| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.3856812933025404, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004618937644341801, | |
| "grad_norm": 10.975909233093262, | |
| "learning_rate": 3.4642032332563515e-07, | |
| "loss": 2.4942, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009237875288683603, | |
| "grad_norm": 9.186948776245117, | |
| "learning_rate": 9.237875288683603e-07, | |
| "loss": 2.7582, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013856812933025405, | |
| "grad_norm": 10.907584190368652, | |
| "learning_rate": 1.443418013856813e-06, | |
| "loss": 2.6814, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.018475750577367205, | |
| "grad_norm": 9.531168937683105, | |
| "learning_rate": 2.0207852193995383e-06, | |
| "loss": 2.6982, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.023094688221709007, | |
| "grad_norm": 14.727725982666016, | |
| "learning_rate": 2.5981524249422633e-06, | |
| "loss": 2.0218, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02771362586605081, | |
| "grad_norm": 8.314309120178223, | |
| "learning_rate": 3.117782909930716e-06, | |
| "loss": 1.5595, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03233256351039261, | |
| "grad_norm": 4.944284915924072, | |
| "learning_rate": 3.6951501154734412e-06, | |
| "loss": 1.02, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03695150115473441, | |
| "grad_norm": 10.882843971252441, | |
| "learning_rate": 4.272517321016166e-06, | |
| "loss": 1.0419, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04157043879907621, | |
| "grad_norm": 12.310320854187012, | |
| "learning_rate": 4.849884526558892e-06, | |
| "loss": 1.0801, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.046189376443418015, | |
| "grad_norm": 6.49992036819458, | |
| "learning_rate": 5.427251732101616e-06, | |
| "loss": 0.8444, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.050808314087759814, | |
| "grad_norm": 2.89493465423584, | |
| "learning_rate": 6.004618937644342e-06, | |
| "loss": 0.8884, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05542725173210162, | |
| "grad_norm": 3.966763734817505, | |
| "learning_rate": 6.581986143187067e-06, | |
| "loss": 0.8672, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06004618937644342, | |
| "grad_norm": 4.442293167114258, | |
| "learning_rate": 7.159353348729793e-06, | |
| "loss": 0.8037, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06466512702078522, | |
| "grad_norm": 2.506918430328369, | |
| "learning_rate": 7.736720554272519e-06, | |
| "loss": 0.724, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06928406466512702, | |
| "grad_norm": 5.733686447143555, | |
| "learning_rate": 8.314087759815242e-06, | |
| "loss": 0.7692, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07390300230946882, | |
| "grad_norm": 4.161188125610352, | |
| "learning_rate": 8.891454965357968e-06, | |
| "loss": 0.7413, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07852193995381063, | |
| "grad_norm": 3.9434962272644043, | |
| "learning_rate": 9.468822170900693e-06, | |
| "loss": 0.7386, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08314087759815242, | |
| "grad_norm": 2.9100701808929443, | |
| "learning_rate": 1.0046189376443418e-05, | |
| "loss": 0.6942, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.08775981524249422, | |
| "grad_norm": 5.367318153381348, | |
| "learning_rate": 1.0623556581986144e-05, | |
| "loss": 0.8011, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09237875288683603, | |
| "grad_norm": 3.1690614223480225, | |
| "learning_rate": 1.1200923787528869e-05, | |
| "loss": 0.6816, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09699769053117784, | |
| "grad_norm": 4.35976505279541, | |
| "learning_rate": 1.1778290993071595e-05, | |
| "loss": 0.7408, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.10161662817551963, | |
| "grad_norm": 3.330937623977661, | |
| "learning_rate": 1.235565819861432e-05, | |
| "loss": 0.7159, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.10623556581986143, | |
| "grad_norm": 5.761129379272461, | |
| "learning_rate": 1.2933025404157046e-05, | |
| "loss": 0.6838, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11085450346420324, | |
| "grad_norm": 7.05668830871582, | |
| "learning_rate": 1.351039260969977e-05, | |
| "loss": 0.7135, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.11547344110854503, | |
| "grad_norm": 3.7135939598083496, | |
| "learning_rate": 1.4087759815242497e-05, | |
| "loss": 0.6385, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.12009237875288684, | |
| "grad_norm": 5.477907657623291, | |
| "learning_rate": 1.4665127020785218e-05, | |
| "loss": 0.6292, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.12471131639722864, | |
| "grad_norm": 6.577059268951416, | |
| "learning_rate": 1.5242494226327944e-05, | |
| "loss": 0.6921, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.12933025404157045, | |
| "grad_norm": 3.6328892707824707, | |
| "learning_rate": 1.581986143187067e-05, | |
| "loss": 0.6621, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.13394919168591224, | |
| "grad_norm": 4.084783554077148, | |
| "learning_rate": 1.6397228637413393e-05, | |
| "loss": 0.6667, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.13856812933025403, | |
| "grad_norm": 3.8719701766967773, | |
| "learning_rate": 1.697459584295612e-05, | |
| "loss": 0.6692, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14318706697459585, | |
| "grad_norm": 7.860931873321533, | |
| "learning_rate": 1.7551963048498846e-05, | |
| "loss": 0.6251, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.14780600461893764, | |
| "grad_norm": 4.381837368011475, | |
| "learning_rate": 1.812933025404157e-05, | |
| "loss": 0.6297, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.15242494226327943, | |
| "grad_norm": 3.7145886421203613, | |
| "learning_rate": 1.8706697459584295e-05, | |
| "loss": 0.6483, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.15704387990762125, | |
| "grad_norm": 2.609006643295288, | |
| "learning_rate": 1.9284064665127023e-05, | |
| "loss": 0.6149, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.16166281755196305, | |
| "grad_norm": 4.774081230163574, | |
| "learning_rate": 1.9861431870669748e-05, | |
| "loss": 0.6034, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.16628175519630484, | |
| "grad_norm": 7.305100440979004, | |
| "learning_rate": 2.0438799076212473e-05, | |
| "loss": 0.6496, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.17090069284064666, | |
| "grad_norm": 5.507181644439697, | |
| "learning_rate": 2.1016166281755197e-05, | |
| "loss": 0.643, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.17551963048498845, | |
| "grad_norm": 4.033135890960693, | |
| "learning_rate": 2.1593533487297922e-05, | |
| "loss": 0.6186, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.18013856812933027, | |
| "grad_norm": 3.903007745742798, | |
| "learning_rate": 2.217090069284065e-05, | |
| "loss": 0.6041, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.18475750577367206, | |
| "grad_norm": 4.785562992095947, | |
| "learning_rate": 2.2748267898383374e-05, | |
| "loss": 0.5527, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.18937644341801385, | |
| "grad_norm": 3.4289231300354004, | |
| "learning_rate": 2.3325635103926096e-05, | |
| "loss": 0.5936, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.19399538106235567, | |
| "grad_norm": 2.384840965270996, | |
| "learning_rate": 2.3903002309468824e-05, | |
| "loss": 0.5421, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.19861431870669746, | |
| "grad_norm": 4.025755882263184, | |
| "learning_rate": 2.448036951501155e-05, | |
| "loss": 0.5839, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.20323325635103925, | |
| "grad_norm": 4.832013130187988, | |
| "learning_rate": 2.5057736720554276e-05, | |
| "loss": 0.5938, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.20785219399538107, | |
| "grad_norm": 3.66886305809021, | |
| "learning_rate": 2.5635103926096998e-05, | |
| "loss": 0.5607, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.21247113163972287, | |
| "grad_norm": 3.7285852432250977, | |
| "learning_rate": 2.6212471131639726e-05, | |
| "loss": 0.5457, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.21709006928406466, | |
| "grad_norm": 3.755711555480957, | |
| "learning_rate": 2.678983833718245e-05, | |
| "loss": 0.5721, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.22170900692840648, | |
| "grad_norm": 4.016116619110107, | |
| "learning_rate": 2.7367205542725178e-05, | |
| "loss": 0.59, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.22632794457274827, | |
| "grad_norm": 6.123377799987793, | |
| "learning_rate": 2.79445727482679e-05, | |
| "loss": 0.6236, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.23094688221709006, | |
| "grad_norm": 3.77093505859375, | |
| "learning_rate": 2.8521939953810624e-05, | |
| "loss": 0.6306, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23556581986143188, | |
| "grad_norm": 5.101199626922607, | |
| "learning_rate": 2.9099307159353352e-05, | |
| "loss": 0.5806, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.24018475750577367, | |
| "grad_norm": 4.425400257110596, | |
| "learning_rate": 2.9676674364896073e-05, | |
| "loss": 0.5644, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.24480369515011546, | |
| "grad_norm": 3.661381244659424, | |
| "learning_rate": 3.02540415704388e-05, | |
| "loss": 0.538, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.24942263279445728, | |
| "grad_norm": 3.271655559539795, | |
| "learning_rate": 3.0831408775981526e-05, | |
| "loss": 0.5927, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2540415704387991, | |
| "grad_norm": 4.603795051574707, | |
| "learning_rate": 3.140877598152425e-05, | |
| "loss": 0.5772, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2586605080831409, | |
| "grad_norm": 3.379786968231201, | |
| "learning_rate": 3.1986143187066975e-05, | |
| "loss": 0.5577, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2632794457274827, | |
| "grad_norm": 3.3702409267425537, | |
| "learning_rate": 3.25635103926097e-05, | |
| "loss": 0.6258, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2678983833718245, | |
| "grad_norm": 3.1498706340789795, | |
| "learning_rate": 3.3140877598152424e-05, | |
| "loss": 0.5631, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.27251732101616627, | |
| "grad_norm": 4.846761703491211, | |
| "learning_rate": 3.3718244803695156e-05, | |
| "loss": 0.5944, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.27713625866050806, | |
| "grad_norm": 4.397818088531494, | |
| "learning_rate": 3.4295612009237874e-05, | |
| "loss": 0.5698, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2817551963048499, | |
| "grad_norm": 2.2724201679229736, | |
| "learning_rate": 3.4872979214780605e-05, | |
| "loss": 0.5468, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2863741339491917, | |
| "grad_norm": 4.73584508895874, | |
| "learning_rate": 3.545034642032333e-05, | |
| "loss": 0.5697, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.2909930715935335, | |
| "grad_norm": 3.557711124420166, | |
| "learning_rate": 3.6027713625866054e-05, | |
| "loss": 0.5195, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2956120092378753, | |
| "grad_norm": 4.273311614990234, | |
| "learning_rate": 3.660508083140878e-05, | |
| "loss": 0.5561, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3002309468822171, | |
| "grad_norm": 3.6489686965942383, | |
| "learning_rate": 3.7182448036951504e-05, | |
| "loss": 0.566, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.30484988452655887, | |
| "grad_norm": 3.4011149406433105, | |
| "learning_rate": 3.775981524249423e-05, | |
| "loss": 0.5468, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3094688221709007, | |
| "grad_norm": 3.517822742462158, | |
| "learning_rate": 3.833718244803695e-05, | |
| "loss": 0.5834, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3140877598152425, | |
| "grad_norm": 2.7577602863311768, | |
| "learning_rate": 3.891454965357968e-05, | |
| "loss": 0.5489, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3187066974595843, | |
| "grad_norm": 2.856598138809204, | |
| "learning_rate": 3.94919168591224e-05, | |
| "loss": 0.4906, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.3233256351039261, | |
| "grad_norm": 4.136464595794678, | |
| "learning_rate": 4.0069284064665133e-05, | |
| "loss": 0.5705, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3279445727482679, | |
| "grad_norm": 4.121008396148682, | |
| "learning_rate": 4.064665127020785e-05, | |
| "loss": 0.6241, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3325635103926097, | |
| "grad_norm": 3.611814498901367, | |
| "learning_rate": 4.122401847575058e-05, | |
| "loss": 0.5034, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3371824480369515, | |
| "grad_norm": 3.4624834060668945, | |
| "learning_rate": 4.18013856812933e-05, | |
| "loss": 0.5643, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.3418013856812933, | |
| "grad_norm": 3.3586976528167725, | |
| "learning_rate": 4.237875288683603e-05, | |
| "loss": 0.5439, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3464203233256351, | |
| "grad_norm": 2.5129313468933105, | |
| "learning_rate": 4.2956120092378757e-05, | |
| "loss": 0.5062, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3510392609699769, | |
| "grad_norm": 3.0052690505981445, | |
| "learning_rate": 4.353348729792148e-05, | |
| "loss": 0.5219, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.3556581986143187, | |
| "grad_norm": 3.7070388793945312, | |
| "learning_rate": 4.4110854503464206e-05, | |
| "loss": 0.5425, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.36027713625866054, | |
| "grad_norm": 2.8378756046295166, | |
| "learning_rate": 4.468822170900693e-05, | |
| "loss": 0.5208, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.3648960739030023, | |
| "grad_norm": 2.3988196849823, | |
| "learning_rate": 4.5265588914549655e-05, | |
| "loss": 0.5149, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3695150115473441, | |
| "grad_norm": 2.2305569648742676, | |
| "learning_rate": 4.584295612009238e-05, | |
| "loss": 0.5028, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3741339491916859, | |
| "grad_norm": 3.0817391872406006, | |
| "learning_rate": 4.6420323325635104e-05, | |
| "loss": 0.5419, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.3787528868360277, | |
| "grad_norm": 2.767381429672241, | |
| "learning_rate": 4.699769053117783e-05, | |
| "loss": 0.5036, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.3833718244803695, | |
| "grad_norm": 2.8563129901885986, | |
| "learning_rate": 4.757505773672056e-05, | |
| "loss": 0.5079, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.38799076212471134, | |
| "grad_norm": 4.459218978881836, | |
| "learning_rate": 4.815242494226328e-05, | |
| "loss": 0.4891, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.39260969976905313, | |
| "grad_norm": 2.825631856918335, | |
| "learning_rate": 4.872979214780601e-05, | |
| "loss": 0.4921, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.3972286374133949, | |
| "grad_norm": 2.835643768310547, | |
| "learning_rate": 4.9307159353348734e-05, | |
| "loss": 0.4736, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4018475750577367, | |
| "grad_norm": 3.237922430038452, | |
| "learning_rate": 4.988452655889146e-05, | |
| "loss": 0.5254, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.4064665127020785, | |
| "grad_norm": 3.913771390914917, | |
| "learning_rate": 4.9999870022388165e-05, | |
| "loss": 0.5212, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4110854503464203, | |
| "grad_norm": 4.357654094696045, | |
| "learning_rate": 4.999934199065641e-05, | |
| "loss": 0.6004, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.41570438799076215, | |
| "grad_norm": 2.1519246101379395, | |
| "learning_rate": 4.999840778977644e-05, | |
| "loss": 0.5106, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.42032332563510394, | |
| "grad_norm": 3.196580410003662, | |
| "learning_rate": 4.9997067434926386e-05, | |
| "loss": 0.4708, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.42494226327944573, | |
| "grad_norm": 3.0933518409729004, | |
| "learning_rate": 4.9995320947883265e-05, | |
| "loss": 0.5228, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.4295612009237875, | |
| "grad_norm": 2.5883092880249023, | |
| "learning_rate": 4.999316835702259e-05, | |
| "loss": 0.54, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.4341801385681293, | |
| "grad_norm": 2.738126277923584, | |
| "learning_rate": 4.9990609697317916e-05, | |
| "loss": 0.4965, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.4387990762124711, | |
| "grad_norm": 2.903364419937134, | |
| "learning_rate": 4.998764501034028e-05, | |
| "loss": 0.4732, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.44341801385681295, | |
| "grad_norm": 1.895765781402588, | |
| "learning_rate": 4.998427434425753e-05, | |
| "loss": 0.4857, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.44803695150115475, | |
| "grad_norm": 5.765718936920166, | |
| "learning_rate": 4.998049775383353e-05, | |
| "loss": 0.4901, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.45265588914549654, | |
| "grad_norm": 3.9257349967956543, | |
| "learning_rate": 4.997631530042727e-05, | |
| "loss": 0.493, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.45727482678983833, | |
| "grad_norm": 2.3068912029266357, | |
| "learning_rate": 4.997172705199189e-05, | |
| "loss": 0.5061, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.4618937644341801, | |
| "grad_norm": 2.966820001602173, | |
| "learning_rate": 4.996673308307355e-05, | |
| "loss": 0.4557, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4665127020785219, | |
| "grad_norm": 3.704120397567749, | |
| "learning_rate": 4.996133347481021e-05, | |
| "loss": 0.5302, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.47113163972286376, | |
| "grad_norm": 2.066206216812134, | |
| "learning_rate": 4.9955528314930376e-05, | |
| "loss": 0.5035, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.47575057736720555, | |
| "grad_norm": 3.7750158309936523, | |
| "learning_rate": 4.9949317697751596e-05, | |
| "loss": 0.5327, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.48036951501154734, | |
| "grad_norm": 2.5127952098846436, | |
| "learning_rate": 4.9942701724178965e-05, | |
| "loss": 0.4929, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.48498845265588914, | |
| "grad_norm": 2.594442367553711, | |
| "learning_rate": 4.9935680501703485e-05, | |
| "loss": 0.4795, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.4896073903002309, | |
| "grad_norm": 4.830276966094971, | |
| "learning_rate": 4.992825414440032e-05, | |
| "loss": 0.5268, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.4942263279445728, | |
| "grad_norm": 2.404269218444824, | |
| "learning_rate": 4.9920422772926933e-05, | |
| "loss": 0.5199, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.49884526558891457, | |
| "grad_norm": 2.622856378555298, | |
| "learning_rate": 4.991218651452114e-05, | |
| "loss": 0.4928, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5034642032332564, | |
| "grad_norm": 2.660313844680786, | |
| "learning_rate": 4.9903545502999014e-05, | |
| "loss": 0.516, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5080831408775982, | |
| "grad_norm": 2.2060108184814453, | |
| "learning_rate": 4.9894499878752744e-05, | |
| "loss": 0.474, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5127020785219399, | |
| "grad_norm": 2.774624824523926, | |
| "learning_rate": 4.988504978874835e-05, | |
| "loss": 0.5053, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5173210161662818, | |
| "grad_norm": 2.0086095333099365, | |
| "learning_rate": 4.987519538652326e-05, | |
| "loss": 0.4478, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5219399538106235, | |
| "grad_norm": 2.7462918758392334, | |
| "learning_rate": 4.986493683218386e-05, | |
| "loss": 0.4872, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.5265588914549654, | |
| "grad_norm": 2.2208642959594727, | |
| "learning_rate": 4.985427429240286e-05, | |
| "loss": 0.4617, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.5311778290993071, | |
| "grad_norm": 3.3714964389801025, | |
| "learning_rate": 4.984320794041662e-05, | |
| "loss": 0.5293, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.535796766743649, | |
| "grad_norm": 2.732804298400879, | |
| "learning_rate": 4.98317379560223e-05, | |
| "loss": 0.4582, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.5404157043879908, | |
| "grad_norm": 3.4411492347717285, | |
| "learning_rate": 4.981986452557495e-05, | |
| "loss": 0.5359, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.5450346420323325, | |
| "grad_norm": 2.5346171855926514, | |
| "learning_rate": 4.9807587841984494e-05, | |
| "loss": 0.438, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.5496535796766744, | |
| "grad_norm": 2.068873882293701, | |
| "learning_rate": 4.9794908104712586e-05, | |
| "loss": 0.4541, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.5542725173210161, | |
| "grad_norm": 3.7652342319488525, | |
| "learning_rate": 4.978182551976939e-05, | |
| "loss": 0.5372, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.558891454965358, | |
| "grad_norm": 2.7133548259735107, | |
| "learning_rate": 4.976834029971017e-05, | |
| "loss": 0.4949, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.5635103926096998, | |
| "grad_norm": 2.7093558311462402, | |
| "learning_rate": 4.975445266363191e-05, | |
| "loss": 0.4484, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.5681293302540416, | |
| "grad_norm": 2.551117420196533, | |
| "learning_rate": 4.974016283716974e-05, | |
| "loss": 0.4865, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.5727482678983834, | |
| "grad_norm": 3.5680108070373535, | |
| "learning_rate": 4.9725471052493225e-05, | |
| "loss": 0.512, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.5773672055427251, | |
| "grad_norm": 3.1270763874053955, | |
| "learning_rate": 4.9710377548302636e-05, | |
| "loss": 0.4658, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.581986143187067, | |
| "grad_norm": 2.7450294494628906, | |
| "learning_rate": 4.9694882569825045e-05, | |
| "loss": 0.5726, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.5866050808314087, | |
| "grad_norm": 2.192270040512085, | |
| "learning_rate": 4.967898636881039e-05, | |
| "loss": 0.456, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5912240184757506, | |
| "grad_norm": 3.5508313179016113, | |
| "learning_rate": 4.9662689203527304e-05, | |
| "loss": 0.4363, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5958429561200924, | |
| "grad_norm": 3.789588212966919, | |
| "learning_rate": 4.964599133875899e-05, | |
| "loss": 0.3894, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6004618937644342, | |
| "grad_norm": 2.185136556625366, | |
| "learning_rate": 4.9628893045798905e-05, | |
| "loss": 0.4388, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.605080831408776, | |
| "grad_norm": 3.9445693492889404, | |
| "learning_rate": 4.961139460244631e-05, | |
| "loss": 0.4761, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6096997690531177, | |
| "grad_norm": 2.027426242828369, | |
| "learning_rate": 4.95934962930018e-05, | |
| "loss": 0.4726, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.6143187066974596, | |
| "grad_norm": 2.555227518081665, | |
| "learning_rate": 4.957519840826268e-05, | |
| "loss": 0.4546, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.6189376443418014, | |
| "grad_norm": 3.800353765487671, | |
| "learning_rate": 4.955650124551823e-05, | |
| "loss": 0.4657, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.6235565819861432, | |
| "grad_norm": 1.877545952796936, | |
| "learning_rate": 4.953740510854485e-05, | |
| "loss": 0.46, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.628175519630485, | |
| "grad_norm": 2.205172538757324, | |
| "learning_rate": 4.9517910307601204e-05, | |
| "loss": 0.4845, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.6327944572748267, | |
| "grad_norm": 4.207221031188965, | |
| "learning_rate": 4.949801715942306e-05, | |
| "loss": 0.4784, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.6374133949191686, | |
| "grad_norm": 2.3010756969451904, | |
| "learning_rate": 4.947772598721828e-05, | |
| "loss": 0.4679, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.6420323325635104, | |
| "grad_norm": 1.776963233947754, | |
| "learning_rate": 4.9457037120661455e-05, | |
| "loss": 0.4405, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.6466512702078522, | |
| "grad_norm": 2.8734822273254395, | |
| "learning_rate": 4.9435950895888604e-05, | |
| "loss": 0.474, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.651270207852194, | |
| "grad_norm": 3.049509286880493, | |
| "learning_rate": 4.9414467655491695e-05, | |
| "loss": 0.4334, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.6558891454965358, | |
| "grad_norm": 3.274083375930786, | |
| "learning_rate": 4.9392587748513105e-05, | |
| "loss": 0.4519, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.6605080831408776, | |
| "grad_norm": 2.149959087371826, | |
| "learning_rate": 4.937031153043991e-05, | |
| "loss": 0.4625, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.6651270207852193, | |
| "grad_norm": 2.3562982082366943, | |
| "learning_rate": 4.934763936319814e-05, | |
| "loss": 0.4487, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.6697459584295612, | |
| "grad_norm": 2.5950613021850586, | |
| "learning_rate": 4.932457161514689e-05, | |
| "loss": 0.4979, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.674364896073903, | |
| "grad_norm": 2.4045233726501465, | |
| "learning_rate": 4.9301108661072315e-05, | |
| "loss": 0.4442, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.6789838337182448, | |
| "grad_norm": 2.833594799041748, | |
| "learning_rate": 4.9277250882181575e-05, | |
| "loss": 0.4408, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.6836027713625866, | |
| "grad_norm": 1.7960981130599976, | |
| "learning_rate": 4.9252998666096625e-05, | |
| "loss": 0.4672, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.6882217090069284, | |
| "grad_norm": 2.261974811553955, | |
| "learning_rate": 4.922835240684792e-05, | |
| "loss": 0.3947, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.6928406466512702, | |
| "grad_norm": 2.966799020767212, | |
| "learning_rate": 4.9203312504867994e-05, | |
| "loss": 0.4546, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6974595842956121, | |
| "grad_norm": 2.1086158752441406, | |
| "learning_rate": 4.9177879366984985e-05, | |
| "loss": 0.4226, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.7020785219399538, | |
| "grad_norm": 2.4200305938720703, | |
| "learning_rate": 4.915205340641601e-05, | |
| "loss": 0.4481, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.7066974595842956, | |
| "grad_norm": 2.486839771270752, | |
| "learning_rate": 4.912583504276045e-05, | |
| "loss": 0.4594, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.7113163972286374, | |
| "grad_norm": 1.7007755041122437, | |
| "learning_rate": 4.9099224701993115e-05, | |
| "loss": 0.4472, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.7159353348729792, | |
| "grad_norm": 1.6047892570495605, | |
| "learning_rate": 4.907222281645739e-05, | |
| "loss": 0.4734, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.7205542725173211, | |
| "grad_norm": 3.141820192337036, | |
| "learning_rate": 4.904482982485813e-05, | |
| "loss": 0.4777, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.7251732101616628, | |
| "grad_norm": 2.6380691528320312, | |
| "learning_rate": 4.901704617225455e-05, | |
| "loss": 0.4438, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.7297921478060047, | |
| "grad_norm": 2.238063335418701, | |
| "learning_rate": 4.898887231005306e-05, | |
| "loss": 0.4424, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.7344110854503464, | |
| "grad_norm": 2.1911888122558594, | |
| "learning_rate": 4.896030869599983e-05, | |
| "loss": 0.437, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.7390300230946882, | |
| "grad_norm": 2.0892059803009033, | |
| "learning_rate": 4.893135579417345e-05, | |
| "loss": 0.4341, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.74364896073903, | |
| "grad_norm": 1.727277398109436, | |
| "learning_rate": 4.8902014074977295e-05, | |
| "loss": 0.4428, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.7482678983833718, | |
| "grad_norm": 2.6759467124938965, | |
| "learning_rate": 4.8872284015131965e-05, | |
| "loss": 0.411, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.7528868360277137, | |
| "grad_norm": 3.616370439529419, | |
| "learning_rate": 4.8842166097667505e-05, | |
| "loss": 0.4376, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.7575057736720554, | |
| "grad_norm": 1.77035653591156, | |
| "learning_rate": 4.881166081191555e-05, | |
| "loss": 0.4217, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.7621247113163973, | |
| "grad_norm": 2.7078516483306885, | |
| "learning_rate": 4.878076865350136e-05, | |
| "loss": 0.4451, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.766743648960739, | |
| "grad_norm": 1.7627867460250854, | |
| "learning_rate": 4.874949012433584e-05, | |
| "loss": 0.4399, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.7713625866050808, | |
| "grad_norm": 1.7273634672164917, | |
| "learning_rate": 4.871782573260729e-05, | |
| "loss": 0.4202, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.7759815242494227, | |
| "grad_norm": 2.0963988304138184, | |
| "learning_rate": 4.868577599277322e-05, | |
| "loss": 0.4447, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.7806004618937644, | |
| "grad_norm": 2.144355297088623, | |
| "learning_rate": 4.865334142555196e-05, | |
| "loss": 0.4387, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.7852193995381063, | |
| "grad_norm": 2.0202884674072266, | |
| "learning_rate": 4.862052255791419e-05, | |
| "loss": 0.4124, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.789838337182448, | |
| "grad_norm": 1.6518304347991943, | |
| "learning_rate": 4.858731992307441e-05, | |
| "loss": 0.4222, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.7944572748267898, | |
| "grad_norm": 1.3399207592010498, | |
| "learning_rate": 4.855373406048226e-05, | |
| "loss": 0.4713, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.7990762124711316, | |
| "grad_norm": 1.7980890274047852, | |
| "learning_rate": 4.8519765515813744e-05, | |
| "loss": 0.4484, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.8036951501154734, | |
| "grad_norm": 3.4402995109558105, | |
| "learning_rate": 4.8485414840962384e-05, | |
| "loss": 0.3777, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.8083140877598153, | |
| "grad_norm": 1.9399069547653198, | |
| "learning_rate": 4.845068259403024e-05, | |
| "loss": 0.3942, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.812933025404157, | |
| "grad_norm": 1.8960610628128052, | |
| "learning_rate": 4.841556933931886e-05, | |
| "loss": 0.3738, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.8175519630484989, | |
| "grad_norm": 2.6203131675720215, | |
| "learning_rate": 4.838007564732008e-05, | |
| "loss": 0.4725, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.8221709006928406, | |
| "grad_norm": 2.441143274307251, | |
| "learning_rate": 4.834420209470679e-05, | |
| "loss": 0.4287, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.8267898383371824, | |
| "grad_norm": 3.31255841255188, | |
| "learning_rate": 4.830794926432355e-05, | |
| "loss": 0.4215, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.8314087759815243, | |
| "grad_norm": 2.7491471767425537, | |
| "learning_rate": 4.827131774517709e-05, | |
| "loss": 0.4301, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.836027713625866, | |
| "grad_norm": 1.9549751281738281, | |
| "learning_rate": 4.8234308132426807e-05, | |
| "loss": 0.4379, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.8406466512702079, | |
| "grad_norm": 2.4840948581695557, | |
| "learning_rate": 4.819692102737505e-05, | |
| "loss": 0.4474, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.8452655889145496, | |
| "grad_norm": 2.865260601043701, | |
| "learning_rate": 4.815915703745734e-05, | |
| "loss": 0.4608, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.8498845265588915, | |
| "grad_norm": 2.3620669841766357, | |
| "learning_rate": 4.812101677623254e-05, | |
| "loss": 0.4311, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.8545034642032333, | |
| "grad_norm": 3.4287946224212646, | |
| "learning_rate": 4.808250086337284e-05, | |
| "loss": 0.4035, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.859122401847575, | |
| "grad_norm": 1.6560109853744507, | |
| "learning_rate": 4.8043609924653745e-05, | |
| "loss": 0.4004, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.8637413394919169, | |
| "grad_norm": 2.310784339904785, | |
| "learning_rate": 4.800434459194386e-05, | |
| "loss": 0.4219, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.8683602771362586, | |
| "grad_norm": 1.660062313079834, | |
| "learning_rate": 4.796470550319465e-05, | |
| "loss": 0.4102, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.8729792147806005, | |
| "grad_norm": 2.578974723815918, | |
| "learning_rate": 4.792469330243007e-05, | |
| "loss": 0.4375, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.8775981524249422, | |
| "grad_norm": 2.095388412475586, | |
| "learning_rate": 4.788430863973607e-05, | |
| "loss": 0.4385, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.8822170900692841, | |
| "grad_norm": 2.4962880611419678, | |
| "learning_rate": 4.7843552171250085e-05, | |
| "loss": 0.4399, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.8868360277136259, | |
| "grad_norm": 2.8000426292419434, | |
| "learning_rate": 4.780242455915036e-05, | |
| "loss": 0.4344, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.8914549653579676, | |
| "grad_norm": 2.0987462997436523, | |
| "learning_rate": 4.776092647164516e-05, | |
| "loss": 0.3961, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.8960739030023095, | |
| "grad_norm": 3.531191825866699, | |
| "learning_rate": 4.771905858296195e-05, | |
| "loss": 0.4463, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.9006928406466512, | |
| "grad_norm": 2.8484129905700684, | |
| "learning_rate": 4.767682157333643e-05, | |
| "loss": 0.4993, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.9053117782909931, | |
| "grad_norm": 2.164492607116699, | |
| "learning_rate": 4.7634216129001453e-05, | |
| "loss": 0.4545, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.9099307159353349, | |
| "grad_norm": 2.37625789642334, | |
| "learning_rate": 4.7591242942175924e-05, | |
| "loss": 0.3965, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.9145496535796767, | |
| "grad_norm": 2.488903284072876, | |
| "learning_rate": 4.7547902711053535e-05, | |
| "loss": 0.4459, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.9191685912240185, | |
| "grad_norm": 3.028672933578491, | |
| "learning_rate": 4.7504196139791405e-05, | |
| "loss": 0.452, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.9237875288683602, | |
| "grad_norm": 2.084078073501587, | |
| "learning_rate": 4.746012393849866e-05, | |
| "loss": 0.4078, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9284064665127021, | |
| "grad_norm": 2.061680555343628, | |
| "learning_rate": 4.741568682322488e-05, | |
| "loss": 0.3691, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.9330254041570438, | |
| "grad_norm": 2.4597408771514893, | |
| "learning_rate": 4.7370885515948486e-05, | |
| "loss": 0.3844, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.9376443418013857, | |
| "grad_norm": 2.8126583099365234, | |
| "learning_rate": 4.732572074456498e-05, | |
| "loss": 0.392, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.9422632794457275, | |
| "grad_norm": 2.3137059211730957, | |
| "learning_rate": 4.728019324287516e-05, | |
| "loss": 0.4237, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.9468822170900693, | |
| "grad_norm": 3.646803855895996, | |
| "learning_rate": 4.723430375057314e-05, | |
| "loss": 0.3857, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.9515011547344111, | |
| "grad_norm": 1.3006259202957153, | |
| "learning_rate": 4.718805301323439e-05, | |
| "loss": 0.4193, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.9561200923787528, | |
| "grad_norm": 3.648193359375, | |
| "learning_rate": 4.71414417823036e-05, | |
| "loss": 0.3896, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.9607390300230947, | |
| "grad_norm": 2.0217714309692383, | |
| "learning_rate": 4.7094470815082443e-05, | |
| "loss": 0.4315, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.9653579676674365, | |
| "grad_norm": 1.8971320390701294, | |
| "learning_rate": 4.704714087471733e-05, | |
| "loss": 0.4643, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.9699769053117783, | |
| "grad_norm": 2.0947530269622803, | |
| "learning_rate": 4.6999452730186966e-05, | |
| "loss": 0.4065, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.9745958429561201, | |
| "grad_norm": 2.21998929977417, | |
| "learning_rate": 4.6951407156289856e-05, | |
| "loss": 0.4102, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.9792147806004619, | |
| "grad_norm": 2.2499032020568848, | |
| "learning_rate": 4.6903004933631745e-05, | |
| "loss": 0.4339, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.9838337182448037, | |
| "grad_norm": 2.322390079498291, | |
| "learning_rate": 4.68542468486129e-05, | |
| "loss": 0.41, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.9884526558891455, | |
| "grad_norm": 2.2508604526519775, | |
| "learning_rate": 4.680513369341539e-05, | |
| "loss": 0.4142, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.9930715935334873, | |
| "grad_norm": 2.1183762550354004, | |
| "learning_rate": 4.675566626599014e-05, | |
| "loss": 0.4064, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.9976905311778291, | |
| "grad_norm": 1.6184558868408203, | |
| "learning_rate": 4.6705845370044015e-05, | |
| "loss": 0.4167, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.002309468822171, | |
| "grad_norm": 2.536665678024292, | |
| "learning_rate": 4.665567181502676e-05, | |
| "loss": 0.3415, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.0069284064665127, | |
| "grad_norm": 1.4247705936431885, | |
| "learning_rate": 4.6605146416117835e-05, | |
| "loss": 0.2922, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.0115473441108545, | |
| "grad_norm": 2.4373302459716797, | |
| "learning_rate": 4.655426999421317e-05, | |
| "loss": 0.346, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.0161662817551964, | |
| "grad_norm": 2.1059627532958984, | |
| "learning_rate": 4.650304337591185e-05, | |
| "loss": 0.3091, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.0207852193995381, | |
| "grad_norm": 1.6682378053665161, | |
| "learning_rate": 4.6451467393502644e-05, | |
| "loss": 0.286, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.0254041570438799, | |
| "grad_norm": 2.1200857162475586, | |
| "learning_rate": 4.639954288495054e-05, | |
| "loss": 0.3246, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.0300230946882216, | |
| "grad_norm": 2.503147602081299, | |
| "learning_rate": 4.634727069388306e-05, | |
| "loss": 0.3594, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.0346420323325636, | |
| "grad_norm": 1.5546462535858154, | |
| "learning_rate": 4.629465166957662e-05, | |
| "loss": 0.3154, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.0392609699769053, | |
| "grad_norm": 1.6011874675750732, | |
| "learning_rate": 4.6241686666942694e-05, | |
| "loss": 0.3309, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.043879907621247, | |
| "grad_norm": 1.319419026374817, | |
| "learning_rate": 4.618837654651393e-05, | |
| "loss": 0.3198, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.048498845265589, | |
| "grad_norm": 1.689467191696167, | |
| "learning_rate": 4.613472217443017e-05, | |
| "loss": 0.3203, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.0531177829099307, | |
| "grad_norm": 3.2122750282287598, | |
| "learning_rate": 4.608072442242439e-05, | |
| "loss": 0.3269, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.0577367205542725, | |
| "grad_norm": 1.627459168434143, | |
| "learning_rate": 4.602638416780851e-05, | |
| "loss": 0.3094, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.0623556581986142, | |
| "grad_norm": 2.7835917472839355, | |
| "learning_rate": 4.5971702293459163e-05, | |
| "loss": 0.3146, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.0669745958429562, | |
| "grad_norm": 1.5927238464355469, | |
| "learning_rate": 4.591667968780336e-05, | |
| "loss": 0.3058, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.071593533487298, | |
| "grad_norm": 1.8358134031295776, | |
| "learning_rate": 4.5861317244804015e-05, | |
| "loss": 0.3079, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.0762124711316396, | |
| "grad_norm": 1.3477319478988647, | |
| "learning_rate": 4.580561586394545e-05, | |
| "loss": 0.3355, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.0808314087759816, | |
| "grad_norm": 2.179304361343384, | |
| "learning_rate": 4.57495764502188e-05, | |
| "loss": 0.3301, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.0854503464203233, | |
| "grad_norm": 1.5659071207046509, | |
| "learning_rate": 4.569319991410725e-05, | |
| "loss": 0.3133, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.090069284064665, | |
| "grad_norm": 1.8552881479263306, | |
| "learning_rate": 4.563648717157131e-05, | |
| "loss": 0.3287, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.0946882217090068, | |
| "grad_norm": 2.5701863765716553, | |
| "learning_rate": 4.557943914403386e-05, | |
| "loss": 0.3219, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.0993071593533488, | |
| "grad_norm": 1.5083070993423462, | |
| "learning_rate": 4.552205675836527e-05, | |
| "loss": 0.317, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.1039260969976905, | |
| "grad_norm": 2.8584365844726562, | |
| "learning_rate": 4.5464340946868256e-05, | |
| "loss": 0.2864, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.1085450346420322, | |
| "grad_norm": 1.630210518836975, | |
| "learning_rate": 4.540629264726278e-05, | |
| "loss": 0.2998, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.1131639722863742, | |
| "grad_norm": 1.8594950437545776, | |
| "learning_rate": 4.53479128026708e-05, | |
| "loss": 0.3156, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.117782909930716, | |
| "grad_norm": 1.5817375183105469, | |
| "learning_rate": 4.528920236160096e-05, | |
| "loss": 0.3201, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.1224018475750577, | |
| "grad_norm": 2.3912880420684814, | |
| "learning_rate": 4.523016227793315e-05, | |
| "loss": 0.2986, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.1270207852193996, | |
| "grad_norm": 1.5563384294509888, | |
| "learning_rate": 4.5170793510903046e-05, | |
| "loss": 0.3021, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.1316397228637414, | |
| "grad_norm": 1.8241164684295654, | |
| "learning_rate": 4.511109702508648e-05, | |
| "loss": 0.3316, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.136258660508083, | |
| "grad_norm": 2.19366192817688, | |
| "learning_rate": 4.505107379038384e-05, | |
| "loss": 0.3255, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.140877598152425, | |
| "grad_norm": 1.471584439277649, | |
| "learning_rate": 4.499072478200421e-05, | |
| "loss": 0.3245, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.1454965357967668, | |
| "grad_norm": 1.6688241958618164, | |
| "learning_rate": 4.493005098044963e-05, | |
| "loss": 0.3336, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.1501154734411085, | |
| "grad_norm": 1.9700522422790527, | |
| "learning_rate": 4.486905337149909e-05, | |
| "loss": 0.3277, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.1547344110854503, | |
| "grad_norm": 2.3884382247924805, | |
| "learning_rate": 4.480773294619255e-05, | |
| "loss": 0.2978, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.1593533487297922, | |
| "grad_norm": 2.134876251220703, | |
| "learning_rate": 4.474609070081483e-05, | |
| "loss": 0.3168, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.163972286374134, | |
| "grad_norm": 1.8444334268569946, | |
| "learning_rate": 4.468412763687942e-05, | |
| "loss": 0.3228, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.1685912240184757, | |
| "grad_norm": 1.6720913648605347, | |
| "learning_rate": 4.4621844761112216e-05, | |
| "loss": 0.3174, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.1732101616628174, | |
| "grad_norm": 1.8607888221740723, | |
| "learning_rate": 4.4559243085435154e-05, | |
| "loss": 0.265, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.1778290993071594, | |
| "grad_norm": 1.9527697563171387, | |
| "learning_rate": 4.449632362694978e-05, | |
| "loss": 0.3143, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.1824480369515011, | |
| "grad_norm": 2.462707757949829, | |
| "learning_rate": 4.443308740792072e-05, | |
| "loss": 0.3007, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.1870669745958429, | |
| "grad_norm": 1.9817255735397339, | |
| "learning_rate": 4.436953545575904e-05, | |
| "loss": 0.3135, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.1916859122401848, | |
| "grad_norm": 2.517148971557617, | |
| "learning_rate": 4.430566880300563e-05, | |
| "loss": 0.3346, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.1963048498845266, | |
| "grad_norm": 2.951472282409668, | |
| "learning_rate": 4.4241488487314365e-05, | |
| "loss": 0.3111, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.2009237875288683, | |
| "grad_norm": 3.675570249557495, | |
| "learning_rate": 4.417699555143523e-05, | |
| "loss": 0.3082, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.2055427251732103, | |
| "grad_norm": 1.6169134378433228, | |
| "learning_rate": 4.411219104319746e-05, | |
| "loss": 0.3247, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.210161662817552, | |
| "grad_norm": 4.280526638031006, | |
| "learning_rate": 4.404707601549244e-05, | |
| "loss": 0.3309, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.2147806004618937, | |
| "grad_norm": 1.2955102920532227, | |
| "learning_rate": 4.398165152625663e-05, | |
| "loss": 0.2845, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.2193995381062355, | |
| "grad_norm": 2.1361026763916016, | |
| "learning_rate": 4.391591863845436e-05, | |
| "loss": 0.3143, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.2240184757505774, | |
| "grad_norm": 2.1552889347076416, | |
| "learning_rate": 4.384987842006059e-05, | |
| "loss": 0.3042, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.2286374133949192, | |
| "grad_norm": 2.199091672897339, | |
| "learning_rate": 4.378353194404352e-05, | |
| "loss": 0.3426, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.233256351039261, | |
| "grad_norm": 2.23543119430542, | |
| "learning_rate": 4.371688028834721e-05, | |
| "loss": 0.3047, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.2378752886836029, | |
| "grad_norm": 1.7668689489364624, | |
| "learning_rate": 4.3649924535873994e-05, | |
| "loss": 0.3269, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.2424942263279446, | |
| "grad_norm": 1.9710248708724976, | |
| "learning_rate": 4.3582665774466946e-05, | |
| "loss": 0.3252, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.2471131639722863, | |
| "grad_norm": 1.9266043901443481, | |
| "learning_rate": 4.351510509689218e-05, | |
| "loss": 0.2913, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.251732101616628, | |
| "grad_norm": 1.4974067211151123, | |
| "learning_rate": 4.34472436008211e-05, | |
| "loss": 0.3071, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 1.25635103926097, | |
| "grad_norm": 1.7189793586730957, | |
| "learning_rate": 4.337908238881258e-05, | |
| "loss": 0.321, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 1.2609699769053118, | |
| "grad_norm": 1.8770833015441895, | |
| "learning_rate": 4.331062256829501e-05, | |
| "loss": 0.3252, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.2655889145496535, | |
| "grad_norm": 2.516841411590576, | |
| "learning_rate": 4.324186525154835e-05, | |
| "loss": 0.2971, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 1.2702078521939955, | |
| "grad_norm": 1.8335446119308472, | |
| "learning_rate": 4.317281155568604e-05, | |
| "loss": 0.3137, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.2748267898383372, | |
| "grad_norm": 1.632881999015808, | |
| "learning_rate": 4.3103462602636845e-05, | |
| "loss": 0.3096, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.279445727482679, | |
| "grad_norm": 1.6128835678100586, | |
| "learning_rate": 4.303381951912661e-05, | |
| "loss": 0.3228, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 1.2840646651270209, | |
| "grad_norm": 2.278327703475952, | |
| "learning_rate": 4.296388343666001e-05, | |
| "loss": 0.2835, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 1.2886836027713626, | |
| "grad_norm": 1.4413753747940063, | |
| "learning_rate": 4.2893655491502105e-05, | |
| "loss": 0.2993, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.2933025404157044, | |
| "grad_norm": 2.016895055770874, | |
| "learning_rate": 4.282313682465992e-05, | |
| "loss": 0.2721, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.2979214780600463, | |
| "grad_norm": 1.687320590019226, | |
| "learning_rate": 4.275232858186388e-05, | |
| "loss": 0.2811, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 1.302540415704388, | |
| "grad_norm": 1.556947112083435, | |
| "learning_rate": 4.268123191354921e-05, | |
| "loss": 0.3164, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.3071593533487298, | |
| "grad_norm": 1.7516964673995972, | |
| "learning_rate": 4.260984797483725e-05, | |
| "loss": 0.2722, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 1.3117782909930715, | |
| "grad_norm": 1.999335765838623, | |
| "learning_rate": 4.2538177925516665e-05, | |
| "loss": 0.289, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 1.3163972286374133, | |
| "grad_norm": 2.1948471069335938, | |
| "learning_rate": 4.2466222930024616e-05, | |
| "loss": 0.2969, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.3210161662817552, | |
| "grad_norm": 3.0124406814575195, | |
| "learning_rate": 4.239398415742784e-05, | |
| "loss": 0.336, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 1.325635103926097, | |
| "grad_norm": 2.1753032207489014, | |
| "learning_rate": 4.232146278140366e-05, | |
| "loss": 0.3086, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 1.3302540415704387, | |
| "grad_norm": 1.9752156734466553, | |
| "learning_rate": 4.224865998022092e-05, | |
| "loss": 0.2998, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.3348729792147807, | |
| "grad_norm": 2.4908149242401123, | |
| "learning_rate": 4.2175576936720805e-05, | |
| "loss": 0.3096, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 1.3394919168591224, | |
| "grad_norm": 1.8155696392059326, | |
| "learning_rate": 4.2102214838297696e-05, | |
| "loss": 0.3046, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.3441108545034641, | |
| "grad_norm": 2.2324860095977783, | |
| "learning_rate": 4.202857487687977e-05, | |
| "loss": 0.3065, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.348729792147806, | |
| "grad_norm": 1.9542967081069946, | |
| "learning_rate": 4.1954658248909764e-05, | |
| "loss": 0.284, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 1.3533487297921478, | |
| "grad_norm": 1.6855888366699219, | |
| "learning_rate": 4.188046615532541e-05, | |
| "loss": 0.2928, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 1.3579676674364896, | |
| "grad_norm": 2.32239031791687, | |
| "learning_rate": 4.180599980154005e-05, | |
| "loss": 0.3303, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.3625866050808315, | |
| "grad_norm": 1.3265233039855957, | |
| "learning_rate": 4.173126039742292e-05, | |
| "loss": 0.2994, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.3672055427251733, | |
| "grad_norm": 1.8445942401885986, | |
| "learning_rate": 4.16562491572796e-05, | |
| "loss": 0.2883, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.371824480369515, | |
| "grad_norm": 1.4618576765060425, | |
| "learning_rate": 4.158096729983222e-05, | |
| "loss": 0.3107, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.376443418013857, | |
| "grad_norm": 1.7024414539337158, | |
| "learning_rate": 4.150541604819969e-05, | |
| "loss": 0.2596, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.3810623556581987, | |
| "grad_norm": 1.2643704414367676, | |
| "learning_rate": 4.142959662987783e-05, | |
| "loss": 0.311, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.3856812933025404, | |
| "grad_norm": 2.466792345046997, | |
| "learning_rate": 4.1353510276719386e-05, | |
| "loss": 0.2991, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 8660, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.252134482698568e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |