + "grad_norm": 2.6628603194444094, + "learning_rate": 1.1881435101418036e-08, + "loss": 0.772, + "num_input_tokens_seen": 173461675, + "step": 8036 + }, + { + "epoch": 0.9663921120663741, + "flos": 67930649616960.0, + "grad_norm": 0.7376506069737556, + "learning_rate": 1.1796802636860003e-08, + "loss": 0.6545, + "num_input_tokens_seen": 173530205, + "step": 8037 + }, + { + "epoch": 0.9665123549570132, + "flos": 26285198582400.0, + "grad_norm": 2.6314227230982747, + "learning_rate": 1.1712471781146316e-08, + "loss": 0.7425, + "num_input_tokens_seen": 173549540, + "step": 8038 + }, + { + "epoch": 0.9666325978476522, + "flos": 43877590417440.0, + "grad_norm": 2.597981505717994, + "learning_rate": 1.1628442547069628e-08, + "loss": 0.6716, + "num_input_tokens_seen": 173571740, + "step": 8039 + }, + { + "epoch": 0.9667528407382914, + "flos": 21512668031520.0, + "grad_norm": 2.9795806881624314, + "learning_rate": 1.1544714947377521e-08, + "loss": 0.7706, + "num_input_tokens_seen": 173589425, + "step": 8040 + }, + { + "epoch": 0.9668730836289304, + "flos": 23844433838400.0, + "grad_norm": 2.487454694785829, + "learning_rate": 1.1461288994770945e-08, + "loss": 0.702, + "num_input_tokens_seen": 173608500, + "step": 8041 + }, + { + "epoch": 0.9669933265195695, + "flos": 28252420386720.0, + "grad_norm": 2.243033007130176, + "learning_rate": 1.1378164701906002e-08, + "loss": 0.7717, + "num_input_tokens_seen": 173630265, + "step": 8042 + }, + { + "epoch": 0.9671135694102087, + "flos": 22422652756800.0, + "grad_norm": 3.162947545405676, + "learning_rate": 1.1295342081392156e-08, + "loss": 0.6655, + "num_input_tokens_seen": 173649625, + "step": 8043 + }, + { + "epoch": 0.9672338123008477, + "flos": 20127170369280.0, + "grad_norm": 2.371070522339122, + "learning_rate": 1.1212821145793804e-08, + "loss": 0.695, + "num_input_tokens_seen": 173669240, + "step": 8044 + }, + { + "epoch": 0.9673540551914868, + "flos": 16953191776800.0, + "grad_norm": 3.006943614353675, + "learning_rate": 1.1130601907629156e-08, + "loss": 0.7885, + "num_input_tokens_seen": 173686970, + "step": 8045 + }, + { + "epoch": 0.9674742980821259, + "flos": 61804593131040.0, + "grad_norm": 0.8148512564703795, + "learning_rate": 1.1048684379370899e-08, + "loss": 0.6449, + "num_input_tokens_seen": 173747655, + "step": 8046 + }, + { + "epoch": 0.967594540972765, + "flos": 18671224544640.0, + "grad_norm": 2.400419925669165, + "learning_rate": 1.0967068573445759e-08, + "loss": 0.7495, + "num_input_tokens_seen": 173765140, + "step": 8047 + }, + { + "epoch": 0.967714783863404, + "flos": 20747779813920.0, + "grad_norm": 3.1675513353530937, + "learning_rate": 1.0885754502234945e-08, + "loss": 0.6551, + "num_input_tokens_seen": 173784800, + "step": 8048 + }, + { + "epoch": 0.9678350267540432, + "flos": 23152335478080.0, + "grad_norm": 2.844359615504772, + "learning_rate": 1.08047421780737e-08, + "loss": 0.7821, + "num_input_tokens_seen": 173803990, + "step": 8049 + }, + { + "epoch": 0.9679552696446823, + "flos": 21695051542080.0, + "grad_norm": 2.779130792648393, + "learning_rate": 1.0724031613251305e-08, + "loss": 0.7382, + "num_input_tokens_seen": 173821890, + "step": 8050 + }, + { + "epoch": 0.9680755125353213, + "flos": 26830490626080.0, + "grad_norm": 2.935836645711531, + "learning_rate": 1.0643622820011744e-08, + "loss": 0.6668, + "num_input_tokens_seen": 173842945, + "step": 8051 + }, + { + "epoch": 0.9681957554259605, + "flos": 28284206265120.0, + "grad_norm": 3.609431939274568, + "learning_rate": 1.0563515810552814e-08, + "loss": 0.6852, + "num_input_tokens_seen": 173859915, + "step": 8052 + }, + { + "epoch": 0.9683159983165995, + "flos": 20528332319040.0, + "grad_norm": 1.7454916920962769, + "learning_rate": 1.0483710597026795e-08, + "loss": 0.7329, + "num_input_tokens_seen": 173879775, + "step": 8053 + }, + { + "epoch": 0.9684362412072386, + "flos": 24173363477280.0, + "grad_norm": 2.862165079971336, + "learning_rate": 1.0404207191540227e-08, + "loss": 0.745, + "num_input_tokens_seen": 173900230, + "step": 8054 + }, + { + "epoch": 0.9685564840978778, + "flos": 22314322875360.0, + "grad_norm": 5.59947925598733, + "learning_rate": 1.0325005606153236e-08, + "loss": 0.748, + "num_input_tokens_seen": 173919690, + "step": 8055 + }, + { + "epoch": 0.9686767269885168, + "flos": 14365843734720.0, + "grad_norm": 2.972521183589723, + "learning_rate": 1.0246105852881104e-08, + "loss": 0.7984, + "num_input_tokens_seen": 173934790, + "step": 8056 + }, + { + "epoch": 0.9687969698791559, + "flos": 21257234475360.0, + "grad_norm": 2.3663930031779383, + "learning_rate": 1.0167507943692476e-08, + "loss": 0.7886, + "num_input_tokens_seen": 173953875, + "step": 8057 + }, + { + "epoch": 0.968917212769795, + "flos": 19800062048640.0, + "grad_norm": 2.961763855458103, + "learning_rate": 1.008921189051093e-08, + "loss": 0.7179, + "num_input_tokens_seen": 173971220, + "step": 8058 + }, + { + "epoch": 0.9690374556604341, + "flos": 21650516436000.0, + "grad_norm": 2.4664456199715525, + "learning_rate": 1.0011217705213848e-08, + "loss": 0.7747, + "num_input_tokens_seen": 173989095, + "step": 8059 + }, + { + "epoch": 0.9691576985510731, + "flos": 32588546321760.0, + "grad_norm": 2.236337186969904, + "learning_rate": 9.933525399632658e-09, + "loss": 0.7437, + "num_input_tokens_seen": 174007330, + "step": 8060 + }, + { + "epoch": 0.9692779414417123, + "flos": 35611852942560.0, + "grad_norm": 2.3763844319885226, + "learning_rate": 9.856134985553488e-09, + "loss": 0.6494, + "num_input_tokens_seen": 174027055, + "step": 8061 + }, + { + "epoch": 0.9693981843323514, + "flos": 28326288167040.0, + "grad_norm": 3.9546508256831188, + "learning_rate": 9.77904647471628e-09, + "loss": 0.7384, + "num_input_tokens_seen": 174050235, + "step": 8062 + }, + { + "epoch": 0.9695184272229904, + "flos": 23589074621760.0, + "grad_norm": 2.6416726827302255, + "learning_rate": 9.702259878815454e-09, + "loss": 0.7393, + "num_input_tokens_seen": 174070560, + "step": 8063 + }, + { + "epoch": 0.9696386701136296, + "flos": 23261371584960.0, + "grad_norm": 2.896875829603947, + "learning_rate": 9.625775209499254e-09, + "loss": 0.7484, + "num_input_tokens_seen": 174090565, + "step": 8064 + }, + { + "epoch": 0.9697589130042686, + "flos": 15100470034080.0, + "grad_norm": 3.267442991172382, + "learning_rate": 9.549592478370172e-09, + "loss": 0.7382, + "num_input_tokens_seen": 174108745, + "step": 8065 + }, + { + "epoch": 0.9698791558949077, + "flos": 18852641641440.0, + "grad_norm": 2.2058064260692345, + "learning_rate": 9.473711696985632e-09, + "loss": 0.7969, + "num_input_tokens_seen": 174128075, + "step": 8066 + }, + { + "epoch": 0.9699993987855468, + "flos": 17905964629440.0, + "grad_norm": 2.4330180116261904, + "learning_rate": 9.398132876856201e-09, + "loss": 0.761, + "num_input_tokens_seen": 174147040, + "step": 8067 + }, + { + "epoch": 0.9701196416761859, + "flos": 67086132306240.0, + "grad_norm": 0.7817253192730043, + "learning_rate": 9.322856029447379e-09, + "loss": 0.6073, + "num_input_tokens_seen": 174208225, + "step": 8068 + }, + { + "epoch": 0.970239884566825, + "flos": 24242993904960.0, + "grad_norm": 4.129127518166435, + "learning_rate": 9.247881166178695e-09, + "loss": 0.8028, + "num_input_tokens_seen": 174226935, + "step": 8069 + }, + { + "epoch": 0.970360127457464, + "flos": 25265471524800.0, + "grad_norm": 3.4849201682537236, + "learning_rate": 9.173208298423274e-09, + "loss": 0.7672, + "num_input_tokens_seen": 174248140, + "step": 8070 + }, + { + "epoch": 0.9704803703481032, + "flos": 29530108544160.0, + "grad_norm": 8.776451021301085, + "learning_rate": 9.09883743750961e-09, + "loss": 