mini_qwen_sft_checkpoint-15461 / trainer_state.json
qiufengqijun's picture
上传权重
066ab31 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9998868183839282,
"eval_steps": 500,
"global_step": 15461,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002587008367355188,
"grad_norm": 3.117795467376709,
"learning_rate": 8.624407072013799e-08,
"loss": 3.7293,
"step": 20
},
{
"epoch": 0.005174016734710376,
"grad_norm": 3.1848440170288086,
"learning_rate": 1.7248814144027598e-07,
"loss": 3.7345,
"step": 40
},
{
"epoch": 0.007761025102065565,
"grad_norm": 3.073965072631836,
"learning_rate": 2.5873221216041403e-07,
"loss": 3.7232,
"step": 60
},
{
"epoch": 0.010348033469420752,
"grad_norm": 2.366281032562256,
"learning_rate": 3.4497628288055197e-07,
"loss": 3.6733,
"step": 80
},
{
"epoch": 0.012935041836775941,
"grad_norm": 1.8071706295013428,
"learning_rate": 4.3122035360069e-07,
"loss": 3.5801,
"step": 100
},
{
"epoch": 0.01552205020413113,
"grad_norm": 1.3455595970153809,
"learning_rate": 5.174644243208281e-07,
"loss": 3.4895,
"step": 120
},
{
"epoch": 0.018109058571486317,
"grad_norm": 0.9497399926185608,
"learning_rate": 6.03708495040966e-07,
"loss": 3.3785,
"step": 140
},
{
"epoch": 0.020696066938841504,
"grad_norm": 0.9015834331512451,
"learning_rate": 6.899525657611039e-07,
"loss": 3.3095,
"step": 160
},
{
"epoch": 0.023283075306196693,
"grad_norm": 0.8537250757217407,
"learning_rate": 7.761966364812419e-07,
"loss": 3.1901,
"step": 180
},
{
"epoch": 0.025870083673551883,
"grad_norm": 0.8126162886619568,
"learning_rate": 8.6244070720138e-07,
"loss": 3.1323,
"step": 200
},
{
"epoch": 0.02845709204090707,
"grad_norm": 0.8042150139808655,
"learning_rate": 9.48684777921518e-07,
"loss": 3.0969,
"step": 220
},
{
"epoch": 0.03104410040826226,
"grad_norm": 0.7180504202842712,
"learning_rate": 1.0349288486416561e-06,
"loss": 3.0284,
"step": 240
},
{
"epoch": 0.033631108775617445,
"grad_norm": 0.7540543675422668,
"learning_rate": 1.1211729193617941e-06,
"loss": 3.043,
"step": 260
},
{
"epoch": 0.036218117142972635,
"grad_norm": 0.720626175403595,
"learning_rate": 1.207416990081932e-06,
"loss": 3.002,
"step": 280
},
{
"epoch": 0.038805125510327824,
"grad_norm": 0.761584997177124,
"learning_rate": 1.29366106080207e-06,
"loss": 3.0019,
"step": 300
},
{
"epoch": 0.04139213387768301,
"grad_norm": 0.7120524644851685,
"learning_rate": 1.3799051315222079e-06,
"loss": 2.9614,
"step": 320
},
{
"epoch": 0.0439791422450382,
"grad_norm": 0.6938987374305725,
"learning_rate": 1.4661492022423459e-06,
"loss": 2.9435,
"step": 340
},
{
"epoch": 0.04656615061239339,
"grad_norm": 0.7532190680503845,
"learning_rate": 1.5523932729624839e-06,
"loss": 2.9194,
"step": 360
},
{
"epoch": 0.049153158979748576,
"grad_norm": 0.7144018411636353,
"learning_rate": 1.6386373436826219e-06,
"loss": 2.909,
"step": 380
},
{
"epoch": 0.051740167347103766,
"grad_norm": 0.7019342184066772,
"learning_rate": 1.72488141440276e-06,
"loss": 2.8771,
"step": 400
},
{
"epoch": 0.05432717571445895,
"grad_norm": 0.7003944516181946,
"learning_rate": 1.8111254851228978e-06,
"loss": 2.892,
"step": 420
},
{
"epoch": 0.05691418408181414,
"grad_norm": 0.7309603095054626,
"learning_rate": 1.897369555843036e-06,
"loss": 2.8603,
"step": 440
},
{
"epoch": 0.05950119244916933,
"grad_norm": 0.6825790405273438,
"learning_rate": 1.983613626563174e-06,
"loss": 2.8596,
"step": 460
},
{
"epoch": 0.06208820081652452,
"grad_norm": 0.6874191164970398,
"learning_rate": 2.0698576972833122e-06,
"loss": 2.8482,
"step": 480
},
{
"epoch": 0.0646752091838797,
"grad_norm": 0.7430766224861145,
"learning_rate": 2.15610176800345e-06,
"loss": 2.8446,
"step": 500
},
{
"epoch": 0.06726221755123489,
"grad_norm": 0.7108047008514404,
"learning_rate": 2.2423458387235882e-06,
"loss": 2.828,
"step": 520
},
{
"epoch": 0.06984922591859008,
"grad_norm": 0.7121191620826721,
"learning_rate": 2.328589909443726e-06,
"loss": 2.8002,
"step": 540
},
{
"epoch": 0.07243623428594527,
"grad_norm": 0.7068054676055908,
"learning_rate": 2.414833980163864e-06,
"loss": 2.8119,
"step": 560
},
{
"epoch": 0.07502324265330046,
"grad_norm": 0.6924973726272583,
"learning_rate": 2.501078050884002e-06,
"loss": 2.7846,
"step": 580
},
{
"epoch": 0.07761025102065565,
"grad_norm": 0.7206343412399292,
"learning_rate": 2.58732212160414e-06,
"loss": 2.7896,
"step": 600
},
{
"epoch": 0.08019725938801084,
"grad_norm": 0.7203959822654724,
"learning_rate": 2.673566192324278e-06,
"loss": 2.7793,
"step": 620
},
{
"epoch": 0.08278426775536601,
"grad_norm": 0.7108203172683716,
"learning_rate": 2.7598102630444157e-06,
"loss": 2.7642,
"step": 640
},
{
"epoch": 0.0853712761227212,
"grad_norm": 0.7053371667861938,
"learning_rate": 2.846054333764554e-06,
"loss": 2.7588,
"step": 660
},
{
"epoch": 0.0879582844900764,
"grad_norm": 0.6986986398696899,
"learning_rate": 2.9322984044846917e-06,
"loss": 2.7453,
"step": 680
},
{
"epoch": 0.09054529285743158,
"grad_norm": 0.7130378484725952,
"learning_rate": 3.01854247520483e-06,
"loss": 2.7249,
"step": 700
},
{
"epoch": 0.09313230122478677,
"grad_norm": 0.7068546414375305,
"learning_rate": 3.1047865459249677e-06,
"loss": 2.7134,
"step": 720
},
{
"epoch": 0.09571930959214196,
"grad_norm": 0.7224396467208862,
"learning_rate": 3.191030616645106e-06,
"loss": 2.718,
"step": 740
},
{
"epoch": 0.09830631795949715,
"grad_norm": 0.7299512624740601,
"learning_rate": 3.2772746873652437e-06,
"loss": 2.7049,
"step": 760
},
{
"epoch": 0.10089332632685234,
"grad_norm": 0.7126713991165161,
"learning_rate": 3.363518758085382e-06,
"loss": 2.7199,
"step": 780
},
{
"epoch": 0.10348033469420753,
"grad_norm": 0.6644526720046997,
"learning_rate": 3.44976282880552e-06,
"loss": 2.7046,
"step": 800
},
{
"epoch": 0.10606734306156272,
"grad_norm": 0.7410451769828796,
"learning_rate": 3.536006899525658e-06,
"loss": 2.7012,
"step": 820
},
{
"epoch": 0.1086543514289179,
"grad_norm": 0.7218043208122253,
"learning_rate": 3.6222509702457957e-06,
"loss": 2.71,
"step": 840
},
{
"epoch": 0.11124135979627309,
"grad_norm": 0.6941894292831421,
"learning_rate": 3.708495040965934e-06,
"loss": 2.7035,
"step": 860
},
{
"epoch": 0.11382836816362828,
"grad_norm": 0.6824482679367065,
"learning_rate": 3.794739111686072e-06,
"loss": 2.6741,
"step": 880
},
{
"epoch": 0.11641537653098347,
"grad_norm": 0.7253878712654114,
"learning_rate": 3.88098318240621e-06,
"loss": 2.6831,
"step": 900
},
{
"epoch": 0.11900238489833866,
"grad_norm": 0.6949134469032288,
"learning_rate": 3.967227253126348e-06,
"loss": 2.6674,
"step": 920
},
{
"epoch": 0.12158939326569385,
"grad_norm": 0.7096832394599915,
"learning_rate": 4.053471323846485e-06,
"loss": 2.674,
"step": 940
},
{
"epoch": 0.12417640163304904,
"grad_norm": 0.6938119530677795,
"learning_rate": 4.1397153945666245e-06,
"loss": 2.6533,
"step": 960
},
{
"epoch": 0.12676341000040423,
"grad_norm": 0.6781166195869446,
"learning_rate": 4.225959465286762e-06,
"loss": 2.6426,
"step": 980
},
{
"epoch": 0.1293504183677594,
"grad_norm": 0.7600920796394348,
"learning_rate": 4.3122035360069e-06,
"loss": 2.6454,
"step": 1000
},
{
"epoch": 0.1319374267351146,
"grad_norm": 0.6884315609931946,
"learning_rate": 4.398447606727037e-06,
"loss": 2.6324,
"step": 1020
},
{
"epoch": 0.13452443510246978,
"grad_norm": 0.7092069387435913,
"learning_rate": 4.4846916774471764e-06,
"loss": 2.6181,
"step": 1040
},
{
"epoch": 0.13711144346982498,
"grad_norm": 0.729020357131958,
"learning_rate": 4.570935748167314e-06,
"loss": 2.6217,
"step": 1060
},
{
"epoch": 0.13969845183718016,
"grad_norm": 0.7104108333587646,
"learning_rate": 4.657179818887452e-06,
"loss": 2.6152,
"step": 1080
},
{
"epoch": 0.14228546020453534,
"grad_norm": 0.8092931509017944,
"learning_rate": 4.743423889607589e-06,
"loss": 2.621,
"step": 1100
},
{
"epoch": 0.14487246857189054,
"grad_norm": 0.6620950102806091,
"learning_rate": 4.829667960327728e-06,
"loss": 2.6166,
"step": 1120
},
{
"epoch": 0.14745947693924571,
"grad_norm": 0.6813467741012573,
"learning_rate": 4.915912031047866e-06,
"loss": 2.6171,
"step": 1140
},
{
"epoch": 0.15004648530660092,
"grad_norm": 0.7140293121337891,
"learning_rate": 5.002156101768004e-06,
"loss": 2.5911,
"step": 1160
},
{
"epoch": 0.1526334936739561,
"grad_norm": 0.7278040051460266,
"learning_rate": 5.088400172488141e-06,
"loss": 2.6055,
"step": 1180
},
{
"epoch": 0.1552205020413113,
"grad_norm": 0.7283148169517517,
"learning_rate": 5.17464424320828e-06,
"loss": 2.6054,
"step": 1200
},
{
"epoch": 0.15780751040866647,
"grad_norm": 0.7109535932540894,
"learning_rate": 5.260888313928419e-06,
"loss": 2.56,
"step": 1220
},
{
"epoch": 0.16039451877602168,
"grad_norm": 0.7203260064125061,
"learning_rate": 5.347132384648556e-06,
"loss": 2.578,
"step": 1240
},
{
"epoch": 0.16298152714337685,
"grad_norm": 0.7385180592536926,
"learning_rate": 5.433376455368694e-06,
"loss": 2.5829,
"step": 1260
},
{
"epoch": 0.16556853551073203,
"grad_norm": 0.7511777281761169,
"learning_rate": 5.5196205260888315e-06,
"loss": 2.5703,
"step": 1280
},
{
"epoch": 0.16815554387808723,
"grad_norm": 0.7461130619049072,
"learning_rate": 5.60586459680897e-06,
"loss": 2.5891,
"step": 1300
},
{
"epoch": 0.1707425522454424,
"grad_norm": 0.7192751169204712,
"learning_rate": 5.692108667529108e-06,
"loss": 2.552,
"step": 1320
},
{
"epoch": 0.1733295606127976,
"grad_norm": 0.7672246694564819,
"learning_rate": 5.778352738249245e-06,
"loss": 2.5451,
"step": 1340
},
{
"epoch": 0.1759165689801528,
"grad_norm": 0.8286859393119812,
"learning_rate": 5.8645968089693835e-06,
"loss": 2.5691,
"step": 1360
},
{
"epoch": 0.178503577347508,
"grad_norm": 0.8903458714485168,
"learning_rate": 5.9508408796895225e-06,
"loss": 2.5717,
"step": 1380
},
{
"epoch": 0.18109058571486317,
"grad_norm": 0.7192072868347168,
"learning_rate": 6.03708495040966e-06,
"loss": 2.547,
"step": 1400
},
{
"epoch": 0.18367759408221837,
"grad_norm": 0.7470182776451111,
"learning_rate": 6.123329021129798e-06,
"loss": 2.5309,
"step": 1420
},
{
"epoch": 0.18626460244957355,
"grad_norm": 0.7186440229415894,
"learning_rate": 6.2095730918499354e-06,
"loss": 2.5433,
"step": 1440
},
{
"epoch": 0.18885161081692872,
"grad_norm": 0.7359221577644348,
"learning_rate": 6.295817162570074e-06,
"loss": 2.5482,
"step": 1460
},
{
"epoch": 0.19143861918428393,
"grad_norm": 0.7859694957733154,
"learning_rate": 6.382061233290212e-06,
"loss": 2.5371,
"step": 1480
},
{
"epoch": 0.1940256275516391,
"grad_norm": 0.7339861392974854,
"learning_rate": 6.468305304010349e-06,
"loss": 2.5217,
"step": 1500
},
{
"epoch": 0.1966126359189943,
"grad_norm": 0.7527260780334473,
"learning_rate": 6.554549374730487e-06,
"loss": 2.5183,
"step": 1520
},
{
"epoch": 0.19919964428634948,
"grad_norm": 0.7345518469810486,
"learning_rate": 6.6407934454506265e-06,
"loss": 2.5249,
"step": 1540
},
{
"epoch": 0.20178665265370468,
"grad_norm": 0.736298680305481,
"learning_rate": 6.727037516170764e-06,
"loss": 2.5184,
"step": 1560
},
{
"epoch": 0.20437366102105986,
"grad_norm": 0.7677698135375977,
"learning_rate": 6.813281586890902e-06,
"loss": 2.5134,
"step": 1580
},
{
"epoch": 0.20696066938841506,
"grad_norm": 0.7627900838851929,
"learning_rate": 6.89952565761104e-06,
"loss": 2.5064,
"step": 1600
},
{
"epoch": 0.20954767775577024,
"grad_norm": 0.7290985584259033,
"learning_rate": 6.985769728331178e-06,
"loss": 2.5124,
"step": 1620
},
{
"epoch": 0.21213468612312544,
"grad_norm": 0.7347148060798645,
"learning_rate": 7.072013799051316e-06,
"loss": 2.4901,
"step": 1640
},
{
"epoch": 0.21472169449048062,
"grad_norm": 0.7257357239723206,
"learning_rate": 7.158257869771453e-06,
"loss": 2.4884,
"step": 1660
},
{
"epoch": 0.2173087028578358,
"grad_norm": 0.7803710699081421,
"learning_rate": 7.244501940491591e-06,
"loss": 2.4899,
"step": 1680
},
{
"epoch": 0.219895711225191,
"grad_norm": 0.6987377405166626,
"learning_rate": 7.33074601121173e-06,
"loss": 2.5001,
"step": 1700
},
{
"epoch": 0.22248271959254617,
"grad_norm": 0.6989637017250061,
"learning_rate": 7.416990081931868e-06,
"loss": 2.4941,
"step": 1720
},
{
"epoch": 0.22506972795990138,
"grad_norm": 0.7391577363014221,
"learning_rate": 7.503234152652006e-06,
"loss": 2.4811,
"step": 1740
},
{
"epoch": 0.22765673632725655,
"grad_norm": 0.7664337754249573,
"learning_rate": 7.589478223372144e-06,
"loss": 2.4806,
"step": 1760
},
{
"epoch": 0.23024374469461176,
"grad_norm": 0.7150381207466125,
"learning_rate": 7.675722294092282e-06,
"loss": 2.4674,
"step": 1780
},
{
"epoch": 0.23283075306196693,
"grad_norm": 0.7978541254997253,
"learning_rate": 7.76196636481242e-06,
"loss": 2.4633,
"step": 1800
},
{
"epoch": 0.23541776142932214,
"grad_norm": 0.7218653559684753,
"learning_rate": 7.848210435532557e-06,
"loss": 2.4589,
"step": 1820
},
{
"epoch": 0.2380047697966773,
"grad_norm": 0.723008930683136,
"learning_rate": 7.934454506252696e-06,
"loss": 2.467,
"step": 1840
},
{
"epoch": 0.2405917781640325,
"grad_norm": 0.7375757098197937,
"learning_rate": 8.020698576972833e-06,
"loss": 2.4812,
"step": 1860
},
{
"epoch": 0.2431787865313877,
"grad_norm": 0.7501986026763916,
"learning_rate": 8.10694264769297e-06,
"loss": 2.4587,
"step": 1880
},
{
"epoch": 0.24576579489874287,
"grad_norm": 0.7394606471061707,
"learning_rate": 8.19318671841311e-06,
"loss": 2.4514,
"step": 1900
},
{
"epoch": 0.24835280326609807,
"grad_norm": 0.7856109738349915,
"learning_rate": 8.279430789133249e-06,
"loss": 2.45,
"step": 1920
},
{
"epoch": 0.25093981163345325,
"grad_norm": 0.7603466510772705,
"learning_rate": 8.365674859853386e-06,
"loss": 2.4341,
"step": 1940
},
{
"epoch": 0.25352682000080845,
"grad_norm": 0.7223484516143799,
"learning_rate": 8.451918930573524e-06,
"loss": 2.4349,
"step": 1960
},
{
"epoch": 0.25611382836816365,
"grad_norm": 0.7488518357276917,
"learning_rate": 8.538163001293663e-06,
"loss": 2.4417,
"step": 1980
},
{
"epoch": 0.2587008367355188,
"grad_norm": 0.83389812707901,
"learning_rate": 8.6244070720138e-06,
"loss": 2.4266,
"step": 2000
},
{
"epoch": 0.261287845102874,
"grad_norm": 0.7574110627174377,
"learning_rate": 8.710651142733937e-06,
"loss": 2.4471,
"step": 2020
},
{
"epoch": 0.2638748534702292,
"grad_norm": 0.7482550740242004,
"learning_rate": 8.796895213454075e-06,
"loss": 2.4368,
"step": 2040
},
{
"epoch": 0.26646186183758436,
"grad_norm": 0.7341257333755493,
"learning_rate": 8.883139284174214e-06,
"loss": 2.4303,
"step": 2060
},
{
"epoch": 0.26904887020493956,
"grad_norm": 0.764855682849884,
"learning_rate": 8.969383354894353e-06,
"loss": 2.4176,
"step": 2080
},
{
"epoch": 0.27163587857229476,
"grad_norm": 0.7559799551963806,
"learning_rate": 9.055627425614489e-06,
"loss": 2.4088,
"step": 2100
},
{
"epoch": 0.27422288693964997,
"grad_norm": 0.7435436844825745,
"learning_rate": 9.141871496334628e-06,
"loss": 2.4241,
"step": 2120
},
{
"epoch": 0.2768098953070051,
"grad_norm": 0.7100109457969666,
"learning_rate": 9.228115567054767e-06,
"loss": 2.4122,
"step": 2140
},
{
"epoch": 0.2793969036743603,
"grad_norm": 0.7357873320579529,
"learning_rate": 9.314359637774904e-06,
"loss": 2.425,
"step": 2160
},
{
"epoch": 0.2819839120417155,
"grad_norm": 0.7281599640846252,
"learning_rate": 9.400603708495041e-06,
"loss": 2.4242,
"step": 2180
},
{
"epoch": 0.28457092040907067,
"grad_norm": 0.7816882729530334,
"learning_rate": 9.486847779215179e-06,
"loss": 2.4115,
"step": 2200
},
{
"epoch": 0.2871579287764259,
"grad_norm": 0.7402174472808838,
"learning_rate": 9.573091849935318e-06,
"loss": 2.4253,
"step": 2220
},
{
"epoch": 0.2897449371437811,
"grad_norm": 0.7378434538841248,
"learning_rate": 9.659335920655457e-06,
"loss": 2.4015,
"step": 2240
},
{
"epoch": 0.2923319455111363,
"grad_norm": 0.7792133688926697,
"learning_rate": 9.745579991375592e-06,
"loss": 2.3947,
"step": 2260
},
{
"epoch": 0.29491895387849143,
"grad_norm": 0.7180085778236389,
"learning_rate": 9.831824062095732e-06,
"loss": 2.3975,
"step": 2280
},
{
"epoch": 0.29750596224584663,
"grad_norm": 0.7536414861679077,
"learning_rate": 9.91806813281587e-06,
"loss": 2.4052,
"step": 2300
},
{
"epoch": 0.30009297061320184,
"grad_norm": 0.7593511939048767,
"learning_rate": 9.999999943356089e-06,
"loss": 2.3856,
"step": 2320
},
{
"epoch": 0.30267997898055704,
"grad_norm": 0.6879323720932007,
"learning_rate": 9.999975020055695e-06,
"loss": 2.4107,
"step": 2340
},
{
"epoch": 0.3052669873479122,
"grad_norm": 0.7193155884742737,
"learning_rate": 9.999904781886476e-06,
"loss": 2.3872,
"step": 2360
},
{
"epoch": 0.3078539957152674,
"grad_norm": 0.7621170282363892,
"learning_rate": 9.999789229485002e-06,
"loss": 2.3844,
"step": 2380
},
{
"epoch": 0.3104410040826226,
"grad_norm": 0.7357332110404968,
"learning_rate": 9.999628363898525e-06,
"loss": 2.3863,
"step": 2400
},
{
"epoch": 0.31302801244997774,
"grad_norm": 0.753011167049408,
"learning_rate": 9.999422186584978e-06,
"loss": 2.3807,
"step": 2420
},
{
"epoch": 0.31561502081733295,
"grad_norm": 0.7547248005867004,
"learning_rate": 9.999170699412942e-06,
"loss": 2.3691,
"step": 2440
},
{
"epoch": 0.31820202918468815,
"grad_norm": 0.7450150847434998,
"learning_rate": 9.998873904661655e-06,
"loss": 2.3495,
"step": 2460
},
{
"epoch": 0.32078903755204335,
"grad_norm": 0.7460102438926697,
"learning_rate": 9.998531805020974e-06,
"loss": 2.3582,
"step": 2480
},
{
"epoch": 0.3233760459193985,
"grad_norm": 0.7281093597412109,
"learning_rate": 9.998144403591352e-06,
"loss": 2.3559,
"step": 2500
},
{
"epoch": 0.3259630542867537,
"grad_norm": 0.7635049223899841,
"learning_rate": 9.99771170388382e-06,
"loss": 2.3668,
"step": 2520
},
{
"epoch": 0.3285500626541089,
"grad_norm": 0.7504300475120544,
"learning_rate": 9.997233709819935e-06,
"loss": 2.3642,
"step": 2540
},
{
"epoch": 0.33113707102146406,
"grad_norm": 0.73244309425354,
"learning_rate": 9.996710425731776e-06,
"loss": 2.3634,
"step": 2560
},
{
"epoch": 0.33372407938881926,
"grad_norm": 0.816100537776947,
"learning_rate": 9.996141856361871e-06,
"loss": 2.3405,
"step": 2580
},
{
"epoch": 0.33631108775617446,
"grad_norm": 0.7125614285469055,
"learning_rate": 9.995528006863175e-06,
"loss": 2.3594,
"step": 2600
},
{
"epoch": 0.33889809612352967,
"grad_norm": 0.7901700735092163,
"learning_rate": 9.994868882799022e-06,
"loss": 2.3636,
"step": 2620
},
{
"epoch": 0.3414851044908848,
"grad_norm": 0.7499126195907593,
"learning_rate": 9.994164490143062e-06,
"loss": 2.36,
"step": 2640
},
{
"epoch": 0.34407211285824,
"grad_norm": 0.78279048204422,
"learning_rate": 9.99341483527922e-06,
"loss": 2.3685,
"step": 2660
},
{
"epoch": 0.3466591212255952,
"grad_norm": 0.7683995366096497,
"learning_rate": 9.992619925001632e-06,
"loss": 2.3532,
"step": 2680
},
{
"epoch": 0.3492461295929504,
"grad_norm": 0.7231781482696533,
"learning_rate": 9.991779766514586e-06,
"loss": 2.377,
"step": 2700
},
{
"epoch": 0.3518331379603056,
"grad_norm": 0.7408224940299988,
"learning_rate": 9.990894367432453e-06,
"loss": 2.3406,
"step": 2720
},
{
"epoch": 0.3544201463276608,
"grad_norm": 0.7654145956039429,
"learning_rate": 9.989963735779623e-06,
"loss": 2.3627,
"step": 2740
},
{
"epoch": 0.357007154695016,
"grad_norm": 0.7166858315467834,
"learning_rate": 9.988987879990428e-06,
"loss": 2.3343,
"step": 2760
},
{
"epoch": 0.35959416306237113,
"grad_norm": 0.79310142993927,
"learning_rate": 9.987966808909069e-06,
"loss": 2.3521,
"step": 2780
},
{
"epoch": 0.36218117142972633,
"grad_norm": 0.7726064324378967,
"learning_rate": 9.98690053178953e-06,
"loss": 2.3408,
"step": 2800
},
{
"epoch": 0.36476817979708154,
"grad_norm": 0.7428059577941895,
"learning_rate": 9.985789058295501e-06,
"loss": 2.3238,
"step": 2820
},
{
"epoch": 0.36735518816443674,
"grad_norm": 0.7808260321617126,
"learning_rate": 9.984632398500289e-06,
"loss": 2.3164,
"step": 2840
},
{
"epoch": 0.3699421965317919,
"grad_norm": 0.7835705280303955,
"learning_rate": 9.983430562886723e-06,
"loss": 2.343,
"step": 2860
},
{
"epoch": 0.3725292048991471,
"grad_norm": 0.7703275680541992,
"learning_rate": 9.982183562347063e-06,
"loss": 2.3354,
"step": 2880
},
{
"epoch": 0.3751162132665023,
"grad_norm": 0.7307262420654297,
"learning_rate": 9.980891408182897e-06,
"loss": 2.3445,
"step": 2900
},
{
"epoch": 0.37770322163385744,
"grad_norm": 0.6830443143844604,
"learning_rate": 9.979554112105045e-06,
"loss": 2.3437,
"step": 2920
},
{
"epoch": 0.38029023000121265,
"grad_norm": 0.7646154761314392,
"learning_rate": 9.978171686233445e-06,
"loss": 2.3316,
"step": 2940
},
{
"epoch": 0.38287723836856785,
"grad_norm": 0.7598294615745544,
"learning_rate": 9.97674414309705e-06,
"loss": 2.3135,
"step": 2960
},
{
"epoch": 0.38546424673592306,
"grad_norm": 0.7368634343147278,
"learning_rate": 9.975271495633709e-06,
"loss": 2.3189,
"step": 2980
},
{
"epoch": 0.3880512551032782,
"grad_norm": 0.7349269390106201,
"learning_rate": 9.973753757190057e-06,
"loss": 2.307,
"step": 3000
},
{
"epoch": 0.3906382634706334,
"grad_norm": 0.7512723803520203,
"learning_rate": 9.972190941521382e-06,
"loss": 2.3195,
"step": 3020
},
{
"epoch": 0.3932252718379886,
"grad_norm": 0.7430227398872375,
"learning_rate": 9.970583062791517e-06,
"loss": 2.3087,
"step": 3040
},
{
"epoch": 0.3958122802053438,
"grad_norm": 0.7576317191123962,
"learning_rate": 9.968930135572694e-06,
"loss": 2.3,
"step": 3060
},
{
"epoch": 0.39839928857269896,
"grad_norm": 0.7394402623176575,
"learning_rate": 9.967232174845426e-06,
"loss": 2.3164,
"step": 3080
},
{
"epoch": 0.40098629694005417,
"grad_norm": 0.7061188220977783,
"learning_rate": 9.965489195998363e-06,
"loss": 2.3187,
"step": 3100
},
{
"epoch": 0.40357330530740937,
"grad_norm": 0.7332024574279785,
"learning_rate": 9.963701214828154e-06,
"loss": 2.3085,
"step": 3120
},
{
"epoch": 0.4061603136747645,
"grad_norm": 0.663925051689148,
"learning_rate": 9.961868247539308e-06,
"loss": 2.2978,
"step": 3140
},
{
"epoch": 0.4087473220421197,
"grad_norm": 0.7256314158439636,
"learning_rate": 9.959990310744042e-06,
"loss": 2.3209,
"step": 3160
},
{
"epoch": 0.4113343304094749,
"grad_norm": 0.711919903755188,
"learning_rate": 9.958067421462133e-06,
"loss": 2.3041,
"step": 3180
},
{
"epoch": 0.41392133877683013,
"grad_norm": 0.714450478553772,
"learning_rate": 9.956099597120762e-06,
"loss": 2.2789,
"step": 3200
},
{
"epoch": 0.4165083471441853,
"grad_norm": 0.7216628193855286,
"learning_rate": 9.95408685555436e-06,
"loss": 2.3056,
"step": 3220
},
{
"epoch": 0.4190953555115405,
"grad_norm": 0.6629658341407776,
"learning_rate": 9.952029215004441e-06,
"loss": 2.3001,
"step": 3240
},
{
"epoch": 0.4216823638788957,
"grad_norm": 0.8131959438323975,
"learning_rate": 9.949926694119443e-06,
"loss": 2.2881,
"step": 3260
},
{
"epoch": 0.4242693722462509,
"grad_norm": 0.7133468389511108,
"learning_rate": 9.94777931195455e-06,
"loss": 2.2902,
"step": 3280
},
{
"epoch": 0.42685638061360603,
"grad_norm": 0.7825130224227905,
"learning_rate": 9.945587087971529e-06,
"loss": 2.2829,
"step": 3300
},
{
"epoch": 0.42944338898096124,
"grad_norm": 0.6878073811531067,
"learning_rate": 9.943350042038545e-06,
"loss": 2.2844,
"step": 3320
},
{
"epoch": 0.43203039734831644,
"grad_norm": 0.7324455976486206,
"learning_rate": 9.941068194429992e-06,
"loss": 2.2932,
"step": 3340
},
{
"epoch": 0.4346174057156716,
"grad_norm": 0.7330353856086731,
"learning_rate": 9.938741565826295e-06,
"loss": 2.2611,
"step": 3360
},
{
"epoch": 0.4372044140830268,
"grad_norm": 0.8474377393722534,
"learning_rate": 9.936370177313737e-06,
"loss": 2.2938,
"step": 3380
},
{
"epoch": 0.439791422450382,
"grad_norm": 0.776228666305542,
"learning_rate": 9.933954050384253e-06,
"loss": 2.292,
"step": 3400
},
{
"epoch": 0.4423784308177372,
"grad_norm": 0.7257934212684631,
"learning_rate": 9.931493206935249e-06,
"loss": 2.2964,
"step": 3420
},
{
"epoch": 0.44496543918509235,
"grad_norm": 0.7738561034202576,
"learning_rate": 9.928987669269397e-06,
"loss": 2.2585,
"step": 3440
},
{
"epoch": 0.44755244755244755,
"grad_norm": 0.6888708472251892,
"learning_rate": 9.926437460094431e-06,
"loss": 2.281,
"step": 3460
},
{
"epoch": 0.45013945591980276,
"grad_norm": 0.7120715975761414,
"learning_rate": 9.923842602522948e-06,
"loss": 2.2861,
"step": 3480
},
{
"epoch": 0.4527264642871579,
"grad_norm": 0.770353376865387,
"learning_rate": 9.92120312007219e-06,
"loss": 2.271,
"step": 3500
},
{
"epoch": 0.4553134726545131,
"grad_norm": 0.753852367401123,
"learning_rate": 9.918519036663835e-06,
"loss": 2.2592,
"step": 3520
},
{
"epoch": 0.4579004810218683,
"grad_norm": 0.7479904294013977,
"learning_rate": 9.915790376623785e-06,
"loss": 2.2691,
"step": 3540
},
{
"epoch": 0.4604874893892235,
"grad_norm": 0.7162041068077087,
"learning_rate": 9.913017164681936e-06,
"loss": 2.2637,
"step": 3560
},
{
"epoch": 0.46307449775657866,
"grad_norm": 0.6859399080276489,
"learning_rate": 9.91019942597196e-06,
"loss": 2.2424,
"step": 3580
},
{
"epoch": 0.46566150612393387,
"grad_norm": 0.7681860327720642,
"learning_rate": 9.907337186031078e-06,
"loss": 2.2758,
"step": 3600
},
{
"epoch": 0.46824851449128907,
"grad_norm": 0.7416828274726868,
"learning_rate": 9.904430470799826e-06,
"loss": 2.2715,
"step": 3620
},
{
"epoch": 0.4708355228586443,
"grad_norm": 0.7957201600074768,
"learning_rate": 9.901479306621818e-06,
"loss": 2.277,
"step": 3640
},
{
"epoch": 0.4734225312259994,
"grad_norm": 0.71152663230896,
"learning_rate": 9.89848372024351e-06,
"loss": 2.2831,
"step": 3660
},
{
"epoch": 0.4760095395933546,
"grad_norm": 0.6950436234474182,
"learning_rate": 9.89544373881396e-06,
"loss": 2.2678,
"step": 3680
},
{
"epoch": 0.47859654796070983,
"grad_norm": 0.7226356267929077,
"learning_rate": 9.89235938988458e-06,
"loss": 2.2549,
"step": 3700
},
{
"epoch": 0.481183556328065,
"grad_norm": 0.7032837867736816,
"learning_rate": 9.88923070140888e-06,
"loss": 2.2669,
"step": 3720
},
{
"epoch": 0.4837705646954202,
"grad_norm": 0.749229371547699,
"learning_rate": 9.886057701742222e-06,
"loss": 2.2421,
"step": 3740
},
{
"epoch": 0.4863575730627754,
"grad_norm": 0.7166919112205505,
"learning_rate": 9.882840419641566e-06,
"loss": 2.245,
"step": 3760
},
{
"epoch": 0.4889445814301306,
"grad_norm": 0.7276502847671509,
"learning_rate": 9.879578884265198e-06,
"loss": 2.2542,
"step": 3780
},
{
"epoch": 0.49153158979748574,
"grad_norm": 0.7321860194206238,
"learning_rate": 9.876273125172476e-06,
"loss": 2.2395,
"step": 3800
},
{
"epoch": 0.49411859816484094,
"grad_norm": 0.6959764361381531,
"learning_rate": 9.872923172323559e-06,
"loss": 2.2588,
"step": 3820
},
{
"epoch": 0.49670560653219614,
"grad_norm": 0.7237563729286194,
"learning_rate": 9.869529056079133e-06,
"loss": 2.2463,
"step": 3840
},
{
"epoch": 0.4992926148995513,
"grad_norm": 0.7536144256591797,
"learning_rate": 9.866090807200135e-06,
"loss": 2.2394,
"step": 3860
},
{
"epoch": 0.5018796232669065,
"grad_norm": 0.6835715770721436,
"learning_rate": 9.862608456847484e-06,
"loss": 2.2447,
"step": 3880
},
{
"epoch": 0.5044666316342616,
"grad_norm": 0.7627236247062683,
"learning_rate": 9.859082036581787e-06,
"loss": 2.2727,
"step": 3900
},
{
"epoch": 0.5070536400016169,
"grad_norm": 0.7260853052139282,
"learning_rate": 9.855511578363057e-06,
"loss": 2.2373,
"step": 3920
},
{
"epoch": 0.509640648368972,
"grad_norm": 0.7366577386856079,
"learning_rate": 9.851897114550423e-06,
"loss": 2.2583,
"step": 3940
},
{
"epoch": 0.5122276567363273,
"grad_norm": 0.718189001083374,
"learning_rate": 9.848238677901844e-06,
"loss": 2.2376,
"step": 3960
},
{
"epoch": 0.5148146651036825,
"grad_norm": 0.6988586783409119,
"learning_rate": 9.844536301573798e-06,
"loss": 2.2413,
"step": 3980
},
{
"epoch": 0.5174016734710376,
"grad_norm": 0.752648115158081,
"learning_rate": 9.840790019120993e-06,
"loss": 2.2346,
"step": 4000
},
{
"epoch": 0.5199886818383929,
"grad_norm": 0.8000548481941223,
"learning_rate": 9.836999864496058e-06,
"loss": 2.2365,
"step": 4020
},
{
"epoch": 0.522575690205748,
"grad_norm": 0.7324941158294678,
"learning_rate": 9.833165872049235e-06,
"loss": 2.2294,
"step": 4040
},
{
"epoch": 0.5251626985731032,
"grad_norm": 0.765869140625,
"learning_rate": 9.829288076528071e-06,
"loss": 2.2418,
"step": 4060
},
{
"epoch": 0.5277497069404584,
"grad_norm": 0.7222108244895935,
"learning_rate": 9.825366513077104e-06,
"loss": 2.2292,
"step": 4080
},
{
"epoch": 0.5303367153078136,
"grad_norm": 0.7359380722045898,
"learning_rate": 9.821401217237535e-06,
"loss": 2.2491,
"step": 4100
},
{
"epoch": 0.5329237236751687,
"grad_norm": 0.7288945317268372,
"learning_rate": 9.817392224946916e-06,
"loss": 2.2404,
"step": 4120
},
{
"epoch": 0.535510732042524,
"grad_norm": 0.7518214583396912,
"learning_rate": 9.813339572538822e-06,
"loss": 2.2381,
"step": 4140
},
{
"epoch": 0.5380977404098791,
"grad_norm": 0.753818154335022,
"learning_rate": 9.809243296742516e-06,
"loss": 2.2236,
"step": 4160
},
{
"epoch": 0.5406847487772344,
"grad_norm": 0.7281918525695801,
"learning_rate": 9.805103434682628e-06,
"loss": 2.2142,
"step": 4180
},
{
"epoch": 0.5432717571445895,
"grad_norm": 0.7605194449424744,
"learning_rate": 9.800920023878803e-06,
"loss": 2.2326,
"step": 4200
},
{
"epoch": 0.5458587655119447,
"grad_norm": 0.7266237139701843,
"learning_rate": 9.796693102245376e-06,
"loss": 2.2144,
"step": 4220
},
{
"epoch": 0.5484457738792999,
"grad_norm": 0.7575150728225708,
"learning_rate": 9.792422708091014e-06,
"loss": 2.2282,
"step": 4240
},
{
"epoch": 0.5510327822466551,
"grad_norm": 0.7097237706184387,
"learning_rate": 9.788108880118383e-06,
"loss": 2.2139,
"step": 4260
},
{
"epoch": 0.5536197906140102,
"grad_norm": 0.7074447870254517,
"learning_rate": 9.783751657423787e-06,
"loss": 2.2169,
"step": 4280
},
{
"epoch": 0.5562067989813655,
"grad_norm": 0.7193735241889954,
"learning_rate": 9.779351079496821e-06,
"loss": 2.2435,
"step": 4300
},
{
"epoch": 0.5587938073487206,
"grad_norm": 0.7104501128196716,
"learning_rate": 9.774907186220005e-06,
"loss": 2.198,
"step": 4320
},
{
"epoch": 0.5613808157160758,
"grad_norm": 0.6711876392364502,
"learning_rate": 9.770420017868426e-06,
"loss": 2.1927,
"step": 4340
},
{
"epoch": 0.563967824083431,
"grad_norm": 0.7272083163261414,
"learning_rate": 9.765889615109379e-06,
"loss": 2.2437,
"step": 4360
},
{
"epoch": 0.5665548324507862,
"grad_norm": 0.7311263680458069,
"learning_rate": 9.761316019001991e-06,
"loss": 2.2106,
"step": 4380
},
{
"epoch": 0.5691418408181413,
"grad_norm": 0.7337877154350281,
"learning_rate": 9.756699270996848e-06,
"loss": 2.2257,
"step": 4400
},
{
"epoch": 0.5717288491854966,
"grad_norm": 0.7337206602096558,
"learning_rate": 9.752039412935627e-06,
"loss": 2.2066,
"step": 4420
},
{
"epoch": 0.5743158575528517,
"grad_norm": 0.8056983947753906,
"learning_rate": 9.74733648705071e-06,
"loss": 2.2049,
"step": 4440
},
{
"epoch": 0.576902865920207,
"grad_norm": 0.7203987240791321,
"learning_rate": 9.742590535964805e-06,
"loss": 2.2279,
"step": 4460
},
{
"epoch": 0.5794898742875622,
"grad_norm": 0.7495117783546448,
"learning_rate": 9.737801602690554e-06,
"loss": 2.1986,
"step": 4480
},
{
"epoch": 0.5820768826549173,
"grad_norm": 0.7343236207962036,
"learning_rate": 9.732969730630153e-06,
"loss": 2.2233,
"step": 4500
},
{
"epoch": 0.5846638910222726,
"grad_norm": 0.7727257609367371,
"learning_rate": 9.728094963574948e-06,
"loss": 2.2042,
"step": 4520
},
{
"epoch": 0.5872508993896277,
"grad_norm": 0.7064406275749207,
"learning_rate": 9.723177345705048e-06,
"loss": 2.2295,
"step": 4540
},
{
"epoch": 0.5898379077569829,
"grad_norm": 0.7306511998176575,
"learning_rate": 9.71821692158892e-06,
"loss": 2.1982,
"step": 4560
},
{
"epoch": 0.5924249161243381,
"grad_norm": 0.7717193365097046,
"learning_rate": 9.713213736182976e-06,
"loss": 2.206,
"step": 4580
},
{
"epoch": 0.5950119244916933,
"grad_norm": 0.7180883288383484,
"learning_rate": 9.708167834831183e-06,
"loss": 2.2044,
"step": 4600
},
{
"epoch": 0.5975989328590484,
"grad_norm": 0.7017503380775452,
"learning_rate": 9.703079263264643e-06,
"loss": 2.192,
"step": 4620
},
{
"epoch": 0.6001859412264037,
"grad_norm": 0.7421537637710571,
"learning_rate": 9.697948067601176e-06,
"loss": 2.199,
"step": 4640
},
{
"epoch": 0.6027729495937588,
"grad_norm": 0.7729679942131042,
"learning_rate": 9.692774294344905e-06,
"loss": 2.2073,
"step": 4660
},
{
"epoch": 0.6053599579611141,
"grad_norm": 0.7517857551574707,
"learning_rate": 9.687557990385836e-06,
"loss": 2.1942,
"step": 4680
},
{
"epoch": 0.6079469663284692,
"grad_norm": 0.7258435487747192,
"learning_rate": 9.682299202999433e-06,
"loss": 2.1916,
"step": 4700
},
{
"epoch": 0.6105339746958244,
"grad_norm": 0.7256997227668762,
"learning_rate": 9.676997979846183e-06,
"loss": 2.1986,
"step": 4720
},
{
"epoch": 0.6131209830631796,
"grad_norm": 0.7594742774963379,
"learning_rate": 9.671654368971176e-06,
"loss": 2.1867,
"step": 4740
},
{
"epoch": 0.6157079914305348,
"grad_norm": 0.7546527981758118,
"learning_rate": 9.666268418803655e-06,
"loss": 2.1999,
"step": 4760
},
{
"epoch": 0.6182949997978899,
"grad_norm": 0.7679339051246643,
"learning_rate": 9.660840178156592e-06,
"loss": 2.1964,
"step": 4780
},
{
"epoch": 0.6208820081652452,
"grad_norm": 0.7719926834106445,
"learning_rate": 9.655369696226235e-06,
"loss": 2.1829,
"step": 4800
},
{
"epoch": 0.6234690165326003,
"grad_norm": 0.7456310987472534,
"learning_rate": 9.649857022591664e-06,
"loss": 2.1725,
"step": 4820
},
{
"epoch": 0.6260560248999555,
"grad_norm": 0.720140814781189,
"learning_rate": 9.644302207214346e-06,
"loss": 2.1806,
"step": 4840
},
{
"epoch": 0.6286430332673107,
"grad_norm": 0.7672199010848999,
"learning_rate": 9.63870530043768e-06,
"loss": 2.1921,
"step": 4860
},
{
"epoch": 0.6312300416346659,
"grad_norm": 0.7888000011444092,
"learning_rate": 9.633066352986538e-06,
"loss": 2.198,
"step": 4880
},
{
"epoch": 0.6338170500020212,
"grad_norm": 0.7024748921394348,
"learning_rate": 9.627385415966807e-06,
"loss": 2.1579,
"step": 4900
},
{
"epoch": 0.6364040583693763,
"grad_norm": 0.7126362919807434,
"learning_rate": 9.62166254086493e-06,
"loss": 2.1882,
"step": 4920
},
{
"epoch": 0.6389910667367315,
"grad_norm": 0.6748985052108765,
"learning_rate": 9.61589777954743e-06,
"loss": 2.1931,
"step": 4940
},
{
"epoch": 0.6415780751040867,
"grad_norm": 0.7744324207305908,
"learning_rate": 9.61009118426045e-06,
"loss": 2.1823,
"step": 4960
},
{
"epoch": 0.6441650834714419,
"grad_norm": 0.6999046802520752,
"learning_rate": 9.604242807629275e-06,
"loss": 2.1598,
"step": 4980
},
{
"epoch": 0.646752091838797,
"grad_norm": 0.7126173973083496,
"learning_rate": 9.59835270265785e-06,
"loss": 2.1661,
"step": 5000
},
{
"epoch": 0.6493391002061523,
"grad_norm": 0.7425235509872437,
"learning_rate": 9.592420922728312e-06,
"loss": 2.1729,
"step": 5020
},
{
"epoch": 0.6519261085735074,
"grad_norm": 0.735068142414093,
"learning_rate": 9.586447521600496e-06,
"loss": 2.1734,
"step": 5040
},
{
"epoch": 0.6545131169408626,
"grad_norm": 0.772686779499054,
"learning_rate": 9.580432553411446e-06,
"loss": 2.1943,
"step": 5060
},
{
"epoch": 0.6571001253082178,
"grad_norm": 0.8012945652008057,
"learning_rate": 9.574376072674936e-06,
"loss": 2.1678,
"step": 5080
},
{
"epoch": 0.659687133675573,
"grad_norm": 0.7072063684463501,
"learning_rate": 9.568278134280966e-06,
"loss": 2.1749,
"step": 5100
},
{
"epoch": 0.6622741420429281,
"grad_norm": 0.687140941619873,
"learning_rate": 9.562138793495268e-06,
"loss": 2.1844,
"step": 5120
},
{
"epoch": 0.6648611504102834,
"grad_norm": 0.7809275984764099,
"learning_rate": 9.555958105958805e-06,
"loss": 2.1719,
"step": 5140
},
{
"epoch": 0.6674481587776385,
"grad_norm": 0.7361642122268677,
"learning_rate": 9.549736127687265e-06,
"loss": 2.1706,
"step": 5160
},
{
"epoch": 0.6700351671449938,
"grad_norm": 0.7150685787200928,
"learning_rate": 9.543472915070555e-06,
"loss": 2.1648,
"step": 5180
},
{
"epoch": 0.6726221755123489,
"grad_norm": 0.7491324543952942,
"learning_rate": 9.537168524872292e-06,
"loss": 2.156,
"step": 5200
},
{
"epoch": 0.6752091838797041,
"grad_norm": 0.7028157711029053,
"learning_rate": 9.530823014229283e-06,
"loss": 2.1588,
"step": 5220
},
{
"epoch": 0.6777961922470593,
"grad_norm": 0.7949670553207397,
"learning_rate": 9.52443644065101e-06,
"loss": 2.1788,
"step": 5240
},
{
"epoch": 0.6803832006144145,
"grad_norm": 0.7003277540206909,
"learning_rate": 9.518008862019116e-06,
"loss": 2.1917,
"step": 5260
},
{
"epoch": 0.6829702089817696,
"grad_norm": 0.7252909541130066,
"learning_rate": 9.511540336586864e-06,
"loss": 2.1709,
"step": 5280
},
{
"epoch": 0.6855572173491249,
"grad_norm": 0.7193975448608398,
"learning_rate": 9.505030922978626e-06,
"loss": 2.185,
"step": 5300
},
{
"epoch": 0.68814422571648,
"grad_norm": 0.7502670288085938,
"learning_rate": 9.49848068018934e-06,
"loss": 2.1787,
"step": 5320
},
{
"epoch": 0.6907312340838352,
"grad_norm": 0.7460989952087402,
"learning_rate": 9.49188966758398e-06,
"loss": 2.1557,
"step": 5340
},
{
"epoch": 0.6933182424511904,
"grad_norm": 0.710442066192627,
"learning_rate": 9.485257944897021e-06,
"loss": 2.1547,
"step": 5360
},
{
"epoch": 0.6959052508185456,
"grad_norm": 0.7363094091415405,
"learning_rate": 9.478585572231891e-06,
"loss": 2.1473,
"step": 5380
},
{
"epoch": 0.6984922591859009,
"grad_norm": 0.7203119397163391,
"learning_rate": 9.47187261006043e-06,
"loss": 2.1743,
"step": 5400
},
{
"epoch": 0.701079267553256,
"grad_norm": 0.715248167514801,
"learning_rate": 9.465119119222346e-06,
"loss": 2.1626,
"step": 5420
},
{
"epoch": 0.7036662759206112,
"grad_norm": 0.7098533511161804,
"learning_rate": 9.458325160924648e-06,
"loss": 2.1807,
"step": 5440
},
{
"epoch": 0.7062532842879664,
"grad_norm": 0.7714352607727051,
"learning_rate": 9.451490796741117e-06,
"loss": 2.1517,
"step": 5460
},
{
"epoch": 0.7088402926553216,
"grad_norm": 0.720448911190033,
"learning_rate": 9.444616088611718e-06,
"loss": 2.1775,
"step": 5480
},
{
"epoch": 0.7114273010226767,
"grad_norm": 0.7580015659332275,
"learning_rate": 9.437701098842067e-06,
"loss": 2.1529,
"step": 5500
},
{
"epoch": 0.714014309390032,
"grad_norm": 0.727325439453125,
"learning_rate": 9.430745890102849e-06,
"loss": 2.1654,
"step": 5520
},
{
"epoch": 0.7166013177573871,
"grad_norm": 0.7433022260665894,
"learning_rate": 9.423750525429248e-06,
"loss": 2.1565,
"step": 5540
},
{
"epoch": 0.7191883261247423,
"grad_norm": 0.728881299495697,
"learning_rate": 9.416715068220393e-06,
"loss": 2.1734,
"step": 5560
},
{
"epoch": 0.7217753344920975,
"grad_norm": 0.7467840909957886,
"learning_rate": 9.409639582238761e-06,
"loss": 2.1502,
"step": 5580
},
{
"epoch": 0.7243623428594527,
"grad_norm": 0.7820594310760498,
"learning_rate": 9.40252413160962e-06,
"loss": 2.1725,
"step": 5600
},
{
"epoch": 0.7269493512268079,
"grad_norm": 0.7077836394309998,
"learning_rate": 9.395368780820433e-06,
"loss": 2.1508,
"step": 5620
},
{
"epoch": 0.7295363595941631,
"grad_norm": 0.7221850752830505,
"learning_rate": 9.388173594720283e-06,
"loss": 2.1649,
"step": 5640
},
{
"epoch": 0.7321233679615182,
"grad_norm": 0.7874971628189087,
"learning_rate": 9.380938638519274e-06,
"loss": 2.158,
"step": 5660
},
{
"epoch": 0.7347103763288735,
"grad_norm": 0.7347155213356018,
"learning_rate": 9.373663977787956e-06,
"loss": 2.1486,
"step": 5680
},
{
"epoch": 0.7372973846962286,
"grad_norm": 0.7876786589622498,
"learning_rate": 9.366349678456717e-06,
"loss": 2.1501,
"step": 5700
},
{
"epoch": 0.7398843930635838,
"grad_norm": 0.678352952003479,
"learning_rate": 9.35899580681519e-06,
"loss": 2.1644,
"step": 5720
},
{
"epoch": 0.742471401430939,
"grad_norm": 0.7279735803604126,
"learning_rate": 9.351602429511655e-06,
"loss": 2.1508,
"step": 5740
},
{
"epoch": 0.7450584097982942,
"grad_norm": 0.7041738629341125,
"learning_rate": 9.344169613552428e-06,
"loss": 2.1617,
"step": 5760
},
{
"epoch": 0.7476454181656493,
"grad_norm": 0.750642716884613,
"learning_rate": 9.336697426301267e-06,
"loss": 2.1409,
"step": 5780
},
{
"epoch": 0.7502324265330046,
"grad_norm": 0.7688580751419067,
"learning_rate": 9.329185935478741e-06,
"loss": 2.1459,
"step": 5800
},
{
"epoch": 0.7528194349003597,
"grad_norm": 0.704578697681427,
"learning_rate": 9.321635209161642e-06,
"loss": 2.1417,
"step": 5820
},
{
"epoch": 0.7554064432677149,
"grad_norm": 0.733323872089386,
"learning_rate": 9.314045315782339e-06,
"loss": 2.1516,
"step": 5840
},
{
"epoch": 0.7579934516350701,
"grad_norm": 0.771930992603302,
"learning_rate": 9.306416324128184e-06,
"loss": 2.1256,
"step": 5860
},
{
"epoch": 0.7605804600024253,
"grad_norm": 0.6910899877548218,
"learning_rate": 9.298748303340871e-06,
"loss": 2.1421,
"step": 5880
},
{
"epoch": 0.7631674683697806,
"grad_norm": 0.7119818329811096,
"learning_rate": 9.291041322915824e-06,
"loss": 2.1631,
"step": 5900
},
{
"epoch": 0.7657544767371357,
"grad_norm": 0.6817770600318909,
"learning_rate": 9.283295452701549e-06,
"loss": 2.143,
"step": 5920
},
{
"epoch": 0.7683414851044909,
"grad_norm": 0.714541494846344,
"learning_rate": 9.275510762899016e-06,
"loss": 2.1546,
"step": 5940
},
{
"epoch": 0.7709284934718461,
"grad_norm": 0.6969371438026428,
"learning_rate": 9.267687324061016e-06,
"loss": 2.1186,
"step": 5960
},
{
"epoch": 0.7735155018392013,
"grad_norm": 0.7325617671012878,
"learning_rate": 9.259825207091526e-06,
"loss": 2.1363,
"step": 5980
},
{
"epoch": 0.7761025102065564,
"grad_norm": 0.743165135383606,
"learning_rate": 9.25192448324506e-06,
"loss": 2.1269,
"step": 6000
},
{
"epoch": 0.7786895185739117,
"grad_norm": 0.7538695931434631,
"learning_rate": 9.243985224126031e-06,
"loss": 2.1237,
"step": 6020
},
{
"epoch": 0.7812765269412668,
"grad_norm": 0.7549970746040344,
"learning_rate": 9.236007501688094e-06,
"loss": 2.1343,
"step": 6040
},
{
"epoch": 0.783863535308622,
"grad_norm": 0.732627272605896,
"learning_rate": 9.2279913882335e-06,
"loss": 2.1465,
"step": 6060
},
{
"epoch": 0.7864505436759772,
"grad_norm": 0.7083766460418701,
"learning_rate": 9.219936956412436e-06,
"loss": 2.1394,
"step": 6080
},
{
"epoch": 0.7890375520433324,
"grad_norm": 0.7239031195640564,
"learning_rate": 9.211844279222376e-06,
"loss": 2.1386,
"step": 6100
},
{
"epoch": 0.7916245604106876,
"grad_norm": 0.7016878128051758,
"learning_rate": 9.2037134300074e-06,
"loss": 2.1313,
"step": 6120
},
{
"epoch": 0.7942115687780428,
"grad_norm": 0.6846844553947449,
"learning_rate": 9.195544482457555e-06,
"loss": 2.1383,
"step": 6140
},
{
"epoch": 0.7967985771453979,
"grad_norm": 0.7442426085472107,
"learning_rate": 9.187337510608168e-06,
"loss": 2.1177,
"step": 6160
},
{
"epoch": 0.7993855855127532,
"grad_norm": 0.7443544268608093,
"learning_rate": 9.179092588839178e-06,
"loss": 2.1391,
"step": 6180
},
{
"epoch": 0.8019725938801083,
"grad_norm": 0.7053080201148987,
"learning_rate": 9.170809791874468e-06,
"loss": 2.1221,
"step": 6200
},
{
"epoch": 0.8045596022474635,
"grad_norm": 0.7665020227432251,
"learning_rate": 9.16248919478119e-06,
"loss": 2.1116,
"step": 6220
},
{
"epoch": 0.8071466106148187,
"grad_norm": 0.7906709909439087,
"learning_rate": 9.154130872969067e-06,
"loss": 2.1363,
"step": 6240
},
{
"epoch": 0.8097336189821739,
"grad_norm": 0.6853694915771484,
"learning_rate": 9.145734902189733e-06,
"loss": 2.149,
"step": 6260
},
{
"epoch": 0.812320627349529,
"grad_norm": 0.7411865592002869,
"learning_rate": 9.137301358536032e-06,
"loss": 2.1355,
"step": 6280
},
{
"epoch": 0.8149076357168843,
"grad_norm": 0.8013186454772949,
"learning_rate": 9.128830318441327e-06,
"loss": 2.1175,
"step": 6300
},
{
"epoch": 0.8174946440842394,
"grad_norm": 0.7919278144836426,
"learning_rate": 9.120321858678817e-06,
"loss": 2.128,
"step": 6320
},
{
"epoch": 0.8200816524515947,
"grad_norm": 0.7355955243110657,
"learning_rate": 9.111776056360838e-06,
"loss": 2.1253,
"step": 6340
},
{
"epoch": 0.8226686608189498,
"grad_norm": 0.7750183343887329,
"learning_rate": 9.103192988938155e-06,
"loss": 2.1225,
"step": 6360
},
{
"epoch": 0.825255669186305,
"grad_norm": 0.7022154927253723,
"learning_rate": 9.094572734199271e-06,
"loss": 2.1193,
"step": 6380
},
{
"epoch": 0.8278426775536603,
"grad_norm": 0.7672535181045532,
"learning_rate": 9.085915370269723e-06,
"loss": 2.1188,
"step": 6400
},
{
"epoch": 0.8304296859210154,
"grad_norm": 0.7107143998146057,
"learning_rate": 9.077220975611363e-06,
"loss": 2.1278,
"step": 6420
},
{
"epoch": 0.8330166942883706,
"grad_norm": 0.7380732297897339,
"learning_rate": 9.068489629021655e-06,
"loss": 2.1374,
"step": 6440
},
{
"epoch": 0.8356037026557258,
"grad_norm": 0.6959198117256165,
"learning_rate": 9.05972140963296e-06,
"loss": 2.129,
"step": 6460
},
{
"epoch": 0.838190711023081,
"grad_norm": 0.742127001285553,
"learning_rate": 9.050916396911818e-06,
"loss": 2.1285,
"step": 6480
},
{
"epoch": 0.8407777193904361,
"grad_norm": 0.6720155477523804,
"learning_rate": 9.042074670658223e-06,
"loss": 2.1172,
"step": 6500
},
{
"epoch": 0.8433647277577914,
"grad_norm": 0.7020736932754517,
"learning_rate": 9.033196311004915e-06,
"loss": 2.1036,
"step": 6520
},
{
"epoch": 0.8459517361251465,
"grad_norm": 0.746281087398529,
"learning_rate": 9.024281398416632e-06,
"loss": 2.1183,
"step": 6540
},
{
"epoch": 0.8485387444925018,
"grad_norm": 0.7673128843307495,
"learning_rate": 9.015330013689396e-06,
"loss": 2.1435,
"step": 6560
},
{
"epoch": 0.8511257528598569,
"grad_norm": 0.7635971903800964,
"learning_rate": 9.006342237949782e-06,
"loss": 2.1018,
"step": 6580
},
{
"epoch": 0.8537127612272121,
"grad_norm": 0.7365761995315552,
"learning_rate": 8.997318152654167e-06,
"loss": 2.116,
"step": 6600
},
{
"epoch": 0.8562997695945673,
"grad_norm": 0.7582866549491882,
"learning_rate": 8.988257839588011e-06,
"loss": 2.1146,
"step": 6620
},
{
"epoch": 0.8588867779619225,
"grad_norm": 0.7569854259490967,
"learning_rate": 8.979161380865104e-06,
"loss": 2.1156,
"step": 6640
},
{
"epoch": 0.8614737863292776,
"grad_norm": 0.6821621656417847,
"learning_rate": 8.970028858926825e-06,
"loss": 2.1134,
"step": 6660
},
{
"epoch": 0.8640607946966329,
"grad_norm": 0.7307649850845337,
"learning_rate": 8.96086035654139e-06,
"loss": 2.124,
"step": 6680
},
{
"epoch": 0.866647803063988,
"grad_norm": 0.7775338292121887,
"learning_rate": 8.951655956803118e-06,
"loss": 2.1052,
"step": 6700
},
{
"epoch": 0.8692348114313432,
"grad_norm": 0.7425001859664917,
"learning_rate": 8.942415743131651e-06,
"loss": 2.106,
"step": 6720
},
{
"epoch": 0.8718218197986984,
"grad_norm": 0.7379507422447205,
"learning_rate": 8.933139799271229e-06,
"loss": 2.0974,
"step": 6740
},
{
"epoch": 0.8744088281660536,
"grad_norm": 0.7932276129722595,
"learning_rate": 8.923828209289904e-06,
"loss": 2.1257,
"step": 6760
},
{
"epoch": 0.8769958365334087,
"grad_norm": 0.7237013578414917,
"learning_rate": 8.914481057578791e-06,
"loss": 2.1033,
"step": 6780
},
{
"epoch": 0.879582844900764,
"grad_norm": 0.7310687899589539,
"learning_rate": 8.905098428851309e-06,
"loss": 2.1194,
"step": 6800
},
{
"epoch": 0.8821698532681191,
"grad_norm": 0.6933685541152954,
"learning_rate": 8.8956804081424e-06,
"loss": 2.0891,
"step": 6820
},
{
"epoch": 0.8847568616354744,
"grad_norm": 0.7352440357208252,
"learning_rate": 8.886227080807762e-06,
"loss": 2.1022,
"step": 6840
},
{
"epoch": 0.8873438700028295,
"grad_norm": 0.7073638439178467,
"learning_rate": 8.876738532523081e-06,
"loss": 2.1065,
"step": 6860
},
{
"epoch": 0.8899308783701847,
"grad_norm": 0.7440693974494934,
"learning_rate": 8.867214849283252e-06,
"loss": 2.1079,
"step": 6880
},
{
"epoch": 0.89251788673754,
"grad_norm": 0.7651025056838989,
"learning_rate": 8.8576561174016e-06,
"loss": 2.121,
"step": 6900
},
{
"epoch": 0.8951048951048951,
"grad_norm": 0.7150782346725464,
"learning_rate": 8.84806242350909e-06,
"loss": 2.1025,
"step": 6920
},
{
"epoch": 0.8976919034722503,
"grad_norm": 0.7346411943435669,
"learning_rate": 8.838433854553555e-06,
"loss": 2.0959,
"step": 6940
},
{
"epoch": 0.9002789118396055,
"grad_norm": 0.7531387209892273,
"learning_rate": 8.828770497798897e-06,
"loss": 2.0901,
"step": 6960
},
{
"epoch": 0.9028659202069607,
"grad_norm": 0.7870126366615295,
"learning_rate": 8.819072440824303e-06,
"loss": 2.1034,
"step": 6980
},
{
"epoch": 0.9054529285743158,
"grad_norm": 0.7796097993850708,
"learning_rate": 8.80933977152345e-06,
"loss": 2.0825,
"step": 7000
},
{
"epoch": 0.9080399369416711,
"grad_norm": 0.7178963422775269,
"learning_rate": 8.799572578103703e-06,
"loss": 2.1252,
"step": 7020
},
{
"epoch": 0.9106269453090262,
"grad_norm": 0.7850649952888489,
"learning_rate": 8.789770949085321e-06,
"loss": 2.0886,
"step": 7040
},
{
"epoch": 0.9132139536763815,
"grad_norm": 0.7312331795692444,
"learning_rate": 8.779934973300657e-06,
"loss": 2.1015,
"step": 7060
},
{
"epoch": 0.9158009620437366,
"grad_norm": 0.7429481148719788,
"learning_rate": 8.770064739893346e-06,
"loss": 2.0791,
"step": 7080
},
{
"epoch": 0.9183879704110918,
"grad_norm": 0.70722895860672,
"learning_rate": 8.7601603383175e-06,
"loss": 2.0992,
"step": 7100
},
{
"epoch": 0.920974978778447,
"grad_norm": 0.7041098475456238,
"learning_rate": 8.750221858336902e-06,
"loss": 2.1027,
"step": 7120
},
{
"epoch": 0.9235619871458022,
"grad_norm": 0.7348500490188599,
"learning_rate": 8.740249390024183e-06,
"loss": 2.0899,
"step": 7140
},
{
"epoch": 0.9261489955131573,
"grad_norm": 0.7186647653579712,
"learning_rate": 8.730243023760012e-06,
"loss": 2.1059,
"step": 7160
},
{
"epoch": 0.9287360038805126,
"grad_norm": 0.7586842775344849,
"learning_rate": 8.720202850232281e-06,
"loss": 2.0929,
"step": 7180
},
{
"epoch": 0.9313230122478677,
"grad_norm": 0.7751206159591675,
"learning_rate": 8.710128960435271e-06,
"loss": 2.0961,
"step": 7200
},
{
"epoch": 0.9339100206152229,
"grad_norm": 0.7438105344772339,
"learning_rate": 8.700021445668839e-06,
"loss": 2.1013,
"step": 7220
},
{
"epoch": 0.9364970289825781,
"grad_norm": 0.7343020439147949,
"learning_rate": 8.68988039753758e-06,
"loss": 2.0904,
"step": 7240
},
{
"epoch": 0.9390840373499333,
"grad_norm": 0.7382264733314514,
"learning_rate": 8.67970590795001e-06,
"loss": 2.1016,
"step": 7260
},
{
"epoch": 0.9416710457172885,
"grad_norm": 0.720942497253418,
"learning_rate": 8.669498069117721e-06,
"loss": 2.1167,
"step": 7280
},
{
"epoch": 0.9442580540846437,
"grad_norm": 0.7112380266189575,
"learning_rate": 8.65925697355455e-06,
"loss": 2.0981,
"step": 7300
},
{
"epoch": 0.9468450624519988,
"grad_norm": 0.7193992137908936,
"learning_rate": 8.648982714075743e-06,
"loss": 2.0853,
"step": 7320
},
{
"epoch": 0.9494320708193541,
"grad_norm": 0.7250093221664429,
"learning_rate": 8.638675383797106e-06,
"loss": 2.0936,
"step": 7340
},
{
"epoch": 0.9520190791867092,
"grad_norm": 0.727990448474884,
"learning_rate": 8.628335076134173e-06,
"loss": 2.0882,
"step": 7360
},
{
"epoch": 0.9546060875540644,
"grad_norm": 0.7733245491981506,
"learning_rate": 8.617961884801346e-06,
"loss": 2.0799,
"step": 7380
},
{
"epoch": 0.9571930959214197,
"grad_norm": 0.7614827752113342,
"learning_rate": 8.60755590381106e-06,
"loss": 2.0935,
"step": 7400
},
{
"epoch": 0.9597801042887748,
"grad_norm": 0.7480348348617554,
"learning_rate": 8.597117227472915e-06,
"loss": 2.0889,
"step": 7420
},
{
"epoch": 0.96236711265613,
"grad_norm": 0.7127954959869385,
"learning_rate": 8.586645950392835e-06,
"loss": 2.0981,
"step": 7440
},
{
"epoch": 0.9649541210234852,
"grad_norm": 0.6996986865997314,
"learning_rate": 8.576142167472204e-06,
"loss": 2.0856,
"step": 7460
},
{
"epoch": 0.9675411293908404,
"grad_norm": 0.757079005241394,
"learning_rate": 8.565605973907006e-06,
"loss": 2.0873,
"step": 7480
},
{
"epoch": 0.9701281377581955,
"grad_norm": 0.7550200819969177,
"learning_rate": 8.555037465186962e-06,
"loss": 2.0817,
"step": 7500
},
{
"epoch": 0.9727151461255508,
"grad_norm": 0.7152219414710999,
"learning_rate": 8.544436737094672e-06,
"loss": 2.0962,
"step": 7520
},
{
"epoch": 0.9753021544929059,
"grad_norm": 0.774013876914978,
"learning_rate": 8.533803885704732e-06,
"loss": 2.0944,
"step": 7540
},
{
"epoch": 0.9778891628602612,
"grad_norm": 0.721843421459198,
"learning_rate": 8.523139007382881e-06,
"loss": 2.0849,
"step": 7560
},
{
"epoch": 0.9804761712276163,
"grad_norm": 0.7735098600387573,
"learning_rate": 8.51244219878511e-06,
"loss": 2.0906,
"step": 7580
},
{
"epoch": 0.9830631795949715,
"grad_norm": 0.7077389359474182,
"learning_rate": 8.501713556856803e-06,
"loss": 2.0665,
"step": 7600
},
{
"epoch": 0.9856501879623267,
"grad_norm": 0.7230038046836853,
"learning_rate": 8.490953178831846e-06,
"loss": 2.0766,
"step": 7620
},
{
"epoch": 0.9882371963296819,
"grad_norm": 0.7460753321647644,
"learning_rate": 8.480161162231747e-06,
"loss": 2.0556,
"step": 7640
},
{
"epoch": 0.990824204697037,
"grad_norm": 0.7653132081031799,
"learning_rate": 8.469337604864759e-06,
"loss": 2.0821,
"step": 7660
},
{
"epoch": 0.9934112130643923,
"grad_norm": 0.7243251204490662,
"learning_rate": 8.458482604824988e-06,
"loss": 2.0926,
"step": 7680
},
{
"epoch": 0.9959982214317474,
"grad_norm": 0.7776849865913391,
"learning_rate": 8.447596260491508e-06,
"loss": 2.0925,
"step": 7700
},
{
"epoch": 0.9985852297991026,
"grad_norm": 0.7500579953193665,
"learning_rate": 8.436678670527463e-06,
"loss": 2.0736,
"step": 7720
},
{
"epoch": 1.0011722381664578,
"grad_norm": 0.8922491669654846,
"learning_rate": 8.42572993387918e-06,
"loss": 2.0694,
"step": 7740
},
{
"epoch": 1.003759246533813,
"grad_norm": 0.7387255430221558,
"learning_rate": 8.414750149775269e-06,
"loss": 2.0478,
"step": 7760
},
{
"epoch": 1.0063462549011681,
"grad_norm": 0.7570685148239136,
"learning_rate": 8.403739417725718e-06,
"loss": 2.0594,
"step": 7780
},
{
"epoch": 1.0089332632685233,
"grad_norm": 0.7447026371955872,
"learning_rate": 8.392697837521006e-06,
"loss": 2.0535,
"step": 7800
},
{
"epoch": 1.0115202716358787,
"grad_norm": 0.7841023802757263,
"learning_rate": 8.381625509231185e-06,
"loss": 2.0437,
"step": 7820
},
{
"epoch": 1.0141072800032338,
"grad_norm": 0.7826112508773804,
"learning_rate": 8.37052253320497e-06,
"loss": 2.052,
"step": 7840
},
{
"epoch": 1.016694288370589,
"grad_norm": 0.7537607550621033,
"learning_rate": 8.359389010068852e-06,
"loss": 2.0726,
"step": 7860
},
{
"epoch": 1.019281296737944,
"grad_norm": 0.730627715587616,
"learning_rate": 8.34822504072616e-06,
"loss": 2.0649,
"step": 7880
},
{
"epoch": 1.0218683051052992,
"grad_norm": 0.7026150822639465,
"learning_rate": 8.337030726356162e-06,
"loss": 2.0518,
"step": 7900
},
{
"epoch": 1.0244553134726546,
"grad_norm": 0.7492315173149109,
"learning_rate": 8.32580616841314e-06,
"loss": 2.0681,
"step": 7920
},
{
"epoch": 1.0270423218400098,
"grad_norm": 0.7300956845283508,
"learning_rate": 8.314551468625478e-06,
"loss": 2.0589,
"step": 7940
},
{
"epoch": 1.029629330207365,
"grad_norm": 0.7457550764083862,
"learning_rate": 8.303266728994736e-06,
"loss": 2.0464,
"step": 7960
},
{
"epoch": 1.03221633857472,
"grad_norm": 0.8049952387809753,
"learning_rate": 8.291952051794721e-06,
"loss": 2.0398,
"step": 7980
},
{
"epoch": 1.0348033469420752,
"grad_norm": 0.7218441963195801,
"learning_rate": 8.28060753957057e-06,
"loss": 2.0454,
"step": 8000
},
{
"epoch": 1.0373903553094304,
"grad_norm": 0.7822824716567993,
"learning_rate": 8.269233295137811e-06,
"loss": 2.0577,
"step": 8020
},
{
"epoch": 1.0399773636767857,
"grad_norm": 0.7328481674194336,
"learning_rate": 8.257829421581442e-06,
"loss": 2.0363,
"step": 8040
},
{
"epoch": 1.0425643720441409,
"grad_norm": 0.7289636135101318,
"learning_rate": 8.246396022254983e-06,
"loss": 2.0604,
"step": 8060
},
{
"epoch": 1.045151380411496,
"grad_norm": 0.7583228349685669,
"learning_rate": 8.23493320077955e-06,
"loss": 2.0358,
"step": 8080
},
{
"epoch": 1.0477383887788512,
"grad_norm": 0.7434881329536438,
"learning_rate": 8.223441061042914e-06,
"loss": 2.042,
"step": 8100
},
{
"epoch": 1.0503253971462063,
"grad_norm": 0.7682307362556458,
"learning_rate": 8.211919707198555e-06,
"loss": 2.0631,
"step": 8120
},
{
"epoch": 1.0529124055135617,
"grad_norm": 0.7827048897743225,
"learning_rate": 8.200369243664722e-06,
"loss": 2.0541,
"step": 8140
},
{
"epoch": 1.0554994138809168,
"grad_norm": 0.7634522914886475,
"learning_rate": 8.188789775123487e-06,
"loss": 2.0401,
"step": 8160
},
{
"epoch": 1.058086422248272,
"grad_norm": 0.7598258852958679,
"learning_rate": 8.177181406519794e-06,
"loss": 2.0394,
"step": 8180
},
{
"epoch": 1.0606734306156271,
"grad_norm": 0.7735849022865295,
"learning_rate": 8.165544243060507e-06,
"loss": 2.0442,
"step": 8200
},
{
"epoch": 1.0632604389829823,
"grad_norm": 0.742406964302063,
"learning_rate": 8.153878390213464e-06,
"loss": 2.0365,
"step": 8220
},
{
"epoch": 1.0658474473503374,
"grad_norm": 0.7898341417312622,
"learning_rate": 8.142183953706506e-06,
"loss": 2.0591,
"step": 8240
},
{
"epoch": 1.0684344557176928,
"grad_norm": 0.7791109681129456,
"learning_rate": 8.130461039526538e-06,
"loss": 2.0352,
"step": 8260
},
{
"epoch": 1.071021464085048,
"grad_norm": 0.6895206570625305,
"learning_rate": 8.118709753918553e-06,
"loss": 2.0434,
"step": 8280
},
{
"epoch": 1.073608472452403,
"grad_norm": 0.7232005000114441,
"learning_rate": 8.106930203384675e-06,
"loss": 2.029,
"step": 8300
},
{
"epoch": 1.0761954808197582,
"grad_norm": 0.7908313870429993,
"learning_rate": 8.095122494683196e-06,
"loss": 2.0635,
"step": 8320
},
{
"epoch": 1.0787824891871134,
"grad_norm": 0.7397485971450806,
"learning_rate": 8.083286734827605e-06,
"loss": 2.0378,
"step": 8340
},
{
"epoch": 1.0813694975544688,
"grad_norm": 0.7352006435394287,
"learning_rate": 8.07142303108562e-06,
"loss": 2.0416,
"step": 8360
},
{
"epoch": 1.083956505921824,
"grad_norm": 0.7397055625915527,
"learning_rate": 8.059531490978212e-06,
"loss": 2.0494,
"step": 8380
},
{
"epoch": 1.086543514289179,
"grad_norm": 0.7299455404281616,
"learning_rate": 8.047612222278639e-06,
"loss": 2.0602,
"step": 8400
},
{
"epoch": 1.0891305226565342,
"grad_norm": 0.7375478744506836,
"learning_rate": 8.035665333011453e-06,
"loss": 2.0331,
"step": 8420
},
{
"epoch": 1.0917175310238894,
"grad_norm": 0.7608847618103027,
"learning_rate": 8.023690931451546e-06,
"loss": 2.0488,
"step": 8440
},
{
"epoch": 1.0943045393912445,
"grad_norm": 0.7570823431015015,
"learning_rate": 8.011689126123145e-06,
"loss": 2.0438,
"step": 8460
},
{
"epoch": 1.0968915477585999,
"grad_norm": 0.71818608045578,
"learning_rate": 7.99966002579884e-06,
"loss": 2.0358,
"step": 8480
},
{
"epoch": 1.099478556125955,
"grad_norm": 0.7376600503921509,
"learning_rate": 7.987603739498595e-06,
"loss": 2.0506,
"step": 8500
},
{
"epoch": 1.1020655644933102,
"grad_norm": 0.7798555493354797,
"learning_rate": 7.975520376488764e-06,
"loss": 2.0435,
"step": 8520
},
{
"epoch": 1.1046525728606653,
"grad_norm": 0.7355314493179321,
"learning_rate": 7.963410046281097e-06,
"loss": 2.0353,
"step": 8540
},
{
"epoch": 1.1072395812280205,
"grad_norm": 0.7146715521812439,
"learning_rate": 7.951272858631745e-06,
"loss": 2.0435,
"step": 8560
},
{
"epoch": 1.1098265895953756,
"grad_norm": 0.7580558657646179,
"learning_rate": 7.939108923540272e-06,
"loss": 2.0284,
"step": 8580
},
{
"epoch": 1.112413597962731,
"grad_norm": 0.7931268811225891,
"learning_rate": 7.926918351248654e-06,
"loss": 2.0384,
"step": 8600
},
{
"epoch": 1.1150006063300861,
"grad_norm": 0.7423412203788757,
"learning_rate": 7.914701252240278e-06,
"loss": 2.0302,
"step": 8620
},
{
"epoch": 1.1175876146974413,
"grad_norm": 0.7477630972862244,
"learning_rate": 7.902457737238946e-06,
"loss": 2.0271,
"step": 8640
},
{
"epoch": 1.1201746230647964,
"grad_norm": 0.7558584809303284,
"learning_rate": 7.89018791720787e-06,
"loss": 2.0456,
"step": 8660
},
{
"epoch": 1.1227616314321516,
"grad_norm": 0.7822209596633911,
"learning_rate": 7.877891903348658e-06,
"loss": 2.0122,
"step": 8680
},
{
"epoch": 1.125348639799507,
"grad_norm": 0.7526381611824036,
"learning_rate": 7.865569807100321e-06,
"loss": 2.0487,
"step": 8700
},
{
"epoch": 1.127935648166862,
"grad_norm": 0.7704318761825562,
"learning_rate": 7.85322174013825e-06,
"loss": 2.0251,
"step": 8720
},
{
"epoch": 1.1305226565342172,
"grad_norm": 0.725405752658844,
"learning_rate": 7.840847814373213e-06,
"loss": 2.0538,
"step": 8740
},
{
"epoch": 1.1331096649015724,
"grad_norm": 0.7347708940505981,
"learning_rate": 7.828448141950334e-06,
"loss": 2.0307,
"step": 8760
},
{
"epoch": 1.1356966732689275,
"grad_norm": 0.7331801652908325,
"learning_rate": 7.81602283524808e-06,
"loss": 2.0442,
"step": 8780
},
{
"epoch": 1.138283681636283,
"grad_norm": 0.7498618364334106,
"learning_rate": 7.803572006877241e-06,
"loss": 2.0394,
"step": 8800
},
{
"epoch": 1.140870690003638,
"grad_norm": 0.8074409365653992,
"learning_rate": 7.791095769679915e-06,
"loss": 2.0308,
"step": 8820
},
{
"epoch": 1.1434576983709932,
"grad_norm": 0.7502855062484741,
"learning_rate": 7.778594236728474e-06,
"loss": 2.0259,
"step": 8840
},
{
"epoch": 1.1460447067383484,
"grad_norm": 0.7721311450004578,
"learning_rate": 7.766067521324552e-06,
"loss": 2.0348,
"step": 8860
},
{
"epoch": 1.1486317151057035,
"grad_norm": 0.8015813231468201,
"learning_rate": 7.753515736998008e-06,
"loss": 2.0174,
"step": 8880
},
{
"epoch": 1.1512187234730586,
"grad_norm": 0.7582475543022156,
"learning_rate": 7.740938997505902e-06,
"loss": 2.0228,
"step": 8900
},
{
"epoch": 1.153805731840414,
"grad_norm": 0.7484523057937622,
"learning_rate": 7.728337416831462e-06,
"loss": 2.0324,
"step": 8920
},
{
"epoch": 1.1563927402077692,
"grad_norm": 0.7139847874641418,
"learning_rate": 7.715711109183057e-06,
"loss": 2.0391,
"step": 8940
},
{
"epoch": 1.1589797485751243,
"grad_norm": 0.7828614711761475,
"learning_rate": 7.703060188993152e-06,
"loss": 2.0075,
"step": 8960
},
{
"epoch": 1.1615667569424795,
"grad_norm": 0.7387328743934631,
"learning_rate": 7.690384770917279e-06,
"loss": 2.0245,
"step": 8980
},
{
"epoch": 1.1641537653098346,
"grad_norm": 0.7702174186706543,
"learning_rate": 7.677684969832991e-06,
"loss": 2.0268,
"step": 9000
},
{
"epoch": 1.1667407736771898,
"grad_norm": 0.7316907048225403,
"learning_rate": 7.664960900838827e-06,
"loss": 2.0245,
"step": 9020
},
{
"epoch": 1.1693277820445451,
"grad_norm": 0.7085281610488892,
"learning_rate": 7.65221267925327e-06,
"loss": 2.0175,
"step": 9040
},
{
"epoch": 1.1719147904119003,
"grad_norm": 0.7116014361381531,
"learning_rate": 7.639440420613691e-06,
"loss": 2.0249,
"step": 9060
},
{
"epoch": 1.1745017987792554,
"grad_norm": 0.7048114538192749,
"learning_rate": 7.626644240675316e-06,
"loss": 2.0059,
"step": 9080
},
{
"epoch": 1.1770888071466106,
"grad_norm": 0.7138361930847168,
"learning_rate": 7.613824255410165e-06,
"loss": 2.0408,
"step": 9100
},
{
"epoch": 1.1796758155139657,
"grad_norm": 0.7742594480514526,
"learning_rate": 7.6009805810060126e-06,
"loss": 2.0221,
"step": 9120
},
{
"epoch": 1.182262823881321,
"grad_norm": 0.8150427341461182,
"learning_rate": 7.5881133338653215e-06,
"loss": 2.0189,
"step": 9140
},
{
"epoch": 1.1848498322486762,
"grad_norm": 0.7460334300994873,
"learning_rate": 7.575222630604198e-06,
"loss": 2.0397,
"step": 9160
},
{
"epoch": 1.1874368406160314,
"grad_norm": 0.7613769769668579,
"learning_rate": 7.56230858805133e-06,
"loss": 2.0386,
"step": 9180
},
{
"epoch": 1.1900238489833865,
"grad_norm": 0.7651441693305969,
"learning_rate": 7.549371323246934e-06,
"loss": 2.0433,
"step": 9200
},
{
"epoch": 1.1926108573507417,
"grad_norm": 0.810110330581665,
"learning_rate": 7.536410953441685e-06,
"loss": 2.0286,
"step": 9220
},
{
"epoch": 1.195197865718097,
"grad_norm": 0.7386390566825867,
"learning_rate": 7.523427596095663e-06,
"loss": 2.0062,
"step": 9240
},
{
"epoch": 1.1977848740854522,
"grad_norm": 0.7521648406982422,
"learning_rate": 7.510421368877281e-06,
"loss": 2.0072,
"step": 9260
},
{
"epoch": 1.2003718824528073,
"grad_norm": 0.7483949661254883,
"learning_rate": 7.497392389662225e-06,
"loss": 2.0281,
"step": 9280
},
{
"epoch": 1.2029588908201625,
"grad_norm": 0.7583929300308228,
"learning_rate": 7.484340776532382e-06,
"loss": 2.0402,
"step": 9300
},
{
"epoch": 1.2055458991875176,
"grad_norm": 0.8147648572921753,
"learning_rate": 7.471266647774771e-06,
"loss": 2.013,
"step": 9320
},
{
"epoch": 1.2081329075548728,
"grad_norm": 0.7282470464706421,
"learning_rate": 7.458170121880468e-06,
"loss": 2.0124,
"step": 9340
},
{
"epoch": 1.210719915922228,
"grad_norm": 0.7271668314933777,
"learning_rate": 7.445051317543538e-06,
"loss": 2.0059,
"step": 9360
},
{
"epoch": 1.2133069242895833,
"grad_norm": 0.7665330171585083,
"learning_rate": 7.4319103536599536e-06,
"loss": 2.0307,
"step": 9380
},
{
"epoch": 1.2158939326569385,
"grad_norm": 0.7498813271522522,
"learning_rate": 7.418747349326523e-06,
"loss": 2.0086,
"step": 9400
},
{
"epoch": 1.2184809410242936,
"grad_norm": 0.7305240035057068,
"learning_rate": 7.4055624238398014e-06,
"loss": 2.0019,
"step": 9420
},
{
"epoch": 1.2210679493916488,
"grad_norm": 0.7301153540611267,
"learning_rate": 7.392355696695023e-06,
"loss": 2.0062,
"step": 9440
},
{
"epoch": 1.223654957759004,
"grad_norm": 0.740593433380127,
"learning_rate": 7.379127287585005e-06,
"loss": 2.0055,
"step": 9460
},
{
"epoch": 1.2262419661263593,
"grad_norm": 0.7132686972618103,
"learning_rate": 7.365877316399072e-06,
"loss": 2.0229,
"step": 9480
},
{
"epoch": 1.2288289744937144,
"grad_norm": 0.7279082536697388,
"learning_rate": 7.3526059032219655e-06,
"loss": 2.009,
"step": 9500
},
{
"epoch": 1.2314159828610696,
"grad_norm": 0.7989427447319031,
"learning_rate": 7.339313168332756e-06,
"loss": 2.0126,
"step": 9520
},
{
"epoch": 1.2340029912284247,
"grad_norm": 0.7120375633239746,
"learning_rate": 7.3259992322037525e-06,
"loss": 2.0057,
"step": 9540
},
{
"epoch": 1.2365899995957799,
"grad_norm": 0.8398416638374329,
"learning_rate": 7.312664215499411e-06,
"loss": 2.0028,
"step": 9560
},
{
"epoch": 1.2391770079631352,
"grad_norm": 0.7388201355934143,
"learning_rate": 7.299308239075243e-06,
"loss": 2.0111,
"step": 9580
},
{
"epoch": 1.2417640163304904,
"grad_norm": 0.707332193851471,
"learning_rate": 7.285931423976717e-06,
"loss": 1.9917,
"step": 9600
},
{
"epoch": 1.2443510246978455,
"grad_norm": 0.7708361744880676,
"learning_rate": 7.272533891438164e-06,
"loss": 2.0042,
"step": 9620
},
{
"epoch": 1.2469380330652007,
"grad_norm": 0.7365394830703735,
"learning_rate": 7.259115762881673e-06,
"loss": 2.0061,
"step": 9640
},
{
"epoch": 1.2495250414325558,
"grad_norm": 0.7249080538749695,
"learning_rate": 7.245677159916002e-06,
"loss": 2.014,
"step": 9660
},
{
"epoch": 1.2521120497999112,
"grad_norm": 0.7352914810180664,
"learning_rate": 7.232218204335465e-06,
"loss": 2.0171,
"step": 9680
},
{
"epoch": 1.2546990581672661,
"grad_norm": 0.7830759286880493,
"learning_rate": 7.218739018118831e-06,
"loss": 2.0057,
"step": 9700
},
{
"epoch": 1.2572860665346215,
"grad_norm": 0.730114758014679,
"learning_rate": 7.205239723428224e-06,
"loss": 2.0264,
"step": 9720
},
{
"epoch": 1.2598730749019766,
"grad_norm": 0.7489242553710938,
"learning_rate": 7.1917204426080065e-06,
"loss": 1.9886,
"step": 9740
},
{
"epoch": 1.2624600832693318,
"grad_norm": 0.7818061709403992,
"learning_rate": 7.17818129818368e-06,
"loss": 2.0001,
"step": 9760
},
{
"epoch": 1.265047091636687,
"grad_norm": 0.733417809009552,
"learning_rate": 7.164622412860766e-06,
"loss": 2.012,
"step": 9780
},
{
"epoch": 1.267634100004042,
"grad_norm": 0.76729816198349,
"learning_rate": 7.151043909523702e-06,
"loss": 2.0066,
"step": 9800
},
{
"epoch": 1.2702211083713975,
"grad_norm": 0.814164936542511,
"learning_rate": 7.137445911234724e-06,
"loss": 2.0132,
"step": 9820
},
{
"epoch": 1.2728081167387526,
"grad_norm": 0.750637948513031,
"learning_rate": 7.1238285412327456e-06,
"loss": 2.0161,
"step": 9840
},
{
"epoch": 1.2753951251061078,
"grad_norm": 0.7567455768585205,
"learning_rate": 7.110191922932252e-06,
"loss": 2.0201,
"step": 9860
},
{
"epoch": 1.277982133473463,
"grad_norm": 0.7351866364479065,
"learning_rate": 7.0965361799221756e-06,
"loss": 2.0183,
"step": 9880
},
{
"epoch": 1.280569141840818,
"grad_norm": 0.7858896851539612,
"learning_rate": 7.082861435964774e-06,
"loss": 2.016,
"step": 9900
},
{
"epoch": 1.2831561502081734,
"grad_norm": 0.7436255216598511,
"learning_rate": 7.069167814994515e-06,
"loss": 2.0116,
"step": 9920
},
{
"epoch": 1.2857431585755286,
"grad_norm": 0.7635684609413147,
"learning_rate": 7.0554554411169414e-06,
"loss": 2.0172,
"step": 9940
},
{
"epoch": 1.2883301669428837,
"grad_norm": 0.759364902973175,
"learning_rate": 7.041724438607562e-06,
"loss": 1.9951,
"step": 9960
},
{
"epoch": 1.2909171753102389,
"grad_norm": 0.7556828856468201,
"learning_rate": 7.027974931910714e-06,
"loss": 2.0098,
"step": 9980
},
{
"epoch": 1.293504183677594,
"grad_norm": 0.7476949095726013,
"learning_rate": 7.014207045638437e-06,
"loss": 2.0051,
"step": 10000
},
{
"epoch": 1.2960911920449494,
"grad_norm": 0.7407810688018799,
"learning_rate": 7.000420904569348e-06,
"loss": 2.0138,
"step": 10020
},
{
"epoch": 1.2986782004123045,
"grad_norm": 0.7346110343933105,
"learning_rate": 6.986616633647505e-06,
"loss": 2.0077,
"step": 10040
},
{
"epoch": 1.3012652087796597,
"grad_norm": 0.7659201622009277,
"learning_rate": 6.9727943579812785e-06,
"loss": 1.9878,
"step": 10060
},
{
"epoch": 1.3038522171470148,
"grad_norm": 0.7628219723701477,
"learning_rate": 6.958954202842216e-06,
"loss": 1.9814,
"step": 10080
},
{
"epoch": 1.30643922551437,
"grad_norm": 0.7482396960258484,
"learning_rate": 6.945096293663908e-06,
"loss": 2.0132,
"step": 10100
},
{
"epoch": 1.3090262338817253,
"grad_norm": 0.7802220582962036,
"learning_rate": 6.931220756040848e-06,
"loss": 2.0065,
"step": 10120
},
{
"epoch": 1.3116132422490803,
"grad_norm": 0.8256868124008179,
"learning_rate": 6.917327715727299e-06,
"loss": 1.9969,
"step": 10140
},
{
"epoch": 1.3142002506164356,
"grad_norm": 0.7396267652511597,
"learning_rate": 6.903417298636148e-06,
"loss": 1.9911,
"step": 10160
},
{
"epoch": 1.3167872589837908,
"grad_norm": 0.7678589224815369,
"learning_rate": 6.889489630837772e-06,
"loss": 2.009,
"step": 10180
},
{
"epoch": 1.319374267351146,
"grad_norm": 0.7238781452178955,
"learning_rate": 6.875544838558888e-06,
"loss": 2.0097,
"step": 10200
},
{
"epoch": 1.321961275718501,
"grad_norm": 0.748767614364624,
"learning_rate": 6.861583048181416e-06,
"loss": 1.9943,
"step": 10220
},
{
"epoch": 1.3245482840858562,
"grad_norm": 0.7102675437927246,
"learning_rate": 6.8476043862413255e-06,
"loss": 1.9959,
"step": 10240
},
{
"epoch": 1.3271352924532116,
"grad_norm": 0.7931022047996521,
"learning_rate": 6.833608979427497e-06,
"loss": 2.0084,
"step": 10260
},
{
"epoch": 1.3297223008205667,
"grad_norm": 0.7266056537628174,
"learning_rate": 6.8195969545805704e-06,
"loss": 1.9942,
"step": 10280
},
{
"epoch": 1.332309309187922,
"grad_norm": 0.7871178984642029,
"learning_rate": 6.805568438691792e-06,
"loss": 2.02,
"step": 10300
},
{
"epoch": 1.334896317555277,
"grad_norm": 0.7746707201004028,
"learning_rate": 6.791523558901869e-06,
"loss": 2.003,
"step": 10320
},
{
"epoch": 1.3374833259226322,
"grad_norm": 0.7147629261016846,
"learning_rate": 6.777462442499814e-06,
"loss": 1.9881,
"step": 10340
},
{
"epoch": 1.3400703342899876,
"grad_norm": 0.75360107421875,
"learning_rate": 6.763385216921794e-06,
"loss": 2.0068,
"step": 10360
},
{
"epoch": 1.3426573426573427,
"grad_norm": 0.7157790660858154,
"learning_rate": 6.749292009749971e-06,
"loss": 1.9808,
"step": 10380
},
{
"epoch": 1.3452443510246979,
"grad_norm": 0.7485711574554443,
"learning_rate": 6.73518294871135e-06,
"loss": 2.0038,
"step": 10400
},
{
"epoch": 1.347831359392053,
"grad_norm": 0.7876157760620117,
"learning_rate": 6.721058161676621e-06,
"loss": 1.9937,
"step": 10420
},
{
"epoch": 1.3504183677594082,
"grad_norm": 0.7066450119018555,
"learning_rate": 6.706917776658997e-06,
"loss": 1.9937,
"step": 10440
},
{
"epoch": 1.3530053761267635,
"grad_norm": 0.7409518957138062,
"learning_rate": 6.692761921813057e-06,
"loss": 1.9938,
"step": 10460
},
{
"epoch": 1.3555923844941187,
"grad_norm": 0.7196327447891235,
"learning_rate": 6.678590725433583e-06,
"loss": 1.9705,
"step": 10480
},
{
"epoch": 1.3581793928614738,
"grad_norm": 0.7340644598007202,
"learning_rate": 6.664404315954397e-06,
"loss": 1.9975,
"step": 10500
},
{
"epoch": 1.360766401228829,
"grad_norm": 0.7806727886199951,
"learning_rate": 6.6502028219472e-06,
"loss": 1.9754,
"step": 10520
},
{
"epoch": 1.3633534095961841,
"grad_norm": 0.7246034741401672,
"learning_rate": 6.635986372120401e-06,
"loss": 1.9734,
"step": 10540
},
{
"epoch": 1.3659404179635395,
"grad_norm": 0.795344352722168,
"learning_rate": 6.621755095317957e-06,
"loss": 1.9765,
"step": 10560
},
{
"epoch": 1.3685274263308944,
"grad_norm": 0.7736918330192566,
"learning_rate": 6.6075091205181986e-06,
"loss": 1.9954,
"step": 10580
},
{
"epoch": 1.3711144346982498,
"grad_norm": 0.7415809035301208,
"learning_rate": 6.59324857683267e-06,
"loss": 1.9701,
"step": 10600
},
{
"epoch": 1.373701443065605,
"grad_norm": 0.764735758304596,
"learning_rate": 6.578973593504947e-06,
"loss": 1.9956,
"step": 10620
},
{
"epoch": 1.37628845143296,
"grad_norm": 0.7368749976158142,
"learning_rate": 6.564684299909476e-06,
"loss": 1.9679,
"step": 10640
},
{
"epoch": 1.3788754598003152,
"grad_norm": 0.7555897235870361,
"learning_rate": 6.550380825550397e-06,
"loss": 1.9974,
"step": 10660
},
{
"epoch": 1.3814624681676704,
"grad_norm": 0.7471804618835449,
"learning_rate": 6.536063300060371e-06,
"loss": 2.0033,
"step": 10680
},
{
"epoch": 1.3840494765350257,
"grad_norm": 0.7835204005241394,
"learning_rate": 6.521731853199406e-06,
"loss": 1.9812,
"step": 10700
},
{
"epoch": 1.386636484902381,
"grad_norm": 0.8036173582077026,
"learning_rate": 6.507386614853674e-06,
"loss": 2.0038,
"step": 10720
},
{
"epoch": 1.389223493269736,
"grad_norm": 0.7817917466163635,
"learning_rate": 6.493027715034346e-06,
"loss": 1.9794,
"step": 10740
},
{
"epoch": 1.3918105016370912,
"grad_norm": 0.7393909692764282,
"learning_rate": 6.478655283876404e-06,
"loss": 1.9752,
"step": 10760
},
{
"epoch": 1.3943975100044463,
"grad_norm": 0.7540735006332397,
"learning_rate": 6.464269451637464e-06,
"loss": 1.9719,
"step": 10780
},
{
"epoch": 1.3969845183718017,
"grad_norm": 0.7721074819564819,
"learning_rate": 6.449870348696599e-06,
"loss": 1.99,
"step": 10800
},
{
"epoch": 1.3995715267391569,
"grad_norm": 0.746705949306488,
"learning_rate": 6.435458105553154e-06,
"loss": 1.9967,
"step": 10820
},
{
"epoch": 1.402158535106512,
"grad_norm": 0.7043789029121399,
"learning_rate": 6.421032852825561e-06,
"loss": 1.9606,
"step": 10840
},
{
"epoch": 1.4047455434738672,
"grad_norm": 0.7369928359985352,
"learning_rate": 6.4065947212501614e-06,
"loss": 1.9975,
"step": 10860
},
{
"epoch": 1.4073325518412223,
"grad_norm": 0.7037030458450317,
"learning_rate": 6.392143841680016e-06,
"loss": 1.9852,
"step": 10880
},
{
"epoch": 1.4099195602085777,
"grad_norm": 0.7750017642974854,
"learning_rate": 6.377680345083721e-06,
"loss": 1.9737,
"step": 10900
},
{
"epoch": 1.4125065685759328,
"grad_norm": 0.7891290783882141,
"learning_rate": 6.3632043625442244e-06,
"loss": 1.9869,
"step": 10920
},
{
"epoch": 1.415093576943288,
"grad_norm": 0.7484495043754578,
"learning_rate": 6.3487160252576275e-06,
"loss": 1.9903,
"step": 10940
},
{
"epoch": 1.4176805853106431,
"grad_norm": 0.7297314405441284,
"learning_rate": 6.33421546453201e-06,
"loss": 1.9951,
"step": 10960
},
{
"epoch": 1.4202675936779983,
"grad_norm": 0.7458155751228333,
"learning_rate": 6.319702811786228e-06,
"loss": 1.9877,
"step": 10980
},
{
"epoch": 1.4228546020453534,
"grad_norm": 0.7403327822685242,
"learning_rate": 6.305178198548732e-06,
"loss": 1.9762,
"step": 11000
},
{
"epoch": 1.4254416104127086,
"grad_norm": 0.7428018450737,
"learning_rate": 6.290641756456365e-06,
"loss": 1.992,
"step": 11020
},
{
"epoch": 1.428028618780064,
"grad_norm": 0.7310479283332825,
"learning_rate": 6.2760936172531815e-06,
"loss": 1.9693,
"step": 11040
},
{
"epoch": 1.430615627147419,
"grad_norm": 0.7689222693443298,
"learning_rate": 6.261533912789244e-06,
"loss": 1.9612,
"step": 11060
},
{
"epoch": 1.4332026355147742,
"grad_norm": 0.7754133343696594,
"learning_rate": 6.246962775019431e-06,
"loss": 1.9787,
"step": 11080
},
{
"epoch": 1.4357896438821294,
"grad_norm": 0.7548643946647644,
"learning_rate": 6.232380336002241e-06,
"loss": 1.9772,
"step": 11100
},
{
"epoch": 1.4383766522494845,
"grad_norm": 0.8060590624809265,
"learning_rate": 6.217786727898595e-06,
"loss": 1.9632,
"step": 11120
},
{
"epoch": 1.44096366061684,
"grad_norm": 0.7445610165596008,
"learning_rate": 6.203182082970643e-06,
"loss": 1.9693,
"step": 11140
},
{
"epoch": 1.443550668984195,
"grad_norm": 0.84158855676651,
"learning_rate": 6.1885665335805614e-06,
"loss": 1.9776,
"step": 11160
},
{
"epoch": 1.4461376773515502,
"grad_norm": 0.7844957113265991,
"learning_rate": 6.173940212189351e-06,
"loss": 1.9969,
"step": 11180
},
{
"epoch": 1.4487246857189053,
"grad_norm": 0.7566760182380676,
"learning_rate": 6.159303251355641e-06,
"loss": 1.9723,
"step": 11200
},
{
"epoch": 1.4513116940862605,
"grad_norm": 0.7104907035827637,
"learning_rate": 6.144655783734485e-06,
"loss": 1.9766,
"step": 11220
},
{
"epoch": 1.4538987024536159,
"grad_norm": 0.6929414868354797,
"learning_rate": 6.129997942076164e-06,
"loss": 1.9835,
"step": 11240
},
{
"epoch": 1.456485710820971,
"grad_norm": 0.7387101054191589,
"learning_rate": 6.115329859224974e-06,
"loss": 1.9799,
"step": 11260
},
{
"epoch": 1.4590727191883262,
"grad_norm": 0.7652817964553833,
"learning_rate": 6.100651668118029e-06,
"loss": 1.9742,
"step": 11280
},
{
"epoch": 1.4616597275556813,
"grad_norm": 0.7251682877540588,
"learning_rate": 6.085963501784057e-06,
"loss": 1.9761,
"step": 11300
},
{
"epoch": 1.4642467359230364,
"grad_norm": 0.7064483761787415,
"learning_rate": 6.071265493342188e-06,
"loss": 1.9839,
"step": 11320
},
{
"epoch": 1.4668337442903918,
"grad_norm": 0.759384036064148,
"learning_rate": 6.056557776000751e-06,
"loss": 1.9927,
"step": 11340
},
{
"epoch": 1.4694207526577467,
"grad_norm": 0.7314841747283936,
"learning_rate": 6.041840483056068e-06,
"loss": 1.9805,
"step": 11360
},
{
"epoch": 1.4720077610251021,
"grad_norm": 0.7411664128303528,
"learning_rate": 6.0271137478912475e-06,
"loss": 1.966,
"step": 11380
},
{
"epoch": 1.4745947693924573,
"grad_norm": 0.725292444229126,
"learning_rate": 6.01237770397497e-06,
"loss": 1.9778,
"step": 11400
},
{
"epoch": 1.4771817777598124,
"grad_norm": 0.7571858763694763,
"learning_rate": 5.997632484860281e-06,
"loss": 1.9805,
"step": 11420
},
{
"epoch": 1.4797687861271676,
"grad_norm": 0.7663134932518005,
"learning_rate": 5.982878224183384e-06,
"loss": 1.9793,
"step": 11440
},
{
"epoch": 1.4823557944945227,
"grad_norm": 0.8204968571662903,
"learning_rate": 5.968115055662424e-06,
"loss": 1.9687,
"step": 11460
},
{
"epoch": 1.484942802861878,
"grad_norm": 0.757935643196106,
"learning_rate": 5.953343113096277e-06,
"loss": 1.9599,
"step": 11480
},
{
"epoch": 1.4875298112292332,
"grad_norm": 0.768320620059967,
"learning_rate": 5.938562530363341e-06,
"loss": 1.9853,
"step": 11500
},
{
"epoch": 1.4901168195965884,
"grad_norm": 0.7116501331329346,
"learning_rate": 5.9237734414203185e-06,
"loss": 1.973,
"step": 11520
},
{
"epoch": 1.4927038279639435,
"grad_norm": 0.7462672591209412,
"learning_rate": 5.9089759803010035e-06,
"loss": 1.97,
"step": 11540
},
{
"epoch": 1.4952908363312987,
"grad_norm": 0.7260856628417969,
"learning_rate": 5.894170281115071e-06,
"loss": 1.9761,
"step": 11560
},
{
"epoch": 1.497877844698654,
"grad_norm": 0.754158079624176,
"learning_rate": 5.879356478046849e-06,
"loss": 1.9823,
"step": 11580
},
{
"epoch": 1.5004648530660092,
"grad_norm": 0.7565730810165405,
"learning_rate": 5.864534705354123e-06,
"loss": 1.9782,
"step": 11600
},
{
"epoch": 1.5030518614333643,
"grad_norm": 0.7671797275543213,
"learning_rate": 5.849705097366898e-06,
"loss": 1.9611,
"step": 11620
},
{
"epoch": 1.5056388698007195,
"grad_norm": 0.7473678588867188,
"learning_rate": 5.834867788486194e-06,
"loss": 1.9544,
"step": 11640
},
{
"epoch": 1.5082258781680746,
"grad_norm": 0.7111749649047852,
"learning_rate": 5.8200229131828225e-06,
"loss": 1.9661,
"step": 11660
},
{
"epoch": 1.51081288653543,
"grad_norm": 0.7859391570091248,
"learning_rate": 5.805170605996173e-06,
"loss": 1.9493,
"step": 11680
},
{
"epoch": 1.513399894902785,
"grad_norm": 0.7364257574081421,
"learning_rate": 5.7903110015329886e-06,
"loss": 1.9703,
"step": 11700
},
{
"epoch": 1.5159869032701403,
"grad_norm": 0.7579001784324646,
"learning_rate": 5.775444234466145e-06,
"loss": 1.9656,
"step": 11720
},
{
"epoch": 1.5185739116374954,
"grad_norm": 0.7483280897140503,
"learning_rate": 5.76057043953344e-06,
"loss": 1.9785,
"step": 11740
},
{
"epoch": 1.5211609200048506,
"grad_norm": 0.7101817727088928,
"learning_rate": 5.745689751536355e-06,
"loss": 1.9794,
"step": 11760
},
{
"epoch": 1.523747928372206,
"grad_norm": 0.7399246096611023,
"learning_rate": 5.730802305338852e-06,
"loss": 1.9798,
"step": 11780
},
{
"epoch": 1.5263349367395609,
"grad_norm": 0.7751443982124329,
"learning_rate": 5.715908235866139e-06,
"loss": 1.9586,
"step": 11800
},
{
"epoch": 1.5289219451069163,
"grad_norm": 0.7285624742507935,
"learning_rate": 5.701007678103449e-06,
"loss": 1.9712,
"step": 11820
},
{
"epoch": 1.5315089534742714,
"grad_norm": 0.72426837682724,
"learning_rate": 5.686100767094826e-06,
"loss": 1.9655,
"step": 11840
},
{
"epoch": 1.5340959618416266,
"grad_norm": 0.8364124894142151,
"learning_rate": 5.671187637941882e-06,
"loss": 1.9664,
"step": 11860
},
{
"epoch": 1.536682970208982,
"grad_norm": 0.6998101472854614,
"learning_rate": 5.656268425802594e-06,
"loss": 1.9811,
"step": 11880
},
{
"epoch": 1.5392699785763369,
"grad_norm": 0.7216737270355225,
"learning_rate": 5.641343265890065e-06,
"loss": 1.9686,
"step": 11900
},
{
"epoch": 1.5418569869436922,
"grad_norm": 0.7465617060661316,
"learning_rate": 5.626412293471304e-06,
"loss": 1.9642,
"step": 11920
},
{
"epoch": 1.5444439953110474,
"grad_norm": 0.7474328279495239,
"learning_rate": 5.6114756438659966e-06,
"loss": 1.9698,
"step": 11940
},
{
"epoch": 1.5470310036784025,
"grad_norm": 0.709169328212738,
"learning_rate": 5.5965334524452856e-06,
"loss": 1.9629,
"step": 11960
},
{
"epoch": 1.5496180120457577,
"grad_norm": 0.778994083404541,
"learning_rate": 5.5815858546305345e-06,
"loss": 1.9922,
"step": 11980
},
{
"epoch": 1.5522050204131128,
"grad_norm": 0.7840653657913208,
"learning_rate": 5.566632985892107e-06,
"loss": 1.9659,
"step": 12000
},
{
"epoch": 1.5547920287804682,
"grad_norm": 0.7363440990447998,
"learning_rate": 5.551674981748137e-06,
"loss": 1.9739,
"step": 12020
},
{
"epoch": 1.557379037147823,
"grad_norm": 0.74187833070755,
"learning_rate": 5.5367119777633025e-06,
"loss": 1.9654,
"step": 12040
},
{
"epoch": 1.5599660455151785,
"grad_norm": 0.7796049118041992,
"learning_rate": 5.521744109547592e-06,
"loss": 1.9442,
"step": 12060
},
{
"epoch": 1.5625530538825336,
"grad_norm": 0.7324556708335876,
"learning_rate": 5.506771512755083e-06,
"loss": 1.9703,
"step": 12080
},
{
"epoch": 1.5651400622498888,
"grad_norm": 0.7358458638191223,
"learning_rate": 5.491794323082701e-06,
"loss": 1.9563,
"step": 12100
},
{
"epoch": 1.5677270706172441,
"grad_norm": 0.725487232208252,
"learning_rate": 5.476812676269003e-06,
"loss": 1.9715,
"step": 12120
},
{
"epoch": 1.570314078984599,
"grad_norm": 0.7749314308166504,
"learning_rate": 5.46182670809294e-06,
"loss": 1.9812,
"step": 12140
},
{
"epoch": 1.5729010873519544,
"grad_norm": 0.7413734793663025,
"learning_rate": 5.446836554372626e-06,
"loss": 1.9621,
"step": 12160
},
{
"epoch": 1.5754880957193096,
"grad_norm": 0.782210111618042,
"learning_rate": 5.431842350964111e-06,
"loss": 1.9639,
"step": 12180
},
{
"epoch": 1.5780751040866647,
"grad_norm": 0.7391452193260193,
"learning_rate": 5.416844233760145e-06,
"loss": 1.9654,
"step": 12200
},
{
"epoch": 1.58066211245402,
"grad_norm": 0.755551815032959,
"learning_rate": 5.401842338688951e-06,
"loss": 1.9746,
"step": 12220
},
{
"epoch": 1.583249120821375,
"grad_norm": 0.8115268349647522,
"learning_rate": 5.386836801712988e-06,
"loss": 1.9772,
"step": 12240
},
{
"epoch": 1.5858361291887304,
"grad_norm": 0.7625905275344849,
"learning_rate": 5.371827758827723e-06,
"loss": 1.9654,
"step": 12260
},
{
"epoch": 1.5884231375560856,
"grad_norm": 0.735562264919281,
"learning_rate": 5.356815346060401e-06,
"loss": 1.9701,
"step": 12280
},
{
"epoch": 1.5910101459234407,
"grad_norm": 0.7974105477333069,
"learning_rate": 5.3417996994688015e-06,
"loss": 1.9849,
"step": 12300
},
{
"epoch": 1.593597154290796,
"grad_norm": 0.7090020179748535,
"learning_rate": 5.326780955140019e-06,
"loss": 1.9872,
"step": 12320
},
{
"epoch": 1.596184162658151,
"grad_norm": 0.80815589427948,
"learning_rate": 5.311759249189213e-06,
"loss": 1.9791,
"step": 12340
},
{
"epoch": 1.5987711710255064,
"grad_norm": 0.8091129660606384,
"learning_rate": 5.296734717758397e-06,
"loss": 1.9725,
"step": 12360
},
{
"epoch": 1.6013581793928615,
"grad_norm": 0.7238413691520691,
"learning_rate": 5.2817074970151815e-06,
"loss": 1.9716,
"step": 12380
},
{
"epoch": 1.6039451877602167,
"grad_norm": 0.7404330372810364,
"learning_rate": 5.266677723151558e-06,
"loss": 1.9559,
"step": 12400
},
{
"epoch": 1.6065321961275718,
"grad_norm": 0.7491504549980164,
"learning_rate": 5.251645532382649e-06,
"loss": 1.9582,
"step": 12420
},
{
"epoch": 1.609119204494927,
"grad_norm": 0.7146170139312744,
"learning_rate": 5.236611060945488e-06,
"loss": 1.957,
"step": 12440
},
{
"epoch": 1.6117062128622823,
"grad_norm": 0.7770988345146179,
"learning_rate": 5.221574445097776e-06,
"loss": 1.975,
"step": 12460
},
{
"epoch": 1.6142932212296373,
"grad_norm": 0.736663281917572,
"learning_rate": 5.2065358211166486e-06,
"loss": 1.9719,
"step": 12480
},
{
"epoch": 1.6168802295969926,
"grad_norm": 0.7258009910583496,
"learning_rate": 5.191495325297441e-06,
"loss": 1.9491,
"step": 12500
},
{
"epoch": 1.6194672379643478,
"grad_norm": 0.739658534526825,
"learning_rate": 5.1764530939524524e-06,
"loss": 1.936,
"step": 12520
},
{
"epoch": 1.622054246331703,
"grad_norm": 0.7602417469024658,
"learning_rate": 5.1614092634097146e-06,
"loss": 1.9631,
"step": 12540
},
{
"epoch": 1.6246412546990583,
"grad_norm": 0.7500277161598206,
"learning_rate": 5.14636397001175e-06,
"loss": 1.9678,
"step": 12560
},
{
"epoch": 1.6272282630664132,
"grad_norm": 0.8021513223648071,
"learning_rate": 5.131317350114335e-06,
"loss": 1.9597,
"step": 12580
},
{
"epoch": 1.6298152714337686,
"grad_norm": 0.7306881546974182,
"learning_rate": 5.116269540085277e-06,
"loss": 1.956,
"step": 12600
},
{
"epoch": 1.6324022798011237,
"grad_norm": 0.786392867565155,
"learning_rate": 5.101220676303161e-06,
"loss": 1.961,
"step": 12620
},
{
"epoch": 1.6349892881684789,
"grad_norm": 0.7665414810180664,
"learning_rate": 5.08617089515613e-06,
"loss": 1.9689,
"step": 12640
},
{
"epoch": 1.6375762965358343,
"grad_norm": 0.7925981283187866,
"learning_rate": 5.0711203330406334e-06,
"loss": 1.965,
"step": 12660
},
{
"epoch": 1.6401633049031892,
"grad_norm": 0.7429537177085876,
"learning_rate": 5.056069126360206e-06,
"loss": 1.9493,
"step": 12680
},
{
"epoch": 1.6427503132705445,
"grad_norm": 0.7412150502204895,
"learning_rate": 5.041017411524217e-06,
"loss": 1.9598,
"step": 12700
},
{
"epoch": 1.6453373216378997,
"grad_norm": 0.7570027112960815,
"learning_rate": 5.025965324946646e-06,
"loss": 1.962,
"step": 12720
},
{
"epoch": 1.6479243300052548,
"grad_norm": 0.7325958013534546,
"learning_rate": 5.010913003044842e-06,
"loss": 1.9606,
"step": 12740
},
{
"epoch": 1.6505113383726102,
"grad_norm": 0.7462648153305054,
"learning_rate": 4.995860582238284e-06,
"loss": 1.943,
"step": 12760
},
{
"epoch": 1.6530983467399651,
"grad_norm": 0.733491063117981,
"learning_rate": 4.980808198947347e-06,
"loss": 1.9519,
"step": 12780
},
{
"epoch": 1.6556853551073205,
"grad_norm": 0.7238721251487732,
"learning_rate": 4.965755989592069e-06,
"loss": 1.9709,
"step": 12800
},
{
"epoch": 1.6582723634746757,
"grad_norm": 0.7215867042541504,
"learning_rate": 4.950704090590908e-06,
"loss": 1.9739,
"step": 12820
},
{
"epoch": 1.6608593718420308,
"grad_norm": 0.7542194128036499,
"learning_rate": 4.935652638359514e-06,
"loss": 1.9684,
"step": 12840
},
{
"epoch": 1.663446380209386,
"grad_norm": 0.7353436350822449,
"learning_rate": 4.920601769309484e-06,
"loss": 1.9678,
"step": 12860
},
{
"epoch": 1.666033388576741,
"grad_norm": 0.7446224689483643,
"learning_rate": 4.9055516198471295e-06,
"loss": 1.9686,
"step": 12880
},
{
"epoch": 1.6686203969440965,
"grad_norm": 0.7474517226219177,
"learning_rate": 4.890502326372244e-06,
"loss": 1.9715,
"step": 12900
},
{
"epoch": 1.6712074053114514,
"grad_norm": 0.7246785163879395,
"learning_rate": 4.87545402527686e-06,
"loss": 1.9641,
"step": 12920
},
{
"epoch": 1.6737944136788068,
"grad_norm": 0.746099591255188,
"learning_rate": 4.8604068529440165e-06,
"loss": 1.984,
"step": 12940
},
{
"epoch": 1.676381422046162,
"grad_norm": 0.7616404294967651,
"learning_rate": 4.845360945746523e-06,
"loss": 1.9708,
"step": 12960
},
{
"epoch": 1.678968430413517,
"grad_norm": 0.7588489055633545,
"learning_rate": 4.830316440045723e-06,
"loss": 1.9497,
"step": 12980
},
{
"epoch": 1.6815554387808724,
"grad_norm": 0.7502963542938232,
"learning_rate": 4.815273472190259e-06,
"loss": 1.9742,
"step": 13000
},
{
"epoch": 1.6841424471482274,
"grad_norm": 0.7508452534675598,
"learning_rate": 4.800232178514834e-06,
"loss": 1.966,
"step": 13020
},
{
"epoch": 1.6867294555155827,
"grad_norm": 0.7516492605209351,
"learning_rate": 4.78519269533898e-06,
"loss": 1.9499,
"step": 13040
},
{
"epoch": 1.6893164638829379,
"grad_norm": 0.7797842025756836,
"learning_rate": 4.770155158965818e-06,
"loss": 1.954,
"step": 13060
},
{
"epoch": 1.691903472250293,
"grad_norm": 0.7264002561569214,
"learning_rate": 4.7551197056808275e-06,
"loss": 1.9431,
"step": 13080
},
{
"epoch": 1.6944904806176484,
"grad_norm": 0.7275776267051697,
"learning_rate": 4.740086471750608e-06,
"loss": 1.9382,
"step": 13100
},
{
"epoch": 1.6970774889850033,
"grad_norm": 0.745893657207489,
"learning_rate": 4.7250555934216446e-06,
"loss": 1.961,
"step": 13120
},
{
"epoch": 1.6996644973523587,
"grad_norm": 0.7499802112579346,
"learning_rate": 4.7100272069190735e-06,
"loss": 1.9462,
"step": 13140
},
{
"epoch": 1.7022515057197138,
"grad_norm": 0.7486584782600403,
"learning_rate": 4.69500144844545e-06,
"loss": 1.9543,
"step": 13160
},
{
"epoch": 1.704838514087069,
"grad_norm": 0.715093195438385,
"learning_rate": 4.679978454179507e-06,
"loss": 1.9484,
"step": 13180
},
{
"epoch": 1.7074255224544241,
"grad_norm": 0.7148420810699463,
"learning_rate": 4.664958360274927e-06,
"loss": 1.9584,
"step": 13200
},
{
"epoch": 1.7100125308217793,
"grad_norm": 0.7597442269325256,
"learning_rate": 4.649941302859109e-06,
"loss": 1.9651,
"step": 13220
},
{
"epoch": 1.7125995391891347,
"grad_norm": 0.7285702228546143,
"learning_rate": 4.63492741803193e-06,
"loss": 1.9514,
"step": 13240
},
{
"epoch": 1.7151865475564898,
"grad_norm": 0.7763605117797852,
"learning_rate": 4.619916841864513e-06,
"loss": 1.9697,
"step": 13260
},
{
"epoch": 1.717773555923845,
"grad_norm": 0.7495293617248535,
"learning_rate": 4.604909710397997e-06,
"loss": 1.9471,
"step": 13280
},
{
"epoch": 1.7203605642912,
"grad_norm": 0.7468627691268921,
"learning_rate": 4.5899061596423e-06,
"loss": 1.9835,
"step": 13300
},
{
"epoch": 1.7229475726585552,
"grad_norm": 0.7560572624206543,
"learning_rate": 4.574906325574888e-06,
"loss": 1.945,
"step": 13320
},
{
"epoch": 1.7255345810259106,
"grad_norm": 0.7571983337402344,
"learning_rate": 4.5599103441395415e-06,
"loss": 1.9437,
"step": 13340
},
{
"epoch": 1.7281215893932655,
"grad_norm": 0.7818071842193604,
"learning_rate": 4.544918351245128e-06,
"loss": 1.9482,
"step": 13360
},
{
"epoch": 1.730708597760621,
"grad_norm": 0.764826238155365,
"learning_rate": 4.529930482764362e-06,
"loss": 1.9408,
"step": 13380
},
{
"epoch": 1.733295606127976,
"grad_norm": 0.7344059348106384,
"learning_rate": 4.514946874532584e-06,
"loss": 1.9487,
"step": 13400
},
{
"epoch": 1.7358826144953312,
"grad_norm": 0.7779365181922913,
"learning_rate": 4.499967662346519e-06,
"loss": 1.9417,
"step": 13420
},
{
"epoch": 1.7384696228626866,
"grad_norm": 0.7346351146697998,
"learning_rate": 4.484992981963053e-06,
"loss": 1.95,
"step": 13440
},
{
"epoch": 1.7410566312300415,
"grad_norm": 0.7388240098953247,
"learning_rate": 4.4700229690979985e-06,
"loss": 1.9404,
"step": 13460
},
{
"epoch": 1.7436436395973969,
"grad_norm": 0.7134066224098206,
"learning_rate": 4.455057759424868e-06,
"loss": 1.9515,
"step": 13480
},
{
"epoch": 1.746230647964752,
"grad_norm": 0.7144001722335815,
"learning_rate": 4.440097488573642e-06,
"loss": 1.9579,
"step": 13500
},
{
"epoch": 1.7488176563321072,
"grad_norm": 0.7423569560050964,
"learning_rate": 4.425142292129539e-06,
"loss": 1.9642,
"step": 13520
},
{
"epoch": 1.7514046646994625,
"grad_norm": 0.7508745789527893,
"learning_rate": 4.410192305631791e-06,
"loss": 1.9495,
"step": 13540
},
{
"epoch": 1.7539916730668175,
"grad_norm": 0.788547933101654,
"learning_rate": 4.395247664572409e-06,
"loss": 1.9626,
"step": 13560
},
{
"epoch": 1.7565786814341728,
"grad_norm": 0.7150464057922363,
"learning_rate": 4.3803085043949585e-06,
"loss": 1.9405,
"step": 13580
},
{
"epoch": 1.759165689801528,
"grad_norm": 0.7496698498725891,
"learning_rate": 4.365374960493335e-06,
"loss": 1.9256,
"step": 13600
},
{
"epoch": 1.7617526981688831,
"grad_norm": 0.72135990858078,
"learning_rate": 4.35044716821053e-06,
"loss": 1.9226,
"step": 13620
},
{
"epoch": 1.7643397065362383,
"grad_norm": 0.717047929763794,
"learning_rate": 4.335525262837409e-06,
"loss": 1.9551,
"step": 13640
},
{
"epoch": 1.7669267149035934,
"grad_norm": 0.8064849376678467,
"learning_rate": 4.3206093796114875e-06,
"loss": 1.9284,
"step": 13660
},
{
"epoch": 1.7695137232709488,
"grad_norm": 0.778425931930542,
"learning_rate": 4.305699653715695e-06,
"loss": 1.9557,
"step": 13680
},
{
"epoch": 1.7721007316383037,
"grad_norm": 0.742351770401001,
"learning_rate": 4.290796220277162e-06,
"loss": 1.9541,
"step": 13700
},
{
"epoch": 1.774687740005659,
"grad_norm": 0.7197045087814331,
"learning_rate": 4.275899214365991e-06,
"loss": 1.962,
"step": 13720
},
{
"epoch": 1.7772747483730142,
"grad_norm": 0.7143533825874329,
"learning_rate": 4.26100877099403e-06,
"loss": 1.9533,
"step": 13740
},
{
"epoch": 1.7798617567403694,
"grad_norm": 0.7439367175102234,
"learning_rate": 4.24612502511365e-06,
"loss": 1.9479,
"step": 13760
},
{
"epoch": 1.7824487651077248,
"grad_norm": 0.7582443952560425,
"learning_rate": 4.231248111616525e-06,
"loss": 1.9495,
"step": 13780
},
{
"epoch": 1.7850357734750797,
"grad_norm": 0.7773938775062561,
"learning_rate": 4.216378165332404e-06,
"loss": 1.9526,
"step": 13800
},
{
"epoch": 1.787622781842435,
"grad_norm": 0.7457807660102844,
"learning_rate": 4.201515321027894e-06,
"loss": 1.9511,
"step": 13820
},
{
"epoch": 1.7902097902097902,
"grad_norm": 0.708568811416626,
"learning_rate": 4.186659713405235e-06,
"loss": 1.9399,
"step": 13840
},
{
"epoch": 1.7927967985771454,
"grad_norm": 0.8040826320648193,
"learning_rate": 4.171811477101083e-06,
"loss": 1.9546,
"step": 13860
},
{
"epoch": 1.7953838069445007,
"grad_norm": 0.8364230990409851,
"learning_rate": 4.156970746685286e-06,
"loss": 1.9466,
"step": 13880
},
{
"epoch": 1.7979708153118557,
"grad_norm": 0.7208890914916992,
"learning_rate": 4.142137656659665e-06,
"loss": 1.9477,
"step": 13900
},
{
"epoch": 1.800557823679211,
"grad_norm": 0.7639915943145752,
"learning_rate": 4.1273123414568e-06,
"loss": 1.9493,
"step": 13920
},
{
"epoch": 1.8031448320465662,
"grad_norm": 0.710559606552124,
"learning_rate": 4.112494935438803e-06,
"loss": 1.9613,
"step": 13940
},
{
"epoch": 1.8057318404139213,
"grad_norm": 0.7264363765716553,
"learning_rate": 4.097685572896107e-06,
"loss": 1.9369,
"step": 13960
},
{
"epoch": 1.8083188487812767,
"grad_norm": 0.7411283254623413,
"learning_rate": 4.082884388046248e-06,
"loss": 1.9311,
"step": 13980
},
{
"epoch": 1.8109058571486316,
"grad_norm": 0.7250226140022278,
"learning_rate": 4.068091515032645e-06,
"loss": 1.9455,
"step": 14000
},
{
"epoch": 1.813492865515987,
"grad_norm": 0.7740228772163391,
"learning_rate": 4.053307087923389e-06,
"loss": 1.9459,
"step": 14020
},
{
"epoch": 1.8160798738833421,
"grad_norm": 0.7399852871894836,
"learning_rate": 4.038531240710024e-06,
"loss": 1.937,
"step": 14040
},
{
"epoch": 1.8186668822506973,
"grad_norm": 0.764217734336853,
"learning_rate": 4.0237641073063335e-06,
"loss": 1.9334,
"step": 14060
},
{
"epoch": 1.8212538906180524,
"grad_norm": 0.68044114112854,
"learning_rate": 4.009005821547131e-06,
"loss": 1.9374,
"step": 14080
},
{
"epoch": 1.8238408989854076,
"grad_norm": 0.7502301335334778,
"learning_rate": 3.994256517187038e-06,
"loss": 1.9443,
"step": 14100
},
{
"epoch": 1.826427907352763,
"grad_norm": 0.8162171840667725,
"learning_rate": 3.979516327899285e-06,
"loss": 1.9696,
"step": 14120
},
{
"epoch": 1.8290149157201179,
"grad_norm": 0.7348920106887817,
"learning_rate": 3.964785387274485e-06,
"loss": 1.9489,
"step": 14140
},
{
"epoch": 1.8316019240874732,
"grad_norm": 0.707190752029419,
"learning_rate": 3.950063828819437e-06,
"loss": 1.9545,
"step": 14160
},
{
"epoch": 1.8341889324548284,
"grad_norm": 0.7493889331817627,
"learning_rate": 3.935351785955901e-06,
"loss": 1.9414,
"step": 14180
},
{
"epoch": 1.8367759408221835,
"grad_norm": 0.730453372001648,
"learning_rate": 3.920649392019404e-06,
"loss": 1.933,
"step": 14200
},
{
"epoch": 1.839362949189539,
"grad_norm": 0.7539755702018738,
"learning_rate": 3.9059567802580205e-06,
"loss": 1.9684,
"step": 14220
},
{
"epoch": 1.8419499575568938,
"grad_norm": 0.7118528485298157,
"learning_rate": 3.8912740838311735e-06,
"loss": 1.9399,
"step": 14240
},
{
"epoch": 1.8445369659242492,
"grad_norm": 0.7369790077209473,
"learning_rate": 3.876601435808419e-06,
"loss": 1.9586,
"step": 14260
},
{
"epoch": 1.8471239742916044,
"grad_norm": 0.7587118744850159,
"learning_rate": 3.861938969168245e-06,
"loss": 1.9373,
"step": 14280
},
{
"epoch": 1.8497109826589595,
"grad_norm": 0.7134782075881958,
"learning_rate": 3.847286816796867e-06,
"loss": 1.936,
"step": 14300
},
{
"epoch": 1.8522979910263149,
"grad_norm": 0.7343664169311523,
"learning_rate": 3.8326451114870204e-06,
"loss": 1.9177,
"step": 14320
},
{
"epoch": 1.8548849993936698,
"grad_norm": 0.779914915561676,
"learning_rate": 3.818013985936759e-06,
"loss": 1.9347,
"step": 14340
},
{
"epoch": 1.8574720077610252,
"grad_norm": 0.7164252400398254,
"learning_rate": 3.8033935727482522e-06,
"loss": 1.9449,
"step": 14360
},
{
"epoch": 1.8600590161283803,
"grad_norm": 0.7624450325965881,
"learning_rate": 3.788784004426584e-06,
"loss": 1.9288,
"step": 14380
},
{
"epoch": 1.8626460244957355,
"grad_norm": 0.6874932646751404,
"learning_rate": 3.7741854133785493e-06,
"loss": 1.9348,
"step": 14400
},
{
"epoch": 1.8652330328630906,
"grad_norm": 0.7472041845321655,
"learning_rate": 3.759597931911458e-06,
"loss": 1.9471,
"step": 14420
},
{
"epoch": 1.8678200412304458,
"grad_norm": 0.7028324007987976,
"learning_rate": 3.7450216922319317e-06,
"loss": 1.9283,
"step": 14440
},
{
"epoch": 1.8704070495978011,
"grad_norm": 0.689389169216156,
"learning_rate": 3.730456826444707e-06,
"loss": 1.924,
"step": 14460
},
{
"epoch": 1.8729940579651563,
"grad_norm": 0.7389649748802185,
"learning_rate": 3.715903466551442e-06,
"loss": 1.9498,
"step": 14480
},
{
"epoch": 1.8755810663325114,
"grad_norm": 0.7347127199172974,
"learning_rate": 3.7013617444495136e-06,
"loss": 1.9788,
"step": 14500
},
{
"epoch": 1.8781680746998666,
"grad_norm": 0.7470489144325256,
"learning_rate": 3.686831791930826e-06,
"loss": 1.9372,
"step": 14520
},
{
"epoch": 1.8807550830672217,
"grad_norm": 0.7183331847190857,
"learning_rate": 3.6723137406806162e-06,
"loss": 1.9303,
"step": 14540
},
{
"epoch": 1.883342091434577,
"grad_norm": 0.7698735594749451,
"learning_rate": 3.6578077222762587e-06,
"loss": 1.947,
"step": 14560
},
{
"epoch": 1.885929099801932,
"grad_norm": 0.7268314957618713,
"learning_rate": 3.643313868186075e-06,
"loss": 1.9219,
"step": 14580
},
{
"epoch": 1.8885161081692874,
"grad_norm": 0.6923518776893616,
"learning_rate": 3.62883230976814e-06,
"loss": 1.9196,
"step": 14600
},
{
"epoch": 1.8911031165366425,
"grad_norm": 0.7996921539306641,
"learning_rate": 3.6143631782690936e-06,
"loss": 1.95,
"step": 14620
},
{
"epoch": 1.8936901249039977,
"grad_norm": 0.7068451642990112,
"learning_rate": 3.5999066048229512e-06,
"loss": 1.9423,
"step": 14640
},
{
"epoch": 1.896277133271353,
"grad_norm": 0.7754804491996765,
"learning_rate": 3.5854627204499114e-06,
"loss": 1.9398,
"step": 14660
},
{
"epoch": 1.898864141638708,
"grad_norm": 0.7194542288780212,
"learning_rate": 3.5710316560551743e-06,
"loss": 1.9494,
"step": 14680
},
{
"epoch": 1.9014511500060634,
"grad_norm": 0.7717624306678772,
"learning_rate": 3.5566135424277514e-06,
"loss": 1.9385,
"step": 14700
},
{
"epoch": 1.9040381583734185,
"grad_norm": 0.750825822353363,
"learning_rate": 3.542208510239281e-06,
"loss": 1.9404,
"step": 14720
},
{
"epoch": 1.9066251667407736,
"grad_norm": 0.7245256900787354,
"learning_rate": 3.5278166900428435e-06,
"loss": 1.9395,
"step": 14740
},
{
"epoch": 1.909212175108129,
"grad_norm": 0.7693631649017334,
"learning_rate": 3.513438212271781e-06,
"loss": 1.9468,
"step": 14760
},
{
"epoch": 1.911799183475484,
"grad_norm": 0.7538326382637024,
"learning_rate": 3.4990732072385096e-06,
"loss": 1.9473,
"step": 14780
},
{
"epoch": 1.9143861918428393,
"grad_norm": 0.7545143961906433,
"learning_rate": 3.4847218051333447e-06,
"loss": 1.9373,
"step": 14800
},
{
"epoch": 1.9169732002101945,
"grad_norm": 0.7644726037979126,
"learning_rate": 3.4703841360233166e-06,
"loss": 1.9446,
"step": 14820
},
{
"epoch": 1.9195602085775496,
"grad_norm": 0.7512372732162476,
"learning_rate": 3.456060329850993e-06,
"loss": 1.9187,
"step": 14840
},
{
"epoch": 1.9221472169449048,
"grad_norm": 0.7338785529136658,
"learning_rate": 3.441750516433304e-06,
"loss": 1.9368,
"step": 14860
},
{
"epoch": 1.92473422531226,
"grad_norm": 0.6979427337646484,
"learning_rate": 3.427454825460358e-06,
"loss": 1.9341,
"step": 14880
},
{
"epoch": 1.9273212336796153,
"grad_norm": 0.7081001996994019,
"learning_rate": 3.4131733864942755e-06,
"loss": 1.9467,
"step": 14900
},
{
"epoch": 1.9299082420469704,
"grad_norm": 0.726633608341217,
"learning_rate": 3.398906328968011e-06,
"loss": 1.9608,
"step": 14920
},
{
"epoch": 1.9324952504143256,
"grad_norm": 0.7110878229141235,
"learning_rate": 3.384653782184178e-06,
"loss": 1.9337,
"step": 14940
},
{
"epoch": 1.9350822587816807,
"grad_norm": 0.7285667061805725,
"learning_rate": 3.3704158753138795e-06,
"loss": 1.9488,
"step": 14960
},
{
"epoch": 1.9376692671490359,
"grad_norm": 0.7054875493049622,
"learning_rate": 3.356192737395536e-06,
"loss": 1.9491,
"step": 14980
},
{
"epoch": 1.9402562755163912,
"grad_norm": 0.7593070268630981,
"learning_rate": 3.3419844973337178e-06,
"loss": 1.9456,
"step": 15000
},
{
"epoch": 1.9428432838837462,
"grad_norm": 0.7805582880973816,
"learning_rate": 3.327791283897975e-06,
"loss": 1.9399,
"step": 15020
},
{
"epoch": 1.9454302922511015,
"grad_norm": 0.6953914761543274,
"learning_rate": 3.313613225721672e-06,
"loss": 1.9398,
"step": 15040
},
{
"epoch": 1.9480173006184567,
"grad_norm": 0.7649853229522705,
"learning_rate": 3.2994504513008185e-06,
"loss": 1.9202,
"step": 15060
},
{
"epoch": 1.9506043089858118,
"grad_norm": 0.7447147965431213,
"learning_rate": 3.2853030889929105e-06,
"loss": 1.9252,
"step": 15080
},
{
"epoch": 1.9531913173531672,
"grad_norm": 0.7669313549995422,
"learning_rate": 3.2711712670157613e-06,
"loss": 1.9228,
"step": 15100
},
{
"epoch": 1.9557783257205221,
"grad_norm": 0.7048401236534119,
"learning_rate": 3.2570551134463426e-06,
"loss": 1.9316,
"step": 15120
},
{
"epoch": 1.9583653340878775,
"grad_norm": 0.7256927490234375,
"learning_rate": 3.242954756219623e-06,
"loss": 1.9201,
"step": 15140
},
{
"epoch": 1.9609523424552326,
"grad_norm": 0.7202728390693665,
"learning_rate": 3.228870323127409e-06,
"loss": 1.9283,
"step": 15160
},
{
"epoch": 1.9635393508225878,
"grad_norm": 0.7406941056251526,
"learning_rate": 3.214801941817185e-06,
"loss": 1.9384,
"step": 15180
},
{
"epoch": 1.9661263591899432,
"grad_norm": 0.7298773527145386,
"learning_rate": 3.2007497397909588e-06,
"loss": 1.9366,
"step": 15200
},
{
"epoch": 1.968713367557298,
"grad_norm": 0.7127735018730164,
"learning_rate": 3.186713844404105e-06,
"loss": 1.919,
"step": 15220
},
{
"epoch": 1.9713003759246535,
"grad_norm": 0.7574893832206726,
"learning_rate": 3.172694382864212e-06,
"loss": 1.9286,
"step": 15240
},
{
"epoch": 1.9738873842920086,
"grad_norm": 0.7244542241096497,
"learning_rate": 3.158691482229925e-06,
"loss": 1.9195,
"step": 15260
},
{
"epoch": 1.9764743926593638,
"grad_norm": 0.7580679059028625,
"learning_rate": 3.1447052694098006e-06,
"loss": 1.9422,
"step": 15280
},
{
"epoch": 1.979061401026719,
"grad_norm": 0.7484192848205566,
"learning_rate": 3.130735871161151e-06,
"loss": 1.9322,
"step": 15300
},
{
"epoch": 1.981648409394074,
"grad_norm": 0.7437182068824768,
"learning_rate": 3.1167834140888992e-06,
"loss": 1.9283,
"step": 15320
},
{
"epoch": 1.9842354177614294,
"grad_norm": 0.7359150052070618,
"learning_rate": 3.1028480246444293e-06,
"loss": 1.9465,
"step": 15340
},
{
"epoch": 1.9868224261287843,
"grad_norm": 0.7357231378555298,
"learning_rate": 3.08892982912444e-06,
"loss": 1.9216,
"step": 15360
},
{
"epoch": 1.9894094344961397,
"grad_norm": 0.8238598704338074,
"learning_rate": 3.075028953669804e-06,
"loss": 1.9368,
"step": 15380
},
{
"epoch": 1.9919964428634949,
"grad_norm": 0.7543041110038757,
"learning_rate": 3.0611455242644193e-06,
"loss": 1.9318,
"step": 15400
},
{
"epoch": 1.99458345123085,
"grad_norm": 0.7391118407249451,
"learning_rate": 3.047279666734072e-06,
"loss": 1.9231,
"step": 15420
},
{
"epoch": 1.9971704595982054,
"grad_norm": 0.7640807628631592,
"learning_rate": 3.0334315067452914e-06,
"loss": 1.932,
"step": 15440
},
{
"epoch": 1.9997574679655603,
"grad_norm": 0.7693647742271423,
"learning_rate": 3.019601169804216e-06,
"loss": 1.9367,
"step": 15460
}
],
"logging_steps": 20,
"max_steps": 23190,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 7.92643128386304e+19,
"train_batch_size": 12,
"trial_name": null,
"trial_params": null
}