0.7623, + "num_input_tokens_seen": 174271030, + "step": 8071 + }, + { + "epoch": 0.9706006132387422, + "flos": 17355208631040.0, + "grad_norm": 1.963159039872348, + "learning_rate": 9.024768594719124e-09, + "loss": 0.8399, + "num_input_tokens_seen": 174289410, + "step": 8072 + }, + { + "epoch": 0.9707208561293813, + "flos": 18160543281120.0, + "grad_norm": 2.5028112817352413, + "learning_rate": 8.95100178128816e-09, + "loss": 0.7265, + "num_input_tokens_seen": 174308550, + "step": 8073 + }, + { + "epoch": 0.9708410990200205, + "flos": 31203494696640.0, + "grad_norm": 2.3550763918398414, + "learning_rate": 8.877537008407321e-09, + "loss": 0.7051, + "num_input_tokens_seen": 174328600, + "step": 8074 + }, + { + "epoch": 0.9709613419106595, + "flos": 30510541431840.0, + "grad_norm": 1.85207311615273, + "learning_rate": 8.804374287221028e-09, + "loss": 0.6872, + "num_input_tokens_seen": 174349835, + "step": 8075 + }, + { + "epoch": 0.9710815848012986, + "flos": 23696140731360.0, + "grad_norm": 1.9799717386714135, + "learning_rate": 8.731513628827958e-09, + "loss": 0.8462, + "num_input_tokens_seen": 174369200, + "step": 8076 + }, + { + "epoch": 0.9712018276919377, + "flos": 23727629251680.0, + "grad_norm": 2.6752220739003776, + "learning_rate": 8.658955044280825e-09, + "loss": 0.8287, + "num_input_tokens_seen": 174388125, + "step": 8077 + }, + { + "epoch": 0.9713220705825768, + "flos": 23297617834560.0, + "grad_norm": 4.869270870763528, + "learning_rate": 8.586698544587268e-09, + "loss": 0.7767, + "num_input_tokens_seen": 174409735, + "step": 8078 + }, + { + "epoch": 0.9714423134732159, + "flos": 22169263537440.0, + "grad_norm": 2.835503326893003, + "learning_rate": 8.514744140707853e-09, + "loss": 0.7424, + "num_input_tokens_seen": 174428875, + "step": 8079 + }, + { + "epoch": 0.971562556363855, + "flos": 20201075319360.0, + "grad_norm": 2.1886834349387407, + "learning_rate": 8.443091843558515e-09, + "loss": 0.7688, + "num_input_tokens_seen": 174447960, + "step": 8080 + }, + { + "epoch": 0.9716827992544941, + "flos": 24934348870080.0, + "grad_norm": 2.8176189039403106, + "learning_rate": 8.37174166400878e-09, + "loss": 0.6501, + "num_input_tokens_seen": 174463535, + "step": 8081 + }, + { + "epoch": 0.9718030421451331, + "flos": 24646163021760.0, + "grad_norm": 4.04242853803347, + "learning_rate": 8.300693612881992e-09, + "loss": 0.8504, + "num_input_tokens_seen": 174483600, + "step": 8082 + }, + { + "epoch": 0.9719232850357723, + "flos": 22059149507520.0, + "grad_norm": 3.1039943433887025, + "learning_rate": 8.22994770095664e-09, + "loss": 0.8164, + "num_input_tokens_seen": 174502175, + "step": 8083 + }, + { + "epoch": 0.9720435279264114, + "flos": 23622979176480.0, + "grad_norm": 2.9588946136859797, + "learning_rate": 8.159503938964585e-09, + "loss": 0.7564, + "num_input_tokens_seen": 174519495, + "step": 8084 + }, + { + "epoch": 0.9721637708170504, + "flos": 28325210244000.0, + "grad_norm": 5.167087148148134, + "learning_rate": 8.089362337592164e-09, + "loss": 0.7017, + "num_input_tokens_seen": 174543120, + "step": 8085 + }, + { + "epoch": 0.9722840137076896, + "flos": 29088946199040.0, + "grad_norm": 6.3524221891871875, + "learning_rate": 8.019522907479536e-09, + "loss": 0.7195, + "num_input_tokens_seen": 174563480, + "step": 8086 + }, + { + "epoch": 0.9724042565983286, + "flos": 19216516588320.0, + "grad_norm": 4.517673964747546, + "learning_rate": 7.949985659221558e-09, + "loss": 0.776, + "num_input_tokens_seen": 174580745, + "step": 8087 + }, + { + "epoch": 0.9725244994889677, + "flos": 23006755763520.0, + "grad_norm": 2.586326270253309, + "learning_rate": 7.880750603366904e-09, + "loss": 0.7915, + "num_input_tokens_seen": 174599615, + "step": 8088 + }, + { + "epoch": 0.9726447423796069, + "flos": 23334049932960.0, + "grad_norm": 2.2628723259687575, + "learning_rate": 7.811817750418282e-09, + "loss": 0.7971, + "num_input_tokens_seen": 174618375, + "step": 8089 + }, + { + "epoch": 0.9727649852702459, + "flos": 26503605324000.0, + "grad_norm": 2.4118219242943364, + "learning_rate": 7.743187110833105e-09, + "loss": 0.8029, + "num_input_tokens_seen": 174641135, + "step": 8090 + }, + { + "epoch": 0.972885228160885, + "flos": 20493089652960.0, + "grad_norm": 2.051534459895012, + "learning_rate": 7.674858695022602e-09, + "loss": 0.8075, + "num_input_tokens_seen": 174659490, + "step": 8091 + }, + { + "epoch": 0.9730054710515241, + "flos": 17541086099040.0, + "grad_norm": 3.013116455319215, + "learning_rate": 7.606832513351591e-09, + "loss": 0.7558, + "num_input_tokens_seen": 174677440, + "step": 8092 + }, + { + "epoch": 0.9731257139421632, + "flos": 68873423804160.0, + "grad_norm": 0.8372059405475588, + "learning_rate": 7.539108576140264e-09, + "loss": 0.6385, + "num_input_tokens_seen": 174741550, + "step": 8093 + }, + { + "epoch": 0.9732459568328022, + "flos": 18452557614720.0, + "grad_norm": 4.047278895898038, + "learning_rate": 7.471686893661732e-09, + "loss": 0.702, + "num_input_tokens_seen": 174759845, + "step": 8094 + }, + { + "epoch": 0.9733661997234414, + "flos": 20853991019040.0, + "grad_norm": 2.113558992478104, + "learning_rate": 7.4045674761442636e-09, + "loss": 0.6429, + "num_input_tokens_seen": 174777175, + "step": 8095 + }, + { + "epoch": 0.9734864426140805, + "flos": 23732721508800.0, + "grad_norm": 3.289796144842653, + "learning_rate": 7.337750333769488e-09, + "loss": 0.742, + "num_input_tokens_seen": 174796980, + "step": 8096 + }, + { + "epoch": 0.9736066855047195, + "flos": 34991986893120.0, + "grad_norm": 3.2929135789695443, + "learning_rate": 7.2712354766737425e-09, + "loss": 0.7339, + "num_input_tokens_seen": 174817310, + "step": 8097 + }, + { + "epoch": 0.9737269283953586, + "flos": 20381451662880.0, + "grad_norm": 1.7445605072697796, + "learning_rate": 7.2050229149469565e-09, + "loss": 0.805, + "num_input_tokens_seen": 174837320, + "step": 8098 + }, + { + "epoch": 0.9738471712859977, + "flos": 28870130590080.0, + "grad_norm": 2.303462229745903, + "learning_rate": 7.139112658633984e-09, + "loss": 0.6371, + "num_input_tokens_seen": 174857470, + "step": 8099 + }, + { + "epoch": 0.9739674141766368, + "flos": 27742482518400.0, + "grad_norm": 3.4762138277577153, + "learning_rate": 7.073504717733048e-09, + "loss": 0.7058, + "num_input_tokens_seen": 174877035, + "step": 8100 + }, + { + "epoch": 0.9740876570672758, + "flos": 68764647885600.0, + "grad_norm": 0.7437328512955589, + "learning_rate": 7.008199102196855e-09, + "loss": 0.5711, + "num_input_tokens_seen": 174938460, + "step": 8101 + }, + { + "epoch": 0.974207899957915, + "flos": 58153354622880.0, + "grad_norm": 0.804241908873749, + "learning_rate": 6.9431958219321464e-09, + "loss": 0.5886, + "num_input_tokens_seen": 174994625, + "step": 8102 + }, + { + "epoch": 0.9743281428485541, + "flos": 22567712094720.0, + "grad_norm": 2.455686653122498, + "learning_rate": 6.878494886800146e-09, + "loss": 0.7766, + "num_input_tokens_seen": 175015400, + "step": 8103 + }, + { + "epoch": 0.9744483857391931, + "flos": 19980215373600.0, + "grad_norm": 2.180575610296056, + "learning_rate": 6.814096306615669e-09, + "loss": 0.7617, + "num_input_tokens_seen": 175033540, + "step": 8104 + }, + { + "epoch": 0.9745686286298323, + "flos": 17650122205920.0, + "grad_norm": 2.9052076213658444, + "learning_rate": 6.750000091148011e-09, + "loss": 0.6536, + "num_input_tokens_seen": 175050835, + "step": 8105 + }, + { + "epoch": 0.9746888715204713, + "flos": 29417950177440.0, + "grad_norm": 2.5529469384769263, + "learning_rate": 6.686206250120729e-09, + "loss": 0.7284, + "num_input_tokens_seen": 175072330, + "step": 8106 + }, + { + "epoch": 0.9748091144111104, + "flos": 18453040821600.0, + "grad_norm": 2.362752757604336, + "learning_rate": 6.622714793210749e-09, + "loss": 0.7465, + "num_input_tokens_seen": 175091250, + "step": 8107 + }, + { + "epoch": 0.9749293573017496, + "flos": 20635584277440.0, + "grad_norm": 2.0379372484468647, + "learning_rate": 6.559525730050364e-09, + "loss": 0.7883, + "num_input_tokens_seen": 175111350, + "step": 8108 + }, + { + "epoch": 0.9750496001923886, + "flos": 18452297426400.0, + "grad_norm": 2.616526050393478, + "learning_rate": 6.496639070224574e-09, + "loss": 0.7603, + "num_input_tokens_seen": 175129835, + "step": 8109 + }, + { + "epoch": 0.9751698430830277, + "flos": 19545074529600.0, + "grad_norm": 2.9277607596584625, + "learning_rate": 6.4340548232739714e-09, + "loss": 0.8417, + "num_input_tokens_seen": 175146035, + "step": 8110 + }, + { + "epoch": 0.9752900859736668, + "flos": 23516656462080.0, + "grad_norm": 6.831525920241083, + "learning_rate": 6.371772998692071e-09, + "loss": 0.7948, + "num_input_tokens_seen": 175165290, + "step": 8111 + }, + { + "epoch": 0.9754103288643059, + "flos": 20274274044000.0, + "grad_norm": 3.679173395158628, + "learning_rate": 6.309793605927094e-09, + "loss": 0.6516, + "num_input_tokens_seen": 175183610, + "step": 8112 + }, + { + "epoch": 0.975530571754945, + "flos": 19322987981760.0, + "grad_norm": 2.243575248182372, + "learning_rate": 6.248116654381297e-09, + "loss": 0.8021, + "num_input_tokens_seen": 175202080, + "step": 8113 + }, + { + "epoch": 0.9756508146455841, + "flos": 23549557433280.0, + "grad_norm": 2.308729242005356, + "learning_rate": 6.186742153410751e-09, + "loss": 0.7249, + "num_input_tokens_seen": 175221575, + "step": 8114 + }, + { + "epoch": 0.9757710575362232, + "flos": 22934114585280.0, + "grad_norm": 2.757985312258841, + "learning_rate": 6.125670112326453e-09, + "loss": 0.8754, + "num_input_tokens_seen": 175240835, + "step": 8115 + }, + { + "epoch": 0.9758913004268622, + "flos": 27925943952000.0, + "grad_norm": 2.318306000761177, + "learning_rate": 6.064900540392548e-09, + "loss": 0.7023, + "num_input_tokens_seen": 175262930, + "step": 8116 + }, + { + "epoch": 0.9760115433175014, + "flos": 22168445802720.0, + "grad_norm": 7.902598385896837, + "learning_rate": 6.0044334468278835e-09, + "loss": 0.7922, + "num_input_tokens_seen": 175282275, + "step": 8117 + }, + { + "epoch": 0.9761317862081405, + "flos": 26213152120320.0, + "grad_norm": 4.423152240379834, + "learning_rate": 5.944268840805345e-09, + "loss": 0.7186, + "num_input_tokens_seen": 175303020, + "step": 8118 + }, + { + "epoch": 0.9762520290987795, + "flos": 26537249690400.0, + "grad_norm": 2.6407890494587893, + "learning_rate": 5.88440673145163e-09, + "loss": 0.644, + "num_input_tokens_seen": 175324070, + "step": 8119 + }, + { + "epoch": 0.9763722719894187, + "flos": 17979460712160.0, + "grad_norm": 2.882399614741596, + "learning_rate": 5.824847127848142e-09, + "loss": 0.8281, + "num_input_tokens_seen": 175342595, + "step": 8120 + }, + { + "epoch": 0.9764925148800577, + "flos": 22423582000800.0, + "grad_norm": 2.1410870575467933, + "learning_rate": 5.765590039029433e-09, + "loss": 0.7879, + "num_input_tokens_seen": 175361915, + "step": 8121 + }, + { + "epoch": 0.9766127577706968, + "flos": 36704146838880.0, + "grad_norm": 2.411322362594109, + "learning_rate": 5.706635473985422e-09, + "loss": 0.7147, + "num_input_tokens_seen": 175385785, + "step": 8122 + }, + { + "epoch": 0.976733000661336, + "flos": 22278113795520.0, + "grad_norm": 2.602790794579757, + "learning_rate": 5.6479834416591764e-09, + "loss": 0.8575, + "num_input_tokens_seen": 175405145, + "step": 8123 + }, + { + "epoch": 0.976853243551975, + "flos": 25773922602720.0, + "grad_norm": 2.4675599139050175, + "learning_rate": 5.589633950947803e-09, + "loss": 0.6864, + "num_input_tokens_seen": 175422645, + "step": 8124 + }, + { + "epoch": 0.9769734864426141, + "flos": 21366902468160.0, + "grad_norm": 2.7623383556731684, + "learning_rate": 5.5315870107035535e-09, + "loss": 0.7044, + "num_input_tokens_seen": 175440695, + "step": 8125 + }, + { + "epoch": 0.9770937293332532, + "flos": 13971223662720.0, + "grad_norm": 2.070190195753578, + "learning_rate": 5.473842629731607e-09, + "loss": 0.7853, + "num_input_tokens_seen": 175459195, + "step": 8126 + }, + { + "epoch": 0.9772139722238923, + "flos": 17906522175840.0, + "grad_norm": 3.2923848459183525, + "learning_rate": 5.416400816792066e-09, + "loss": 0.782, + "num_input_tokens_seen": 175476710, + "step": 8127 + }, + { + "epoch": 0.9773342151145313, + "flos": 20416917347520.0, + "grad_norm": 8.946781913973926, + "learning_rate": 5.359261580598407e-09, + "loss": 0.7783, + "num_input_tokens_seen": 175492550, + "step": 8128 + }, + { + "epoch": 0.9774544580051704, + "flos": 11820651934080.0, + "grad_norm": 3.4496373303844043, + "learning_rate": 5.302424929819027e-09, + "loss": 0.7811, + "num_input_tokens_seen": 175506560, + "step": 8129 + }, + { + "epoch": 0.9775747008958096, + "flos": 13460914096800.0, + "grad_norm": 3.2362638845842073, + "learning_rate": 5.24589087307592e-09, + "loss": 0.7305, + "num_input_tokens_seen": 175524180, + "step": 8130 + }, + { + "epoch": 0.9776949437864486, + "flos": 59446511134080.0, + "grad_norm": 1.7681801460973183, + "learning_rate": 5.189659418944891e-09, + "loss": 0.6543, + "num_input_tokens_seen": 175554355, + "step": 8131 + }, + { + "epoch": 0.9778151866770877, + "flos": 21148384217280.0, + "grad_norm": 4.740677971943147, + "learning_rate": 5.133730575956674e-09, + "loss": 0.787, + "num_input_tokens_seen": 175574025, + "step": 8132 + }, + { + "epoch": 0.9779354295677268, + "flos": 20857001769600.0, + "grad_norm": 2.705699510506665, + "learning_rate": 5.0781043525953696e-09, + "loss": 0.7218, + "num_input_tokens_seen": 175592090, + "step": 8133 + }, + { + "epoch": 0.9780556724583659, + "flos": 23407099978560.0, + "grad_norm": 8.928214845348819, + "learning_rate": 5.0227807572995605e-09, + "loss": 0.7421, + "num_input_tokens_seen": 175614065, + "step": 8134 + }, + { + "epoch": 0.9781759153490049, + "flos": 20893396698240.0, + "grad_norm": 2.727069272052911, + "learning_rate": 4.967759798461646e-09, + "loss": 0.6726, + "num_input_tokens_seen": 175632410, + "step": 8135 + }, + { + "epoch": 0.9782961582396441, + "flos": 28252606235520.0, + "grad_norm": 5.18314666084815, + "learning_rate": 4.913041484428282e-09, + "loss": 0.7538, + "num_input_tokens_seen": 175652885, + "step": 8136 + }, + { + "epoch": 0.9784164011302832, + "flos": 25516221691200.0, + "grad_norm": 2.4378792883363833, + "learning_rate": 4.858625823500384e-09, + "loss": 0.7411, + "num_input_tokens_seen": 175670585, + "step": 8137 + }, + { + "epoch": 0.9785366440209222, + "flos": 29922721449120.0, + "grad_norm": 2.933296930202823, + "learning_rate": 4.80451282393246e-09, + "loss": 0.7342, + "num_input_tokens_seen": 175690570, + "step": 8138 + }, + { + "epoch": 0.9786568869115614, + "flos": 32297126704320.0, + "grad_norm": 3.693366217644706, + "learning_rate": 4.750702493933722e-09, + "loss": 0.6753, + "num_input_tokens_seen": 175710455, + "step": 8139 + }, + { + "epoch": 0.9787771298022004, + "flos": 23298063871680.0, + "grad_norm": 2.220534990130761, + "learning_rate": 4.697194841666974e-09, + "loss": 0.8529, + "num_input_tokens_seen": 175729380, + "step": 8140 + }, + { + "epoch": 0.9788973726928395, + "flos": 21437833837440.0, + "grad_norm": 2.4147946133361256, + "learning_rate": 4.6439898752492764e-09, + "loss": 0.8203, + "num_input_tokens_seen": 175749520, + "step": 8141 + }, + { + "epoch": 0.9790176155834787, + "flos": 68798924137920.0, + "grad_norm": 0.7601787160001946, + "learning_rate": 4.591087602751731e-09, + "loss": 0.6348, + "num_input_tokens_seen": 175811380, + "step": 8142 + }, + { + "epoch": 0.9791378584741177, + "flos": 21400175136960.0, + "grad_norm": 1.7919413629543244, + "learning_rate": 4.538488032199916e-09, + "loss": 0.7182, + "num_input_tokens_seen": 175829480, + "step": 8143 + }, + { + "epoch": 0.9792581013647568, + "flos": 20124159618720.0, + "grad_norm": 2.976434339432524, + "learning_rate": 4.486191171572784e-09, + "loss": 0.6946, + "num_input_tokens_seen": 175846750, + "step": 8144 + }, + { + "epoch": 0.9793783442553959, + "flos": 23694096394560.0, + "grad_norm": 1.83512620247658, + "learning_rate": 4.434197028803766e-09, + "loss": 0.7744, + "num_input_tokens_seen": 175865975, + "step": 8145 + }, + { + "epoch": 0.979498587146035, + "flos": 23005789349760.0, + "grad_norm": 2.8144539721220965, + "learning_rate": 4.3825056117805514e-09, + "loss": 0.8199, + "num_input_tokens_seen": 175881860, + "step": 8146 + }, + { + "epoch": 0.979618830036674, + "flos": 14298592171680.0, + "grad_norm": 11.06913842629535, + "learning_rate": 4.331116928344425e-09, + "loss": 0.7982, + "num_input_tokens_seen": 175898175, + "step": 8147 + }, + { + "epoch": 0.9797390729273132, + "flos": 16703779721760.0, + "grad_norm": 2.8433571009594734, + "learning_rate": 4.28003098629115e-09, + "loss": 0.6298, + "num_input_tokens_seen": 175914310, + "step": 8148 + }, + { + "epoch": 0.9798593158179523, + "flos": 24497461047360.0, + "grad_norm": 2.6697857814391055, + "learning_rate": 4.229247793370305e-09, + "loss": 0.7872, + "num_input_tokens_seen": 175933435, + "step": 8149 + }, + { + "epoch": 0.9799795587085913, + "flos": 27269682973920.0, + "grad_norm": 2.176761207903083, + "learning_rate": 4.178767357285951e-09, + "loss": 0.707, + "num_input_tokens_seen": 175955065, + "step": 8150 + }, + { + "epoch": 0.9800998015992305, + "flos": 26248915163040.0, + "grad_norm": 2.972754738738858, + "learning_rate": 4.128589685695516e-09, + "loss": 0.7156, + "num_input_tokens_seen": 175975490, + "step": 8151 + }, + { + "epoch": 0.9802200444898695, + "flos": 16699765387680.0, + "grad_norm": 3.4831814624831043, + "learning_rate": 4.078714786211135e-09, + "loss": 0.8458, + "num_input_tokens_seen": 175991340, + "step": 8152 + }, + { + "epoch": 0.9803402873805086, + "flos": 24864829951680.0, + "grad_norm": 2.403022682581354, + "learning_rate": 4.029142666398977e-09, + "loss": 0.7689, + "num_input_tokens_seen": 176011505, + "step": 8153 + }, + { + "epoch": 0.9804605302711478, + "flos": 22532692447200.0, + "grad_norm": 2.2552139481848283, + "learning_rate": 3.979873333778805e-09, + "loss": 0.7967, + "num_input_tokens_seen": 176029680, + "step": 8154 + }, + { + "epoch": 0.9805807731617868, + "flos": 38849143103520.0, + "grad_norm": 2.298437508298458, + "learning_rate": 3.930906795824862e-09, + "loss": 0.7466, + "num_input_tokens_seen": 176050355, + "step": 8155 + }, + { + "epoch": 0.9807010160524259, + "flos": 17796668334240.0, + "grad_norm": 2.4258856267172972, + "learning_rate": 3.882243059965207e-09, + "loss": 0.7696, + "num_input_tokens_seen": 176068070, + "step": 8156 + }, + { + "epoch": 0.980821258943065, + "flos": 13533852633120.0, + "grad_norm": 3.4020878271823474, + "learning_rate": 3.833882133582156e-09, + "loss": 0.6601, + "num_input_tokens_seen": 176083730, + "step": 8157 + }, + { + "epoch": 0.9809415018337041, + "flos": 21658842462240.0, + "grad_norm": 2.2693887951472833, + "learning_rate": 3.785824024012285e-09, + "loss": 0.7836, + "num_input_tokens_seen": 176102730, + "step": 8158 + }, + { + "epoch": 0.9810617447243432, + "flos": 23261557433760.0, + "grad_norm": 1.8137513149264595, + "learning_rate": 3.738068738545541e-09, + "loss": 0.7875, + "num_input_tokens_seen": 176121815, + "step": 8159 + }, + { + "epoch": 0.9811819876149822, + "flos": 18305937146880.0, + "grad_norm": 3.2184618478326863, + "learning_rate": 3.6906162844265733e-09, + "loss": 0.7907, + "num_input_tokens_seen": 176138900, + "step": 8160 + }, + { + "epoch": 0.9813022305056214, + "flos": 22569682092000.0, + "grad_norm": 2.7038573226756473, + "learning_rate": 3.643466668853845e-09, + "loss": 0.713, + "num_input_tokens_seen": 176156925, + "step": 8161 + }, + { + "epoch": 0.9814224733962604, + "flos": 25377035175360.0, + "grad_norm": 2.897915610654425, + "learning_rate": 3.59661989898008e-09, + "loss": 0.755, + "num_input_tokens_seen": 176177980, + "step": 8162 + }, + { + "epoch": 0.9815427162868995, + "flos": 24971301345120.0, + "grad_norm": 2.6456520038454907, + "learning_rate": 3.5500759819115934e-09, + "loss": 0.7666, + "num_input_tokens_seen": 176198345, + "step": 8163 + }, + { + "epoch": 0.9816629591775387, + "flos": 20632945224480.0, + "grad_norm": 2.2140541351095733, + "learning_rate": 3.5038349247094034e-09, + "loss": 0.8113, + "num_input_tokens_seen": 176215605, + "step": 8164 + }, + { + "epoch": 0.9817832020681777, + "flos": 17687037511200.0, + "grad_norm": 2.5629562343949277, + "learning_rate": 3.4578967343878994e-09, + "loss": 0.771, + "num_input_tokens_seen": 176231810, + "step": 8165 + }, + { + "epoch": 0.9819034449588168, + "flos": 22497895818240.0, + "grad_norm": 2.344386510248691, + "learning_rate": 3.4122614179161733e-09, + "loss": 0.8171, + "num_input_tokens_seen": 176251770, + "step": 8166 + }, + { + "epoch": 0.9820236878494559, + "flos": 19982705747520.0, + "grad_norm": 2.3048921138645895, + "learning_rate": 3.36692898221691e-09, + "loss": 0.7826, + "num_input_tokens_seen": 176270135, + "step": 8167 + }, + { + "epoch": 0.982143930740095, + "flos": 18780558009600.0, + "grad_norm": 2.0295452300052856, + "learning_rate": 3.3218994341668305e-09, + "loss": 0.7356, + "num_input_tokens_seen": 176289065, + "step": 8168 + }, + { + "epoch": 0.982264173630734, + "flos": 26539925913120.0, + "grad_norm": 2.2743933067832462, + "learning_rate": 3.2771727805971373e-09, + "loss": 0.7573, + "num_input_tokens_seen": 176310450, + "step": 8169 + }, + { + "epoch": 0.9823844165213732, + "flos": 21982902862560.0, + "grad_norm": 2.1164846907759722, + "learning_rate": 3.232749028292847e-09, + "loss": 0.7737, + "num_input_tokens_seen": 176327415, + "step": 8170 + }, + { + "epoch": 0.9825046594120123, + "flos": 21876877506240.0, + "grad_norm": 2.3452613535265563, + "learning_rate": 3.188628183992792e-09, + "loss": 0.8864, + "num_input_tokens_seen": 176347680, + "step": 8171 + }, + { + "epoch": 0.9826249023026513, + "flos": 59410147641600.0, + "grad_norm": 0.7614598152017503, + "learning_rate": 3.1448102543902844e-09, + "loss": 0.6248, + "num_input_tokens_seen": 176411595, + "step": 8172 + }, + { + "epoch": 0.9827451451932905, + "flos": 16048262138880.0, + "grad_norm": 2.80458642464003, + "learning_rate": 3.1012952461324515e-09, + "loss": 0.6758, + "num_input_tokens_seen": 176430570, + "step": 8173 + }, + { + "epoch": 0.9828653880839295, + "flos": 20233753272000.0, + "grad_norm": 2.7832791930394354, + "learning_rate": 3.0580831658204575e-09, + "loss": 0.7434, + "num_input_tokens_seen": 176448150, + "step": 8174 + }, + { + "epoch": 0.9829856309745686, + "flos": 21585272040000.0, + "grad_norm": 2.150492403287936, + "learning_rate": 3.015174020009281e-09, + "loss": 0.7786, + "num_input_tokens_seen": 176467475, + "step": 8175 + }, + { + "epoch": 0.9831058738652078, + "flos": 23730045286080.0, + "grad_norm": 2.403371104324111, + "learning_rate": 2.9725678152086043e-09, + "loss": 0.7521, + "num_input_tokens_seen": 176486045, + "step": 8176 + }, + { + "epoch": 0.9832261167558468, + "flos": 11304804073920.0, + "grad_norm": 3.4944558510512462, + "learning_rate": 2.930264557881257e-09, + "loss": 0.829, + "num_input_tokens_seen": 176501230, + "step": 8177 + }, + { + "epoch": 0.9833463596464859, + "flos": 59914918913280.0, + "grad_norm": 0.8697210195635883, + "learning_rate": 2.8882642544452163e-09, + "loss": 0.5812, + "num_input_tokens_seen": 176565955, + "step": 8178 + }, + { + "epoch": 0.983466602537125, + "flos": 13607497394880.0, + "grad_norm": 4.029465627904023, + "learning_rate": 2.8465669112716083e-09, + "loss": 0.744, + "num_input_tokens_seen": 176583480, + "step": 8179 + }, + { + "epoch": 0.9835868454277641, + "flos": 22893965510880.0, + "grad_norm": 2.779450423716521, + "learning_rate": 2.8051725346858177e-09, + "loss": 0.7658, + "num_input_tokens_seen": 176603740, + "step": 8180 + }, + { + "epoch": 0.9837070883184031, + "flos": 27633743769600.0, + "grad_norm": 2.3807656423817343, + "learning_rate": 2.7640811309674883e-09, + "loss": 0.7082, + "num_input_tokens_seen": 176623240, + "step": 8181 + }, + { + "epoch": 0.9838273312090423, + "flos": 29199989472960.0, + "grad_norm": 2.108916056120465, + "learning_rate": 2.7232927063498557e-09, + "loss": 0.8057, + "num_input_tokens_seen": 176643615, + "step": 8182 + }, + { + "epoch": 0.9839475740996814, + "flos": 40052257255200.0, + "grad_norm": 3.808257196668387, + "learning_rate": 2.682807267020859e-09, + "loss": 0.6915, + "num_input_tokens_seen": 176666375, + "step": 8183 + }, + { + "epoch": 0.9840678169903204, + "flos": 24134775532800.0, + "grad_norm": 1.9502517928883087, + "learning_rate": 2.642624819121808e-09, + "loss": 0.6245, + "num_input_tokens_seen": 176686525, + "step": 8184 + }, + { + "epoch": 0.9841880598809596, + "flos": 14663730890400.0, + "grad_norm": 2.3362565348999653, + "learning_rate": 2.6027453687487154e-09, + "loss": 0.6189, + "num_input_tokens_seen": 176703615, + "step": 8185 + }, + { + "epoch": 0.9843083027715986, + "flos": 22312798915200.0, + "grad_norm": 12.965376302522639, + "learning_rate": 2.5631689219509643e-09, + "loss": 0.5419, + "num_input_tokens_seen": 176722295, + "step": 8186 + }, + { + "epoch": 0.9844285456622377, + "flos": 21769030831680.0, + "grad_norm": 2.7707625262934057, + "learning_rate": 2.523895484732197e-09, + "loss": 0.8346, + "num_input_tokens_seen": 176741750, + "step": 8187 + }, + { + "epoch": 0.9845487885528769, + "flos": 18009536781600.0, + "grad_norm": 17.148174115289397, + "learning_rate": 2.4849250630505357e-09, + "loss": 0.7503, + "num_input_tokens_seen": 176759995, + "step": 8188 + }, + { + "epoch": 0.9846690314435159, + "flos": 25192495818720.0, + "grad_norm": 2.495901563048262, + "learning_rate": 2.4462576628172528e-09, + "loss": 0.7282, + "num_input_tokens_seen": 176778625, + "step": 8189 + }, + { + "epoch": 0.984789274334155, + "flos": 18525793509120.0, + "grad_norm": 3.7504991834909402, + "learning_rate": 2.407893289898766e-09, + "loss": 0.7458, + "num_input_tokens_seen": 176797655, + "step": 8190 + }, + { + "epoch": 0.984909517224794, + "flos": 27306263751360.0, + "grad_norm": 2.470580569156539, + "learning_rate": 2.3698319501144202e-09, + "loss": 0.8408, + "num_input_tokens_seen": 176818640, + "step": 8191 + }, + { + "epoch": 0.9850297601154332, + "flos": 18707396454720.0, + "grad_norm": 5.493743459027143, + "learning_rate": 2.3320736492382644e-09, + "loss": 0.7337, + "num_input_tokens_seen": 176837475, + "step": 8192 + }, + { + "epoch": 0.9851500030060723, + "flos": 22276329647040.0, + "grad_norm": 2.24499431736828, + "learning_rate": 2.29461839299816e-09, + "loss": 0.6806, + "num_input_tokens_seen": 176857190, + "step": 8193 + }, + { + "epoch": 0.9852702458967113, + "flos": 26315720688960.0, + "grad_norm": 1.904839603650175, + "learning_rate": 2.257466187076229e-09, + "loss": 0.7985, + "num_input_tokens_seen": 176876395, + "step": 8194 + }, + { + "epoch": 0.9853904887873505, + "flos": 20853693660960.0, + "grad_norm": 2.3120795202736537, + "learning_rate": 2.2206170371081854e-09, + "loss": 0.7138, + "num_input_tokens_seen": 176894980, + "step": 8195 + }, + { + "epoch": 0.9855107316779895, + "flos": 25227218108160.0, + "grad_norm": 1.9533093174748046, + "learning_rate": 2.1840709486842247e-09, + "loss": 0.8477, + "num_input_tokens_seen": 176914600, + "step": 8196 + }, + { + "epoch": 0.9856309745686286, + "flos": 19034950812480.0, + "grad_norm": 2.465301906366129, + "learning_rate": 2.1478279273481335e-09, + "loss": 0.7915, + "num_input_tokens_seen": 176933085, + "step": 8197 + }, + { + "epoch": 0.9857512174592677, + "flos": 34081667640000.0, + "grad_norm": 2.7956756918780323, + "learning_rate": 2.1118879785981815e-09, + "loss": 0.8096, + "num_input_tokens_seen": 176953225, + "step": 8198 + }, + { + "epoch": 0.9858714603499068, + "flos": 25957458375840.0, + "grad_norm": 2.7250636662036505, + "learning_rate": 2.0762511078862288e-09, + "loss": 0.7953, + "num_input_tokens_seen": 176973570, + "step": 8199 + }, + { + "epoch": 0.9859917032405459, + "flos": 23662347685920.0, + "grad_norm": 2.210575141463119, + "learning_rate": 2.0409173206186183e-09, + "loss": 0.649, + "num_input_tokens_seen": 176992810, + "step": 8200 + }, + { + "epoch": 0.986111946131185, + "flos": 19909878720480.0, + "grad_norm": 3.264808370157069, + "learning_rate": 2.0058866221550617e-09, + "loss": 0.8743, + "num_input_tokens_seen": 177011050, + "step": 8201 + }, + { + "epoch": 0.9862321890218241, + "flos": 19800582425280.0, + "grad_norm": 2.559298492517766, + "learning_rate": 1.971159017809976e-09, + "loss": 0.7537, + "num_input_tokens_seen": 177029340, + "step": 8202 + }, + { + "epoch": 0.9863524319124631, + "flos": 21621109422240.0, + "grad_norm": 3.1188839381016, + "learning_rate": 1.93673451285159e-09, + "loss": 0.7767, + "num_input_tokens_seen": 177048390, + "step": 8203 + }, + { + "epoch": 0.9864726748031023, + "flos": 52694895891840.0, + "grad_norm": 0.7428934053069522, + "learning_rate": 1.9026131125019495e-09, + "loss": 0.5633, + "num_input_tokens_seen": 177105760, + "step": 8204 + }, + { + "epoch": 0.9865929176937414, + "flos": 23331113521920.0, + "grad_norm": 2.1428463958289083, + "learning_rate": 1.8687948219371363e-09, + "loss": 0.8716, + "num_input_tokens_seen": 177124985, + "step": 8205 + }, + { + "epoch": 0.9867131605843804, + "flos": 21585829586400.0, + "grad_norm": 3.1014341236518392, + "learning_rate": 1.835279646287491e-09, + "loss": 0.8859, + "num_input_tokens_seen": 177142995, + "step": 8206 + }, + { + "epoch": 0.9868334034750196, + "flos": 22239525851040.0, + "grad_norm": 3.0201473816536137, + "learning_rate": 1.8020675906371685e-09, + "loss": 0.7684, + "num_input_tokens_seen": 177162390, + "step": 8207 + }, + { + "epoch": 0.9869536463656586, + "flos": 25772510151840.0, + "grad_norm": 2.4683596145429343, + "learning_rate": 1.7691586600243612e-09, + "loss": 0.7527, + "num_input_tokens_seen": 177181290, + "step": 8208 + }, + { + "epoch": 0.9870738892562977, + "flos": 16375221780480.0, + "grad_norm": 3.824128973026897, + "learning_rate": 1.7365528594415202e-09, + "loss": 0.8678, + "num_input_tokens_seen": 177195360, + "step": 8209 + }, + { + "epoch": 0.9871941321469369, + "flos": 35430324336480.0, + "grad_norm": 2.2608143212034233, + "learning_rate": 1.7042501938346888e-09, + "loss": 0.6721, + "num_input_tokens_seen": 177218090, + "step": 8210 + }, + { + "epoch": 0.9873143750375759, + "flos": 21403631924640.0, + "grad_norm": 5.351969682559144, + "learning_rate": 1.6722506681043913e-09, + "loss": 0.766, + "num_input_tokens_seen": 177237050, + "step": 8211 + }, + { + "epoch": 0.987434617928215, + "flos": 16303361167200.0, + "grad_norm": 2.8291533060184415, + "learning_rate": 1.640554287104745e-09, + "loss": 0.6917, + "num_input_tokens_seen": 177255325, + "step": 8212 + }, + { + "epoch": 0.9875548608188541, + "flos": 17826372706080.0, + "grad_norm": 2.914992278728434, + "learning_rate": 1.609161055644348e-09, + "loss": 0.8022, + "num_input_tokens_seen": 177271250, + "step": 8213 + }, + { + "epoch": 0.9876751037094932, + "flos": 26095158101280.0, + "grad_norm": 3.0183509009987652, + "learning_rate": 1.5780709784849467e-09, + "loss": 0.683, + "num_input_tokens_seen": 177288270, + "step": 8214 + }, + { + "epoch": 0.9877953466001322, + "flos": 15968447196960.0, + "grad_norm": 2.5322084849190816, + "learning_rate": 1.5472840603436565e-09, + "loss": 0.8257, + "num_input_tokens_seen": 177305500, + "step": 8215 + }, + { + "epoch": 0.9879155894907714, + "flos": 18780000463200.0, + "grad_norm": 2.2857991675385807, + "learning_rate": 1.5168003058900757e-09, + "loss": 0.7802, + "num_input_tokens_seen": 177323500, + "step": 8216 + }, + { + "epoch": 0.9880358323814105, + "flos": 22350048748320.0, + "grad_norm": 2.760907648915009, + "learning_rate": 1.4866197197491715e-09, + "loss": 0.9231, + "num_input_tokens_seen": 177342860, + "step": 8217 + }, + { + "epoch": 0.9881560752720495, + "flos": 15646616982240.0, + "grad_norm": 4.5959029569681045, + "learning_rate": 1.4567423064988371e-09, + "loss": 0.7883, + "num_input_tokens_seen": 177360210, + "step": 8218 + }, + { + "epoch": 0.9882763181626887, + "flos": 21469173678720.0, + "grad_norm": 2.5896752764419233, + "learning_rate": 1.4271680706718913e-09, + "loss": 0.7797, + "num_input_tokens_seen": 177377885, + "step": 8219 + }, + { + "epoch": 0.9883965610533277, + "flos": 27994273438080.0, + "grad_norm": 2.350621538958427, + "learning_rate": 1.3978970167543013e-09, + "loss": 0.8262, + "num_input_tokens_seen": 177398065, + "step": 8220 + }, + { + "epoch": 0.9885168039439668, + "flos": 14078624300160.0, + "grad_norm": 4.0914225477644655, + "learning_rate": 1.3689291491867372e-09, + "loss": 0.7754, + "num_input_tokens_seen": 177416380, + "step": 8221 + }, + { + "epoch": 0.988637046834606, + "flos": 26394123180000.0, + "grad_norm": 2.5133199615864834, + "learning_rate": 1.3402644723636836e-09, + "loss": 0.7344, + "num_input_tokens_seen": 177438410, + "step": 8222 + }, + { + "epoch": 0.988757289725245, + "flos": 25193016195360.0, + "grad_norm": 2.846623237272623, + "learning_rate": 1.311902990633218e-09, + "loss": 0.8397, + "num_input_tokens_seen": 177457375, + "step": 8223 + }, + { + "epoch": 0.9888775326158841, + "flos": 26321258983200.0, + "grad_norm": 2.195353225927567, + "learning_rate": 1.2838447082978987e-09, + "loss": 0.7143, + "num_input_tokens_seen": 177478530, + "step": 8224 + }, + { + "epoch": 0.9889977755065231, + "flos": 24281321661120.0, + "grad_norm": 4.7595039455406685, + "learning_rate": 1.2560896296143208e-09, + "loss": 0.8346, + "num_input_tokens_seen": 177496995, + "step": 8225 + }, + { + "epoch": 0.9891180183971623, + "flos": 18924019047840.0, + "grad_norm": 2.6018652754085774, + "learning_rate": 1.2286377587926722e-09, + "loss": 0.8211, + "num_input_tokens_seen": 177513785, + "step": 8226 + }, + { + "epoch": 0.9892382612878013, + "flos": 26139061321440.0, + "grad_norm": 4.497847784144218, + "learning_rate": 1.2014890999973992e-09, + "loss": 0.7539, + "num_input_tokens_seen": 177530705, + "step": 8227 + }, + { + "epoch": 0.9893585041784404, + "flos": 25413244255200.0, + "grad_norm": 7.502604477385956, + "learning_rate": 1.1746436573472073e-09, + "loss": 0.7857, + "num_input_tokens_seen": 177552440, + "step": 8228 + }, + { + "epoch": 0.9894787470690796, + "flos": 20160368698560.0, + "grad_norm": 3.1099199747172768, + "learning_rate": 1.1481014349141726e-09, + "loss": 0.691, + "num_input_tokens_seen": 177569660, + "step": 8229 + }, + { + "epoch": 0.9895989899597186, + "flos": 24609024697920.0, + "grad_norm": 3.2887509304902207, + "learning_rate": 1.121862436724852e-09, + "loss": 0.8461, + "num_input_tokens_seen": 177588500, + "step": 8230 + }, + { + "epoch": 0.9897192328503577, + "flos": 21767915738880.0, + "grad_norm": 2.412868364900558, + "learning_rate": 1.0959266667598388e-09, + "loss": 0.7035, + "num_input_tokens_seen": 177607315, + "step": 8231 + }, + { + "epoch": 0.9898394757409968, + "flos": 21294818836320.0, + "grad_norm": 2.4215282561130294, + "learning_rate": 1.0702941289533196e-09, + "loss": 0.748, + "num_input_tokens_seen": 177625990, + "step": 8232 + }, + { + "epoch": 0.9899597186316359, + "flos": 18518396726880.0, + "grad_norm": 2.749148670548896, + "learning_rate": 1.0449648271939615e-09, + "loss": 0.8888, + "num_input_tokens_seen": 177642335, + "step": 8233 + }, + { + "epoch": 0.990079961522275, + "flos": 23732647169280.0, + "grad_norm": 5.022214749611066, + "learning_rate": 1.0199387653240243e-09, + "loss": 0.7297, + "num_input_tokens_seen": 177662025, + "step": 8234 + }, + { + "epoch": 0.9902002044129141, + "flos": 16375928005920.0, + "grad_norm": 4.410110838312476, + "learning_rate": 9.952159471400267e-10, + "loss": 0.7102, + "num_input_tokens_seen": 177679065, + "step": 8235 + }, + { + "epoch": 0.9903204473035532, + "flos": 22527228492480.0, + "grad_norm": 2.4538637251692808, + "learning_rate": 9.707963763923022e-10, + "loss": 0.8451, + "num_input_tokens_seen": 177698115, + "step": 8236 + }, + { + "epoch": 0.9904406901941922, + "flos": 16121126335680.0, + "grad_norm": 2.9683161368310125, + "learning_rate": 9.466800567854427e-10, + "loss": 0.7925, + "num_input_tokens_seen": 177716715, + "step": 8237 + }, + { + "epoch": 0.9905609330848314, + "flos": 26613310486560.0, + "grad_norm": 9.178211733103337, + "learning_rate": 9.228669919778553e-10, + "loss": 0.6825, + "num_input_tokens_seen": 177735640, + "step": 8238 + }, + { + "epoch": 0.9906811759754705, + "flos": 23078765055840.0, + "grad_norm": 2.52234656783162, + "learning_rate": 8.993571855817617e-10, + "loss": 0.7919, + "num_input_tokens_seen": 177754470, + "step": 8239 + }, + { + "epoch": 0.9908014188661095, + "flos": 22059037998240.0, + "grad_norm": 6.956433004352372, + "learning_rate": 8.761506411638642e-10, + "loss": 0.7529, + "num_input_tokens_seen": 177773935, + "step": 8240 + }, + { + "epoch": 0.9909216617567487, + "flos": 19215475835040.0, + "grad_norm": 2.286727182720081, + "learning_rate": 8.53247362244236e-10, + "loss": 0.7383, + "num_input_tokens_seen": 177792745, + "step": 8241 + }, + { + "epoch": 0.9910419046473877, + "flos": 23587364812800.0, + "grad_norm": 15.17657504761141, + "learning_rate": 8.306473522976532e-10, + "loss": 0.6835, + "num_input_tokens_seen": 177812150, + "step": 8242 + }, + { + "epoch": 0.9911621475380268, + "flos": 22678941217440.0, + "grad_norm": 2.381303953826099, + "learning_rate": 8.083506147522623e-10, + "loss": 0.7201, + "num_input_tokens_seen": 177831545, + "step": 8243 + }, + { + "epoch": 0.991282390428666, + "flos": 13497903741600.0, + "grad_norm": 2.41032520363545, + "learning_rate": 7.863571529906909e-10, + "loss": 0.857, + "num_input_tokens_seen": 177847130, + "step": 8244 + }, + { + "epoch": 0.991402633319305, + "flos": 61742656843680.0, + "grad_norm": 0.7310310688122124, + "learning_rate": 7.646669703489372e-10, + "loss": 0.5954, + "num_input_tokens_seen": 177911440, + "step": 8245 + }, + { + "epoch": 0.9915228762099441, + "flos": 18743419685760.0, + "grad_norm": 3.045939923120413, + "learning_rate": 7.432800701177023e-10, + "loss": 0.5752, + "num_input_tokens_seen": 177929440, + "step": 8246 + }, + { + "epoch": 0.9916431191005832, + "flos": 65842237194240.0, + "grad_norm": 0.8220701007838214, + "learning_rate": 7.221964555415017e-10, + "loss": 0.5768, + "num_input_tokens_seen": 177989100, + "step": 8247 + }, + { + "epoch": 0.9917633619912223, + "flos": 16558051328160.0, + "grad_norm": 4.379212231796737, + "learning_rate": 7.01416129818222e-10, + "loss": 0.7481, + "num_input_tokens_seen": 178006720, + "step": 8248 + }, + { + "epoch": 0.9918836048818613, + "flos": 25374730650240.0, + "grad_norm": 2.2763408985035505, + "learning_rate": 6.809390961006745e-10, + "loss": 0.588, + "num_input_tokens_seen": 178026850, + "step": 8249 + }, + { + "epoch": 0.9920038477725005, + "flos": 25011041552160.0, + "grad_norm": 2.2636540752036916, + "learning_rate": 6.607653574948191e-10, + "loss": 0.6823, + "num_input_tokens_seen": 178047630, + "step": 8250 + }, + { + "epoch": 0.9921240906631396, + "flos": 21798400675680.0, + "grad_norm": 3.470477249789803, + "learning_rate": 6.408949170613187e-10, + "loss": 0.8154, + "num_input_tokens_seen": 178066895, + "step": 8251 + }, + { + "epoch": 0.9922443335537786, + "flos": 24828658041600.0, + "grad_norm": 7.893996321096208, + "learning_rate": 6.213277778144288e-10, + "loss": 0.8181, + "num_input_tokens_seen": 178088540, + "step": 8252 + }, + { + "epoch": 0.9923645764444178, + "flos": 21582112610400.0, + "grad_norm": 2.956319867008118, + "learning_rate": 6.020639427224416e-10, + "loss": 0.6687, + "num_input_tokens_seen": 178106185, + "step": 8253 + }, + { + "epoch": 0.9924848193350568, + "flos": 24965725881120.0, + "grad_norm": 3.2888102310099474, + "learning_rate": 5.831034147076864e-10, + "loss": 0.7323, + "num_input_tokens_seen": 178123435, + "step": 8254 + }, + { + "epoch": 0.9926050622256959, + "flos": 68814535437120.0, + "grad_norm": 0.6972141193418894, + "learning_rate": 5.644461966463065e-10, + "loss": 0.5559, + "num_input_tokens_seen": 178191045, + "step": 8255 + }, + { + "epoch": 0.9927253051163349, + "flos": 20893842735360.0, + "grad_norm": 2.6520669649452584, + "learning_rate": 5.460922913687049e-10, + "loss": 0.7556, + "num_input_tokens_seen": 178210135, + "step": 8256 + }, + { + "epoch": 0.9928455480069741, + "flos": 22276292477280.0, + "grad_norm": 2.494841498804382, + "learning_rate": 5.280417016593208e-10, + "loss": 0.7552, + "num_input_tokens_seen": 178229925, + "step": 8257 + }, + { + "epoch": 0.9929657908976132, + "flos": 17359185795360.0, + "grad_norm": 2.385603074254614, + "learning_rate": 5.102944302559642e-10, + "loss": 0.7478, + "num_input_tokens_seen": 178247420, + "step": 8258 + }, + { + "epoch": 0.9930860337882522, + "flos": 22642286100480.0, + "grad_norm": 2.659878574840955, + "learning_rate": 4.9285047985137e-10, + "loss": 0.799, + "num_input_tokens_seen": 178266840, + "step": 8259 + }, + { + "epoch": 0.9932062766788914, + "flos": 28107435388320.0, + "grad_norm": 3.313702634906743, + "learning_rate": 4.757098530916436e-10, + "loss": 0.7435, + "num_input_tokens_seen": 178287285, + "step": 8260 + }, + { + "epoch": 0.9933265195695304, + "flos": 20127913764480.0, + "grad_norm": 3.7004757188717603, + "learning_rate": 4.5887255257670563e-10, + "loss": 0.7794, + "num_input_tokens_seen": 178304325, + "step": 8261 + }, + { + "epoch": 0.9934467624601695, + "flos": 21331585462560.0, + "grad_norm": 9.392092978801953, + "learning_rate": 4.4233858086117906e-10, + "loss": 0.7709, + "num_input_tokens_seen": 178322850, + "step": 8262 + }, + { + "epoch": 0.9935670053508087, + "flos": 19727904077280.0, + "grad_norm": 3.2037434794954587, + "learning_rate": 4.261079404528356e-10, + "loss": 0.6762, + "num_input_tokens_seen": 178342330, + "step": 8263 + }, + { + "epoch": 0.9936872482414477, + "flos": 21947362838400.0, + "grad_norm": 2.492677837673613, + "learning_rate": 4.1018063381437205e-10, + "loss": 0.6905, + "num_input_tokens_seen": 178362715, + "step": 8264 + }, + { + "epoch": 0.9938074911320868, + "flos": 69710841690720.0, + "grad_norm": 0.8717189307314351, + "learning_rate": 3.9455666336141167e-10, + "loss": 0.6091, + "num_input_tokens_seen": 178426495, + "step": 8265 + }, + { + "epoch": 0.9939277340227259, + "flos": 15062774163840.0, + "grad_norm": 3.9134171309805765, + "learning_rate": 3.7923603146450267e-10, + "loss": 0.8331, + "num_input_tokens_seen": 178442555, + "step": 8266 + }, + { + "epoch": 0.994047976913365, + "flos": 17687186190240.0, + "grad_norm": 2.618108149055605, + "learning_rate": 3.642187404473418e-10, + "loss": 0.8102, + "num_input_tokens_seen": 178460715, + "step": 8267 + }, + { + "epoch": 0.994168219804004, + "flos": 19144247107680.0, + "grad_norm": 2.7035617365306126, + "learning_rate": 3.495047925885508e-10, + "loss": 0.8595, + "num_input_tokens_seen": 178479050, + "step": 8268 + }, + { + "epoch": 0.9942884626946432, + "flos": 17825963838720.0, + "grad_norm": 2.4291593216729943, + "learning_rate": 3.350941901199e-10, + "loss": 0.8312, + "num_input_tokens_seen": 178495970, + "step": 8269 + }, + { + "epoch": 0.9944087055852823, + "flos": 18769444251360.0, + "grad_norm": 3.0105545029989647, + "learning_rate": 3.2098693522764066e-10, + "loss": 0.8361, + "num_input_tokens_seen": 178509170, + "step": 8270 + }, + { + "epoch": 0.9945289484759213, + "flos": 20967041460000.0, + "grad_norm": 2.7818686836563495, + "learning_rate": 3.071830300516165e-10, + "loss": 0.8115, + "num_input_tokens_seen": 178527000, + "step": 8271 + }, + { + "epoch": 0.9946491913665605, + "flos": 14736074710560.0, + "grad_norm": 3.1023819847315726, + "learning_rate": 2.9368247668615234e-10, + "loss": 0.7121, + "num_input_tokens_seen": 178545590, + "step": 8272 + }, + { + "epoch": 0.9947694342571995, + "flos": 12651825300960.0, + "grad_norm": 3.8301860632658355, + "learning_rate": 2.804852771789434e-10, + "loss": 0.6154, + "num_input_tokens_seen": 178558890, + "step": 8273 + }, + { + "epoch": 0.9948896771478386, + "flos": 18816618410400.0, + "grad_norm": 2.858117165653812, + "learning_rate": 2.675914335321661e-10, + "loss": 0.5597, + "num_input_tokens_seen": 178577645, + "step": 8274 + }, + { + "epoch": 0.9950099200384778, + "flos": 24864569763360.0, + "grad_norm": 3.370705309221827, + "learning_rate": 2.550009477018111e-10, + "loss": 0.7902, + "num_input_tokens_seen": 178596415, + "step": 8275 + }, + { + "epoch": 0.9951301629291168, + "flos": 23699709028320.0, + "grad_norm": 3.0381757133754195, + "learning_rate": 2.4271382159790634e-10, + "loss": 0.6305, + "num_input_tokens_seen": 178613845, + "step": 8276 + }, + { + "epoch": 0.9952504058197559, + "flos": 22205138089440.0, + "grad_norm": 15.130093308064634, + "learning_rate": 2.3073005708429406e-10, + "loss": 0.8614, + "num_input_tokens_seen": 178633490, + "step": 8277 + }, + { + "epoch": 0.995370648710395, + "flos": 21180616132800.0, + "grad_norm": 2.3164129401845313, + "learning_rate": 2.190496559788535e-10, + "loss": 0.7211, + "num_input_tokens_seen": 178651775, + "step": 8278 + }, + { + "epoch": 0.9954908916010341, + "flos": 14845705533600.0, + "grad_norm": 3.3842367297217626, + "learning_rate": 2.0767262005372265e-10, + "loss": 0.7595, + "num_input_tokens_seen": 178669290, + "step": 8279 + }, + { + "epoch": 0.9956111344916732, + "flos": 19180902224640.0, + "grad_norm": 2.1297885571332436, + "learning_rate": 1.965989510346322e-10, + "loss": 0.7561, + "num_input_tokens_seen": 178688410, + "step": 8280 + }, + { + "epoch": 0.9957313773823123, + "flos": 20018468790240.0, + "grad_norm": 2.3692738620562315, + "learning_rate": 1.8582865060134955e-10, + "loss": 0.7131, + "num_input_tokens_seen": 178706600, + "step": 8281 + }, + { + "epoch": 0.9958516202729514, + "flos": 57401550160800.0, + "grad_norm": 0.7869901353806749, + "learning_rate": 1.7536172038790098e-10, + "loss": 0.5562, + "num_input_tokens_seen": 178766910, + "step": 8282 + }, + { + "epoch": 0.9959718631635904, + "flos": 27743077234560.0, + "grad_norm": 4.4929313242877, + "learning_rate": 1.651981619819054e-10, + "loss": 0.6936, + "num_input_tokens_seen": 178784060, + "step": 8283 + }, + { + "epoch": 0.9960921060542296, + "flos": 23989716194880.0, + "grad_norm": 2.9932137664103147, + "learning_rate": 1.5533797692546257e-10, + "loss": 0.7092, + "num_input_tokens_seen": 178802795, + "step": 8284 + }, + { + "epoch": 0.9962123489448687, + "flos": 18670778507520.0, + "grad_norm": 2.875412621120527, + "learning_rate": 1.4578116671404296e-10, + "loss": 0.8396, + "num_input_tokens_seen": 178821075, + "step": 8285 + }, + { + "epoch": 0.9963325918355077, + "flos": 19982222540640.0, + "grad_norm": 3.46005534001013, + "learning_rate": 1.3652773279759777e-10, + "loss": 0.7141, + "num_input_tokens_seen": 178837725, + "step": 8286 + }, + { + "epoch": 0.9964528347261468, + "flos": 33060862659360.0, + "grad_norm": 2.0845533486262813, + "learning_rate": 1.2757767657989305e-10, + "loss": 0.6279, + "num_input_tokens_seen": 178861515, + "step": 8287 + }, + { + "epoch": 0.9965730776167859, + "flos": 23076200342400.0, + "grad_norm": 2.761211429601138, + "learning_rate": 1.1893099941850948e-10, + "loss": 0.8715, + "num_input_tokens_seen": 178880410, + "step": 8288 + }, + { + "epoch": 0.996693320507425, + "flos": 22932999492480.0, + "grad_norm": 2.864110142321456, + "learning_rate": 1.105877026252866e-10, + "loss": 0.7766, + "num_input_tokens_seen": 178898740, + "step": 8289 + }, + { + "epoch": 0.996813563398064, + "flos": 13204030920000.0, + "grad_norm": 3.0511977611103456, + "learning_rate": 1.0254778746565663e-10, + "loss": 0.7261, + "num_input_tokens_seen": 178914015, + "step": 8290 + }, + { + "epoch": 0.9969338062887032, + "flos": 14627150112960.0, + "grad_norm": 3.120054375101317, + "learning_rate": 9.481125515953259e-11, + "loss": 0.7355, + "num_input_tokens_seen": 178932075, + "step": 8291 + }, + { + "epoch": 0.9970540491793423, + "flos": 25698753880800.0, + "grad_norm": 2.423525733743444, + "learning_rate": 8.737810688064228e-11, + "loss": 0.7995, + "num_input_tokens_seen": 178951910, + "step": 8292 + }, + { + "epoch": 0.9971742920699813, + "flos": 21439580816160.0, + "grad_norm": 2.507876671656331, + "learning_rate": 8.024834375608414e-11, + "loss": 0.7946, + "num_input_tokens_seen": 178970500, + "step": 8293 + }, + { + "epoch": 0.9972945349606205, + "flos": 72107628875040.0, + "grad_norm": 0.8358715066975131, + "learning_rate": 7.342196686788149e-11, + "loss": 0.627, + "num_input_tokens_seen": 179023665, + "step": 8294 + }, + { + "epoch": 0.9974147778512595, + "flos": 19650096302400.0, + "grad_norm": 4.1408458722782555, + "learning_rate": 6.689897725142834e-11, + "loss": 0.6899, + "num_input_tokens_seen": 179043140, + "step": 8295 + }, + { + "epoch": 0.9975350207418986, + "flos": 15938185278720.0, + "grad_norm": 4.341069334287106, + "learning_rate": 6.067937589615545e-11, + "loss": 0.8868, + "num_input_tokens_seen": 179061215, + "step": 8296 + }, + { + "epoch": 0.9976552636325378, + "flos": 59876479647840.0, + "grad_norm": 0.7528865420872142, + "learning_rate": 5.476316374575241e-11, + "loss": 0.5756, + "num_input_tokens_seen": 179124700, + "step": 8297 + }, + { + "epoch": 0.9977755065231768, + "flos": 22455590897760.0, + "grad_norm": 2.8262567408539105, + "learning_rate": 4.9150341697723476e-11, + "loss": 0.7245, + "num_input_tokens_seen": 179144590, + "step": 8298 + }, + { + "epoch": 0.9978957494138159, + "flos": 25992961230240.0, + "grad_norm": 1.9253273326449167, + "learning_rate": 4.384091060338768e-11, + "loss": 0.6672, + "num_input_tokens_seen": 179165060, + "step": 8299 + }, + { + "epoch": 0.998015992304455, + "flos": 22605482304480.0, + "grad_norm": 3.2310141088094437, + "learning_rate": 3.883487126810081e-11, + "loss": 0.7377, + "num_input_tokens_seen": 179184320, + "step": 8300 + }, + { + "epoch": 0.9981362351950941, + "flos": 18197532925920.0, + "grad_norm": 1.9036219882819776, + "learning_rate": 3.41322244516995e-11, + "loss": 0.7934, + "num_input_tokens_seen": 179202265, + "step": 8301 + }, + { + "epoch": 0.9982564780857331, + "flos": 33425964208320.0, + "grad_norm": 1.9124228168944788, + "learning_rate": 2.9732970866946925e-11, + "loss": 0.6317, + "num_input_tokens_seen": 179222145, + "step": 8302 + }, + { + "epoch": 0.9983767209763723, + "flos": 15501483304800.0, + "grad_norm": 2.7336658539138416, + "learning_rate": 2.563711118175327e-11, + "loss": 0.7852, + "num_input_tokens_seen": 179239030, + "step": 8303 + }, + { + "epoch": 0.9984969638670114, + "flos": 19945901951520.0, + "grad_norm": 2.6938054833595046, + "learning_rate": 2.184464601717728e-11, + "loss": 0.8386, + "num_input_tokens_seen": 179257345, + "step": 8304 + }, + { + "epoch": 0.9986172067576504, + "flos": 20346692203680.0, + "grad_norm": 3.116550466509845, + "learning_rate": 1.8355575948758585e-11, + "loss": 0.7763, + "num_input_tokens_seen": 179275330, + "step": 8305 + }, + { + "epoch": 0.9987374496482896, + "flos": 23693799036480.0, + "grad_norm": 4.042638026177134, + "learning_rate": 1.5169901505407424e-11, + "loss": 0.7338, + "num_input_tokens_seen": 179292395, + "step": 8306 + }, + { + "epoch": 0.9988576925389286, + "flos": 25008105141120.0, + "grad_norm": 2.375578057061921, + "learning_rate": 1.228762317073695e-11, + "loss": 0.7443, + "num_input_tokens_seen": 179311225, + "step": 8307 + }, + { + "epoch": 0.9989779354295677, + "flos": 31240707360000.0, + "grad_norm": 2.558091348724511, + "learning_rate": 9.70874138195299e-12, + "loss": 0.7903, + "num_input_tokens_seen": 179333135, + "step": 8308 + }, + { + "epoch": 0.9990981783202069, + "flos": 19545520566720.0, + "grad_norm": 2.40990497385939, + "learning_rate": 7.433256530076093e-12, + "loss": 0.7481, + "num_input_tokens_seen": 179353090, + "step": 8309 + }, + { + "epoch": 0.9992184212108459, + "flos": 17175166815360.0, + "grad_norm": 2.6529551097811965, + "learning_rate": 5.46116896038562e-12, + "loss": 0.758, + "num_input_tokens_seen": 179367500, + "step": 8310 + }, + { + "epoch": 0.999338664101485, + "flos": 46430699376960.0, + "grad_norm": 2.9570767236637248, + "learning_rate": 3.792478972197699e-12, + "loss": 0.6218, + "num_input_tokens_seen": 179388085, + "step": 8311 + }, + { + "epoch": 0.9994589069921241, + "flos": 15136790623200.0, + "grad_norm": 2.8784818711762683, + "learning_rate": 2.4271868181990895e-12, + "loss": 0.7055, + "num_input_tokens_seen": 179405250, + "step": 8312 + }, + { + "epoch": 0.9995791498827632, + "flos": 12513828217440.0, + "grad_norm": 2.993000925917248, + "learning_rate": 1.3652927060014973e-12, + "loss": 0.8121, + "num_input_tokens_seen": 179420845, + "step": 8313 + }, + { + "epoch": 0.9996993927734023, + "flos": 19217780360160.0, + "grad_norm": 5.894311211930189, + "learning_rate": 6.067967965872612e-13, + "loss": 0.6358, + "num_input_tokens_seen": 179440455, + "step": 8314 + }, + { + "epoch": 0.9998196356640414, + "flos": 62861687264640.0, + "grad_norm": 1.882088169679754, + "learning_rate": 1.5169920497548615e-13, + "loss": 0.7712, + "num_input_tokens_seen": 179465075, + "step": 8315 + }, + { + "epoch": 0.9999398785546805, + "flos": 50850463005600.0, + "grad_norm": 1.189765074966066, + "learning_rate": 0.0, + "loss": 0.5484, + "num_input_tokens_seen": 179513955, + "step": 8316 + }, + { + "epoch": 0.9999398785546805, + "num_input_tokens_seen": 179513955, + "step": 8316, + "total_flos": 6.986777266843812e+17, + "train_loss": 0.7679832751623938, + "train_runtime": 91171.0892, + "train_samples_per_second": 3.649, + "train_steps_per_second": 0.091 + } + ], + "logging_steps": 1.0, + "max_steps": 8316, + "num_input_tokens_seen": 179513955, + "num_train_epochs": 1, + "save_steps": 832, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.986777266843812e+17, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +}