Files
TR_SFT1/last-checkpoint/trainer_state.json
ModelHub XC 85af7d3826 初始化项目,由ModelHub XC社区提供模型
Model: Ba2han/TR_SFT1
Source: Original Platform
2026-10-08 03:17:19 +08:00

8559 lines
213 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 485,
"global_step": 2424,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008251676121712223,
"grad_norm": 4.05064582824707,
"learning_rate": 6.84931506849315e-06,
"loss": 4.001155376434326,
"step": 2
},
{
"epoch": 0.0016503352243424445,
"grad_norm": 2.6356191635131836,
"learning_rate": 2.0547945205479453e-05,
"loss": 2.9222776889801025,
"step": 4
},
{
"epoch": 0.002475502836513667,
"grad_norm": 3.0713953971862793,
"learning_rate": 3.424657534246575e-05,
"loss": 3.390582799911499,
"step": 6
},
{
"epoch": 0.003300670448684889,
"grad_norm": 1.3838273286819458,
"learning_rate": 4.794520547945205e-05,
"loss": 3.1828839778900146,
"step": 8
},
{
"epoch": 0.0041258380608561115,
"grad_norm": 1.7452095746994019,
"learning_rate": 6.164383561643835e-05,
"loss": 2.813962936401367,
"step": 10
},
{
"epoch": 0.004951005673027334,
"grad_norm": 1.5893399715423584,
"learning_rate": 7.534246575342465e-05,
"loss": 2.9807863235473633,
"step": 12
},
{
"epoch": 0.005776173285198556,
"grad_norm": 3.5792675018310547,
"learning_rate": 8.904109589041096e-05,
"loss": 3.5560591220855713,
"step": 14
},
{
"epoch": 0.006601340897369778,
"grad_norm": 1.8478755950927734,
"learning_rate": 0.00010273972602739725,
"loss": 3.3866186141967773,
"step": 16
},
{
"epoch": 0.007426508509541001,
"grad_norm": 2.1352012157440186,
"learning_rate": 0.00011643835616438356,
"loss": 3.1586074829101562,
"step": 18
},
{
"epoch": 0.008251676121712223,
"grad_norm": 1.7606539726257324,
"learning_rate": 0.00013013698630136986,
"loss": 2.9098258018493652,
"step": 20
},
{
"epoch": 0.009076843733883445,
"grad_norm": 3.2620508670806885,
"learning_rate": 0.00014383561643835615,
"loss": 3.561354160308838,
"step": 22
},
{
"epoch": 0.009902011346054668,
"grad_norm": 2.023899555206299,
"learning_rate": 0.00015753424657534247,
"loss": 3.592134952545166,
"step": 24
},
{
"epoch": 0.01072717895822589,
"grad_norm": 3.6786065101623535,
"learning_rate": 0.00017123287671232877,
"loss": 3.442713499069214,
"step": 26
},
{
"epoch": 0.011552346570397111,
"grad_norm": 2.0674569606781006,
"learning_rate": 0.00018493150684931506,
"loss": 3.746337652206421,
"step": 28
},
{
"epoch": 0.012377514182568335,
"grad_norm": 1.5185879468917847,
"learning_rate": 0.00019863013698630136,
"loss": 3.570563793182373,
"step": 30
},
{
"epoch": 0.013202681794739556,
"grad_norm": 2.360107421875,
"learning_rate": 0.00021232876712328765,
"loss": 3.3992598056793213,
"step": 32
},
{
"epoch": 0.01402784940691078,
"grad_norm": 1.5070563554763794,
"learning_rate": 0.00022602739726027398,
"loss": 3.5061898231506348,
"step": 34
},
{
"epoch": 0.014853017019082001,
"grad_norm": 2.7158215045928955,
"learning_rate": 0.00023972602739726027,
"loss": 3.3630847930908203,
"step": 36
},
{
"epoch": 0.015678184631253225,
"grad_norm": 2.2133257389068604,
"learning_rate": 0.0002534246575342466,
"loss": 3.105353832244873,
"step": 38
},
{
"epoch": 0.016503352243424446,
"grad_norm": 2.5868213176727295,
"learning_rate": 0.0002671232876712329,
"loss": 3.7335898876190186,
"step": 40
},
{
"epoch": 0.017328519855595668,
"grad_norm": 1.559792160987854,
"learning_rate": 0.0002808219178082192,
"loss": 2.9675424098968506,
"step": 42
},
{
"epoch": 0.01815368746776689,
"grad_norm": 1.5991140604019165,
"learning_rate": 0.0002945205479452055,
"loss": 3.0212419033050537,
"step": 44
},
{
"epoch": 0.01897885507993811,
"grad_norm": 1.1146661043167114,
"learning_rate": 0.00030821917808219177,
"loss": 3.196434736251831,
"step": 46
},
{
"epoch": 0.019804022692109336,
"grad_norm": 1.2657647132873535,
"learning_rate": 0.0003219178082191781,
"loss": 2.97530198097229,
"step": 48
},
{
"epoch": 0.020629190304280558,
"grad_norm": 0.6638312935829163,
"learning_rate": 0.0003356164383561644,
"loss": 2.9245412349700928,
"step": 50
},
{
"epoch": 0.02145435791645178,
"grad_norm": 1.4437497854232788,
"learning_rate": 0.0003493150684931507,
"loss": 3.0601556301116943,
"step": 52
},
{
"epoch": 0.022279525528623,
"grad_norm": 0.7249438166618347,
"learning_rate": 0.000363013698630137,
"loss": 3.048365831375122,
"step": 54
},
{
"epoch": 0.023104693140794223,
"grad_norm": 2.0431630611419678,
"learning_rate": 0.0003767123287671233,
"loss": 3.2283480167388916,
"step": 56
},
{
"epoch": 0.023929860752965448,
"grad_norm": 1.012539029121399,
"learning_rate": 0.0003904109589041096,
"loss": 3.074009895324707,
"step": 58
},
{
"epoch": 0.02475502836513667,
"grad_norm": 1.1344226598739624,
"learning_rate": 0.0004041095890410959,
"loss": 2.8867099285125732,
"step": 60
},
{
"epoch": 0.02558019597730789,
"grad_norm": 1.0907769203186035,
"learning_rate": 0.0004178082191780822,
"loss": 2.8230175971984863,
"step": 62
},
{
"epoch": 0.026405363589479112,
"grad_norm": 0.9112602472305298,
"learning_rate": 0.0004315068493150685,
"loss": 2.894538640975952,
"step": 64
},
{
"epoch": 0.027230531201650334,
"grad_norm": 0.885581374168396,
"learning_rate": 0.00044520547945205483,
"loss": 2.7423620223999023,
"step": 66
},
{
"epoch": 0.02805569881382156,
"grad_norm": 1.8324520587921143,
"learning_rate": 0.0004589041095890411,
"loss": 3.282146453857422,
"step": 68
},
{
"epoch": 0.02888086642599278,
"grad_norm": 0.9547421336174011,
"learning_rate": 0.0004726027397260274,
"loss": 3.039111614227295,
"step": 70
},
{
"epoch": 0.029706034038164002,
"grad_norm": 0.7706640362739563,
"learning_rate": 0.0004863013698630137,
"loss": 2.978135347366333,
"step": 72
},
{
"epoch": 0.030531201650335224,
"grad_norm": 0.6767997145652771,
"learning_rate": 0.0005,
"loss": 2.643972396850586,
"step": 74
},
{
"epoch": 0.03135636926250645,
"grad_norm": 1.3300039768218994,
"learning_rate": 0.0005,
"loss": 2.976358413696289,
"step": 76
},
{
"epoch": 0.03218153687467767,
"grad_norm": 1.024826169013977,
"learning_rate": 0.0005,
"loss": 2.843690872192383,
"step": 78
},
{
"epoch": 0.03300670448684889,
"grad_norm": 1.0081740617752075,
"learning_rate": 0.0005,
"loss": 2.936100959777832,
"step": 80
},
{
"epoch": 0.03383187209902011,
"grad_norm": 0.8867328763008118,
"learning_rate": 0.0005,
"loss": 2.9175610542297363,
"step": 82
},
{
"epoch": 0.034657039711191336,
"grad_norm": 0.8508431911468506,
"learning_rate": 0.0005,
"loss": 2.5730743408203125,
"step": 84
},
{
"epoch": 0.03548220732336256,
"grad_norm": 1.068881869316101,
"learning_rate": 0.0005,
"loss": 2.7962703704833984,
"step": 86
},
{
"epoch": 0.03630737493553378,
"grad_norm": 1.2468552589416504,
"learning_rate": 0.0005,
"loss": 3.0881192684173584,
"step": 88
},
{
"epoch": 0.037132542547705004,
"grad_norm": 1.1098690032958984,
"learning_rate": 0.0005,
"loss": 2.9157602787017822,
"step": 90
},
{
"epoch": 0.03795771015987622,
"grad_norm": 0.8747031092643738,
"learning_rate": 0.0005,
"loss": 2.8779027462005615,
"step": 92
},
{
"epoch": 0.03878287777204745,
"grad_norm": 1.1960872411727905,
"learning_rate": 0.0005,
"loss": 2.7578372955322266,
"step": 94
},
{
"epoch": 0.03960804538421867,
"grad_norm": 1.5936269760131836,
"learning_rate": 0.0005,
"loss": 2.854313611984253,
"step": 96
},
{
"epoch": 0.04043321299638989,
"grad_norm": 1.427709937095642,
"learning_rate": 0.0005,
"loss": 2.656238555908203,
"step": 98
},
{
"epoch": 0.041258380608561115,
"grad_norm": 1.1925889253616333,
"learning_rate": 0.0005,
"loss": 2.8787596225738525,
"step": 100
},
{
"epoch": 0.042083548220732334,
"grad_norm": 1.4617969989776611,
"learning_rate": 0.0005,
"loss": 2.7878036499023438,
"step": 102
},
{
"epoch": 0.04290871583290356,
"grad_norm": 1.3126530647277832,
"learning_rate": 0.0005,
"loss": 2.7677032947540283,
"step": 104
},
{
"epoch": 0.043733883445074784,
"grad_norm": 0.9550370573997498,
"learning_rate": 0.0005,
"loss": 2.496546745300293,
"step": 106
},
{
"epoch": 0.044559051057246,
"grad_norm": 0.9018437266349792,
"learning_rate": 0.0005,
"loss": 2.6963155269622803,
"step": 108
},
{
"epoch": 0.04538421866941723,
"grad_norm": 1.5497634410858154,
"learning_rate": 0.0005,
"loss": 2.4322242736816406,
"step": 110
},
{
"epoch": 0.046209386281588445,
"grad_norm": 0.9431282877922058,
"learning_rate": 0.0005,
"loss": 2.835155963897705,
"step": 112
},
{
"epoch": 0.04703455389375967,
"grad_norm": 0.9947378635406494,
"learning_rate": 0.0005,
"loss": 2.3825430870056152,
"step": 114
},
{
"epoch": 0.047859721505930895,
"grad_norm": 0.6523154973983765,
"learning_rate": 0.0005,
"loss": 2.450000524520874,
"step": 116
},
{
"epoch": 0.04868488911810211,
"grad_norm": 0.8956325650215149,
"learning_rate": 0.0005,
"loss": 2.425685167312622,
"step": 118
},
{
"epoch": 0.04951005673027334,
"grad_norm": 1.2782790660858154,
"learning_rate": 0.0005,
"loss": 2.6027579307556152,
"step": 120
},
{
"epoch": 0.05033522434244456,
"grad_norm": 1.0693634748458862,
"learning_rate": 0.0005,
"loss": 2.4755916595458984,
"step": 122
},
{
"epoch": 0.05116039195461578,
"grad_norm": 1.3470968008041382,
"learning_rate": 0.0005,
"loss": 2.6689329147338867,
"step": 124
},
{
"epoch": 0.05198555956678701,
"grad_norm": 1.2954727411270142,
"learning_rate": 0.0005,
"loss": 2.4596047401428223,
"step": 126
},
{
"epoch": 0.052810727178958225,
"grad_norm": 1.1295737028121948,
"learning_rate": 0.0005,
"loss": 2.714158058166504,
"step": 128
},
{
"epoch": 0.05363589479112945,
"grad_norm": 0.5364153385162354,
"learning_rate": 0.0005,
"loss": 2.4372639656066895,
"step": 130
},
{
"epoch": 0.05446106240330067,
"grad_norm": 0.5712692141532898,
"learning_rate": 0.0005,
"loss": 2.6134886741638184,
"step": 132
},
{
"epoch": 0.05528623001547189,
"grad_norm": 0.5640398859977722,
"learning_rate": 0.0005,
"loss": 2.5374679565429688,
"step": 134
},
{
"epoch": 0.05611139762764312,
"grad_norm": 0.6347931623458862,
"learning_rate": 0.0005,
"loss": 2.3327348232269287,
"step": 136
},
{
"epoch": 0.05693656523981434,
"grad_norm": 0.8012130856513977,
"learning_rate": 0.0005,
"loss": 2.5947751998901367,
"step": 138
},
{
"epoch": 0.05776173285198556,
"grad_norm": 0.7999157309532166,
"learning_rate": 0.0005,
"loss": 2.4820642471313477,
"step": 140
},
{
"epoch": 0.05858690046415678,
"grad_norm": 0.7587979435920715,
"learning_rate": 0.0005,
"loss": 2.509951114654541,
"step": 142
},
{
"epoch": 0.059412068076328005,
"grad_norm": 1.1788878440856934,
"learning_rate": 0.0005,
"loss": 2.5577077865600586,
"step": 144
},
{
"epoch": 0.06023723568849922,
"grad_norm": 0.5499687194824219,
"learning_rate": 0.0005,
"loss": 2.3385090827941895,
"step": 146
},
{
"epoch": 0.06106240330067045,
"grad_norm": 1.1547513008117676,
"learning_rate": 0.0005,
"loss": 2.2788970470428467,
"step": 148
},
{
"epoch": 0.06188757091284167,
"grad_norm": 0.5423986911773682,
"learning_rate": 0.0005,
"loss": 2.26416277885437,
"step": 150
},
{
"epoch": 0.0627127385250129,
"grad_norm": 1.07978355884552,
"learning_rate": 0.0005,
"loss": 2.109246253967285,
"step": 152
},
{
"epoch": 0.06353790613718412,
"grad_norm": 1.0858091115951538,
"learning_rate": 0.0005,
"loss": 2.4146788120269775,
"step": 154
},
{
"epoch": 0.06436307374935533,
"grad_norm": 1.2315399646759033,
"learning_rate": 0.0005,
"loss": 2.4686007499694824,
"step": 156
},
{
"epoch": 0.06518824136152657,
"grad_norm": 1.1423070430755615,
"learning_rate": 0.0005,
"loss": 2.235238552093506,
"step": 158
},
{
"epoch": 0.06601340897369778,
"grad_norm": 0.6924282908439636,
"learning_rate": 0.0005,
"loss": 2.2394375801086426,
"step": 160
},
{
"epoch": 0.066838576585869,
"grad_norm": 0.6953537464141846,
"learning_rate": 0.0005,
"loss": 2.2416467666625977,
"step": 162
},
{
"epoch": 0.06766374419804022,
"grad_norm": 1.0658975839614868,
"learning_rate": 0.0005,
"loss": 2.3240959644317627,
"step": 164
},
{
"epoch": 0.06848891181021145,
"grad_norm": 1.0153752565383911,
"learning_rate": 0.0005,
"loss": 2.3121490478515625,
"step": 166
},
{
"epoch": 0.06931407942238267,
"grad_norm": 0.7716209292411804,
"learning_rate": 0.0005,
"loss": 2.3162643909454346,
"step": 168
},
{
"epoch": 0.07013924703455389,
"grad_norm": 0.7019490003585815,
"learning_rate": 0.0005,
"loss": 2.0751984119415283,
"step": 170
},
{
"epoch": 0.07096441464672512,
"grad_norm": 0.8617009520530701,
"learning_rate": 0.0005,
"loss": 2.2369861602783203,
"step": 172
},
{
"epoch": 0.07178958225889634,
"grad_norm": 0.9819467663764954,
"learning_rate": 0.0005,
"loss": 2.1822731494903564,
"step": 174
},
{
"epoch": 0.07261474987106756,
"grad_norm": 0.9508243203163147,
"learning_rate": 0.0005,
"loss": 2.303833484649658,
"step": 176
},
{
"epoch": 0.07343991748323879,
"grad_norm": 0.7223880290985107,
"learning_rate": 0.0005,
"loss": 2.270758628845215,
"step": 178
},
{
"epoch": 0.07426508509541001,
"grad_norm": 0.5906323194503784,
"learning_rate": 0.0005,
"loss": 2.12703275680542,
"step": 180
},
{
"epoch": 0.07509025270758123,
"grad_norm": 1.011453628540039,
"learning_rate": 0.0005,
"loss": 2.1704506874084473,
"step": 182
},
{
"epoch": 0.07591542031975244,
"grad_norm": 1.1180076599121094,
"learning_rate": 0.0005,
"loss": 2.135690450668335,
"step": 184
},
{
"epoch": 0.07674058793192368,
"grad_norm": 0.6078499555587769,
"learning_rate": 0.0005,
"loss": 2.0526981353759766,
"step": 186
},
{
"epoch": 0.0775657555440949,
"grad_norm": 0.6462226510047913,
"learning_rate": 0.0005,
"loss": 2.0516133308410645,
"step": 188
},
{
"epoch": 0.07839092315626611,
"grad_norm": 0.6533961296081543,
"learning_rate": 0.0005,
"loss": 2.1560380458831787,
"step": 190
},
{
"epoch": 0.07921609076843734,
"grad_norm": 0.4791254699230194,
"learning_rate": 0.0005,
"loss": 2.2701005935668945,
"step": 192
},
{
"epoch": 0.08004125838060856,
"grad_norm": 0.6959493160247803,
"learning_rate": 0.0005,
"loss": 2.1827073097229004,
"step": 194
},
{
"epoch": 0.08086642599277978,
"grad_norm": 0.5518482327461243,
"learning_rate": 0.0005,
"loss": 2.0163729190826416,
"step": 196
},
{
"epoch": 0.081691593604951,
"grad_norm": 0.6662017107009888,
"learning_rate": 0.0005,
"loss": 2.2138278484344482,
"step": 198
},
{
"epoch": 0.08251676121712223,
"grad_norm": 0.5524933338165283,
"learning_rate": 0.0005,
"loss": 2.0688583850860596,
"step": 200
},
{
"epoch": 0.08334192882929345,
"grad_norm": 0.6075630187988281,
"learning_rate": 0.0005,
"loss": 2.1106863021850586,
"step": 202
},
{
"epoch": 0.08416709644146467,
"grad_norm": 0.6824833154678345,
"learning_rate": 0.0005,
"loss": 1.9529650211334229,
"step": 204
},
{
"epoch": 0.0849922640536359,
"grad_norm": 0.7032500505447388,
"learning_rate": 0.0005,
"loss": 2.0248687267303467,
"step": 206
},
{
"epoch": 0.08581743166580712,
"grad_norm": 0.7788276672363281,
"learning_rate": 0.0005,
"loss": 2.039890766143799,
"step": 208
},
{
"epoch": 0.08664259927797834,
"grad_norm": 1.1545436382293701,
"learning_rate": 0.0005,
"loss": 2.0932705402374268,
"step": 210
},
{
"epoch": 0.08746776689014957,
"grad_norm": 0.6618006825447083,
"learning_rate": 0.0005,
"loss": 1.8879495859146118,
"step": 212
},
{
"epoch": 0.08829293450232079,
"grad_norm": 1.069516658782959,
"learning_rate": 0.0005,
"loss": 2.060929775238037,
"step": 214
},
{
"epoch": 0.089118102114492,
"grad_norm": 0.7089823484420776,
"learning_rate": 0.0005,
"loss": 2.086873769760132,
"step": 216
},
{
"epoch": 0.08994326972666322,
"grad_norm": 0.8543140292167664,
"learning_rate": 0.0005,
"loss": 1.8984220027923584,
"step": 218
},
{
"epoch": 0.09076843733883445,
"grad_norm": 0.7802907824516296,
"learning_rate": 0.0005,
"loss": 1.944888949394226,
"step": 220
},
{
"epoch": 0.09159360495100567,
"grad_norm": 0.7008336186408997,
"learning_rate": 0.0005,
"loss": 1.9289171695709229,
"step": 222
},
{
"epoch": 0.09241877256317689,
"grad_norm": 0.9346032738685608,
"learning_rate": 0.0005,
"loss": 1.9859205484390259,
"step": 224
},
{
"epoch": 0.09324394017534812,
"grad_norm": 0.681209921836853,
"learning_rate": 0.0005,
"loss": 2.0038955211639404,
"step": 226
},
{
"epoch": 0.09406910778751934,
"grad_norm": 0.6827794313430786,
"learning_rate": 0.0005,
"loss": 2.0127346515655518,
"step": 228
},
{
"epoch": 0.09489427539969056,
"grad_norm": 1.0419977903366089,
"learning_rate": 0.0005,
"loss": 1.6082828044891357,
"step": 230
},
{
"epoch": 0.09571944301186179,
"grad_norm": 0.9894194602966309,
"learning_rate": 0.0005,
"loss": 1.8612134456634521,
"step": 232
},
{
"epoch": 0.09654461062403301,
"grad_norm": 0.509680449962616,
"learning_rate": 0.0005,
"loss": 1.8543392419815063,
"step": 234
},
{
"epoch": 0.09736977823620423,
"grad_norm": 0.7202760577201843,
"learning_rate": 0.0005,
"loss": 1.8081979751586914,
"step": 236
},
{
"epoch": 0.09819494584837545,
"grad_norm": 0.7833833694458008,
"learning_rate": 0.0005,
"loss": 2.022106647491455,
"step": 238
},
{
"epoch": 0.09902011346054668,
"grad_norm": 0.4884537160396576,
"learning_rate": 0.0005,
"loss": 1.8525521755218506,
"step": 240
},
{
"epoch": 0.0998452810727179,
"grad_norm": 0.7342460751533508,
"learning_rate": 0.0005,
"loss": 1.8639775514602661,
"step": 242
},
{
"epoch": 0.10067044868488911,
"grad_norm": 0.7737444639205933,
"learning_rate": 0.0005,
"loss": 2.0264413356781006,
"step": 244
},
{
"epoch": 0.10149561629706035,
"grad_norm": 0.8027350902557373,
"learning_rate": 0.0005,
"loss": 1.8213112354278564,
"step": 246
},
{
"epoch": 0.10232078390923156,
"grad_norm": 0.7654951214790344,
"learning_rate": 0.0005,
"loss": 1.882145643234253,
"step": 248
},
{
"epoch": 0.10314595152140278,
"grad_norm": 0.7303850650787354,
"learning_rate": 0.0005,
"loss": 1.977327823638916,
"step": 250
},
{
"epoch": 0.10397111913357401,
"grad_norm": 0.5510226488113403,
"learning_rate": 0.0005,
"loss": 1.9279447793960571,
"step": 252
},
{
"epoch": 0.10479628674574523,
"grad_norm": 1.20448899269104,
"learning_rate": 0.0005,
"loss": 1.7835114002227783,
"step": 254
},
{
"epoch": 0.10562145435791645,
"grad_norm": 1.0279170274734497,
"learning_rate": 0.0005,
"loss": 1.5774431228637695,
"step": 256
},
{
"epoch": 0.10644662197008767,
"grad_norm": 0.6120237708091736,
"learning_rate": 0.0005,
"loss": 1.9870978593826294,
"step": 258
},
{
"epoch": 0.1072717895822589,
"grad_norm": 0.8543900847434998,
"learning_rate": 0.0005,
"loss": 1.7867107391357422,
"step": 260
},
{
"epoch": 0.10809695719443012,
"grad_norm": 0.5745635628700256,
"learning_rate": 0.0005,
"loss": 1.6808240413665771,
"step": 262
},
{
"epoch": 0.10892212480660134,
"grad_norm": 0.5888904929161072,
"learning_rate": 0.0005,
"loss": 1.6171236038208008,
"step": 264
},
{
"epoch": 0.10974729241877257,
"grad_norm": 1.1007593870162964,
"learning_rate": 0.0005,
"loss": 1.6148968935012817,
"step": 266
},
{
"epoch": 0.11057246003094379,
"grad_norm": 0.6041141152381897,
"learning_rate": 0.0005,
"loss": 1.463856816291809,
"step": 268
},
{
"epoch": 0.111397627643115,
"grad_norm": 0.6426807641983032,
"learning_rate": 0.0005,
"loss": 1.719922661781311,
"step": 270
},
{
"epoch": 0.11222279525528624,
"grad_norm": 1.201188564300537,
"learning_rate": 0.0005,
"loss": 1.6529117822647095,
"step": 272
},
{
"epoch": 0.11304796286745745,
"grad_norm": 0.8614847660064697,
"learning_rate": 0.0005,
"loss": 1.7518805265426636,
"step": 274
},
{
"epoch": 0.11387313047962867,
"grad_norm": 0.869533121585846,
"learning_rate": 0.0005,
"loss": 1.6316742897033691,
"step": 276
},
{
"epoch": 0.11469829809179989,
"grad_norm": 1.023616909980774,
"learning_rate": 0.0005,
"loss": 1.7410283088684082,
"step": 278
},
{
"epoch": 0.11552346570397112,
"grad_norm": 0.4677686393260956,
"learning_rate": 0.0005,
"loss": 1.838667392730713,
"step": 280
},
{
"epoch": 0.11634863331614234,
"grad_norm": 0.7932203412055969,
"learning_rate": 0.0005,
"loss": 1.8326950073242188,
"step": 282
},
{
"epoch": 0.11717380092831356,
"grad_norm": 0.6294450759887695,
"learning_rate": 0.0005,
"loss": 1.6988697052001953,
"step": 284
},
{
"epoch": 0.11799896854048479,
"grad_norm": 0.6818397641181946,
"learning_rate": 0.0005,
"loss": 1.7579114437103271,
"step": 286
},
{
"epoch": 0.11882413615265601,
"grad_norm": 1.1268901824951172,
"learning_rate": 0.0005,
"loss": 1.6934425830841064,
"step": 288
},
{
"epoch": 0.11964930376482723,
"grad_norm": 1.1641162633895874,
"learning_rate": 0.0005,
"loss": 1.480327844619751,
"step": 290
},
{
"epoch": 0.12047447137699845,
"grad_norm": 0.9022796750068665,
"learning_rate": 0.0005,
"loss": 1.4163341522216797,
"step": 292
},
{
"epoch": 0.12129963898916968,
"grad_norm": 0.6819456815719604,
"learning_rate": 0.0005,
"loss": 1.4166662693023682,
"step": 294
},
{
"epoch": 0.1221248066013409,
"grad_norm": 0.8714990019798279,
"learning_rate": 0.0005,
"loss": 1.642352819442749,
"step": 296
},
{
"epoch": 0.12294997421351211,
"grad_norm": 0.6552049517631531,
"learning_rate": 0.0005,
"loss": 1.8244096040725708,
"step": 298
},
{
"epoch": 0.12377514182568335,
"grad_norm": 0.6223987936973572,
"learning_rate": 0.0005,
"loss": 1.5980300903320312,
"step": 300
},
{
"epoch": 0.12460030943785456,
"grad_norm": 0.7065282464027405,
"learning_rate": 0.0005,
"loss": 1.418002724647522,
"step": 302
},
{
"epoch": 0.1254254770500258,
"grad_norm": 0.5670342445373535,
"learning_rate": 0.0005,
"loss": 1.7273845672607422,
"step": 304
},
{
"epoch": 0.12625064466219701,
"grad_norm": 0.9579104781150818,
"learning_rate": 0.0005,
"loss": 1.513790488243103,
"step": 306
},
{
"epoch": 0.12707581227436823,
"grad_norm": 0.6429880261421204,
"learning_rate": 0.0005,
"loss": 1.7069754600524902,
"step": 308
},
{
"epoch": 0.12790097988653945,
"grad_norm": 0.8442540764808655,
"learning_rate": 0.0005,
"loss": 1.6215064525604248,
"step": 310
},
{
"epoch": 0.12872614749871067,
"grad_norm": 1.2151155471801758,
"learning_rate": 0.0005,
"loss": 1.5524238348007202,
"step": 312
},
{
"epoch": 0.1295513151108819,
"grad_norm": 0.6252835392951965,
"learning_rate": 0.0005,
"loss": 1.595234751701355,
"step": 314
},
{
"epoch": 0.13037648272305313,
"grad_norm": 0.5894163250923157,
"learning_rate": 0.0005,
"loss": 1.6958962678909302,
"step": 316
},
{
"epoch": 0.13120165033522435,
"grad_norm": 0.6581782102584839,
"learning_rate": 0.0005,
"loss": 1.4035930633544922,
"step": 318
},
{
"epoch": 0.13202681794739557,
"grad_norm": 0.542962372303009,
"learning_rate": 0.0005,
"loss": 1.771968960762024,
"step": 320
},
{
"epoch": 0.1328519855595668,
"grad_norm": 0.5399895310401917,
"learning_rate": 0.0005,
"loss": 1.8783142566680908,
"step": 322
},
{
"epoch": 0.133677153171738,
"grad_norm": 0.7295765280723572,
"learning_rate": 0.0005,
"loss": 1.4890893697738647,
"step": 324
},
{
"epoch": 0.13450232078390922,
"grad_norm": 0.45351487398147583,
"learning_rate": 0.0005,
"loss": 1.6354293823242188,
"step": 326
},
{
"epoch": 0.13532748839608044,
"grad_norm": 1.0010162591934204,
"learning_rate": 0.0005,
"loss": 1.0240691900253296,
"step": 328
},
{
"epoch": 0.1361526560082517,
"grad_norm": 0.8697746396064758,
"learning_rate": 0.0005,
"loss": 1.691316843032837,
"step": 330
},
{
"epoch": 0.1369778236204229,
"grad_norm": 0.7309480309486389,
"learning_rate": 0.0005,
"loss": 1.5780876874923706,
"step": 332
},
{
"epoch": 0.13780299123259412,
"grad_norm": 0.7008673548698425,
"learning_rate": 0.0005,
"loss": 1.2456053495407104,
"step": 334
},
{
"epoch": 0.13862815884476534,
"grad_norm": 0.9308040142059326,
"learning_rate": 0.0005,
"loss": 1.4136176109313965,
"step": 336
},
{
"epoch": 0.13945332645693656,
"grad_norm": 0.4375256299972534,
"learning_rate": 0.0005,
"loss": 1.7965747117996216,
"step": 338
},
{
"epoch": 0.14027849406910778,
"grad_norm": 0.6675146818161011,
"learning_rate": 0.0005,
"loss": 1.2955620288848877,
"step": 340
},
{
"epoch": 0.141103661681279,
"grad_norm": 0.7654633522033691,
"learning_rate": 0.0005,
"loss": 1.521423578262329,
"step": 342
},
{
"epoch": 0.14192882929345024,
"grad_norm": 0.5734323263168335,
"learning_rate": 0.0005,
"loss": 1.5977249145507812,
"step": 344
},
{
"epoch": 0.14275399690562146,
"grad_norm": 0.9441736340522766,
"learning_rate": 0.0005,
"loss": 1.2087130546569824,
"step": 346
},
{
"epoch": 0.14357916451779268,
"grad_norm": 0.6317110061645508,
"learning_rate": 0.0005,
"loss": 1.308985710144043,
"step": 348
},
{
"epoch": 0.1444043321299639,
"grad_norm": 0.6477823853492737,
"learning_rate": 0.0005,
"loss": 1.2664823532104492,
"step": 350
},
{
"epoch": 0.14522949974213512,
"grad_norm": 0.7252758741378784,
"learning_rate": 0.0005,
"loss": 1.548954725265503,
"step": 352
},
{
"epoch": 0.14605466735430633,
"grad_norm": 0.6475541591644287,
"learning_rate": 0.0005,
"loss": 1.3655939102172852,
"step": 354
},
{
"epoch": 0.14687983496647758,
"grad_norm": 1.03873610496521,
"learning_rate": 0.0005,
"loss": 1.167594075202942,
"step": 356
},
{
"epoch": 0.1477050025786488,
"grad_norm": 0.5233055949211121,
"learning_rate": 0.0005,
"loss": 1.6682946681976318,
"step": 358
},
{
"epoch": 0.14853017019082002,
"grad_norm": 0.45791342854499817,
"learning_rate": 0.0005,
"loss": 1.7371954917907715,
"step": 360
},
{
"epoch": 0.14935533780299123,
"grad_norm": 0.48204395174980164,
"learning_rate": 0.0005,
"loss": 1.6079514026641846,
"step": 362
},
{
"epoch": 0.15018050541516245,
"grad_norm": 0.5738229155540466,
"learning_rate": 0.0005,
"loss": 1.3978792428970337,
"step": 364
},
{
"epoch": 0.15100567302733367,
"grad_norm": 1.272986650466919,
"learning_rate": 0.0005,
"loss": 1.3077946901321411,
"step": 366
},
{
"epoch": 0.1518308406395049,
"grad_norm": 0.6108178496360779,
"learning_rate": 0.0005,
"loss": 1.6558018922805786,
"step": 368
},
{
"epoch": 0.15265600825167613,
"grad_norm": 0.5422089099884033,
"learning_rate": 0.0005,
"loss": 1.2281129360198975,
"step": 370
},
{
"epoch": 0.15348117586384735,
"grad_norm": 0.6628785729408264,
"learning_rate": 0.0005,
"loss": 1.4119315147399902,
"step": 372
},
{
"epoch": 0.15430634347601857,
"grad_norm": 0.4428359270095825,
"learning_rate": 0.0005,
"loss": 1.416772723197937,
"step": 374
},
{
"epoch": 0.1551315110881898,
"grad_norm": 1.1590735912322998,
"learning_rate": 0.0005,
"loss": 1.2648556232452393,
"step": 376
},
{
"epoch": 0.155956678700361,
"grad_norm": 0.5401804447174072,
"learning_rate": 0.0005,
"loss": 1.3954392671585083,
"step": 378
},
{
"epoch": 0.15678184631253222,
"grad_norm": 0.6105924844741821,
"learning_rate": 0.0005,
"loss": 1.529508352279663,
"step": 380
},
{
"epoch": 0.15760701392470344,
"grad_norm": 0.6930528879165649,
"learning_rate": 0.0005,
"loss": 1.2204009294509888,
"step": 382
},
{
"epoch": 0.1584321815368747,
"grad_norm": 1.0452812910079956,
"learning_rate": 0.0005,
"loss": 1.0644744634628296,
"step": 384
},
{
"epoch": 0.1592573491490459,
"grad_norm": 0.6817145347595215,
"learning_rate": 0.0005,
"loss": 1.3310097455978394,
"step": 386
},
{
"epoch": 0.16008251676121713,
"grad_norm": 0.6954227685928345,
"learning_rate": 0.0005,
"loss": 1.250698447227478,
"step": 388
},
{
"epoch": 0.16090768437338834,
"grad_norm": 0.5326175689697266,
"learning_rate": 0.0005,
"loss": 1.4613834619522095,
"step": 390
},
{
"epoch": 0.16173285198555956,
"grad_norm": 0.43815600872039795,
"learning_rate": 0.0005,
"loss": 1.557671070098877,
"step": 392
},
{
"epoch": 0.16255801959773078,
"grad_norm": 0.4776799976825714,
"learning_rate": 0.0005,
"loss": 1.5681153535842896,
"step": 394
},
{
"epoch": 0.163383187209902,
"grad_norm": 0.906951904296875,
"learning_rate": 0.0005,
"loss": 1.061201810836792,
"step": 396
},
{
"epoch": 0.16420835482207324,
"grad_norm": 0.6454389691352844,
"learning_rate": 0.0005,
"loss": 1.5125031471252441,
"step": 398
},
{
"epoch": 0.16503352243424446,
"grad_norm": 0.45719677209854126,
"learning_rate": 0.0005,
"loss": 1.5531091690063477,
"step": 400
},
{
"epoch": 0.16585869004641568,
"grad_norm": 0.8183549642562866,
"learning_rate": 0.0005,
"loss": 1.5622446537017822,
"step": 402
},
{
"epoch": 0.1666838576585869,
"grad_norm": 0.4710141122341156,
"learning_rate": 0.0005,
"loss": 1.4995447397232056,
"step": 404
},
{
"epoch": 0.16750902527075812,
"grad_norm": 0.42731228470802307,
"learning_rate": 0.0005,
"loss": 1.4743670225143433,
"step": 406
},
{
"epoch": 0.16833419288292933,
"grad_norm": 0.5082319378852844,
"learning_rate": 0.0005,
"loss": 1.5029337406158447,
"step": 408
},
{
"epoch": 0.16915936049510058,
"grad_norm": 0.8508438467979431,
"learning_rate": 0.0005,
"loss": 0.7607795596122742,
"step": 410
},
{
"epoch": 0.1699845281072718,
"grad_norm": 0.6989767551422119,
"learning_rate": 0.0005,
"loss": 1.1386264562606812,
"step": 412
},
{
"epoch": 0.17080969571944302,
"grad_norm": 0.8017994165420532,
"learning_rate": 0.0005,
"loss": 1.2077910900115967,
"step": 414
},
{
"epoch": 0.17163486333161423,
"grad_norm": 0.5111270546913147,
"learning_rate": 0.0005,
"loss": 1.4858042001724243,
"step": 416
},
{
"epoch": 0.17246003094378545,
"grad_norm": 0.44866520166397095,
"learning_rate": 0.0005,
"loss": 1.493560791015625,
"step": 418
},
{
"epoch": 0.17328519855595667,
"grad_norm": 0.5808163285255432,
"learning_rate": 0.0005,
"loss": 1.1808210611343384,
"step": 420
},
{
"epoch": 0.1741103661681279,
"grad_norm": 0.5781272649765015,
"learning_rate": 0.0005,
"loss": 1.2493834495544434,
"step": 422
},
{
"epoch": 0.17493553378029914,
"grad_norm": 0.7005312442779541,
"learning_rate": 0.0005,
"loss": 1.2910864353179932,
"step": 424
},
{
"epoch": 0.17576070139247035,
"grad_norm": 0.9330860376358032,
"learning_rate": 0.0005,
"loss": 1.0809887647628784,
"step": 426
},
{
"epoch": 0.17658586900464157,
"grad_norm": 0.5688995718955994,
"learning_rate": 0.0005,
"loss": 1.4833179712295532,
"step": 428
},
{
"epoch": 0.1774110366168128,
"grad_norm": 0.4832324683666229,
"learning_rate": 0.0005,
"loss": 1.7395341396331787,
"step": 430
},
{
"epoch": 0.178236204228984,
"grad_norm": 0.6329236626625061,
"learning_rate": 0.0005,
"loss": 0.9554226398468018,
"step": 432
},
{
"epoch": 0.17906137184115523,
"grad_norm": 0.523956298828125,
"learning_rate": 0.0005,
"loss": 1.1947340965270996,
"step": 434
},
{
"epoch": 0.17988653945332644,
"grad_norm": 0.7409864068031311,
"learning_rate": 0.0005,
"loss": 1.2951687574386597,
"step": 436
},
{
"epoch": 0.1807117070654977,
"grad_norm": 0.560711681842804,
"learning_rate": 0.0005,
"loss": 1.3486658334732056,
"step": 438
},
{
"epoch": 0.1815368746776689,
"grad_norm": 0.65752774477005,
"learning_rate": 0.0005,
"loss": 1.2709574699401855,
"step": 440
},
{
"epoch": 0.18236204228984013,
"grad_norm": 0.6136254668235779,
"learning_rate": 0.0005,
"loss": 1.168439507484436,
"step": 442
},
{
"epoch": 0.18318720990201134,
"grad_norm": 0.6078649759292603,
"learning_rate": 0.0005,
"loss": 1.2218141555786133,
"step": 444
},
{
"epoch": 0.18401237751418256,
"grad_norm": 0.5812203288078308,
"learning_rate": 0.0005,
"loss": 1.5448240041732788,
"step": 446
},
{
"epoch": 0.18483754512635378,
"grad_norm": 0.7549668550491333,
"learning_rate": 0.0005,
"loss": 1.484325647354126,
"step": 448
},
{
"epoch": 0.18566271273852503,
"grad_norm": 0.5121369361877441,
"learning_rate": 0.0005,
"loss": 1.2179449796676636,
"step": 450
},
{
"epoch": 0.18648788035069624,
"grad_norm": 0.6205900311470032,
"learning_rate": 0.0005,
"loss": 1.2502247095108032,
"step": 452
},
{
"epoch": 0.18731304796286746,
"grad_norm": 0.8578463792800903,
"learning_rate": 0.0005,
"loss": 0.7786793112754822,
"step": 454
},
{
"epoch": 0.18813821557503868,
"grad_norm": 0.5800952911376953,
"learning_rate": 0.0005,
"loss": 1.3457585573196411,
"step": 456
},
{
"epoch": 0.1889633831872099,
"grad_norm": 0.6097033023834229,
"learning_rate": 0.0005,
"loss": 1.402034878730774,
"step": 458
},
{
"epoch": 0.18978855079938112,
"grad_norm": 0.7255859971046448,
"learning_rate": 0.0005,
"loss": 1.3664675951004028,
"step": 460
},
{
"epoch": 0.19061371841155234,
"grad_norm": 0.7353418469429016,
"learning_rate": 0.0005,
"loss": 1.2583798170089722,
"step": 462
},
{
"epoch": 0.19143888602372358,
"grad_norm": 0.5628712773323059,
"learning_rate": 0.0005,
"loss": 1.3858345746994019,
"step": 464
},
{
"epoch": 0.1922640536358948,
"grad_norm": 0.5797217488288879,
"learning_rate": 0.0005,
"loss": 1.0521609783172607,
"step": 466
},
{
"epoch": 0.19308922124806602,
"grad_norm": 0.6121135950088501,
"learning_rate": 0.0005,
"loss": 1.0847803354263306,
"step": 468
},
{
"epoch": 0.19391438886023724,
"grad_norm": 0.4923863708972931,
"learning_rate": 0.0005,
"loss": 1.081227421760559,
"step": 470
},
{
"epoch": 0.19473955647240845,
"grad_norm": 0.8833675384521484,
"learning_rate": 0.0005,
"loss": 0.9793055057525635,
"step": 472
},
{
"epoch": 0.19556472408457967,
"grad_norm": 0.403506338596344,
"learning_rate": 0.0005,
"loss": 1.6908206939697266,
"step": 474
},
{
"epoch": 0.1963898916967509,
"grad_norm": 0.5630794763565063,
"learning_rate": 0.0005,
"loss": 1.206038236618042,
"step": 476
},
{
"epoch": 0.19721505930892214,
"grad_norm": 0.6680169701576233,
"learning_rate": 0.0005,
"loss": 1.1213828325271606,
"step": 478
},
{
"epoch": 0.19804022692109335,
"grad_norm": 0.9035966396331787,
"learning_rate": 0.0005,
"loss": 0.7148669362068176,
"step": 480
},
{
"epoch": 0.19886539453326457,
"grad_norm": 0.820414662361145,
"learning_rate": 0.0005,
"loss": 0.9385108947753906,
"step": 482
},
{
"epoch": 0.1996905621454358,
"grad_norm": 0.46335649490356445,
"learning_rate": 0.0005,
"loss": 1.353764295578003,
"step": 484
},
{
"epoch": 0.2001031459515214,
"eval_loss": 0.9871540665626526,
"eval_runtime": 8.3235,
"eval_samples_per_second": 94.191,
"eval_steps_per_second": 11.774,
"step": 485
},
{
"epoch": 0.200515729757607,
"grad_norm": 0.5083919167518616,
"learning_rate": 0.0005,
"loss": 1.2146344184875488,
"step": 486
},
{
"epoch": 0.20134089736977823,
"grad_norm": 0.5267910361289978,
"learning_rate": 0.0005,
"loss": 1.1352005004882812,
"step": 488
},
{
"epoch": 0.20216606498194944,
"grad_norm": 0.5246739983558655,
"learning_rate": 0.0005,
"loss": 1.3788514137268066,
"step": 490
},
{
"epoch": 0.2029912325941207,
"grad_norm": 0.6588636040687561,
"learning_rate": 0.0005,
"loss": 1.2196016311645508,
"step": 492
},
{
"epoch": 0.2038164002062919,
"grad_norm": 0.4793865382671356,
"learning_rate": 0.0005,
"loss": 1.5185762643814087,
"step": 494
},
{
"epoch": 0.20464156781846313,
"grad_norm": 0.46975746750831604,
"learning_rate": 0.0005,
"loss": 1.0828098058700562,
"step": 496
},
{
"epoch": 0.20546673543063435,
"grad_norm": 0.8231362700462341,
"learning_rate": 0.0005,
"loss": 1.0690791606903076,
"step": 498
},
{
"epoch": 0.20629190304280556,
"grad_norm": 0.4743996262550354,
"learning_rate": 0.0005,
"loss": 1.3010034561157227,
"step": 500
},
{
"epoch": 0.20711707065497678,
"grad_norm": 1.073641061782837,
"learning_rate": 0.0005,
"loss": 0.998904824256897,
"step": 502
},
{
"epoch": 0.20794223826714803,
"grad_norm": 0.653026819229126,
"learning_rate": 0.0005,
"loss": 1.1743905544281006,
"step": 504
},
{
"epoch": 0.20876740587931925,
"grad_norm": 0.5757710337638855,
"learning_rate": 0.0005,
"loss": 1.3736557960510254,
"step": 506
},
{
"epoch": 0.20959257349149046,
"grad_norm": 1.029820203781128,
"learning_rate": 0.0005,
"loss": 0.9813966751098633,
"step": 508
},
{
"epoch": 0.21041774110366168,
"grad_norm": 0.44173821806907654,
"learning_rate": 0.0005,
"loss": 1.4481749534606934,
"step": 510
},
{
"epoch": 0.2112429087158329,
"grad_norm": 0.9574601054191589,
"learning_rate": 0.0005,
"loss": 1.1396448612213135,
"step": 512
},
{
"epoch": 0.21206807632800412,
"grad_norm": 0.8778027296066284,
"learning_rate": 0.0005,
"loss": 1.1219650506973267,
"step": 514
},
{
"epoch": 0.21289324394017534,
"grad_norm": 0.6104379296302795,
"learning_rate": 0.0005,
"loss": 1.2482566833496094,
"step": 516
},
{
"epoch": 0.21371841155234658,
"grad_norm": 0.5913075804710388,
"learning_rate": 0.0005,
"loss": 1.3111701011657715,
"step": 518
},
{
"epoch": 0.2145435791645178,
"grad_norm": 0.8201984167098999,
"learning_rate": 0.0005,
"loss": 1.2918312549591064,
"step": 520
},
{
"epoch": 0.21536874677668902,
"grad_norm": 1.0435935258865356,
"learning_rate": 0.0005,
"loss": 0.7960109114646912,
"step": 522
},
{
"epoch": 0.21619391438886024,
"grad_norm": 0.5125344395637512,
"learning_rate": 0.0005,
"loss": 1.7052984237670898,
"step": 524
},
{
"epoch": 0.21701908200103145,
"grad_norm": 0.4323374032974243,
"learning_rate": 0.0005,
"loss": 1.375638723373413,
"step": 526
},
{
"epoch": 0.21784424961320267,
"grad_norm": 0.9039424657821655,
"learning_rate": 0.0005,
"loss": 0.8595261573791504,
"step": 528
},
{
"epoch": 0.2186694172253739,
"grad_norm": 1.0073490142822266,
"learning_rate": 0.0005,
"loss": 0.5834345817565918,
"step": 530
},
{
"epoch": 0.21949458483754514,
"grad_norm": 0.42047837376594543,
"learning_rate": 0.0005,
"loss": 1.409820556640625,
"step": 532
},
{
"epoch": 0.22031975244971636,
"grad_norm": 0.8841076493263245,
"learning_rate": 0.0005,
"loss": 0.6091881394386292,
"step": 534
},
{
"epoch": 0.22114492006188757,
"grad_norm": 0.5746383666992188,
"learning_rate": 0.0005,
"loss": 1.3961387872695923,
"step": 536
},
{
"epoch": 0.2219700876740588,
"grad_norm": 0.5998536944389343,
"learning_rate": 0.0005,
"loss": 1.1204264163970947,
"step": 538
},
{
"epoch": 0.22279525528623,
"grad_norm": 0.49772873520851135,
"learning_rate": 0.0005,
"loss": 1.467596411705017,
"step": 540
},
{
"epoch": 0.22362042289840123,
"grad_norm": 0.7317692041397095,
"learning_rate": 0.0005,
"loss": 0.6860278248786926,
"step": 542
},
{
"epoch": 0.22444559051057247,
"grad_norm": 0.8186198472976685,
"learning_rate": 0.0005,
"loss": 0.7754762172698975,
"step": 544
},
{
"epoch": 0.2252707581227437,
"grad_norm": 0.4967374801635742,
"learning_rate": 0.0005,
"loss": 1.2723290920257568,
"step": 546
},
{
"epoch": 0.2260959257349149,
"grad_norm": 0.5389532446861267,
"learning_rate": 0.0005,
"loss": 1.099130630493164,
"step": 548
},
{
"epoch": 0.22692109334708613,
"grad_norm": 0.5076627731323242,
"learning_rate": 0.0005,
"loss": 1.195350170135498,
"step": 550
},
{
"epoch": 0.22774626095925735,
"grad_norm": 0.7378993034362793,
"learning_rate": 0.0005,
"loss": 1.1034424304962158,
"step": 552
},
{
"epoch": 0.22857142857142856,
"grad_norm": 0.3536158800125122,
"learning_rate": 0.0005,
"loss": 1.1557471752166748,
"step": 554
},
{
"epoch": 0.22939659618359978,
"grad_norm": 0.6966869831085205,
"learning_rate": 0.0005,
"loss": 1.3733209371566772,
"step": 556
},
{
"epoch": 0.23022176379577103,
"grad_norm": 0.5721921920776367,
"learning_rate": 0.0005,
"loss": 1.2296565771102905,
"step": 558
},
{
"epoch": 0.23104693140794225,
"grad_norm": 0.5838218927383423,
"learning_rate": 0.0005,
"loss": 1.1840265989303589,
"step": 560
},
{
"epoch": 0.23187209902011346,
"grad_norm": 0.584754467010498,
"learning_rate": 0.0005,
"loss": 1.4932587146759033,
"step": 562
},
{
"epoch": 0.23269726663228468,
"grad_norm": 0.42717620730400085,
"learning_rate": 0.0005,
"loss": 1.5163227319717407,
"step": 564
},
{
"epoch": 0.2335224342444559,
"grad_norm": 0.6021290421485901,
"learning_rate": 0.0005,
"loss": 1.4346580505371094,
"step": 566
},
{
"epoch": 0.23434760185662712,
"grad_norm": 0.5628776550292969,
"learning_rate": 0.0005,
"loss": 1.1077497005462646,
"step": 568
},
{
"epoch": 0.23517276946879834,
"grad_norm": 0.6452749371528625,
"learning_rate": 0.0005,
"loss": 1.2200727462768555,
"step": 570
},
{
"epoch": 0.23599793708096958,
"grad_norm": 0.5786890983581543,
"learning_rate": 0.0005,
"loss": 1.272033452987671,
"step": 572
},
{
"epoch": 0.2368231046931408,
"grad_norm": 0.4542114734649658,
"learning_rate": 0.0005,
"loss": 0.9179959893226624,
"step": 574
},
{
"epoch": 0.23764827230531202,
"grad_norm": 0.39080727100372314,
"learning_rate": 0.0005,
"loss": 1.2384065389633179,
"step": 576
},
{
"epoch": 0.23847343991748324,
"grad_norm": 0.6351068615913391,
"learning_rate": 0.0005,
"loss": 0.8556203842163086,
"step": 578
},
{
"epoch": 0.23929860752965446,
"grad_norm": 0.7682400941848755,
"learning_rate": 0.0005,
"loss": 0.9543220400810242,
"step": 580
},
{
"epoch": 0.24012377514182567,
"grad_norm": 0.92158043384552,
"learning_rate": 0.0005,
"loss": 0.549217700958252,
"step": 582
},
{
"epoch": 0.2409489427539969,
"grad_norm": 0.4430788457393646,
"learning_rate": 0.0005,
"loss": 1.034724473953247,
"step": 584
},
{
"epoch": 0.24177411036616814,
"grad_norm": 0.6903814077377319,
"learning_rate": 0.0005,
"loss": 1.4142558574676514,
"step": 586
},
{
"epoch": 0.24259927797833936,
"grad_norm": 0.43257200717926025,
"learning_rate": 0.0005,
"loss": 1.1588163375854492,
"step": 588
},
{
"epoch": 0.24342444559051057,
"grad_norm": 0.43681901693344116,
"learning_rate": 0.0005,
"loss": 1.2648508548736572,
"step": 590
},
{
"epoch": 0.2442496132026818,
"grad_norm": 0.6315423250198364,
"learning_rate": 0.0005,
"loss": 1.160179615020752,
"step": 592
},
{
"epoch": 0.245074780814853,
"grad_norm": 0.40824076533317566,
"learning_rate": 0.0005,
"loss": 1.290668249130249,
"step": 594
},
{
"epoch": 0.24589994842702423,
"grad_norm": 0.4974735677242279,
"learning_rate": 0.0005,
"loss": 1.0398313999176025,
"step": 596
},
{
"epoch": 0.24672511603919547,
"grad_norm": 0.6924368143081665,
"learning_rate": 0.0005,
"loss": 1.3205965757369995,
"step": 598
},
{
"epoch": 0.2475502836513667,
"grad_norm": 0.4788479804992676,
"learning_rate": 0.0005,
"loss": 1.587035894393921,
"step": 600
},
{
"epoch": 0.2483754512635379,
"grad_norm": 0.5854333639144897,
"learning_rate": 0.0005,
"loss": 0.7899203896522522,
"step": 602
},
{
"epoch": 0.24920061887570913,
"grad_norm": 0.5335381627082825,
"learning_rate": 0.0005,
"loss": 1.1879031658172607,
"step": 604
},
{
"epoch": 0.2500257864878804,
"grad_norm": 0.47977879643440247,
"learning_rate": 0.0005,
"loss": 1.0355863571166992,
"step": 606
},
{
"epoch": 0.2508509541000516,
"grad_norm": 0.9655616283416748,
"learning_rate": 0.0005,
"loss": 0.6792604923248291,
"step": 608
},
{
"epoch": 0.2516761217122228,
"grad_norm": 0.5150128602981567,
"learning_rate": 0.0005,
"loss": 1.0021084547042847,
"step": 610
},
{
"epoch": 0.25250128932439403,
"grad_norm": 0.4458533823490143,
"learning_rate": 0.0005,
"loss": 1.3821619749069214,
"step": 612
},
{
"epoch": 0.25332645693656525,
"grad_norm": 0.4895058274269104,
"learning_rate": 0.0005,
"loss": 1.1168134212493896,
"step": 614
},
{
"epoch": 0.25415162454873647,
"grad_norm": 0.7227627038955688,
"learning_rate": 0.0005,
"loss": 0.8979606032371521,
"step": 616
},
{
"epoch": 0.2549767921609077,
"grad_norm": 0.9172181487083435,
"learning_rate": 0.0005,
"loss": 0.8193933963775635,
"step": 618
},
{
"epoch": 0.2558019597730789,
"grad_norm": 0.7418314814567566,
"learning_rate": 0.0005,
"loss": 0.8683919310569763,
"step": 620
},
{
"epoch": 0.2566271273852501,
"grad_norm": 0.4302678406238556,
"learning_rate": 0.0005,
"loss": 1.5036029815673828,
"step": 622
},
{
"epoch": 0.25745229499742134,
"grad_norm": 0.4392130672931671,
"learning_rate": 0.0005,
"loss": 1.2528417110443115,
"step": 624
},
{
"epoch": 0.25827746260959256,
"grad_norm": 0.613135039806366,
"learning_rate": 0.0005,
"loss": 1.3085262775421143,
"step": 626
},
{
"epoch": 0.2591026302217638,
"grad_norm": 0.4622304141521454,
"learning_rate": 0.0005,
"loss": 1.4244141578674316,
"step": 628
},
{
"epoch": 0.259927797833935,
"grad_norm": 0.7210198640823364,
"learning_rate": 0.0005,
"loss": 0.9631842970848083,
"step": 630
},
{
"epoch": 0.26075296544610627,
"grad_norm": 0.7242471575737,
"learning_rate": 0.0005,
"loss": 0.4878099858760834,
"step": 632
},
{
"epoch": 0.2615781330582775,
"grad_norm": 0.514824628829956,
"learning_rate": 0.0005,
"loss": 1.0283039808273315,
"step": 634
},
{
"epoch": 0.2624033006704487,
"grad_norm": 0.5972491502761841,
"learning_rate": 0.0005,
"loss": 1.2710590362548828,
"step": 636
},
{
"epoch": 0.2632284682826199,
"grad_norm": 0.5197654366493225,
"learning_rate": 0.0005,
"loss": 1.0061873197555542,
"step": 638
},
{
"epoch": 0.26405363589479114,
"grad_norm": 0.7594143748283386,
"learning_rate": 0.0005,
"loss": 0.9267264008522034,
"step": 640
},
{
"epoch": 0.26487880350696236,
"grad_norm": 0.4109676480293274,
"learning_rate": 0.0005,
"loss": 1.4799742698669434,
"step": 642
},
{
"epoch": 0.2657039711191336,
"grad_norm": 0.39466243982315063,
"learning_rate": 0.0005,
"loss": 1.013569712638855,
"step": 644
},
{
"epoch": 0.2665291387313048,
"grad_norm": 0.670261561870575,
"learning_rate": 0.0005,
"loss": 0.689892053604126,
"step": 646
},
{
"epoch": 0.267354306343476,
"grad_norm": 0.5405585765838623,
"learning_rate": 0.0005,
"loss": 1.3047176599502563,
"step": 648
},
{
"epoch": 0.26817947395564723,
"grad_norm": 0.5078087449073792,
"learning_rate": 0.0005,
"loss": 1.262884497642517,
"step": 650
},
{
"epoch": 0.26900464156781845,
"grad_norm": 0.47059938311576843,
"learning_rate": 0.0005,
"loss": 0.8305864930152893,
"step": 652
},
{
"epoch": 0.26982980917998967,
"grad_norm": 0.4846727252006531,
"learning_rate": 0.0005,
"loss": 1.2409794330596924,
"step": 654
},
{
"epoch": 0.2706549767921609,
"grad_norm": 0.4383109509944916,
"learning_rate": 0.0005,
"loss": 1.2484523057937622,
"step": 656
},
{
"epoch": 0.27148014440433216,
"grad_norm": 0.37556707859039307,
"learning_rate": 0.0005,
"loss": 0.9433695077896118,
"step": 658
},
{
"epoch": 0.2723053120165034,
"grad_norm": 0.37254199385643005,
"learning_rate": 0.0005,
"loss": 1.4334774017333984,
"step": 660
},
{
"epoch": 0.2731304796286746,
"grad_norm": 0.5509696006774902,
"learning_rate": 0.0005,
"loss": 0.8734132051467896,
"step": 662
},
{
"epoch": 0.2739556472408458,
"grad_norm": 0.4884193539619446,
"learning_rate": 0.0005,
"loss": 0.9699646830558777,
"step": 664
},
{
"epoch": 0.27478081485301703,
"grad_norm": 0.5946614146232605,
"learning_rate": 0.0005,
"loss": 0.7786047458648682,
"step": 666
},
{
"epoch": 0.27560598246518825,
"grad_norm": 0.4739646911621094,
"learning_rate": 0.0005,
"loss": 0.9495128393173218,
"step": 668
},
{
"epoch": 0.27643115007735947,
"grad_norm": 0.5007510781288147,
"learning_rate": 0.0005,
"loss": 1.351648211479187,
"step": 670
},
{
"epoch": 0.2772563176895307,
"grad_norm": 0.5150129795074463,
"learning_rate": 0.0005,
"loss": 0.761278510093689,
"step": 672
},
{
"epoch": 0.2780814853017019,
"grad_norm": 0.6678033471107483,
"learning_rate": 0.0005,
"loss": 1.4205937385559082,
"step": 674
},
{
"epoch": 0.2789066529138731,
"grad_norm": 0.7818036079406738,
"learning_rate": 0.0005,
"loss": 0.44894301891326904,
"step": 676
},
{
"epoch": 0.27973182052604434,
"grad_norm": 0.5766839981079102,
"learning_rate": 0.0005,
"loss": 1.0723761320114136,
"step": 678
},
{
"epoch": 0.28055698813821556,
"grad_norm": 0.44825315475463867,
"learning_rate": 0.0005,
"loss": 1.4418954849243164,
"step": 680
},
{
"epoch": 0.2813821557503868,
"grad_norm": 0.42877197265625,
"learning_rate": 0.0005,
"loss": 1.3415015935897827,
"step": 682
},
{
"epoch": 0.282207323362558,
"grad_norm": 0.5869501829147339,
"learning_rate": 0.0005,
"loss": 0.9233855605125427,
"step": 684
},
{
"epoch": 0.28303249097472927,
"grad_norm": 0.4123280644416809,
"learning_rate": 0.0005,
"loss": 1.285361886024475,
"step": 686
},
{
"epoch": 0.2838576585869005,
"grad_norm": 0.5483129620552063,
"learning_rate": 0.0005,
"loss": 1.0504131317138672,
"step": 688
},
{
"epoch": 0.2846828261990717,
"grad_norm": 0.43682846426963806,
"learning_rate": 0.0005,
"loss": 1.1810169219970703,
"step": 690
},
{
"epoch": 0.2855079938112429,
"grad_norm": 0.5283734202384949,
"learning_rate": 0.0005,
"loss": 1.3038040399551392,
"step": 692
},
{
"epoch": 0.28633316142341414,
"grad_norm": 0.41053348779678345,
"learning_rate": 0.0005,
"loss": 1.5715042352676392,
"step": 694
},
{
"epoch": 0.28715832903558536,
"grad_norm": 0.5874834060668945,
"learning_rate": 0.0005,
"loss": 1.2950150966644287,
"step": 696
},
{
"epoch": 0.2879834966477566,
"grad_norm": 0.5786132216453552,
"learning_rate": 0.0005,
"loss": 1.431294322013855,
"step": 698
},
{
"epoch": 0.2888086642599278,
"grad_norm": 0.4537035822868347,
"learning_rate": 0.0005,
"loss": 1.3521438837051392,
"step": 700
},
{
"epoch": 0.289633831872099,
"grad_norm": 0.5288688540458679,
"learning_rate": 0.0005,
"loss": 1.2051489353179932,
"step": 702
},
{
"epoch": 0.29045899948427023,
"grad_norm": 0.6778690218925476,
"learning_rate": 0.0005,
"loss": 1.0220110416412354,
"step": 704
},
{
"epoch": 0.29128416709644145,
"grad_norm": 0.7159928679466248,
"learning_rate": 0.0005,
"loss": 1.0520648956298828,
"step": 706
},
{
"epoch": 0.29210933470861267,
"grad_norm": 0.7548841834068298,
"learning_rate": 0.0005,
"loss": 0.6303867101669312,
"step": 708
},
{
"epoch": 0.2929345023207839,
"grad_norm": 0.6287117600440979,
"learning_rate": 0.0005,
"loss": 0.8950468897819519,
"step": 710
},
{
"epoch": 0.29375966993295516,
"grad_norm": 0.5618225336074829,
"learning_rate": 0.0005,
"loss": 1.0949610471725464,
"step": 712
},
{
"epoch": 0.2945848375451264,
"grad_norm": 0.6126213669776917,
"learning_rate": 0.0005,
"loss": 1.1194016933441162,
"step": 714
},
{
"epoch": 0.2954100051572976,
"grad_norm": 1.0764331817626953,
"learning_rate": 0.0005,
"loss": 0.8703246116638184,
"step": 716
},
{
"epoch": 0.2962351727694688,
"grad_norm": 0.6652147769927979,
"learning_rate": 0.0005,
"loss": 0.7871020436286926,
"step": 718
},
{
"epoch": 0.29706034038164003,
"grad_norm": 0.7677785754203796,
"learning_rate": 0.0005,
"loss": 0.9459787607192993,
"step": 720
},
{
"epoch": 0.29788550799381125,
"grad_norm": 0.6401911377906799,
"learning_rate": 0.0005,
"loss": 1.242273211479187,
"step": 722
},
{
"epoch": 0.29871067560598247,
"grad_norm": 0.7156005501747131,
"learning_rate": 0.0005,
"loss": 0.4272935390472412,
"step": 724
},
{
"epoch": 0.2995358432181537,
"grad_norm": 0.515862226486206,
"learning_rate": 0.0005,
"loss": 0.7296764254570007,
"step": 726
},
{
"epoch": 0.3003610108303249,
"grad_norm": 0.4735647439956665,
"learning_rate": 0.0005,
"loss": 1.0296608209609985,
"step": 728
},
{
"epoch": 0.3011861784424961,
"grad_norm": 0.4547019302845001,
"learning_rate": 0.0005,
"loss": 0.7701424956321716,
"step": 730
},
{
"epoch": 0.30201134605466734,
"grad_norm": 0.709504246711731,
"learning_rate": 0.0005,
"loss": 0.8757145404815674,
"step": 732
},
{
"epoch": 0.30283651366683856,
"grad_norm": 0.4775760769844055,
"learning_rate": 0.0005,
"loss": 1.422513723373413,
"step": 734
},
{
"epoch": 0.3036616812790098,
"grad_norm": 0.5066021680831909,
"learning_rate": 0.0005,
"loss": 1.1779431104660034,
"step": 736
},
{
"epoch": 0.304486848891181,
"grad_norm": 0.37824875116348267,
"learning_rate": 0.0005,
"loss": 1.396673560142517,
"step": 738
},
{
"epoch": 0.30531201650335227,
"grad_norm": 0.47177132964134216,
"learning_rate": 0.0005,
"loss": 0.6887521743774414,
"step": 740
},
{
"epoch": 0.3061371841155235,
"grad_norm": 0.4736272692680359,
"learning_rate": 0.0005,
"loss": 0.951183021068573,
"step": 742
},
{
"epoch": 0.3069623517276947,
"grad_norm": 0.5575990080833435,
"learning_rate": 0.0005,
"loss": 0.5598281621932983,
"step": 744
},
{
"epoch": 0.3077875193398659,
"grad_norm": 0.6142449378967285,
"learning_rate": 0.0005,
"loss": 0.8059773445129395,
"step": 746
},
{
"epoch": 0.30861268695203714,
"grad_norm": 0.3264317512512207,
"learning_rate": 0.0005,
"loss": 1.2799454927444458,
"step": 748
},
{
"epoch": 0.30943785456420836,
"grad_norm": 0.6371709704399109,
"learning_rate": 0.0005,
"loss": 0.9574125409126282,
"step": 750
},
{
"epoch": 0.3102630221763796,
"grad_norm": 0.3501775860786438,
"learning_rate": 0.0005,
"loss": 1.5243263244628906,
"step": 752
},
{
"epoch": 0.3110881897885508,
"grad_norm": 0.7664023041725159,
"learning_rate": 0.0005,
"loss": 0.8435171246528625,
"step": 754
},
{
"epoch": 0.311913357400722,
"grad_norm": 0.47338634729385376,
"learning_rate": 0.0005,
"loss": 1.2635228633880615,
"step": 756
},
{
"epoch": 0.31273852501289323,
"grad_norm": 0.5893837213516235,
"learning_rate": 0.0005,
"loss": 1.2268116474151611,
"step": 758
},
{
"epoch": 0.31356369262506445,
"grad_norm": 0.3971027433872223,
"learning_rate": 0.0005,
"loss": 0.9376073479652405,
"step": 760
},
{
"epoch": 0.31438886023723567,
"grad_norm": 0.4974158704280853,
"learning_rate": 0.0005,
"loss": 1.1054447889328003,
"step": 762
},
{
"epoch": 0.3152140278494069,
"grad_norm": 0.4881732165813446,
"learning_rate": 0.0005,
"loss": 1.2376474142074585,
"step": 764
},
{
"epoch": 0.31603919546157816,
"grad_norm": 0.4697839915752411,
"learning_rate": 0.0005,
"loss": 0.9591520428657532,
"step": 766
},
{
"epoch": 0.3168643630737494,
"grad_norm": 0.5661628842353821,
"learning_rate": 0.0005,
"loss": 0.5725570321083069,
"step": 768
},
{
"epoch": 0.3176895306859206,
"grad_norm": 0.41950663924217224,
"learning_rate": 0.0005,
"loss": 0.8325156569480896,
"step": 770
},
{
"epoch": 0.3185146982980918,
"grad_norm": 0.4491046071052551,
"learning_rate": 0.0005,
"loss": 0.8748528957366943,
"step": 772
},
{
"epoch": 0.31933986591026303,
"grad_norm": 0.6069282293319702,
"learning_rate": 0.0005,
"loss": 0.6700723171234131,
"step": 774
},
{
"epoch": 0.32016503352243425,
"grad_norm": 0.387180894613266,
"learning_rate": 0.0005,
"loss": 1.208810806274414,
"step": 776
},
{
"epoch": 0.32099020113460547,
"grad_norm": 0.3678168058395386,
"learning_rate": 0.0005,
"loss": 0.9034237861633301,
"step": 778
},
{
"epoch": 0.3218153687467767,
"grad_norm": 0.39267754554748535,
"learning_rate": 0.0005,
"loss": 1.106095790863037,
"step": 780
},
{
"epoch": 0.3226405363589479,
"grad_norm": 0.5397934317588806,
"learning_rate": 0.0005,
"loss": 1.2158617973327637,
"step": 782
},
{
"epoch": 0.3234657039711191,
"grad_norm": 0.4539080560207367,
"learning_rate": 0.0005,
"loss": 1.0038114786148071,
"step": 784
},
{
"epoch": 0.32429087158329034,
"grad_norm": 0.49578386545181274,
"learning_rate": 0.0005,
"loss": 1.2935315370559692,
"step": 786
},
{
"epoch": 0.32511603919546156,
"grad_norm": 0.3788098394870758,
"learning_rate": 0.0005,
"loss": 1.3695635795593262,
"step": 788
},
{
"epoch": 0.3259412068076328,
"grad_norm": 0.5304418206214905,
"learning_rate": 0.0005,
"loss": 1.2912182807922363,
"step": 790
},
{
"epoch": 0.326766374419804,
"grad_norm": 0.5684431195259094,
"learning_rate": 0.0005,
"loss": 1.1963849067687988,
"step": 792
},
{
"epoch": 0.32759154203197527,
"grad_norm": 0.5429754853248596,
"learning_rate": 0.0005,
"loss": 0.9575740098953247,
"step": 794
},
{
"epoch": 0.3284167096441465,
"grad_norm": 0.6165698170661926,
"learning_rate": 0.0005,
"loss": 0.6333395838737488,
"step": 796
},
{
"epoch": 0.3292418772563177,
"grad_norm": 0.6064788699150085,
"learning_rate": 0.0005,
"loss": 0.8153479099273682,
"step": 798
},
{
"epoch": 0.3300670448684889,
"grad_norm": 0.4232187271118164,
"learning_rate": 0.0005,
"loss": 1.4265881776809692,
"step": 800
},
{
"epoch": 0.33089221248066014,
"grad_norm": 0.4814676344394684,
"learning_rate": 0.0005,
"loss": 1.1406737565994263,
"step": 802
},
{
"epoch": 0.33171738009283136,
"grad_norm": 0.419049471616745,
"learning_rate": 0.0005,
"loss": 1.4567514657974243,
"step": 804
},
{
"epoch": 0.3325425477050026,
"grad_norm": 0.7689643502235413,
"learning_rate": 0.0005,
"loss": 0.9619349241256714,
"step": 806
},
{
"epoch": 0.3333677153171738,
"grad_norm": 0.4174315631389618,
"learning_rate": 0.0005,
"loss": 0.917432963848114,
"step": 808
},
{
"epoch": 0.334192882929345,
"grad_norm": 0.5049794316291809,
"learning_rate": 0.0005,
"loss": 1.0135047435760498,
"step": 810
},
{
"epoch": 0.33501805054151623,
"grad_norm": 0.39736947417259216,
"learning_rate": 0.0005,
"loss": 1.2304755449295044,
"step": 812
},
{
"epoch": 0.33584321815368745,
"grad_norm": 0.5136732459068298,
"learning_rate": 0.0005,
"loss": 0.6235193014144897,
"step": 814
},
{
"epoch": 0.33666838576585867,
"grad_norm": 0.6358846426010132,
"learning_rate": 0.0005,
"loss": 0.9476469159126282,
"step": 816
},
{
"epoch": 0.3374935533780299,
"grad_norm": 0.36760398745536804,
"learning_rate": 0.0005,
"loss": 1.1756913661956787,
"step": 818
},
{
"epoch": 0.33831872099020116,
"grad_norm": 0.6921237111091614,
"learning_rate": 0.0005,
"loss": 0.5435713529586792,
"step": 820
},
{
"epoch": 0.3391438886023724,
"grad_norm": 0.6190113425254822,
"learning_rate": 0.0005,
"loss": 0.6276876926422119,
"step": 822
},
{
"epoch": 0.3399690562145436,
"grad_norm": 0.5015466809272766,
"learning_rate": 0.0005,
"loss": 1.2893644571304321,
"step": 824
},
{
"epoch": 0.3407942238267148,
"grad_norm": 0.6742082834243774,
"learning_rate": 0.0005,
"loss": 0.9958200454711914,
"step": 826
},
{
"epoch": 0.34161939143888603,
"grad_norm": 0.43738895654678345,
"learning_rate": 0.0005,
"loss": 0.8957383632659912,
"step": 828
},
{
"epoch": 0.34244455905105725,
"grad_norm": 0.40049517154693604,
"learning_rate": 0.0005,
"loss": 0.86797696352005,
"step": 830
},
{
"epoch": 0.34326972666322847,
"grad_norm": 0.678524911403656,
"learning_rate": 0.0005,
"loss": 1.1833425760269165,
"step": 832
},
{
"epoch": 0.3440948942753997,
"grad_norm": 0.3560514450073242,
"learning_rate": 0.0005,
"loss": 1.247701644897461,
"step": 834
},
{
"epoch": 0.3449200618875709,
"grad_norm": 0.7766836881637573,
"learning_rate": 0.0005,
"loss": 0.6366620063781738,
"step": 836
},
{
"epoch": 0.3457452294997421,
"grad_norm": 0.3957395553588867,
"learning_rate": 0.0005,
"loss": 1.452673316001892,
"step": 838
},
{
"epoch": 0.34657039711191334,
"grad_norm": 0.6949273347854614,
"learning_rate": 0.0005,
"loss": 0.6712929010391235,
"step": 840
},
{
"epoch": 0.34739556472408456,
"grad_norm": 0.44859611988067627,
"learning_rate": 0.0005,
"loss": 0.9374242424964905,
"step": 842
},
{
"epoch": 0.3482207323362558,
"grad_norm": 0.44476670026779175,
"learning_rate": 0.0005,
"loss": 0.9708834290504456,
"step": 844
},
{
"epoch": 0.34904589994842705,
"grad_norm": 0.5429829955101013,
"learning_rate": 0.0005,
"loss": 1.1306025981903076,
"step": 846
},
{
"epoch": 0.34987106756059827,
"grad_norm": 0.5670883655548096,
"learning_rate": 0.0005,
"loss": 0.9808698296546936,
"step": 848
},
{
"epoch": 0.3506962351727695,
"grad_norm": 0.4450462758541107,
"learning_rate": 0.0005,
"loss": 1.2327816486358643,
"step": 850
},
{
"epoch": 0.3515214027849407,
"grad_norm": 0.44823941588401794,
"learning_rate": 0.0005,
"loss": 1.214652180671692,
"step": 852
},
{
"epoch": 0.3523465703971119,
"grad_norm": 0.46802523732185364,
"learning_rate": 0.0005,
"loss": 0.8434693813323975,
"step": 854
},
{
"epoch": 0.35317173800928314,
"grad_norm": 0.4221685528755188,
"learning_rate": 0.0005,
"loss": 1.385030746459961,
"step": 856
},
{
"epoch": 0.35399690562145436,
"grad_norm": 0.3876509368419647,
"learning_rate": 0.0005,
"loss": 1.3161131143569946,
"step": 858
},
{
"epoch": 0.3548220732336256,
"grad_norm": 0.465462327003479,
"learning_rate": 0.0005,
"loss": 1.246232032775879,
"step": 860
},
{
"epoch": 0.3556472408457968,
"grad_norm": 0.38140279054641724,
"learning_rate": 0.0005,
"loss": 1.1987196207046509,
"step": 862
},
{
"epoch": 0.356472408457968,
"grad_norm": 0.6231861710548401,
"learning_rate": 0.0005,
"loss": 0.7024447917938232,
"step": 864
},
{
"epoch": 0.35729757607013923,
"grad_norm": 0.8739637136459351,
"learning_rate": 0.0005,
"loss": 0.9236916899681091,
"step": 866
},
{
"epoch": 0.35812274368231045,
"grad_norm": 0.8303645253181458,
"learning_rate": 0.0005,
"loss": 0.7618856430053711,
"step": 868
},
{
"epoch": 0.35894791129448167,
"grad_norm": 0.44744884967803955,
"learning_rate": 0.0005,
"loss": 1.1455423831939697,
"step": 870
},
{
"epoch": 0.3597730789066529,
"grad_norm": 0.44626978039741516,
"learning_rate": 0.0005,
"loss": 1.1497100591659546,
"step": 872
},
{
"epoch": 0.36059824651882416,
"grad_norm": 0.47076717019081116,
"learning_rate": 0.0005,
"loss": 1.0223735570907593,
"step": 874
},
{
"epoch": 0.3614234141309954,
"grad_norm": 0.5181264281272888,
"learning_rate": 0.0005,
"loss": 0.8119293451309204,
"step": 876
},
{
"epoch": 0.3622485817431666,
"grad_norm": 0.5308315753936768,
"learning_rate": 0.0005,
"loss": 0.8662136793136597,
"step": 878
},
{
"epoch": 0.3630737493553378,
"grad_norm": 0.3599519729614258,
"learning_rate": 0.0005,
"loss": 1.0991331338882446,
"step": 880
},
{
"epoch": 0.36389891696750903,
"grad_norm": 0.450533002614975,
"learning_rate": 0.0005,
"loss": 1.199340581893921,
"step": 882
},
{
"epoch": 0.36472408457968025,
"grad_norm": 0.41602176427841187,
"learning_rate": 0.0005,
"loss": 0.8881515860557556,
"step": 884
},
{
"epoch": 0.36554925219185147,
"grad_norm": 0.3758910894393921,
"learning_rate": 0.0005,
"loss": 1.0691791772842407,
"step": 886
},
{
"epoch": 0.3663744198040227,
"grad_norm": 0.49511483311653137,
"learning_rate": 0.0005,
"loss": 0.84291672706604,
"step": 888
},
{
"epoch": 0.3671995874161939,
"grad_norm": 0.37525108456611633,
"learning_rate": 0.0005,
"loss": 0.787752091884613,
"step": 890
},
{
"epoch": 0.3680247550283651,
"grad_norm": 0.6520910859107971,
"learning_rate": 0.0005,
"loss": 0.7890579700469971,
"step": 892
},
{
"epoch": 0.36884992264053634,
"grad_norm": 0.4799257218837738,
"learning_rate": 0.0005,
"loss": 0.8910934925079346,
"step": 894
},
{
"epoch": 0.36967509025270756,
"grad_norm": 0.3718203604221344,
"learning_rate": 0.0005,
"loss": 1.2266496419906616,
"step": 896
},
{
"epoch": 0.3705002578648788,
"grad_norm": 0.7900125980377197,
"learning_rate": 0.0005,
"loss": 0.808830738067627,
"step": 898
},
{
"epoch": 0.37132542547705005,
"grad_norm": 0.3988662660121918,
"learning_rate": 0.0005,
"loss": 1.3982020616531372,
"step": 900
},
{
"epoch": 0.37215059308922127,
"grad_norm": 0.4183046519756317,
"learning_rate": 0.0005,
"loss": 0.945640504360199,
"step": 902
},
{
"epoch": 0.3729757607013925,
"grad_norm": 0.4638324975967407,
"learning_rate": 0.0005,
"loss": 0.9022805690765381,
"step": 904
},
{
"epoch": 0.3738009283135637,
"grad_norm": 0.4181121587753296,
"learning_rate": 0.0005,
"loss": 1.0504035949707031,
"step": 906
},
{
"epoch": 0.3746260959257349,
"grad_norm": 0.5623753666877747,
"learning_rate": 0.0005,
"loss": 0.885411262512207,
"step": 908
},
{
"epoch": 0.37545126353790614,
"grad_norm": 0.6528892517089844,
"learning_rate": 0.0005,
"loss": 0.4613204598426819,
"step": 910
},
{
"epoch": 0.37627643115007736,
"grad_norm": 0.43880847096443176,
"learning_rate": 0.0005,
"loss": 0.6581244468688965,
"step": 912
},
{
"epoch": 0.3771015987622486,
"grad_norm": 0.6024404764175415,
"learning_rate": 0.0005,
"loss": 0.5904361009597778,
"step": 914
},
{
"epoch": 0.3779267663744198,
"grad_norm": 0.4631766676902771,
"learning_rate": 0.0005,
"loss": 0.8536958694458008,
"step": 916
},
{
"epoch": 0.378751933986591,
"grad_norm": 0.6748937964439392,
"learning_rate": 0.0005,
"loss": 1.154845952987671,
"step": 918
},
{
"epoch": 0.37957710159876223,
"grad_norm": 0.5579289793968201,
"learning_rate": 0.0005,
"loss": 1.3011925220489502,
"step": 920
},
{
"epoch": 0.38040226921093345,
"grad_norm": 0.4682430326938629,
"learning_rate": 0.0005,
"loss": 1.104500651359558,
"step": 922
},
{
"epoch": 0.38122743682310467,
"grad_norm": 0.6007438898086548,
"learning_rate": 0.0005,
"loss": 1.0890264511108398,
"step": 924
},
{
"epoch": 0.3820526044352759,
"grad_norm": 0.3681229054927826,
"learning_rate": 0.0005,
"loss": 1.0576262474060059,
"step": 926
},
{
"epoch": 0.38287777204744716,
"grad_norm": 0.3867166340351105,
"learning_rate": 0.0005,
"loss": 1.4854787588119507,
"step": 928
},
{
"epoch": 0.3837029396596184,
"grad_norm": 0.6359395384788513,
"learning_rate": 0.0005,
"loss": 0.3282276690006256,
"step": 930
},
{
"epoch": 0.3845281072717896,
"grad_norm": 0.46864011883735657,
"learning_rate": 0.0005,
"loss": 1.2231162786483765,
"step": 932
},
{
"epoch": 0.3853532748839608,
"grad_norm": 0.42163488268852234,
"learning_rate": 0.0005,
"loss": 0.799576997756958,
"step": 934
},
{
"epoch": 0.38617844249613204,
"grad_norm": 0.5348471403121948,
"learning_rate": 0.0005,
"loss": 1.215818166732788,
"step": 936
},
{
"epoch": 0.38700361010830325,
"grad_norm": 0.47862353920936584,
"learning_rate": 0.0005,
"loss": 1.119133472442627,
"step": 938
},
{
"epoch": 0.38782877772047447,
"grad_norm": 0.6671462059020996,
"learning_rate": 0.0005,
"loss": 0.6009505987167358,
"step": 940
},
{
"epoch": 0.3886539453326457,
"grad_norm": 0.654689610004425,
"learning_rate": 0.0005,
"loss": 0.9069526195526123,
"step": 942
},
{
"epoch": 0.3894791129448169,
"grad_norm": 0.6140965223312378,
"learning_rate": 0.0005,
"loss": 1.1941167116165161,
"step": 944
},
{
"epoch": 0.3903042805569881,
"grad_norm": 0.4021962881088257,
"learning_rate": 0.0005,
"loss": 1.2709530591964722,
"step": 946
},
{
"epoch": 0.39112944816915934,
"grad_norm": 0.4146452844142914,
"learning_rate": 0.0005,
"loss": 1.1812446117401123,
"step": 948
},
{
"epoch": 0.39195461578133056,
"grad_norm": 0.6948921084403992,
"learning_rate": 0.0005,
"loss": 0.7930907011032104,
"step": 950
},
{
"epoch": 0.3927797833935018,
"grad_norm": 0.4594041705131531,
"learning_rate": 0.0005,
"loss": 1.1281040906906128,
"step": 952
},
{
"epoch": 0.39360495100567305,
"grad_norm": 0.6090090274810791,
"learning_rate": 0.0005,
"loss": 0.9690212607383728,
"step": 954
},
{
"epoch": 0.39443011861784427,
"grad_norm": 0.49895620346069336,
"learning_rate": 0.0005,
"loss": 0.9293103814125061,
"step": 956
},
{
"epoch": 0.3952552862300155,
"grad_norm": 0.5032721757888794,
"learning_rate": 0.0005,
"loss": 1.3478631973266602,
"step": 958
},
{
"epoch": 0.3960804538421867,
"grad_norm": 0.44382214546203613,
"learning_rate": 0.0005,
"loss": 1.455615520477295,
"step": 960
},
{
"epoch": 0.3969056214543579,
"grad_norm": 0.4999854266643524,
"learning_rate": 0.0005,
"loss": 1.059862732887268,
"step": 962
},
{
"epoch": 0.39773078906652914,
"grad_norm": 0.4809070825576782,
"learning_rate": 0.0005,
"loss": 0.7847961187362671,
"step": 964
},
{
"epoch": 0.39855595667870036,
"grad_norm": 0.677364706993103,
"learning_rate": 0.0005,
"loss": 0.3548208475112915,
"step": 966
},
{
"epoch": 0.3993811242908716,
"grad_norm": 0.35379257798194885,
"learning_rate": 0.0005,
"loss": 1.1698503494262695,
"step": 968
},
{
"epoch": 0.4002062919030428,
"grad_norm": 0.47841575741767883,
"learning_rate": 0.0005,
"loss": 0.6756126284599304,
"step": 970
},
{
"epoch": 0.4002062919030428,
"eval_loss": 0.7067621946334839,
"eval_runtime": 7.3635,
"eval_samples_per_second": 106.471,
"eval_steps_per_second": 13.309,
"step": 970
},
{
"epoch": 0.401031459515214,
"grad_norm": 0.6655641794204712,
"learning_rate": 0.0005,
"loss": 0.3434591293334961,
"step": 972
},
{
"epoch": 0.40185662712738524,
"grad_norm": 0.5978572368621826,
"learning_rate": 0.0005,
"loss": 0.35598450899124146,
"step": 974
},
{
"epoch": 0.40268179473955645,
"grad_norm": 0.6890511512756348,
"learning_rate": 0.0005,
"loss": 0.9411051869392395,
"step": 976
},
{
"epoch": 0.40350696235172767,
"grad_norm": 0.5624321699142456,
"learning_rate": 0.0005,
"loss": 0.3418905436992645,
"step": 978
},
{
"epoch": 0.4043321299638989,
"grad_norm": 0.47039034962654114,
"learning_rate": 0.0005,
"loss": 1.0206353664398193,
"step": 980
},
{
"epoch": 0.40515729757607016,
"grad_norm": 0.40128687024116516,
"learning_rate": 0.0005,
"loss": 0.8043688535690308,
"step": 982
},
{
"epoch": 0.4059824651882414,
"grad_norm": 0.38942623138427734,
"learning_rate": 0.0005,
"loss": 1.4332058429718018,
"step": 984
},
{
"epoch": 0.4068076328004126,
"grad_norm": 0.37515854835510254,
"learning_rate": 0.0005,
"loss": 1.1731629371643066,
"step": 986
},
{
"epoch": 0.4076328004125838,
"grad_norm": 0.47619712352752686,
"learning_rate": 0.0005,
"loss": 0.9455140829086304,
"step": 988
},
{
"epoch": 0.40845796802475504,
"grad_norm": 0.4121805429458618,
"learning_rate": 0.0005,
"loss": 0.9333594441413879,
"step": 990
},
{
"epoch": 0.40928313563692625,
"grad_norm": 0.37878715991973877,
"learning_rate": 0.0005,
"loss": 1.0717476606369019,
"step": 992
},
{
"epoch": 0.41010830324909747,
"grad_norm": 0.45636799931526184,
"learning_rate": 0.0005,
"loss": 0.5954750776290894,
"step": 994
},
{
"epoch": 0.4109334708612687,
"grad_norm": 0.3710258901119232,
"learning_rate": 0.0005,
"loss": 1.2599620819091797,
"step": 996
},
{
"epoch": 0.4117586384734399,
"grad_norm": 0.5332334637641907,
"learning_rate": 0.0005,
"loss": 1.068055510520935,
"step": 998
},
{
"epoch": 0.4125838060856111,
"grad_norm": 0.5794641375541687,
"learning_rate": 0.0005,
"loss": 0.7776126861572266,
"step": 1000
},
{
"epoch": 0.41340897369778234,
"grad_norm": 0.4480159878730774,
"learning_rate": 0.0005,
"loss": 0.6035796999931335,
"step": 1002
},
{
"epoch": 0.41423414130995356,
"grad_norm": 0.3095519244670868,
"learning_rate": 0.0005,
"loss": 0.9357731938362122,
"step": 1004
},
{
"epoch": 0.4150593089221248,
"grad_norm": 0.45328325033187866,
"learning_rate": 0.0005,
"loss": 0.9649761915206909,
"step": 1006
},
{
"epoch": 0.41588447653429605,
"grad_norm": 0.45143985748291016,
"learning_rate": 0.0005,
"loss": 1.0865952968597412,
"step": 1008
},
{
"epoch": 0.4167096441464673,
"grad_norm": 0.4858875572681427,
"learning_rate": 0.0005,
"loss": 0.81813645362854,
"step": 1010
},
{
"epoch": 0.4175348117586385,
"grad_norm": 0.3973652124404907,
"learning_rate": 0.0005,
"loss": 1.021775484085083,
"step": 1012
},
{
"epoch": 0.4183599793708097,
"grad_norm": 0.7302273511886597,
"learning_rate": 0.0005,
"loss": 0.6906617283821106,
"step": 1014
},
{
"epoch": 0.4191851469829809,
"grad_norm": 0.3600204586982727,
"learning_rate": 0.0005,
"loss": 1.3078669309616089,
"step": 1016
},
{
"epoch": 0.42001031459515215,
"grad_norm": 0.5048308372497559,
"learning_rate": 0.0005,
"loss": 0.4761854112148285,
"step": 1018
},
{
"epoch": 0.42083548220732336,
"grad_norm": 0.6642930507659912,
"learning_rate": 0.0005,
"loss": 0.588760495185852,
"step": 1020
},
{
"epoch": 0.4216606498194946,
"grad_norm": 0.6172478795051575,
"learning_rate": 0.0005,
"loss": 0.5218207836151123,
"step": 1022
},
{
"epoch": 0.4224858174316658,
"grad_norm": 0.5850370526313782,
"learning_rate": 0.0005,
"loss": 0.5745161771774292,
"step": 1024
},
{
"epoch": 0.423310985043837,
"grad_norm": 0.4626556634902954,
"learning_rate": 0.0005,
"loss": 0.6343902349472046,
"step": 1026
},
{
"epoch": 0.42413615265600824,
"grad_norm": 0.46811240911483765,
"learning_rate": 0.0005,
"loss": 0.7721204161643982,
"step": 1028
},
{
"epoch": 0.42496132026817945,
"grad_norm": 0.4241119921207428,
"learning_rate": 0.0005,
"loss": 1.2199113368988037,
"step": 1030
},
{
"epoch": 0.4257864878803507,
"grad_norm": 0.7176966071128845,
"learning_rate": 0.0005,
"loss": 0.6248391270637512,
"step": 1032
},
{
"epoch": 0.42661165549252195,
"grad_norm": 0.38848671317100525,
"learning_rate": 0.0005,
"loss": 1.1501232385635376,
"step": 1034
},
{
"epoch": 0.42743682310469316,
"grad_norm": 0.6330976486206055,
"learning_rate": 0.0005,
"loss": 0.314798504114151,
"step": 1036
},
{
"epoch": 0.4282619907168644,
"grad_norm": 0.5879555344581604,
"learning_rate": 0.0005,
"loss": 0.6930589079856873,
"step": 1038
},
{
"epoch": 0.4290871583290356,
"grad_norm": 0.5130351185798645,
"learning_rate": 0.0005,
"loss": 0.6479557156562805,
"step": 1040
},
{
"epoch": 0.4299123259412068,
"grad_norm": 0.6524402499198914,
"learning_rate": 0.0005,
"loss": 0.3105769455432892,
"step": 1042
},
{
"epoch": 0.43073749355337804,
"grad_norm": 0.4507187306880951,
"learning_rate": 0.0005,
"loss": 0.5807090997695923,
"step": 1044
},
{
"epoch": 0.43156266116554926,
"grad_norm": 0.4378560781478882,
"learning_rate": 0.0005,
"loss": 1.1415951251983643,
"step": 1046
},
{
"epoch": 0.4323878287777205,
"grad_norm": 0.41059407591819763,
"learning_rate": 0.0005,
"loss": 1.325959324836731,
"step": 1048
},
{
"epoch": 0.4332129963898917,
"grad_norm": 0.6772822141647339,
"learning_rate": 0.0005,
"loss": 0.9165737628936768,
"step": 1050
},
{
"epoch": 0.4340381640020629,
"grad_norm": 0.6387878656387329,
"learning_rate": 0.0005,
"loss": 0.33348408341407776,
"step": 1052
},
{
"epoch": 0.4348633316142341,
"grad_norm": 0.47842156887054443,
"learning_rate": 0.0005,
"loss": 1.3105005025863647,
"step": 1054
},
{
"epoch": 0.43568849922640535,
"grad_norm": 0.46261030435562134,
"learning_rate": 0.0005,
"loss": 0.7326333522796631,
"step": 1056
},
{
"epoch": 0.43651366683857656,
"grad_norm": 0.3358602523803711,
"learning_rate": 0.0005,
"loss": 1.3044193983078003,
"step": 1058
},
{
"epoch": 0.4373388344507478,
"grad_norm": 0.4453638195991516,
"learning_rate": 0.0005,
"loss": 1.1385118961334229,
"step": 1060
},
{
"epoch": 0.43816400206291906,
"grad_norm": 0.48419469594955444,
"learning_rate": 0.0005,
"loss": 0.8217204809188843,
"step": 1062
},
{
"epoch": 0.4389891696750903,
"grad_norm": 0.33709651231765747,
"learning_rate": 0.0005,
"loss": 1.392715334892273,
"step": 1064
},
{
"epoch": 0.4398143372872615,
"grad_norm": 0.5190922617912292,
"learning_rate": 0.0005,
"loss": 0.43600592017173767,
"step": 1066
},
{
"epoch": 0.4406395048994327,
"grad_norm": 0.46601980924606323,
"learning_rate": 0.0005,
"loss": 0.6399708986282349,
"step": 1068
},
{
"epoch": 0.44146467251160393,
"grad_norm": 0.46281567215919495,
"learning_rate": 0.0005,
"loss": 0.963161289691925,
"step": 1070
},
{
"epoch": 0.44228984012377515,
"grad_norm": 0.3844955563545227,
"learning_rate": 0.0005,
"loss": 1.2005760669708252,
"step": 1072
},
{
"epoch": 0.44311500773594636,
"grad_norm": 0.5524375438690186,
"learning_rate": 0.0005,
"loss": 0.6903715133666992,
"step": 1074
},
{
"epoch": 0.4439401753481176,
"grad_norm": 0.42640113830566406,
"learning_rate": 0.0005,
"loss": 1.2428464889526367,
"step": 1076
},
{
"epoch": 0.4447653429602888,
"grad_norm": 0.5613577961921692,
"learning_rate": 0.0005,
"loss": 0.27432385087013245,
"step": 1078
},
{
"epoch": 0.44559051057246,
"grad_norm": 0.49683111906051636,
"learning_rate": 0.0005,
"loss": 0.7725480794906616,
"step": 1080
},
{
"epoch": 0.44641567818463124,
"grad_norm": 0.6654202938079834,
"learning_rate": 0.0005,
"loss": 0.6296190619468689,
"step": 1082
},
{
"epoch": 0.44724084579680246,
"grad_norm": 0.3712174594402313,
"learning_rate": 0.0005,
"loss": 1.3912023305892944,
"step": 1084
},
{
"epoch": 0.4480660134089737,
"grad_norm": 0.4844004809856415,
"learning_rate": 0.0005,
"loss": 1.1226129531860352,
"step": 1086
},
{
"epoch": 0.44889118102114495,
"grad_norm": 0.4575996994972229,
"learning_rate": 0.0005,
"loss": 1.1576647758483887,
"step": 1088
},
{
"epoch": 0.44971634863331617,
"grad_norm": 0.4895000457763672,
"learning_rate": 0.0005,
"loss": 1.2015182971954346,
"step": 1090
},
{
"epoch": 0.4505415162454874,
"grad_norm": 0.48811668157577515,
"learning_rate": 0.0005,
"loss": 0.6746483445167542,
"step": 1092
},
{
"epoch": 0.4513666838576586,
"grad_norm": 0.4339450001716614,
"learning_rate": 0.0005,
"loss": 0.6415848731994629,
"step": 1094
},
{
"epoch": 0.4521918514698298,
"grad_norm": 0.3784640431404114,
"learning_rate": 0.0005,
"loss": 1.0874764919281006,
"step": 1096
},
{
"epoch": 0.45301701908200104,
"grad_norm": 0.584389865398407,
"learning_rate": 0.0005,
"loss": 1.1541180610656738,
"step": 1098
},
{
"epoch": 0.45384218669417226,
"grad_norm": 0.681960940361023,
"learning_rate": 0.0005,
"loss": 0.48588109016418457,
"step": 1100
},
{
"epoch": 0.4546673543063435,
"grad_norm": 0.5292614102363586,
"learning_rate": 0.0005,
"loss": 0.8904744386672974,
"step": 1102
},
{
"epoch": 0.4554925219185147,
"grad_norm": 0.471851646900177,
"learning_rate": 0.0005,
"loss": 0.7160428762435913,
"step": 1104
},
{
"epoch": 0.4563176895306859,
"grad_norm": 0.520013153553009,
"learning_rate": 0.0005,
"loss": 0.8014139533042908,
"step": 1106
},
{
"epoch": 0.45714285714285713,
"grad_norm": 0.4514811336994171,
"learning_rate": 0.0005,
"loss": 0.683321475982666,
"step": 1108
},
{
"epoch": 0.45796802475502835,
"grad_norm": 0.44530627131462097,
"learning_rate": 0.0005,
"loss": 0.8450584411621094,
"step": 1110
},
{
"epoch": 0.45879319236719956,
"grad_norm": 0.32030922174453735,
"learning_rate": 0.0005,
"loss": 1.2461864948272705,
"step": 1112
},
{
"epoch": 0.4596183599793708,
"grad_norm": 0.39658382534980774,
"learning_rate": 0.0005,
"loss": 0.8875741362571716,
"step": 1114
},
{
"epoch": 0.46044352759154206,
"grad_norm": 0.4232991933822632,
"learning_rate": 0.0005,
"loss": 1.1723307371139526,
"step": 1116
},
{
"epoch": 0.4612686952037133,
"grad_norm": 0.483547180891037,
"learning_rate": 0.0005,
"loss": 0.6569406986236572,
"step": 1118
},
{
"epoch": 0.4620938628158845,
"grad_norm": 0.5539224743843079,
"learning_rate": 0.0005,
"loss": 1.1091156005859375,
"step": 1120
},
{
"epoch": 0.4629190304280557,
"grad_norm": 0.2994230091571808,
"learning_rate": 0.0005,
"loss": 0.9592258334159851,
"step": 1122
},
{
"epoch": 0.46374419804022693,
"grad_norm": 0.3562132716178894,
"learning_rate": 0.0005,
"loss": 0.8727736473083496,
"step": 1124
},
{
"epoch": 0.46456936565239815,
"grad_norm": 0.4182264804840088,
"learning_rate": 0.0005,
"loss": 0.9040244817733765,
"step": 1126
},
{
"epoch": 0.46539453326456937,
"grad_norm": 0.36431142687797546,
"learning_rate": 0.0005,
"loss": 1.1870160102844238,
"step": 1128
},
{
"epoch": 0.4662197008767406,
"grad_norm": 0.3644370138645172,
"learning_rate": 0.0005,
"loss": 0.9849762916564941,
"step": 1130
},
{
"epoch": 0.4670448684889118,
"grad_norm": 0.39381295442581177,
"learning_rate": 0.0005,
"loss": 0.8918764591217041,
"step": 1132
},
{
"epoch": 0.467870036101083,
"grad_norm": 0.5089045763015747,
"learning_rate": 0.0005,
"loss": 1.0649911165237427,
"step": 1134
},
{
"epoch": 0.46869520371325424,
"grad_norm": 0.5415642261505127,
"learning_rate": 0.0005,
"loss": 0.7683084607124329,
"step": 1136
},
{
"epoch": 0.46952037132542546,
"grad_norm": 0.46666574478149414,
"learning_rate": 0.0005,
"loss": 0.8655437231063843,
"step": 1138
},
{
"epoch": 0.4703455389375967,
"grad_norm": 0.5969123840332031,
"learning_rate": 0.0005,
"loss": 0.44099390506744385,
"step": 1140
},
{
"epoch": 0.47117070654976795,
"grad_norm": 0.37121525406837463,
"learning_rate": 0.0005,
"loss": 1.081122636795044,
"step": 1142
},
{
"epoch": 0.47199587416193917,
"grad_norm": 0.47111591696739197,
"learning_rate": 0.0005,
"loss": 0.6740312576293945,
"step": 1144
},
{
"epoch": 0.4728210417741104,
"grad_norm": 0.36166858673095703,
"learning_rate": 0.0005,
"loss": 1.0292555093765259,
"step": 1146
},
{
"epoch": 0.4736462093862816,
"grad_norm": 0.4886690080165863,
"learning_rate": 0.0005,
"loss": 0.9834333658218384,
"step": 1148
},
{
"epoch": 0.4744713769984528,
"grad_norm": 0.6094414591789246,
"learning_rate": 0.0005,
"loss": 0.7775858640670776,
"step": 1150
},
{
"epoch": 0.47529654461062404,
"grad_norm": 0.6761363744735718,
"learning_rate": 0.0005,
"loss": 0.6978498697280884,
"step": 1152
},
{
"epoch": 0.47612171222279526,
"grad_norm": 0.6768798232078552,
"learning_rate": 0.0005,
"loss": 0.33388641476631165,
"step": 1154
},
{
"epoch": 0.4769468798349665,
"grad_norm": 0.3761197030544281,
"learning_rate": 0.0005,
"loss": 1.0626983642578125,
"step": 1156
},
{
"epoch": 0.4777720474471377,
"grad_norm": 0.5958622097969055,
"learning_rate": 0.0005,
"loss": 0.5583434104919434,
"step": 1158
},
{
"epoch": 0.4785972150593089,
"grad_norm": 0.39165687561035156,
"learning_rate": 0.0005,
"loss": 1.1808472871780396,
"step": 1160
},
{
"epoch": 0.47942238267148013,
"grad_norm": 0.4028123617172241,
"learning_rate": 0.0005,
"loss": 1.024972915649414,
"step": 1162
},
{
"epoch": 0.48024755028365135,
"grad_norm": 0.6045606732368469,
"learning_rate": 0.0005,
"loss": 1.0229902267456055,
"step": 1164
},
{
"epoch": 0.48107271789582257,
"grad_norm": 0.3716071546077728,
"learning_rate": 0.0005,
"loss": 1.5719668865203857,
"step": 1166
},
{
"epoch": 0.4818978855079938,
"grad_norm": 0.4329681098461151,
"learning_rate": 0.0005,
"loss": 1.0816712379455566,
"step": 1168
},
{
"epoch": 0.48272305312016506,
"grad_norm": 0.34830883145332336,
"learning_rate": 0.0005,
"loss": 1.2980117797851562,
"step": 1170
},
{
"epoch": 0.4835482207323363,
"grad_norm": 0.4658760130405426,
"learning_rate": 0.0005,
"loss": 1.01692795753479,
"step": 1172
},
{
"epoch": 0.4843733883445075,
"grad_norm": 0.3759547173976898,
"learning_rate": 0.0005,
"loss": 0.8392401933670044,
"step": 1174
},
{
"epoch": 0.4851985559566787,
"grad_norm": 0.43001633882522583,
"learning_rate": 0.0005,
"loss": 0.9360501170158386,
"step": 1176
},
{
"epoch": 0.48602372356884993,
"grad_norm": 0.4443817734718323,
"learning_rate": 0.0005,
"loss": 0.9738578200340271,
"step": 1178
},
{
"epoch": 0.48684889118102115,
"grad_norm": 0.4765748679637909,
"learning_rate": 0.0005,
"loss": 0.9294180274009705,
"step": 1180
},
{
"epoch": 0.48767405879319237,
"grad_norm": 0.3932465612888336,
"learning_rate": 0.0005,
"loss": 1.2581766843795776,
"step": 1182
},
{
"epoch": 0.4884992264053636,
"grad_norm": 0.39340582489967346,
"learning_rate": 0.0005,
"loss": 1.1049050092697144,
"step": 1184
},
{
"epoch": 0.4893243940175348,
"grad_norm": 0.4255252480506897,
"learning_rate": 0.0005,
"loss": 1.0915213823318481,
"step": 1186
},
{
"epoch": 0.490149561629706,
"grad_norm": 0.3390955328941345,
"learning_rate": 0.0005,
"loss": 1.3654221296310425,
"step": 1188
},
{
"epoch": 0.49097472924187724,
"grad_norm": 0.47792261838912964,
"learning_rate": 0.0005,
"loss": 0.8833599090576172,
"step": 1190
},
{
"epoch": 0.49179989685404846,
"grad_norm": 0.3805239200592041,
"learning_rate": 0.0005,
"loss": 1.1383686065673828,
"step": 1192
},
{
"epoch": 0.4926250644662197,
"grad_norm": 0.5304467678070068,
"learning_rate": 0.0005,
"loss": 0.5999901294708252,
"step": 1194
},
{
"epoch": 0.49345023207839095,
"grad_norm": 0.39166152477264404,
"learning_rate": 0.0005,
"loss": 0.6269493103027344,
"step": 1196
},
{
"epoch": 0.49427539969056217,
"grad_norm": 0.4381334185600281,
"learning_rate": 0.0005,
"loss": 1.045685052871704,
"step": 1198
},
{
"epoch": 0.4951005673027334,
"grad_norm": 0.6005203127861023,
"learning_rate": 0.0005,
"loss": 0.7196356654167175,
"step": 1200
},
{
"epoch": 0.4959257349149046,
"grad_norm": 0.5085334777832031,
"learning_rate": 0.0005,
"loss": 0.8923790454864502,
"step": 1202
},
{
"epoch": 0.4967509025270758,
"grad_norm": 0.46505412459373474,
"learning_rate": 0.0005,
"loss": 1.1734912395477295,
"step": 1204
},
{
"epoch": 0.49757607013924704,
"grad_norm": 0.35354235768318176,
"learning_rate": 0.0005,
"loss": 1.4248706102371216,
"step": 1206
},
{
"epoch": 0.49840123775141826,
"grad_norm": 0.518605649471283,
"learning_rate": 0.0005,
"loss": 0.7615759372711182,
"step": 1208
},
{
"epoch": 0.4992264053635895,
"grad_norm": 0.47183889150619507,
"learning_rate": 0.0005,
"loss": 1.002921462059021,
"step": 1210
},
{
"epoch": 0.5000515729757607,
"grad_norm": 0.5179519057273865,
"learning_rate": 0.0005,
"loss": 0.9206326007843018,
"step": 1212
},
{
"epoch": 0.500876740587932,
"grad_norm": 0.3909929394721985,
"learning_rate": 0.0005,
"loss": 1.098214864730835,
"step": 1214
},
{
"epoch": 0.5017019082001032,
"grad_norm": 0.4962225556373596,
"learning_rate": 0.0005,
"loss": 1.0459990501403809,
"step": 1216
},
{
"epoch": 0.5025270758122744,
"grad_norm": 0.3090408742427826,
"learning_rate": 0.0005,
"loss": 0.8205084800720215,
"step": 1218
},
{
"epoch": 0.5033522434244456,
"grad_norm": 0.4188939034938812,
"learning_rate": 0.0005,
"loss": 1.2620233297348022,
"step": 1220
},
{
"epoch": 0.5041774110366168,
"grad_norm": 0.5601072311401367,
"learning_rate": 0.0005,
"loss": 0.7695064544677734,
"step": 1222
},
{
"epoch": 0.5050025786487881,
"grad_norm": 0.49764370918273926,
"learning_rate": 0.0005,
"loss": 1.0200297832489014,
"step": 1224
},
{
"epoch": 0.5058277462609593,
"grad_norm": 0.33202609419822693,
"learning_rate": 0.0005,
"loss": 1.0644361972808838,
"step": 1226
},
{
"epoch": 0.5066529138731305,
"grad_norm": 0.5339445471763611,
"learning_rate": 0.0005,
"loss": 0.9100528955459595,
"step": 1228
},
{
"epoch": 0.5074780814853017,
"grad_norm": 0.6016252636909485,
"learning_rate": 0.0005,
"loss": 0.7232542037963867,
"step": 1230
},
{
"epoch": 0.5083032490974729,
"grad_norm": 0.4585024118423462,
"learning_rate": 0.0005,
"loss": 1.1162054538726807,
"step": 1232
},
{
"epoch": 0.5091284167096441,
"grad_norm": 0.4517193138599396,
"learning_rate": 0.0005,
"loss": 0.9637409448623657,
"step": 1234
},
{
"epoch": 0.5099535843218154,
"grad_norm": 0.3881579637527466,
"learning_rate": 0.0005,
"loss": 0.7728964686393738,
"step": 1236
},
{
"epoch": 0.5107787519339866,
"grad_norm": 0.39411264657974243,
"learning_rate": 0.0005,
"loss": 1.161026120185852,
"step": 1238
},
{
"epoch": 0.5116039195461578,
"grad_norm": 0.5072040557861328,
"learning_rate": 0.0005,
"loss": 0.8273135423660278,
"step": 1240
},
{
"epoch": 0.512429087158329,
"grad_norm": 0.31434544920921326,
"learning_rate": 0.0005,
"loss": 1.194826602935791,
"step": 1242
},
{
"epoch": 0.5132542547705002,
"grad_norm": 0.4009983539581299,
"learning_rate": 0.0005,
"loss": 1.4253584146499634,
"step": 1244
},
{
"epoch": 0.5140794223826715,
"grad_norm": 0.3834805488586426,
"learning_rate": 0.0005,
"loss": 1.2035033702850342,
"step": 1246
},
{
"epoch": 0.5149045899948427,
"grad_norm": 0.35332781076431274,
"learning_rate": 0.0005,
"loss": 1.150169849395752,
"step": 1248
},
{
"epoch": 0.5157297576070139,
"grad_norm": 0.5810186266899109,
"learning_rate": 0.0005,
"loss": 0.6812407970428467,
"step": 1250
},
{
"epoch": 0.5165549252191851,
"grad_norm": 0.40086501836776733,
"learning_rate": 0.0005,
"loss": 0.8425931334495544,
"step": 1252
},
{
"epoch": 0.5173800928313563,
"grad_norm": 0.31276389956474304,
"learning_rate": 0.0005,
"loss": 1.1845734119415283,
"step": 1254
},
{
"epoch": 0.5182052604435275,
"grad_norm": 0.5723806023597717,
"learning_rate": 0.0005,
"loss": 0.8858379125595093,
"step": 1256
},
{
"epoch": 0.5190304280556988,
"grad_norm": 0.5262450575828552,
"learning_rate": 0.0005,
"loss": 0.6109469532966614,
"step": 1258
},
{
"epoch": 0.51985559566787,
"grad_norm": 0.4611845910549164,
"learning_rate": 0.0005,
"loss": 0.9106293320655823,
"step": 1260
},
{
"epoch": 0.5206807632800412,
"grad_norm": 0.4317900538444519,
"learning_rate": 0.0005,
"loss": 1.214951515197754,
"step": 1262
},
{
"epoch": 0.5215059308922125,
"grad_norm": 0.39260217547416687,
"learning_rate": 0.0005,
"loss": 1.2795312404632568,
"step": 1264
},
{
"epoch": 0.5223310985043838,
"grad_norm": 0.44173046946525574,
"learning_rate": 0.0005,
"loss": 0.9159354567527771,
"step": 1266
},
{
"epoch": 0.523156266116555,
"grad_norm": 0.38811400532722473,
"learning_rate": 0.0005,
"loss": 0.9809489846229553,
"step": 1268
},
{
"epoch": 0.5239814337287262,
"grad_norm": 0.39664143323898315,
"learning_rate": 0.0005,
"loss": 1.0067797899246216,
"step": 1270
},
{
"epoch": 0.5248066013408974,
"grad_norm": 0.5188791155815125,
"learning_rate": 0.0005,
"loss": 1.0113688707351685,
"step": 1272
},
{
"epoch": 0.5256317689530686,
"grad_norm": 0.38724851608276367,
"learning_rate": 0.0005,
"loss": 0.9794062972068787,
"step": 1274
},
{
"epoch": 0.5264569365652398,
"grad_norm": 0.49774831533432007,
"learning_rate": 0.0005,
"loss": 0.3858436346054077,
"step": 1276
},
{
"epoch": 0.5272821041774111,
"grad_norm": 0.42279985547065735,
"learning_rate": 0.0005,
"loss": 1.0161031484603882,
"step": 1278
},
{
"epoch": 0.5281072717895823,
"grad_norm": 0.35690587759017944,
"learning_rate": 0.0005,
"loss": 1.36087167263031,
"step": 1280
},
{
"epoch": 0.5289324394017535,
"grad_norm": 0.4029795527458191,
"learning_rate": 0.0005,
"loss": 1.071509599685669,
"step": 1282
},
{
"epoch": 0.5297576070139247,
"grad_norm": 0.41630151867866516,
"learning_rate": 0.0005,
"loss": 0.9624576568603516,
"step": 1284
},
{
"epoch": 0.5305827746260959,
"grad_norm": 0.34391558170318604,
"learning_rate": 0.0005,
"loss": 1.1163986921310425,
"step": 1286
},
{
"epoch": 0.5314079422382672,
"grad_norm": 0.6571336984634399,
"learning_rate": 0.0004991220368744513,
"loss": 0.4417751729488373,
"step": 1288
},
{
"epoch": 0.5322331098504384,
"grad_norm": 0.6130812168121338,
"learning_rate": 0.0004982440737489026,
"loss": 0.36391234397888184,
"step": 1290
},
{
"epoch": 0.5330582774626096,
"grad_norm": 0.4626736342906952,
"learning_rate": 0.0004973661106233539,
"loss": 0.3085653483867645,
"step": 1292
},
{
"epoch": 0.5338834450747808,
"grad_norm": 0.5308783650398254,
"learning_rate": 0.0004964881474978051,
"loss": 0.7092111706733704,
"step": 1294
},
{
"epoch": 0.534708612686952,
"grad_norm": 0.3677787184715271,
"learning_rate": 0.0004956101843722563,
"loss": 1.089616298675537,
"step": 1296
},
{
"epoch": 0.5355337802991232,
"grad_norm": 0.3380015194416046,
"learning_rate": 0.0004947322212467076,
"loss": 1.402726173400879,
"step": 1298
},
{
"epoch": 0.5363589479112945,
"grad_norm": 0.40598276257514954,
"learning_rate": 0.000493854258121159,
"loss": 0.8838700652122498,
"step": 1300
},
{
"epoch": 0.5371841155234657,
"grad_norm": 0.46119776368141174,
"learning_rate": 0.0004929762949956102,
"loss": 0.9069271683692932,
"step": 1302
},
{
"epoch": 0.5380092831356369,
"grad_norm": 0.41092023253440857,
"learning_rate": 0.0004920983318700614,
"loss": 1.1770268678665161,
"step": 1304
},
{
"epoch": 0.5388344507478081,
"grad_norm": 0.3025120496749878,
"learning_rate": 0.0004912203687445128,
"loss": 1.1601327657699585,
"step": 1306
},
{
"epoch": 0.5396596183599793,
"grad_norm": 0.5360481142997742,
"learning_rate": 0.000490342405618964,
"loss": 0.8521667718887329,
"step": 1308
},
{
"epoch": 0.5404847859721505,
"grad_norm": 0.5028820633888245,
"learning_rate": 0.0004894644424934153,
"loss": 0.6387576460838318,
"step": 1310
},
{
"epoch": 0.5413099535843218,
"grad_norm": 0.3791085183620453,
"learning_rate": 0.0004885864793678665,
"loss": 1.1344711780548096,
"step": 1312
},
{
"epoch": 0.542135121196493,
"grad_norm": 0.35359784960746765,
"learning_rate": 0.00048770851624231783,
"loss": 1.1713953018188477,
"step": 1314
},
{
"epoch": 0.5429602888086643,
"grad_norm": 0.3026987612247467,
"learning_rate": 0.00048683055311676913,
"loss": 1.1644753217697144,
"step": 1316
},
{
"epoch": 0.5437854564208355,
"grad_norm": 0.4331860840320587,
"learning_rate": 0.0004859525899912204,
"loss": 0.9334622621536255,
"step": 1318
},
{
"epoch": 0.5446106240330068,
"grad_norm": 0.3967585265636444,
"learning_rate": 0.0004850746268656717,
"loss": 1.1278331279754639,
"step": 1320
},
{
"epoch": 0.545435791645178,
"grad_norm": 0.43664076924324036,
"learning_rate": 0.0004841966637401229,
"loss": 0.5641858577728271,
"step": 1322
},
{
"epoch": 0.5462609592573492,
"grad_norm": 0.3973194360733032,
"learning_rate": 0.0004833187006145742,
"loss": 1.2403768301010132,
"step": 1324
},
{
"epoch": 0.5470861268695204,
"grad_norm": 0.4980064332485199,
"learning_rate": 0.00048244073748902546,
"loss": 0.2895811200141907,
"step": 1326
},
{
"epoch": 0.5479112944816916,
"grad_norm": 0.34680861234664917,
"learning_rate": 0.0004815627743634767,
"loss": 1.3070834875106812,
"step": 1328
},
{
"epoch": 0.5487364620938628,
"grad_norm": 0.42698851227760315,
"learning_rate": 0.000480684811237928,
"loss": 0.9220706224441528,
"step": 1330
},
{
"epoch": 0.5495616297060341,
"grad_norm": 0.5145670175552368,
"learning_rate": 0.0004798068481123793,
"loss": 0.37659546732902527,
"step": 1332
},
{
"epoch": 0.5503867973182053,
"grad_norm": 0.3329300880432129,
"learning_rate": 0.00047892888498683056,
"loss": 0.7529263496398926,
"step": 1334
},
{
"epoch": 0.5512119649303765,
"grad_norm": 0.49658384919166565,
"learning_rate": 0.00047805092186128186,
"loss": 1.0749951601028442,
"step": 1336
},
{
"epoch": 0.5520371325425477,
"grad_norm": 0.3877248167991638,
"learning_rate": 0.00047717295873573316,
"loss": 1.0551838874816895,
"step": 1338
},
{
"epoch": 0.5528623001547189,
"grad_norm": 0.4503857493400574,
"learning_rate": 0.00047629499561018435,
"loss": 1.128148078918457,
"step": 1340
},
{
"epoch": 0.5536874677668902,
"grad_norm": 0.4377855062484741,
"learning_rate": 0.00047541703248463565,
"loss": 1.1315715312957764,
"step": 1342
},
{
"epoch": 0.5545126353790614,
"grad_norm": 0.614337682723999,
"learning_rate": 0.0004745390693590869,
"loss": 1.1050728559494019,
"step": 1344
},
{
"epoch": 0.5553378029912326,
"grad_norm": 0.317623496055603,
"learning_rate": 0.0004736611062335382,
"loss": 1.042557954788208,
"step": 1346
},
{
"epoch": 0.5561629706034038,
"grad_norm": 0.49475234746932983,
"learning_rate": 0.0004727831431079895,
"loss": 1.0620388984680176,
"step": 1348
},
{
"epoch": 0.556988138215575,
"grad_norm": 0.39456936717033386,
"learning_rate": 0.00047190517998244074,
"loss": 1.0948154926300049,
"step": 1350
},
{
"epoch": 0.5578133058277462,
"grad_norm": 0.5954630374908447,
"learning_rate": 0.00047102721685689204,
"loss": 0.5793792009353638,
"step": 1352
},
{
"epoch": 0.5586384734399175,
"grad_norm": 0.41103798151016235,
"learning_rate": 0.00047014925373134334,
"loss": 0.7669075131416321,
"step": 1354
},
{
"epoch": 0.5594636410520887,
"grad_norm": 0.43824896216392517,
"learning_rate": 0.00046927129060579453,
"loss": 0.830961287021637,
"step": 1356
},
{
"epoch": 0.5602888086642599,
"grad_norm": 0.5695056915283203,
"learning_rate": 0.00046839332748024583,
"loss": 0.6444680690765381,
"step": 1358
},
{
"epoch": 0.5611139762764311,
"grad_norm": 0.514701247215271,
"learning_rate": 0.00046751536435469713,
"loss": 0.5221894979476929,
"step": 1360
},
{
"epoch": 0.5619391438886023,
"grad_norm": 0.3853067457675934,
"learning_rate": 0.0004666374012291484,
"loss": 0.8628042936325073,
"step": 1362
},
{
"epoch": 0.5627643115007736,
"grad_norm": 0.4907379746437073,
"learning_rate": 0.0004657594381035997,
"loss": 0.5828011631965637,
"step": 1364
},
{
"epoch": 0.5635894791129448,
"grad_norm": 0.4802544116973877,
"learning_rate": 0.0004648814749780509,
"loss": 0.7528935670852661,
"step": 1366
},
{
"epoch": 0.564414646725116,
"grad_norm": 0.4288300573825836,
"learning_rate": 0.0004640035118525022,
"loss": 1.4232219457626343,
"step": 1368
},
{
"epoch": 0.5652398143372873,
"grad_norm": 0.5881210565567017,
"learning_rate": 0.0004631255487269535,
"loss": 0.916156530380249,
"step": 1370
},
{
"epoch": 0.5660649819494585,
"grad_norm": 0.5717863440513611,
"learning_rate": 0.0004622475856014047,
"loss": 0.5835635662078857,
"step": 1372
},
{
"epoch": 0.5668901495616298,
"grad_norm": 0.42259716987609863,
"learning_rate": 0.000461369622475856,
"loss": 0.5653491020202637,
"step": 1374
},
{
"epoch": 0.567715317173801,
"grad_norm": 0.3683032989501953,
"learning_rate": 0.0004604916593503073,
"loss": 1.0970102548599243,
"step": 1376
},
{
"epoch": 0.5685404847859722,
"grad_norm": 0.5160171389579773,
"learning_rate": 0.00045961369622475856,
"loss": 0.966075599193573,
"step": 1378
},
{
"epoch": 0.5693656523981434,
"grad_norm": 0.469651460647583,
"learning_rate": 0.00045873573309920986,
"loss": 0.3504354953765869,
"step": 1380
},
{
"epoch": 0.5701908200103146,
"grad_norm": 0.43254896998405457,
"learning_rate": 0.00045785776997366116,
"loss": 0.5941451787948608,
"step": 1382
},
{
"epoch": 0.5710159876224858,
"grad_norm": 0.4175528287887573,
"learning_rate": 0.0004569798068481124,
"loss": 0.5728800892829895,
"step": 1384
},
{
"epoch": 0.5718411552346571,
"grad_norm": 0.3921774625778198,
"learning_rate": 0.00045610184372256365,
"loss": 0.9127808213233948,
"step": 1386
},
{
"epoch": 0.5726663228468283,
"grad_norm": 0.3250313997268677,
"learning_rate": 0.0004552238805970149,
"loss": 1.175377368927002,
"step": 1388
},
{
"epoch": 0.5734914904589995,
"grad_norm": 0.41904330253601074,
"learning_rate": 0.0004543459174714662,
"loss": 0.604532778263092,
"step": 1390
},
{
"epoch": 0.5743166580711707,
"grad_norm": 0.528682291507721,
"learning_rate": 0.0004534679543459175,
"loss": 1.0776746273040771,
"step": 1392
},
{
"epoch": 0.5751418256833419,
"grad_norm": 0.4170067608356476,
"learning_rate": 0.00045258999122036874,
"loss": 0.6638918519020081,
"step": 1394
},
{
"epoch": 0.5759669932955132,
"grad_norm": 0.4988243877887726,
"learning_rate": 0.00045171202809482004,
"loss": 0.30006930232048035,
"step": 1396
},
{
"epoch": 0.5767921609076844,
"grad_norm": 0.4658001661300659,
"learning_rate": 0.00045083406496927134,
"loss": 0.8113343119621277,
"step": 1398
},
{
"epoch": 0.5776173285198556,
"grad_norm": 0.3265015482902527,
"learning_rate": 0.0004499561018437226,
"loss": 0.6260623931884766,
"step": 1400
},
{
"epoch": 0.5784424961320268,
"grad_norm": 0.3959124982357025,
"learning_rate": 0.00044907813871817383,
"loss": 0.6220636963844299,
"step": 1402
},
{
"epoch": 0.579267663744198,
"grad_norm": 0.4251529276371002,
"learning_rate": 0.00044820017559262513,
"loss": 0.8875057697296143,
"step": 1404
},
{
"epoch": 0.5800928313563692,
"grad_norm": 0.40155768394470215,
"learning_rate": 0.0004473222124670764,
"loss": 0.7960425019264221,
"step": 1406
},
{
"epoch": 0.5809179989685405,
"grad_norm": 0.4251236021518707,
"learning_rate": 0.0004464442493415277,
"loss": 0.7926419377326965,
"step": 1408
},
{
"epoch": 0.5817431665807117,
"grad_norm": 0.3045961558818817,
"learning_rate": 0.0004455662862159789,
"loss": 1.3861045837402344,
"step": 1410
},
{
"epoch": 0.5825683341928829,
"grad_norm": 0.34726616740226746,
"learning_rate": 0.0004446883230904302,
"loss": 1.2533764839172363,
"step": 1412
},
{
"epoch": 0.5833935018050541,
"grad_norm": 0.5045084953308105,
"learning_rate": 0.0004438103599648815,
"loss": 0.31275373697280884,
"step": 1414
},
{
"epoch": 0.5842186694172253,
"grad_norm": 0.415457546710968,
"learning_rate": 0.0004429323968393327,
"loss": 1.1401708126068115,
"step": 1416
},
{
"epoch": 0.5850438370293966,
"grad_norm": 0.5500905513763428,
"learning_rate": 0.000442054433713784,
"loss": 0.8216848969459534,
"step": 1418
},
{
"epoch": 0.5858690046415678,
"grad_norm": 0.4708525538444519,
"learning_rate": 0.0004411764705882353,
"loss": 0.857657790184021,
"step": 1420
},
{
"epoch": 0.586694172253739,
"grad_norm": 0.5442027449607849,
"learning_rate": 0.00044029850746268656,
"loss": 0.8346497416496277,
"step": 1422
},
{
"epoch": 0.5875193398659103,
"grad_norm": 0.33515867590904236,
"learning_rate": 0.00043942054433713786,
"loss": 1.307684302330017,
"step": 1424
},
{
"epoch": 0.5883445074780815,
"grad_norm": 0.41137146949768066,
"learning_rate": 0.00043854258121158915,
"loss": 0.7234978079795837,
"step": 1426
},
{
"epoch": 0.5891696750902528,
"grad_norm": 0.511323094367981,
"learning_rate": 0.0004376646180860404,
"loss": 0.7082851529121399,
"step": 1428
},
{
"epoch": 0.589994842702424,
"grad_norm": 0.33430615067481995,
"learning_rate": 0.0004367866549604917,
"loss": 0.6920802593231201,
"step": 1430
},
{
"epoch": 0.5908200103145952,
"grad_norm": 0.5513045787811279,
"learning_rate": 0.0004359086918349429,
"loss": 0.2706233859062195,
"step": 1432
},
{
"epoch": 0.5916451779267664,
"grad_norm": 0.37472429871559143,
"learning_rate": 0.0004350307287093942,
"loss": 1.054629921913147,
"step": 1434
},
{
"epoch": 0.5924703455389376,
"grad_norm": 0.4750807583332062,
"learning_rate": 0.0004341527655838455,
"loss": 0.8383529186248779,
"step": 1436
},
{
"epoch": 0.5932955131511088,
"grad_norm": 0.526509165763855,
"learning_rate": 0.00043327480245829674,
"loss": 0.971068799495697,
"step": 1438
},
{
"epoch": 0.5941206807632801,
"grad_norm": 0.5157666802406311,
"learning_rate": 0.00043239683933274804,
"loss": 0.8728978633880615,
"step": 1440
},
{
"epoch": 0.5949458483754513,
"grad_norm": 0.47111940383911133,
"learning_rate": 0.00043151887620719934,
"loss": 1.1358085870742798,
"step": 1442
},
{
"epoch": 0.5957710159876225,
"grad_norm": 0.5207206010818481,
"learning_rate": 0.0004306409130816506,
"loss": 0.8365025520324707,
"step": 1444
},
{
"epoch": 0.5965961835997937,
"grad_norm": 0.40927866101264954,
"learning_rate": 0.0004297629499561019,
"loss": 1.1129674911499023,
"step": 1446
},
{
"epoch": 0.5974213512119649,
"grad_norm": 0.34705039858818054,
"learning_rate": 0.00042888498683055313,
"loss": 1.3167273998260498,
"step": 1448
},
{
"epoch": 0.5982465188241362,
"grad_norm": 0.36796677112579346,
"learning_rate": 0.0004280070237050044,
"loss": 1.2987061738967896,
"step": 1450
},
{
"epoch": 0.5990716864363074,
"grad_norm": 0.4247080385684967,
"learning_rate": 0.0004271290605794557,
"loss": 1.227928638458252,
"step": 1452
},
{
"epoch": 0.5998968540484786,
"grad_norm": 0.36866629123687744,
"learning_rate": 0.0004262510974539069,
"loss": 1.4059418439865112,
"step": 1454
},
{
"epoch": 0.6003094378545643,
"eval_loss": 0.6598258018493652,
"eval_runtime": 7.494,
"eval_samples_per_second": 104.617,
"eval_steps_per_second": 13.077,
"step": 1455
},
{
"epoch": 0.6007220216606498,
"grad_norm": 0.4746299386024475,
"learning_rate": 0.0004253731343283582,
"loss": 0.3093152940273285,
"step": 1456
},
{
"epoch": 0.601547189272821,
"grad_norm": 0.4315734803676605,
"learning_rate": 0.0004244951712028095,
"loss": 0.7119120359420776,
"step": 1458
},
{
"epoch": 0.6023723568849922,
"grad_norm": 0.3615897297859192,
"learning_rate": 0.00042361720807726076,
"loss": 0.6444117426872253,
"step": 1460
},
{
"epoch": 0.6031975244971635,
"grad_norm": 0.4430631995201111,
"learning_rate": 0.000422739244951712,
"loss": 0.6780407428741455,
"step": 1462
},
{
"epoch": 0.6040226921093347,
"grad_norm": 0.37095242738723755,
"learning_rate": 0.0004218612818261633,
"loss": 0.6757611036300659,
"step": 1464
},
{
"epoch": 0.6048478597215059,
"grad_norm": 0.4821498692035675,
"learning_rate": 0.00042098331870061456,
"loss": 0.7575255632400513,
"step": 1466
},
{
"epoch": 0.6056730273336771,
"grad_norm": 0.49552884697914124,
"learning_rate": 0.00042010535557506585,
"loss": 0.30565327405929565,
"step": 1468
},
{
"epoch": 0.6064981949458483,
"grad_norm": 0.4212145507335663,
"learning_rate": 0.00041922739244951715,
"loss": 0.9876704216003418,
"step": 1470
},
{
"epoch": 0.6073233625580196,
"grad_norm": 0.3872186839580536,
"learning_rate": 0.0004183494293239684,
"loss": 0.9778451323509216,
"step": 1472
},
{
"epoch": 0.6081485301701908,
"grad_norm": 0.449233740568161,
"learning_rate": 0.0004174714661984197,
"loss": 1.023714303970337,
"step": 1474
},
{
"epoch": 0.608973697782362,
"grad_norm": 0.39826643466949463,
"learning_rate": 0.00041659350307287095,
"loss": 1.184454321861267,
"step": 1476
},
{
"epoch": 0.6097988653945333,
"grad_norm": 0.3940098285675049,
"learning_rate": 0.0004157155399473222,
"loss": 1.5466281175613403,
"step": 1478
},
{
"epoch": 0.6106240330067045,
"grad_norm": 0.5309575796127319,
"learning_rate": 0.0004148375768217735,
"loss": 0.595586895942688,
"step": 1480
},
{
"epoch": 0.6114492006188758,
"grad_norm": 0.38009849190711975,
"learning_rate": 0.00041395961369622474,
"loss": 1.0831810235977173,
"step": 1482
},
{
"epoch": 0.612274368231047,
"grad_norm": 0.3619714379310608,
"learning_rate": 0.00041308165057067604,
"loss": 0.68986576795578,
"step": 1484
},
{
"epoch": 0.6130995358432182,
"grad_norm": 0.473264217376709,
"learning_rate": 0.00041220368744512734,
"loss": 0.4819551408290863,
"step": 1486
},
{
"epoch": 0.6139247034553894,
"grad_norm": 0.34965449571609497,
"learning_rate": 0.0004113257243195786,
"loss": 0.8660916090011597,
"step": 1488
},
{
"epoch": 0.6147498710675606,
"grad_norm": 0.5898658633232117,
"learning_rate": 0.0004104477611940299,
"loss": 0.707758367061615,
"step": 1490
},
{
"epoch": 0.6155750386797318,
"grad_norm": 0.3754606544971466,
"learning_rate": 0.0004095697980684812,
"loss": 1.2457656860351562,
"step": 1492
},
{
"epoch": 0.6164002062919031,
"grad_norm": 0.30546990036964417,
"learning_rate": 0.0004086918349429324,
"loss": 1.0335997343063354,
"step": 1494
},
{
"epoch": 0.6172253739040743,
"grad_norm": 0.49503740668296814,
"learning_rate": 0.0004078138718173837,
"loss": 0.8972082138061523,
"step": 1496
},
{
"epoch": 0.6180505415162455,
"grad_norm": 0.44761160016059875,
"learning_rate": 0.0004069359086918349,
"loss": 0.5387427806854248,
"step": 1498
},
{
"epoch": 0.6188757091284167,
"grad_norm": 0.3960210680961609,
"learning_rate": 0.0004060579455662862,
"loss": 0.6668367385864258,
"step": 1500
},
{
"epoch": 0.6197008767405879,
"grad_norm": 0.3856101930141449,
"learning_rate": 0.0004051799824407375,
"loss": 0.7248905897140503,
"step": 1502
},
{
"epoch": 0.6205260443527592,
"grad_norm": 0.5122077465057373,
"learning_rate": 0.00040430201931518876,
"loss": 0.8312437534332275,
"step": 1504
},
{
"epoch": 0.6213512119649304,
"grad_norm": 0.47769659757614136,
"learning_rate": 0.00040342405618964006,
"loss": 0.6435129046440125,
"step": 1506
},
{
"epoch": 0.6221763795771016,
"grad_norm": 0.33235400915145874,
"learning_rate": 0.0004025460930640913,
"loss": 0.7319386005401611,
"step": 1508
},
{
"epoch": 0.6230015471892728,
"grad_norm": 0.3593953549861908,
"learning_rate": 0.00040166812993854255,
"loss": 0.5382250547409058,
"step": 1510
},
{
"epoch": 0.623826714801444,
"grad_norm": 0.45222756266593933,
"learning_rate": 0.00040079016681299385,
"loss": 0.8383983373641968,
"step": 1512
},
{
"epoch": 0.6246518824136152,
"grad_norm": 0.4340486526489258,
"learning_rate": 0.00039991220368744515,
"loss": 0.9945090413093567,
"step": 1514
},
{
"epoch": 0.6254770500257865,
"grad_norm": 0.4003313481807709,
"learning_rate": 0.0003990342405618964,
"loss": 1.2847504615783691,
"step": 1516
},
{
"epoch": 0.6263022176379577,
"grad_norm": 0.49628981947898865,
"learning_rate": 0.0003981562774363477,
"loss": 0.8268569707870483,
"step": 1518
},
{
"epoch": 0.6271273852501289,
"grad_norm": 0.3184366524219513,
"learning_rate": 0.00039727831431079895,
"loss": 0.864170253276825,
"step": 1520
},
{
"epoch": 0.6279525528623001,
"grad_norm": 0.5141156315803528,
"learning_rate": 0.00039640035118525025,
"loss": 0.7300211787223816,
"step": 1522
},
{
"epoch": 0.6287777204744713,
"grad_norm": 0.6197386980056763,
"learning_rate": 0.0003955223880597015,
"loss": 0.2955779731273651,
"step": 1524
},
{
"epoch": 0.6296028880866426,
"grad_norm": 0.420742928981781,
"learning_rate": 0.00039464442493415274,
"loss": 0.9051149487495422,
"step": 1526
},
{
"epoch": 0.6304280556988138,
"grad_norm": 0.340239942073822,
"learning_rate": 0.00039376646180860404,
"loss": 1.0363712310791016,
"step": 1528
},
{
"epoch": 0.631253223310985,
"grad_norm": 0.49181580543518066,
"learning_rate": 0.00039288849868305534,
"loss": 0.6771790981292725,
"step": 1530
},
{
"epoch": 0.6320783909231563,
"grad_norm": 0.43757620453834534,
"learning_rate": 0.0003920105355575066,
"loss": 1.0031658411026,
"step": 1532
},
{
"epoch": 0.6329035585353275,
"grad_norm": 0.43937456607818604,
"learning_rate": 0.0003911325724319579,
"loss": 0.4636398255825043,
"step": 1534
},
{
"epoch": 0.6337287261474988,
"grad_norm": 0.42807430028915405,
"learning_rate": 0.0003902546093064092,
"loss": 0.8181231617927551,
"step": 1536
},
{
"epoch": 0.63455389375967,
"grad_norm": 0.45582592487335205,
"learning_rate": 0.00038937664618086043,
"loss": 1.0056265592575073,
"step": 1538
},
{
"epoch": 0.6353790613718412,
"grad_norm": 0.38186997175216675,
"learning_rate": 0.00038849868305531167,
"loss": 1.0932308435440063,
"step": 1540
},
{
"epoch": 0.6362042289840124,
"grad_norm": 0.5423994064331055,
"learning_rate": 0.0003876207199297629,
"loss": 0.34196802973747253,
"step": 1542
},
{
"epoch": 0.6370293965961836,
"grad_norm": 0.3649578094482422,
"learning_rate": 0.0003867427568042142,
"loss": 0.6473023891448975,
"step": 1544
},
{
"epoch": 0.6378545642083548,
"grad_norm": 0.40749239921569824,
"learning_rate": 0.0003858647936786655,
"loss": 0.7299573421478271,
"step": 1546
},
{
"epoch": 0.6386797318205261,
"grad_norm": 0.4330870509147644,
"learning_rate": 0.00038498683055311676,
"loss": 0.8700272440910339,
"step": 1548
},
{
"epoch": 0.6395048994326973,
"grad_norm": 0.4576455056667328,
"learning_rate": 0.00038410886742756806,
"loss": 0.4570147693157196,
"step": 1550
},
{
"epoch": 0.6403300670448685,
"grad_norm": 0.45697763562202454,
"learning_rate": 0.00038323090430201936,
"loss": 0.6018993854522705,
"step": 1552
},
{
"epoch": 0.6411552346570397,
"grad_norm": 0.41962113976478577,
"learning_rate": 0.00038235294117647055,
"loss": 1.1331027746200562,
"step": 1554
},
{
"epoch": 0.6419804022692109,
"grad_norm": 0.38749533891677856,
"learning_rate": 0.00038147497805092185,
"loss": 1.2504525184631348,
"step": 1556
},
{
"epoch": 0.6428055698813822,
"grad_norm": 0.4868486225605011,
"learning_rate": 0.00038059701492537315,
"loss": 0.7167837619781494,
"step": 1558
},
{
"epoch": 0.6436307374935534,
"grad_norm": 0.4041600227355957,
"learning_rate": 0.0003797190517998244,
"loss": 0.8274481892585754,
"step": 1560
},
{
"epoch": 0.6444559051057246,
"grad_norm": 0.45088425278663635,
"learning_rate": 0.0003788410886742757,
"loss": 0.8142085671424866,
"step": 1562
},
{
"epoch": 0.6452810727178958,
"grad_norm": 0.4727039635181427,
"learning_rate": 0.00037796312554872695,
"loss": 0.6186519861221313,
"step": 1564
},
{
"epoch": 0.646106240330067,
"grad_norm": 0.3529300093650818,
"learning_rate": 0.00037708516242317825,
"loss": 0.978136420249939,
"step": 1566
},
{
"epoch": 0.6469314079422382,
"grad_norm": 0.3013465702533722,
"learning_rate": 0.00037620719929762955,
"loss": 1.2747105360031128,
"step": 1568
},
{
"epoch": 0.6477565755544095,
"grad_norm": 0.4476848840713501,
"learning_rate": 0.00037532923617208074,
"loss": 0.8842909932136536,
"step": 1570
},
{
"epoch": 0.6485817431665807,
"grad_norm": 0.3449991047382355,
"learning_rate": 0.00037445127304653204,
"loss": 1.4411609172821045,
"step": 1572
},
{
"epoch": 0.6494069107787519,
"grad_norm": 0.3587144613265991,
"learning_rate": 0.00037357330992098334,
"loss": 1.3855810165405273,
"step": 1574
},
{
"epoch": 0.6502320783909231,
"grad_norm": 0.6220622658729553,
"learning_rate": 0.0003726953467954346,
"loss": 0.6865560412406921,
"step": 1576
},
{
"epoch": 0.6510572460030943,
"grad_norm": 0.4296424984931946,
"learning_rate": 0.0003718173836698859,
"loss": 1.1647059917449951,
"step": 1578
},
{
"epoch": 0.6518824136152656,
"grad_norm": 0.3671064078807831,
"learning_rate": 0.0003709394205443372,
"loss": 1.117335319519043,
"step": 1580
},
{
"epoch": 0.6527075812274368,
"grad_norm": 0.5410109758377075,
"learning_rate": 0.0003700614574187884,
"loss": 0.546983540058136,
"step": 1582
},
{
"epoch": 0.653532748839608,
"grad_norm": 0.3807181119918823,
"learning_rate": 0.0003691834942932397,
"loss": 1.3175835609436035,
"step": 1584
},
{
"epoch": 0.6543579164517793,
"grad_norm": 0.37390604615211487,
"learning_rate": 0.0003683055311676909,
"loss": 1.3647856712341309,
"step": 1586
},
{
"epoch": 0.6551830840639505,
"grad_norm": 0.41414088010787964,
"learning_rate": 0.0003674275680421422,
"loss": 0.7055715322494507,
"step": 1588
},
{
"epoch": 0.6560082516761218,
"grad_norm": 0.4180362820625305,
"learning_rate": 0.0003665496049165935,
"loss": 1.1773333549499512,
"step": 1590
},
{
"epoch": 0.656833419288293,
"grad_norm": 0.4382783770561218,
"learning_rate": 0.00036567164179104476,
"loss": 0.8065241575241089,
"step": 1592
},
{
"epoch": 0.6576585869004642,
"grad_norm": 0.36340877413749695,
"learning_rate": 0.00036479367866549606,
"loss": 1.352949857711792,
"step": 1594
},
{
"epoch": 0.6584837545126354,
"grad_norm": 0.371106892824173,
"learning_rate": 0.00036391571553994736,
"loss": 1.4077990055084229,
"step": 1596
},
{
"epoch": 0.6593089221248066,
"grad_norm": 0.44986051321029663,
"learning_rate": 0.0003630377524143986,
"loss": 0.7200274467468262,
"step": 1598
},
{
"epoch": 0.6601340897369778,
"grad_norm": 0.47581765055656433,
"learning_rate": 0.00036215978928884985,
"loss": 0.7127492427825928,
"step": 1600
},
{
"epoch": 0.6609592573491491,
"grad_norm": 0.37528514862060547,
"learning_rate": 0.00036128182616330115,
"loss": 0.9523314833641052,
"step": 1602
},
{
"epoch": 0.6617844249613203,
"grad_norm": 0.408311665058136,
"learning_rate": 0.0003604038630377524,
"loss": 1.0211224555969238,
"step": 1604
},
{
"epoch": 0.6626095925734915,
"grad_norm": 0.47178107500076294,
"learning_rate": 0.0003595258999122037,
"loss": 1.0567493438720703,
"step": 1606
},
{
"epoch": 0.6634347601856627,
"grad_norm": 0.6007169485092163,
"learning_rate": 0.00035864793678665495,
"loss": 0.7269392013549805,
"step": 1608
},
{
"epoch": 0.6642599277978339,
"grad_norm": 0.3889978528022766,
"learning_rate": 0.00035776997366110625,
"loss": 1.1036947965621948,
"step": 1610
},
{
"epoch": 0.6650850954100052,
"grad_norm": 0.3837811052799225,
"learning_rate": 0.00035689201053555754,
"loss": 1.3856288194656372,
"step": 1612
},
{
"epoch": 0.6659102630221764,
"grad_norm": 0.5089377164840698,
"learning_rate": 0.0003560140474100088,
"loss": 0.4937981963157654,
"step": 1614
},
{
"epoch": 0.6667354306343476,
"grad_norm": 0.34067216515541077,
"learning_rate": 0.00035513608428446004,
"loss": 0.8237130045890808,
"step": 1616
},
{
"epoch": 0.6675605982465188,
"grad_norm": 0.40171846747398376,
"learning_rate": 0.00035425812115891134,
"loss": 1.0685899257659912,
"step": 1618
},
{
"epoch": 0.66838576585869,
"grad_norm": 0.2767162322998047,
"learning_rate": 0.0003533801580333626,
"loss": 1.099470615386963,
"step": 1620
},
{
"epoch": 0.6692109334708612,
"grad_norm": 0.4199186861515045,
"learning_rate": 0.0003525021949078139,
"loss": 0.6647270321846008,
"step": 1622
},
{
"epoch": 0.6700361010830325,
"grad_norm": 0.36644843220710754,
"learning_rate": 0.0003516242317822652,
"loss": 0.937100887298584,
"step": 1624
},
{
"epoch": 0.6708612686952037,
"grad_norm": 0.4060082733631134,
"learning_rate": 0.0003507462686567164,
"loss": 1.304785966873169,
"step": 1626
},
{
"epoch": 0.6716864363073749,
"grad_norm": 0.39758309721946716,
"learning_rate": 0.0003498683055311677,
"loss": 0.8108911514282227,
"step": 1628
},
{
"epoch": 0.6725116039195461,
"grad_norm": 0.39537230134010315,
"learning_rate": 0.0003489903424056189,
"loss": 0.7461697459220886,
"step": 1630
},
{
"epoch": 0.6733367715317173,
"grad_norm": 0.4885866343975067,
"learning_rate": 0.0003481123792800702,
"loss": 0.4592432677745819,
"step": 1632
},
{
"epoch": 0.6741619391438886,
"grad_norm": 0.502941906452179,
"learning_rate": 0.0003472344161545215,
"loss": 0.7971556186676025,
"step": 1634
},
{
"epoch": 0.6749871067560598,
"grad_norm": 0.5918352007865906,
"learning_rate": 0.00034635645302897276,
"loss": 0.48820215463638306,
"step": 1636
},
{
"epoch": 0.675812274368231,
"grad_norm": 0.3900318145751953,
"learning_rate": 0.00034547848990342406,
"loss": 1.0401685237884521,
"step": 1638
},
{
"epoch": 0.6766374419804023,
"grad_norm": 0.5158812403678894,
"learning_rate": 0.00034460052677787536,
"loss": 0.6606639623641968,
"step": 1640
},
{
"epoch": 0.6774626095925735,
"grad_norm": 0.39379486441612244,
"learning_rate": 0.0003437225636523266,
"loss": 0.6211991906166077,
"step": 1642
},
{
"epoch": 0.6782877772047448,
"grad_norm": 0.4580170810222626,
"learning_rate": 0.0003428446005267779,
"loss": 1.0115219354629517,
"step": 1644
},
{
"epoch": 0.679112944816916,
"grad_norm": 0.33610036969184875,
"learning_rate": 0.00034196663740122915,
"loss": 1.1665886640548706,
"step": 1646
},
{
"epoch": 0.6799381124290872,
"grad_norm": 0.618628740310669,
"learning_rate": 0.0003410886742756804,
"loss": 0.607356071472168,
"step": 1648
},
{
"epoch": 0.6807632800412584,
"grad_norm": 0.3346913754940033,
"learning_rate": 0.0003402107111501317,
"loss": 0.7926020622253418,
"step": 1650
},
{
"epoch": 0.6815884476534296,
"grad_norm": 0.42327773571014404,
"learning_rate": 0.00033933274802458295,
"loss": 0.8403385877609253,
"step": 1652
},
{
"epoch": 0.6824136152656008,
"grad_norm": 0.42058059573173523,
"learning_rate": 0.00033845478489903424,
"loss": 1.2506749629974365,
"step": 1654
},
{
"epoch": 0.6832387828777721,
"grad_norm": 0.38172900676727295,
"learning_rate": 0.00033757682177348554,
"loss": 1.2832468748092651,
"step": 1656
},
{
"epoch": 0.6840639504899433,
"grad_norm": 0.38277432322502136,
"learning_rate": 0.0003366988586479368,
"loss": 1.2047661542892456,
"step": 1658
},
{
"epoch": 0.6848891181021145,
"grad_norm": 0.4411187171936035,
"learning_rate": 0.0003358208955223881,
"loss": 1.0236732959747314,
"step": 1660
},
{
"epoch": 0.6857142857142857,
"grad_norm": 0.43182462453842163,
"learning_rate": 0.00033494293239683934,
"loss": 0.6213672757148743,
"step": 1662
},
{
"epoch": 0.6865394533264569,
"grad_norm": 0.30232101678848267,
"learning_rate": 0.0003340649692712906,
"loss": 1.1094951629638672,
"step": 1664
},
{
"epoch": 0.6873646209386282,
"grad_norm": 0.3877616822719574,
"learning_rate": 0.0003331870061457419,
"loss": 1.1295530796051025,
"step": 1666
},
{
"epoch": 0.6881897885507994,
"grad_norm": 0.5489928126335144,
"learning_rate": 0.0003323090430201932,
"loss": 0.5290967226028442,
"step": 1668
},
{
"epoch": 0.6890149561629706,
"grad_norm": 0.35781222581863403,
"learning_rate": 0.0003314310798946444,
"loss": 1.3475923538208008,
"step": 1670
},
{
"epoch": 0.6898401237751418,
"grad_norm": 0.534935474395752,
"learning_rate": 0.0003305531167690957,
"loss": 0.3332616090774536,
"step": 1672
},
{
"epoch": 0.690665291387313,
"grad_norm": 0.42341673374176025,
"learning_rate": 0.00032967515364354697,
"loss": 1.0092079639434814,
"step": 1674
},
{
"epoch": 0.6914904589994842,
"grad_norm": 0.4100952744483948,
"learning_rate": 0.0003287971905179982,
"loss": 0.5497714877128601,
"step": 1676
},
{
"epoch": 0.6923156266116555,
"grad_norm": 0.38641801476478577,
"learning_rate": 0.0003279192273924495,
"loss": 0.49007177352905273,
"step": 1678
},
{
"epoch": 0.6931407942238267,
"grad_norm": 0.36416521668434143,
"learning_rate": 0.00032704126426690076,
"loss": 0.9782978296279907,
"step": 1680
},
{
"epoch": 0.6939659618359979,
"grad_norm": 0.40659424662590027,
"learning_rate": 0.00032616330114135206,
"loss": 1.1877061128616333,
"step": 1682
},
{
"epoch": 0.6947911294481691,
"grad_norm": 0.35465723276138306,
"learning_rate": 0.00032528533801580336,
"loss": 0.7303415536880493,
"step": 1684
},
{
"epoch": 0.6956162970603403,
"grad_norm": 0.3550800383090973,
"learning_rate": 0.0003244073748902546,
"loss": 0.5755535364151001,
"step": 1686
},
{
"epoch": 0.6964414646725116,
"grad_norm": 0.3488351106643677,
"learning_rate": 0.0003235294117647059,
"loss": 1.2392137050628662,
"step": 1688
},
{
"epoch": 0.6972666322846828,
"grad_norm": 0.31888994574546814,
"learning_rate": 0.0003226514486391572,
"loss": 0.8368427157402039,
"step": 1690
},
{
"epoch": 0.6980917998968541,
"grad_norm": 0.5488599538803101,
"learning_rate": 0.0003217734855136084,
"loss": 1.1260064840316772,
"step": 1692
},
{
"epoch": 0.6989169675090253,
"grad_norm": 0.4838300347328186,
"learning_rate": 0.0003208955223880597,
"loss": 0.7826201319694519,
"step": 1694
},
{
"epoch": 0.6997421351211965,
"grad_norm": 0.4394646883010864,
"learning_rate": 0.00032001755926251094,
"loss": 0.9872143864631653,
"step": 1696
},
{
"epoch": 0.7005673027333678,
"grad_norm": 0.3630361557006836,
"learning_rate": 0.00031913959613696224,
"loss": 0.8093552589416504,
"step": 1698
},
{
"epoch": 0.701392470345539,
"grad_norm": 0.3823564052581787,
"learning_rate": 0.00031826163301141354,
"loss": 1.2268786430358887,
"step": 1700
},
{
"epoch": 0.7022176379577102,
"grad_norm": 0.5077764391899109,
"learning_rate": 0.0003173836698858648,
"loss": 1.0341436862945557,
"step": 1702
},
{
"epoch": 0.7030428055698814,
"grad_norm": 0.29783451557159424,
"learning_rate": 0.0003165057067603161,
"loss": 1.2967486381530762,
"step": 1704
},
{
"epoch": 0.7038679731820526,
"grad_norm": 0.35147911310195923,
"learning_rate": 0.0003156277436347674,
"loss": 1.4181314706802368,
"step": 1706
},
{
"epoch": 0.7046931407942238,
"grad_norm": 0.47153785824775696,
"learning_rate": 0.0003147497805092186,
"loss": 0.7683252692222595,
"step": 1708
},
{
"epoch": 0.7055183084063951,
"grad_norm": 0.6050301790237427,
"learning_rate": 0.0003138718173836699,
"loss": 0.7639537453651428,
"step": 1710
},
{
"epoch": 0.7063434760185663,
"grad_norm": 0.3844204545021057,
"learning_rate": 0.0003129938542581212,
"loss": 0.9077352285385132,
"step": 1712
},
{
"epoch": 0.7071686436307375,
"grad_norm": 0.46985238790512085,
"learning_rate": 0.0003121158911325724,
"loss": 0.8675673007965088,
"step": 1714
},
{
"epoch": 0.7079938112429087,
"grad_norm": 0.36089178919792175,
"learning_rate": 0.0003112379280070237,
"loss": 0.7602687478065491,
"step": 1716
},
{
"epoch": 0.7088189788550799,
"grad_norm": 0.4392194449901581,
"learning_rate": 0.00031035996488147497,
"loss": 1.0270740985870361,
"step": 1718
},
{
"epoch": 0.7096441464672512,
"grad_norm": 0.4287470877170563,
"learning_rate": 0.00030948200175592627,
"loss": 0.9357495903968811,
"step": 1720
},
{
"epoch": 0.7104693140794224,
"grad_norm": 0.37422382831573486,
"learning_rate": 0.0003086040386303775,
"loss": 0.6922284364700317,
"step": 1722
},
{
"epoch": 0.7112944816915936,
"grad_norm": 0.3758017122745514,
"learning_rate": 0.00030772607550482876,
"loss": 1.2530040740966797,
"step": 1724
},
{
"epoch": 0.7121196493037648,
"grad_norm": 0.4269058108329773,
"learning_rate": 0.00030684811237928006,
"loss": 0.8867902755737305,
"step": 1726
},
{
"epoch": 0.712944816915936,
"grad_norm": 0.41424843668937683,
"learning_rate": 0.00030597014925373136,
"loss": 0.573106586933136,
"step": 1728
},
{
"epoch": 0.7137699845281072,
"grad_norm": 0.3800395131111145,
"learning_rate": 0.0003050921861281826,
"loss": 1.011023759841919,
"step": 1730
},
{
"epoch": 0.7145951521402785,
"grad_norm": 0.4355084002017975,
"learning_rate": 0.0003042142230026339,
"loss": 1.2095154523849487,
"step": 1732
},
{
"epoch": 0.7154203197524497,
"grad_norm": 0.40257638692855835,
"learning_rate": 0.0003033362598770852,
"loss": 1.2243540287017822,
"step": 1734
},
{
"epoch": 0.7162454873646209,
"grad_norm": 0.31051409244537354,
"learning_rate": 0.00030245829675153645,
"loss": 1.3747313022613525,
"step": 1736
},
{
"epoch": 0.7170706549767921,
"grad_norm": 0.49823421239852905,
"learning_rate": 0.0003015803336259877,
"loss": 0.6394418478012085,
"step": 1738
},
{
"epoch": 0.7178958225889633,
"grad_norm": 0.36289334297180176,
"learning_rate": 0.00030070237050043894,
"loss": 0.6176121234893799,
"step": 1740
},
{
"epoch": 0.7187209902011346,
"grad_norm": 0.36476245522499084,
"learning_rate": 0.00029982440737489024,
"loss": 1.3954023122787476,
"step": 1742
},
{
"epoch": 0.7195461578133058,
"grad_norm": 0.3851822018623352,
"learning_rate": 0.00029894644424934154,
"loss": 0.6782857179641724,
"step": 1744
},
{
"epoch": 0.7203713254254771,
"grad_norm": 0.4556499123573303,
"learning_rate": 0.0002980684811237928,
"loss": 1.230463981628418,
"step": 1746
},
{
"epoch": 0.7211964930376483,
"grad_norm": 0.34713202714920044,
"learning_rate": 0.0002971905179982441,
"loss": 1.0126994848251343,
"step": 1748
},
{
"epoch": 0.7220216606498195,
"grad_norm": 0.34959739446640015,
"learning_rate": 0.0002963125548726954,
"loss": 1.1754891872406006,
"step": 1750
},
{
"epoch": 0.7228468282619908,
"grad_norm": 0.31646373867988586,
"learning_rate": 0.00029543459174714664,
"loss": 0.763278603553772,
"step": 1752
},
{
"epoch": 0.723671995874162,
"grad_norm": 0.4951009452342987,
"learning_rate": 0.0002945566286215979,
"loss": 0.5303061604499817,
"step": 1754
},
{
"epoch": 0.7244971634863332,
"grad_norm": 0.30071818828582764,
"learning_rate": 0.0002936786654960492,
"loss": 1.154544472694397,
"step": 1756
},
{
"epoch": 0.7253223310985044,
"grad_norm": 0.38215145468711853,
"learning_rate": 0.0002928007023705004,
"loss": 0.9662572741508484,
"step": 1758
},
{
"epoch": 0.7261474987106756,
"grad_norm": 0.3595885634422302,
"learning_rate": 0.0002919227392449517,
"loss": 0.7524693608283997,
"step": 1760
},
{
"epoch": 0.7269726663228468,
"grad_norm": 0.3811556398868561,
"learning_rate": 0.00029104477611940297,
"loss": 1.029892921447754,
"step": 1762
},
{
"epoch": 0.7277978339350181,
"grad_norm": 0.4552609920501709,
"learning_rate": 0.00029016681299385427,
"loss": 0.9844744801521301,
"step": 1764
},
{
"epoch": 0.7286230015471893,
"grad_norm": 0.43894925713539124,
"learning_rate": 0.00028928884986830557,
"loss": 0.9094060063362122,
"step": 1766
},
{
"epoch": 0.7294481691593605,
"grad_norm": 0.36925697326660156,
"learning_rate": 0.00028841088674275676,
"loss": 1.037130355834961,
"step": 1768
},
{
"epoch": 0.7302733367715317,
"grad_norm": 0.39265406131744385,
"learning_rate": 0.00028753292361720806,
"loss": 1.284218430519104,
"step": 1770
},
{
"epoch": 0.7310985043837029,
"grad_norm": 0.5375381112098694,
"learning_rate": 0.00028665496049165936,
"loss": 0.8548294901847839,
"step": 1772
},
{
"epoch": 0.7319236719958742,
"grad_norm": 0.3060007691383362,
"learning_rate": 0.0002857769973661106,
"loss": 1.1197073459625244,
"step": 1774
},
{
"epoch": 0.7327488396080454,
"grad_norm": 0.375448614358902,
"learning_rate": 0.0002848990342405619,
"loss": 0.7334431409835815,
"step": 1776
},
{
"epoch": 0.7335740072202166,
"grad_norm": 0.5063633322715759,
"learning_rate": 0.0002840210711150132,
"loss": 0.6901007294654846,
"step": 1778
},
{
"epoch": 0.7343991748323878,
"grad_norm": 0.3879646956920624,
"learning_rate": 0.00028314310798946445,
"loss": 0.780268669128418,
"step": 1780
},
{
"epoch": 0.735224342444559,
"grad_norm": 0.392336368560791,
"learning_rate": 0.00028226514486391575,
"loss": 1.0370898246765137,
"step": 1782
},
{
"epoch": 0.7360495100567302,
"grad_norm": 0.35631754994392395,
"learning_rate": 0.00028138718173836694,
"loss": 1.2187612056732178,
"step": 1784
},
{
"epoch": 0.7368746776689015,
"grad_norm": 0.4401189982891083,
"learning_rate": 0.00028050921861281824,
"loss": 1.017791986465454,
"step": 1786
},
{
"epoch": 0.7376998452810727,
"grad_norm": 0.6243374943733215,
"learning_rate": 0.00027963125548726954,
"loss": 0.6062961220741272,
"step": 1788
},
{
"epoch": 0.7385250128932439,
"grad_norm": 0.5046449303627014,
"learning_rate": 0.0002787532923617208,
"loss": 0.31797459721565247,
"step": 1790
},
{
"epoch": 0.7393501805054151,
"grad_norm": 0.3611700236797333,
"learning_rate": 0.0002778753292361721,
"loss": 1.1861761808395386,
"step": 1792
},
{
"epoch": 0.7401753481175863,
"grad_norm": 0.5414829850196838,
"learning_rate": 0.0002769973661106234,
"loss": 0.5539573431015015,
"step": 1794
},
{
"epoch": 0.7410005157297576,
"grad_norm": 0.5342025756835938,
"learning_rate": 0.00027611940298507464,
"loss": 0.5310317873954773,
"step": 1796
},
{
"epoch": 0.7418256833419288,
"grad_norm": 0.34159955382347107,
"learning_rate": 0.00027524143985952593,
"loss": 1.4940531253814697,
"step": 1798
},
{
"epoch": 0.7426508509541001,
"grad_norm": 0.3083161413669586,
"learning_rate": 0.0002743634767339772,
"loss": 0.656245231628418,
"step": 1800
},
{
"epoch": 0.7434760185662713,
"grad_norm": 0.41366398334503174,
"learning_rate": 0.0002734855136084284,
"loss": 0.6663851141929626,
"step": 1802
},
{
"epoch": 0.7443011861784425,
"grad_norm": 0.3679063022136688,
"learning_rate": 0.0002726075504828797,
"loss": 0.7742401361465454,
"step": 1804
},
{
"epoch": 0.7451263537906138,
"grad_norm": 0.31927743554115295,
"learning_rate": 0.00027172958735733097,
"loss": 0.9118162393569946,
"step": 1806
},
{
"epoch": 0.745951521402785,
"grad_norm": 0.3575168550014496,
"learning_rate": 0.00027085162423178227,
"loss": 0.9703124165534973,
"step": 1808
},
{
"epoch": 0.7467766890149562,
"grad_norm": 0.49206969141960144,
"learning_rate": 0.00026997366110623357,
"loss": 0.756108283996582,
"step": 1810
},
{
"epoch": 0.7476018566271274,
"grad_norm": 0.36436766386032104,
"learning_rate": 0.0002690956979806848,
"loss": 1.3029776811599731,
"step": 1812
},
{
"epoch": 0.7484270242392986,
"grad_norm": 0.4405703544616699,
"learning_rate": 0.00026821773485513606,
"loss": 0.5966270565986633,
"step": 1814
},
{
"epoch": 0.7492521918514699,
"grad_norm": 0.4038497507572174,
"learning_rate": 0.00026733977172958736,
"loss": 0.9639216065406799,
"step": 1816
},
{
"epoch": 0.7500773594636411,
"grad_norm": 0.35064542293548584,
"learning_rate": 0.0002664618086040386,
"loss": 1.124794840812683,
"step": 1818
},
{
"epoch": 0.7509025270758123,
"grad_norm": 0.383252888917923,
"learning_rate": 0.0002655838454784899,
"loss": 0.9272993803024292,
"step": 1820
},
{
"epoch": 0.7517276946879835,
"grad_norm": 0.4175947904586792,
"learning_rate": 0.0002647058823529412,
"loss": 0.9488388299942017,
"step": 1822
},
{
"epoch": 0.7525528623001547,
"grad_norm": 0.34488824009895325,
"learning_rate": 0.00026382791922739245,
"loss": 1.2201396226882935,
"step": 1824
},
{
"epoch": 0.7533780299123259,
"grad_norm": 0.3640938997268677,
"learning_rate": 0.00026294995610184375,
"loss": 0.831686794757843,
"step": 1826
},
{
"epoch": 0.7542031975244972,
"grad_norm": 0.4400560259819031,
"learning_rate": 0.00026207199297629505,
"loss": 0.7985442280769348,
"step": 1828
},
{
"epoch": 0.7550283651366684,
"grad_norm": 0.41823992133140564,
"learning_rate": 0.00026119402985074624,
"loss": 0.9638607501983643,
"step": 1830
},
{
"epoch": 0.7558535327488396,
"grad_norm": 0.41185373067855835,
"learning_rate": 0.00026031606672519754,
"loss": 1.1180473566055298,
"step": 1832
},
{
"epoch": 0.7566787003610108,
"grad_norm": 0.43015286326408386,
"learning_rate": 0.0002594381035996488,
"loss": 0.570551872253418,
"step": 1834
},
{
"epoch": 0.757503867973182,
"grad_norm": 0.37049224972724915,
"learning_rate": 0.0002585601404741001,
"loss": 1.0918923616409302,
"step": 1836
},
{
"epoch": 0.7583290355853533,
"grad_norm": 0.45099660754203796,
"learning_rate": 0.0002576821773485514,
"loss": 0.9413292407989502,
"step": 1838
},
{
"epoch": 0.7591542031975245,
"grad_norm": 0.36844146251678467,
"learning_rate": 0.00025680421422300263,
"loss": 1.3153712749481201,
"step": 1840
},
{
"epoch": 0.7599793708096957,
"grad_norm": 0.44805243611335754,
"learning_rate": 0.00025592625109745393,
"loss": 1.0876983404159546,
"step": 1842
},
{
"epoch": 0.7608045384218669,
"grad_norm": 0.3581131100654602,
"learning_rate": 0.00025504828797190523,
"loss": 1.2112014293670654,
"step": 1844
},
{
"epoch": 0.7616297060340381,
"grad_norm": 0.4142184853553772,
"learning_rate": 0.0002541703248463564,
"loss": 0.6445860862731934,
"step": 1846
},
{
"epoch": 0.7624548736462093,
"grad_norm": 0.42709341645240784,
"learning_rate": 0.0002532923617208077,
"loss": 0.5973528623580933,
"step": 1848
},
{
"epoch": 0.7632800412583806,
"grad_norm": 0.5776807069778442,
"learning_rate": 0.00025241439859525897,
"loss": 0.8826550245285034,
"step": 1850
},
{
"epoch": 0.7641052088705518,
"grad_norm": 0.3040612041950226,
"learning_rate": 0.00025153643546971027,
"loss": 0.8516554236412048,
"step": 1852
},
{
"epoch": 0.7649303764827231,
"grad_norm": 0.5194029211997986,
"learning_rate": 0.00025065847234416157,
"loss": 0.8658687472343445,
"step": 1854
},
{
"epoch": 0.7657555440948943,
"grad_norm": 0.42637261748313904,
"learning_rate": 0.0002497805092186128,
"loss": 0.7141196131706238,
"step": 1856
},
{
"epoch": 0.7665807117070655,
"grad_norm": 0.6934012770652771,
"learning_rate": 0.0002489025460930641,
"loss": 0.5010101795196533,
"step": 1858
},
{
"epoch": 0.7674058793192368,
"grad_norm": 0.3282501995563507,
"learning_rate": 0.00024802458296751536,
"loss": 1.0175241231918335,
"step": 1860
},
{
"epoch": 0.768231046931408,
"grad_norm": 0.37478360533714294,
"learning_rate": 0.00024714661984196666,
"loss": 1.3197461366653442,
"step": 1862
},
{
"epoch": 0.7690562145435792,
"grad_norm": 0.4044327139854431,
"learning_rate": 0.0002462686567164179,
"loss": 0.873219907283783,
"step": 1864
},
{
"epoch": 0.7698813821557504,
"grad_norm": 0.5854364633560181,
"learning_rate": 0.0002453906935908692,
"loss": 0.9082726240158081,
"step": 1866
},
{
"epoch": 0.7707065497679216,
"grad_norm": 0.43806132674217224,
"learning_rate": 0.00024451273046532045,
"loss": 1.21761953830719,
"step": 1868
},
{
"epoch": 0.7715317173800929,
"grad_norm": 0.3324369788169861,
"learning_rate": 0.00024363476733977175,
"loss": 0.8433029055595398,
"step": 1870
},
{
"epoch": 0.7723568849922641,
"grad_norm": 0.4753143787384033,
"learning_rate": 0.000242756804214223,
"loss": 0.5376313328742981,
"step": 1872
},
{
"epoch": 0.7731820526044353,
"grad_norm": 0.5203450918197632,
"learning_rate": 0.00024187884108867427,
"loss": 0.28149574995040894,
"step": 1874
},
{
"epoch": 0.7740072202166065,
"grad_norm": 0.41968461871147156,
"learning_rate": 0.00024100087796312557,
"loss": 0.6044559478759766,
"step": 1876
},
{
"epoch": 0.7748323878287777,
"grad_norm": 0.3454659879207611,
"learning_rate": 0.00024012291483757684,
"loss": 0.9694072604179382,
"step": 1878
},
{
"epoch": 0.7756575554409489,
"grad_norm": 0.3888348937034607,
"learning_rate": 0.0002392449517120281,
"loss": 1.192838430404663,
"step": 1880
},
{
"epoch": 0.7764827230531202,
"grad_norm": 0.49342063069343567,
"learning_rate": 0.00023836698858647936,
"loss": 0.2729596793651581,
"step": 1882
},
{
"epoch": 0.7773078906652914,
"grad_norm": 0.4026685953140259,
"learning_rate": 0.00023748902546093066,
"loss": 0.9652158617973328,
"step": 1884
},
{
"epoch": 0.7781330582774626,
"grad_norm": 0.3335443139076233,
"learning_rate": 0.00023661106233538193,
"loss": 1.2335366010665894,
"step": 1886
},
{
"epoch": 0.7789582258896338,
"grad_norm": 0.48160237073898315,
"learning_rate": 0.00023573309920983318,
"loss": 0.5310620665550232,
"step": 1888
},
{
"epoch": 0.779783393501805,
"grad_norm": 0.4103797972202301,
"learning_rate": 0.00023485513608428445,
"loss": 0.5694519281387329,
"step": 1890
},
{
"epoch": 0.7806085611139763,
"grad_norm": 0.3229621648788452,
"learning_rate": 0.00023397717295873575,
"loss": 0.8229124546051025,
"step": 1892
},
{
"epoch": 0.7814337287261475,
"grad_norm": 0.5029621720314026,
"learning_rate": 0.000233099209833187,
"loss": 0.566665768623352,
"step": 1894
},
{
"epoch": 0.7822588963383187,
"grad_norm": 0.3402746021747589,
"learning_rate": 0.00023222124670763827,
"loss": 0.813495934009552,
"step": 1896
},
{
"epoch": 0.7830840639504899,
"grad_norm": 0.37816622853279114,
"learning_rate": 0.00023134328358208957,
"loss": 0.8660256862640381,
"step": 1898
},
{
"epoch": 0.7839092315626611,
"grad_norm": 0.3423992097377777,
"learning_rate": 0.00023046532045654084,
"loss": 1.261060118675232,
"step": 1900
},
{
"epoch": 0.7847343991748323,
"grad_norm": 0.35645005106925964,
"learning_rate": 0.0002295873573309921,
"loss": 1.2247262001037598,
"step": 1902
},
{
"epoch": 0.7855595667870036,
"grad_norm": 0.5156469345092773,
"learning_rate": 0.00022870939420544336,
"loss": 0.8600709438323975,
"step": 1904
},
{
"epoch": 0.7863847343991748,
"grad_norm": 0.3836282193660736,
"learning_rate": 0.00022783143107989466,
"loss": 0.648573637008667,
"step": 1906
},
{
"epoch": 0.7872099020113461,
"grad_norm": 0.34423956274986267,
"learning_rate": 0.00022695346795434593,
"loss": 0.819477915763855,
"step": 1908
},
{
"epoch": 0.7880350696235173,
"grad_norm": 0.4140184223651886,
"learning_rate": 0.00022607550482879718,
"loss": 0.762831449508667,
"step": 1910
},
{
"epoch": 0.7888602372356885,
"grad_norm": 0.36245521903038025,
"learning_rate": 0.00022519754170324845,
"loss": 1.1872693300247192,
"step": 1912
},
{
"epoch": 0.7896854048478598,
"grad_norm": 0.4501936137676239,
"learning_rate": 0.00022431957857769975,
"loss": 0.6167159676551819,
"step": 1914
},
{
"epoch": 0.790510572460031,
"grad_norm": 0.43863949179649353,
"learning_rate": 0.00022344161545215103,
"loss": 0.8569579124450684,
"step": 1916
},
{
"epoch": 0.7913357400722022,
"grad_norm": 0.4526144564151764,
"learning_rate": 0.00022256365232660227,
"loss": 0.95778489112854,
"step": 1918
},
{
"epoch": 0.7921609076843734,
"grad_norm": 0.35098332166671753,
"learning_rate": 0.00022168568920105357,
"loss": 1.1015641689300537,
"step": 1920
},
{
"epoch": 0.7929860752965446,
"grad_norm": 0.48924732208251953,
"learning_rate": 0.00022080772607550484,
"loss": 0.7116357088088989,
"step": 1922
},
{
"epoch": 0.7938112429087159,
"grad_norm": 0.4620846211910248,
"learning_rate": 0.00021992976294995612,
"loss": 0.6092314720153809,
"step": 1924
},
{
"epoch": 0.7946364105208871,
"grad_norm": 0.473129540681839,
"learning_rate": 0.00021905179982440736,
"loss": 1.1279802322387695,
"step": 1926
},
{
"epoch": 0.7954615781330583,
"grad_norm": 0.37501075863838196,
"learning_rate": 0.00021817383669885866,
"loss": 0.7749414443969727,
"step": 1928
},
{
"epoch": 0.7962867457452295,
"grad_norm": 0.304704487323761,
"learning_rate": 0.00021729587357330993,
"loss": 0.9998986721038818,
"step": 1930
},
{
"epoch": 0.7971119133574007,
"grad_norm": 0.3354075849056244,
"learning_rate": 0.00021641791044776118,
"loss": 1.2239255905151367,
"step": 1932
},
{
"epoch": 0.7979370809695719,
"grad_norm": 0.4298786222934723,
"learning_rate": 0.00021553994732221245,
"loss": 0.8664978742599487,
"step": 1934
},
{
"epoch": 0.7987622485817432,
"grad_norm": 0.45839351415634155,
"learning_rate": 0.00021466198419666375,
"loss": 1.1027792692184448,
"step": 1936
},
{
"epoch": 0.7995874161939144,
"grad_norm": 0.34594669938087463,
"learning_rate": 0.00021378402107111503,
"loss": 1.3883459568023682,
"step": 1938
},
{
"epoch": 0.8004125838060856,
"grad_norm": 0.35496780276298523,
"learning_rate": 0.00021290605794556627,
"loss": 0.7482577562332153,
"step": 1940
},
{
"epoch": 0.8004125838060856,
"eval_loss": 0.6474518179893494,
"eval_runtime": 7.3783,
"eval_samples_per_second": 106.258,
"eval_steps_per_second": 13.282,
"step": 1940
},
{
"epoch": 0.8012377514182568,
"grad_norm": 0.3918442726135254,
"learning_rate": 0.00021202809482001757,
"loss": 0.9577019810676575,
"step": 1942
},
{
"epoch": 0.802062919030428,
"grad_norm": 0.2961651086807251,
"learning_rate": 0.00021115013169446884,
"loss": 0.8763103485107422,
"step": 1944
},
{
"epoch": 0.8028880866425993,
"grad_norm": 0.44731009006500244,
"learning_rate": 0.00021027216856892012,
"loss": 0.8207913041114807,
"step": 1946
},
{
"epoch": 0.8037132542547705,
"grad_norm": 0.3657285273075104,
"learning_rate": 0.00020939420544337136,
"loss": 0.9430095553398132,
"step": 1948
},
{
"epoch": 0.8045384218669417,
"grad_norm": 0.49153050780296326,
"learning_rate": 0.00020851624231782266,
"loss": 0.4028320014476776,
"step": 1950
},
{
"epoch": 0.8053635894791129,
"grad_norm": 0.3464409112930298,
"learning_rate": 0.00020763827919227393,
"loss": 0.9057855606079102,
"step": 1952
},
{
"epoch": 0.8061887570912841,
"grad_norm": 0.3299136459827423,
"learning_rate": 0.0002067603160667252,
"loss": 1.3503923416137695,
"step": 1954
},
{
"epoch": 0.8070139247034553,
"grad_norm": 0.40845128893852234,
"learning_rate": 0.00020588235294117645,
"loss": 0.7318932414054871,
"step": 1956
},
{
"epoch": 0.8078390923156266,
"grad_norm": 0.381439208984375,
"learning_rate": 0.00020500438981562775,
"loss": 0.9772282242774963,
"step": 1958
},
{
"epoch": 0.8086642599277978,
"grad_norm": 0.377944678068161,
"learning_rate": 0.00020412642669007903,
"loss": 1.0666216611862183,
"step": 1960
},
{
"epoch": 0.8094894275399691,
"grad_norm": 0.45510539412498474,
"learning_rate": 0.0002032484635645303,
"loss": 0.8491358160972595,
"step": 1962
},
{
"epoch": 0.8103145951521403,
"grad_norm": 0.4647512137889862,
"learning_rate": 0.00020237050043898157,
"loss": 0.6544166207313538,
"step": 1964
},
{
"epoch": 0.8111397627643115,
"grad_norm": 0.46371912956237793,
"learning_rate": 0.00020149253731343284,
"loss": 0.6601966619491577,
"step": 1966
},
{
"epoch": 0.8119649303764828,
"grad_norm": 0.61481112241745,
"learning_rate": 0.00020061457418788412,
"loss": 0.2898620069026947,
"step": 1968
},
{
"epoch": 0.812790097988654,
"grad_norm": 0.43525785207748413,
"learning_rate": 0.0001997366110623354,
"loss": 0.8058556914329529,
"step": 1970
},
{
"epoch": 0.8136152656008252,
"grad_norm": 0.3856489956378937,
"learning_rate": 0.00019885864793678666,
"loss": 0.7819163203239441,
"step": 1972
},
{
"epoch": 0.8144404332129964,
"grad_norm": 0.563132643699646,
"learning_rate": 0.00019798068481123793,
"loss": 0.3033660650253296,
"step": 1974
},
{
"epoch": 0.8152656008251676,
"grad_norm": 0.39100125432014465,
"learning_rate": 0.0001971027216856892,
"loss": 0.881723940372467,
"step": 1976
},
{
"epoch": 0.8160907684373389,
"grad_norm": 0.36505672335624695,
"learning_rate": 0.00019622475856014045,
"loss": 0.5532952547073364,
"step": 1978
},
{
"epoch": 0.8169159360495101,
"grad_norm": 0.32876351475715637,
"learning_rate": 0.00019534679543459175,
"loss": 1.186223030090332,
"step": 1980
},
{
"epoch": 0.8177411036616813,
"grad_norm": 0.29075101017951965,
"learning_rate": 0.00019446883230904303,
"loss": 1.1906335353851318,
"step": 1982
},
{
"epoch": 0.8185662712738525,
"grad_norm": 0.4109480082988739,
"learning_rate": 0.0001935908691834943,
"loss": 1.085921287536621,
"step": 1984
},
{
"epoch": 0.8193914388860237,
"grad_norm": 0.30707886815071106,
"learning_rate": 0.00019271290605794557,
"loss": 1.0752849578857422,
"step": 1986
},
{
"epoch": 0.8202166064981949,
"grad_norm": 0.4031044542789459,
"learning_rate": 0.00019183494293239684,
"loss": 0.5591620206832886,
"step": 1988
},
{
"epoch": 0.8210417741103662,
"grad_norm": 0.45132431387901306,
"learning_rate": 0.00019095697980684812,
"loss": 0.42883580923080444,
"step": 1990
},
{
"epoch": 0.8218669417225374,
"grad_norm": 0.400478720664978,
"learning_rate": 0.0001900790166812994,
"loss": 0.8203611969947815,
"step": 1992
},
{
"epoch": 0.8226921093347086,
"grad_norm": 0.3984661400318146,
"learning_rate": 0.00018920105355575066,
"loss": 0.7499722838401794,
"step": 1994
},
{
"epoch": 0.8235172769468798,
"grad_norm": 0.4407826066017151,
"learning_rate": 0.00018832309043020193,
"loss": 1.2055046558380127,
"step": 1996
},
{
"epoch": 0.824342444559051,
"grad_norm": 0.4140898585319519,
"learning_rate": 0.0001874451273046532,
"loss": 1.2413440942764282,
"step": 1998
},
{
"epoch": 0.8251676121712223,
"grad_norm": 0.3528973162174225,
"learning_rate": 0.00018656716417910448,
"loss": 1.2663402557373047,
"step": 2000
},
{
"epoch": 0.8259927797833935,
"grad_norm": 0.3932705819606781,
"learning_rate": 0.00018568920105355575,
"loss": 1.2502458095550537,
"step": 2002
},
{
"epoch": 0.8268179473955647,
"grad_norm": 0.3215526044368744,
"learning_rate": 0.00018481123792800702,
"loss": 0.9158849120140076,
"step": 2004
},
{
"epoch": 0.8276431150077359,
"grad_norm": 0.34103333950042725,
"learning_rate": 0.0001839332748024583,
"loss": 1.1533005237579346,
"step": 2006
},
{
"epoch": 0.8284682826199071,
"grad_norm": 0.32145535945892334,
"learning_rate": 0.0001830553116769096,
"loss": 1.060155987739563,
"step": 2008
},
{
"epoch": 0.8292934502320783,
"grad_norm": 0.5072720646858215,
"learning_rate": 0.00018217734855136084,
"loss": 0.6592322587966919,
"step": 2010
},
{
"epoch": 0.8301186178442496,
"grad_norm": 0.3372536897659302,
"learning_rate": 0.00018129938542581212,
"loss": 0.8511307835578918,
"step": 2012
},
{
"epoch": 0.8309437854564208,
"grad_norm": 0.379126638174057,
"learning_rate": 0.0001804214223002634,
"loss": 0.5848067402839661,
"step": 2014
},
{
"epoch": 0.8317689530685921,
"grad_norm": 0.35454440116882324,
"learning_rate": 0.0001795434591747147,
"loss": 1.2567951679229736,
"step": 2016
},
{
"epoch": 0.8325941206807633,
"grad_norm": 0.5818988680839539,
"learning_rate": 0.00017866549604916593,
"loss": 0.9290669560432434,
"step": 2018
},
{
"epoch": 0.8334192882929345,
"grad_norm": 0.33638861775398254,
"learning_rate": 0.0001777875329236172,
"loss": 0.7085928320884705,
"step": 2020
},
{
"epoch": 0.8342444559051058,
"grad_norm": 0.35702192783355713,
"learning_rate": 0.00017690956979806848,
"loss": 0.8658257126808167,
"step": 2022
},
{
"epoch": 0.835069623517277,
"grad_norm": 0.443119615316391,
"learning_rate": 0.00017603160667251975,
"loss": 0.9122428894042969,
"step": 2024
},
{
"epoch": 0.8358947911294482,
"grad_norm": 0.4142494797706604,
"learning_rate": 0.00017515364354697102,
"loss": 1.1117370128631592,
"step": 2026
},
{
"epoch": 0.8367199587416194,
"grad_norm": 0.3291090726852417,
"learning_rate": 0.0001742756804214223,
"loss": 0.7454751133918762,
"step": 2028
},
{
"epoch": 0.8375451263537906,
"grad_norm": 0.42654404044151306,
"learning_rate": 0.0001733977172958736,
"loss": 0.8326446413993835,
"step": 2030
},
{
"epoch": 0.8383702939659619,
"grad_norm": 0.5846657156944275,
"learning_rate": 0.00017251975417032484,
"loss": 0.3306165337562561,
"step": 2032
},
{
"epoch": 0.8391954615781331,
"grad_norm": 0.4307144582271576,
"learning_rate": 0.00017164179104477612,
"loss": 0.6682844161987305,
"step": 2034
},
{
"epoch": 0.8400206291903043,
"grad_norm": 0.39662232995033264,
"learning_rate": 0.0001707638279192274,
"loss": 0.6227077841758728,
"step": 2036
},
{
"epoch": 0.8408457968024755,
"grad_norm": 0.41666364669799805,
"learning_rate": 0.0001698858647936787,
"loss": 0.9306868314743042,
"step": 2038
},
{
"epoch": 0.8416709644146467,
"grad_norm": 0.5066844820976257,
"learning_rate": 0.00016900790166812993,
"loss": 0.5432884097099304,
"step": 2040
},
{
"epoch": 0.842496132026818,
"grad_norm": 0.4285769462585449,
"learning_rate": 0.0001681299385425812,
"loss": 1.1380990743637085,
"step": 2042
},
{
"epoch": 0.8433212996389892,
"grad_norm": 0.37431302666664124,
"learning_rate": 0.00016725197541703248,
"loss": 1.2066806554794312,
"step": 2044
},
{
"epoch": 0.8441464672511604,
"grad_norm": 0.37849363684654236,
"learning_rate": 0.00016637401229148378,
"loss": 0.9622324705123901,
"step": 2046
},
{
"epoch": 0.8449716348633316,
"grad_norm": 0.41041675209999084,
"learning_rate": 0.00016549604916593502,
"loss": 0.5616909265518188,
"step": 2048
},
{
"epoch": 0.8457968024755028,
"grad_norm": 0.4685586392879486,
"learning_rate": 0.0001646180860403863,
"loss": 0.7619725465774536,
"step": 2050
},
{
"epoch": 0.846621970087674,
"grad_norm": 0.4099452495574951,
"learning_rate": 0.0001637401229148376,
"loss": 0.6912955641746521,
"step": 2052
},
{
"epoch": 0.8474471376998453,
"grad_norm": 0.31624022126197815,
"learning_rate": 0.00016286215978928887,
"loss": 1.2275352478027344,
"step": 2054
},
{
"epoch": 0.8482723053120165,
"grad_norm": 0.41531193256378174,
"learning_rate": 0.00016198419666374012,
"loss": 0.42400866746902466,
"step": 2056
},
{
"epoch": 0.8490974729241877,
"grad_norm": 0.3571903705596924,
"learning_rate": 0.0001611062335381914,
"loss": 0.4795610308647156,
"step": 2058
},
{
"epoch": 0.8499226405363589,
"grad_norm": 0.3806760609149933,
"learning_rate": 0.0001602282704126427,
"loss": 1.0110211372375488,
"step": 2060
},
{
"epoch": 0.8507478081485301,
"grad_norm": 0.3004569411277771,
"learning_rate": 0.00015935030728709393,
"loss": 1.134873867034912,
"step": 2062
},
{
"epoch": 0.8515729757607013,
"grad_norm": 0.3739044964313507,
"learning_rate": 0.0001584723441615452,
"loss": 1.0498672723770142,
"step": 2064
},
{
"epoch": 0.8523981433728726,
"grad_norm": 0.3717358708381653,
"learning_rate": 0.00015759438103599648,
"loss": 0.6477480530738831,
"step": 2066
},
{
"epoch": 0.8532233109850439,
"grad_norm": 0.38525673747062683,
"learning_rate": 0.00015671641791044778,
"loss": 1.067115306854248,
"step": 2068
},
{
"epoch": 0.8540484785972151,
"grad_norm": 0.5054396390914917,
"learning_rate": 0.00015583845478489902,
"loss": 0.3837727904319763,
"step": 2070
},
{
"epoch": 0.8548736462093863,
"grad_norm": 0.4015854597091675,
"learning_rate": 0.0001549604916593503,
"loss": 1.11061692237854,
"step": 2072
},
{
"epoch": 0.8556988138215575,
"grad_norm": 0.32830098271369934,
"learning_rate": 0.0001540825285338016,
"loss": 1.2500406503677368,
"step": 2074
},
{
"epoch": 0.8565239814337288,
"grad_norm": 0.3559308648109436,
"learning_rate": 0.00015320456540825287,
"loss": 0.8874732255935669,
"step": 2076
},
{
"epoch": 0.8573491490459,
"grad_norm": 0.5171499848365784,
"learning_rate": 0.00015232660228270412,
"loss": 0.5555412173271179,
"step": 2078
},
{
"epoch": 0.8581743166580712,
"grad_norm": 0.34617871046066284,
"learning_rate": 0.0001514486391571554,
"loss": 1.2626279592514038,
"step": 2080
},
{
"epoch": 0.8589994842702424,
"grad_norm": 0.37653273344039917,
"learning_rate": 0.0001505706760316067,
"loss": 0.7151405811309814,
"step": 2082
},
{
"epoch": 0.8598246518824136,
"grad_norm": 0.5051701068878174,
"learning_rate": 0.00014969271290605796,
"loss": 0.3551837205886841,
"step": 2084
},
{
"epoch": 0.8606498194945849,
"grad_norm": 0.2960555851459503,
"learning_rate": 0.0001488147497805092,
"loss": 0.8339341282844543,
"step": 2086
},
{
"epoch": 0.8614749871067561,
"grad_norm": 0.3677196800708771,
"learning_rate": 0.00014793678665496048,
"loss": 1.050702691078186,
"step": 2088
},
{
"epoch": 0.8623001547189273,
"grad_norm": 0.35735470056533813,
"learning_rate": 0.00014705882352941178,
"loss": 1.3122944831848145,
"step": 2090
},
{
"epoch": 0.8631253223310985,
"grad_norm": 0.3939901888370514,
"learning_rate": 0.00014618086040386305,
"loss": 0.8045898675918579,
"step": 2092
},
{
"epoch": 0.8639504899432697,
"grad_norm": 0.5576947331428528,
"learning_rate": 0.0001453028972783143,
"loss": 0.6548949480056763,
"step": 2094
},
{
"epoch": 0.864775657555441,
"grad_norm": 0.49302399158477783,
"learning_rate": 0.0001444249341527656,
"loss": 0.5889261364936829,
"step": 2096
},
{
"epoch": 0.8656008251676122,
"grad_norm": 0.42709454894065857,
"learning_rate": 0.00014354697102721687,
"loss": 0.8608667254447937,
"step": 2098
},
{
"epoch": 0.8664259927797834,
"grad_norm": 0.49461597204208374,
"learning_rate": 0.00014266900790166814,
"loss": 0.2737064063549042,
"step": 2100
},
{
"epoch": 0.8672511603919546,
"grad_norm": 0.43027442693710327,
"learning_rate": 0.0001417910447761194,
"loss": 0.9382959008216858,
"step": 2102
},
{
"epoch": 0.8680763280041258,
"grad_norm": 0.41945138573646545,
"learning_rate": 0.0001409130816505707,
"loss": 0.7548843026161194,
"step": 2104
},
{
"epoch": 0.868901495616297,
"grad_norm": 0.3771017789840698,
"learning_rate": 0.00014003511852502196,
"loss": 1.207047700881958,
"step": 2106
},
{
"epoch": 0.8697266632284683,
"grad_norm": 0.4375125765800476,
"learning_rate": 0.0001391571553994732,
"loss": 1.170303225517273,
"step": 2108
},
{
"epoch": 0.8705518308406395,
"grad_norm": 0.3707104027271271,
"learning_rate": 0.00013827919227392448,
"loss": 1.1709308624267578,
"step": 2110
},
{
"epoch": 0.8713769984528107,
"grad_norm": 0.38274121284484863,
"learning_rate": 0.00013740122914837578,
"loss": 1.1782824993133545,
"step": 2112
},
{
"epoch": 0.8722021660649819,
"grad_norm": 0.4029725193977356,
"learning_rate": 0.00013652326602282705,
"loss": 0.5443870425224304,
"step": 2114
},
{
"epoch": 0.8730273336771531,
"grad_norm": 0.34103474020957947,
"learning_rate": 0.0001356453028972783,
"loss": 1.1713083982467651,
"step": 2116
},
{
"epoch": 0.8738525012893243,
"grad_norm": 0.4495161175727844,
"learning_rate": 0.0001347673397717296,
"loss": 1.0647896528244019,
"step": 2118
},
{
"epoch": 0.8746776689014956,
"grad_norm": 0.39537158608436584,
"learning_rate": 0.00013388937664618087,
"loss": 0.7871707081794739,
"step": 2120
},
{
"epoch": 0.8755028365136669,
"grad_norm": 0.49032217264175415,
"learning_rate": 0.00013301141352063214,
"loss": 0.8567002415657043,
"step": 2122
},
{
"epoch": 0.8763280041258381,
"grad_norm": 0.531928300857544,
"learning_rate": 0.0001321334503950834,
"loss": 0.42335572838783264,
"step": 2124
},
{
"epoch": 0.8771531717380093,
"grad_norm": 0.4072137773036957,
"learning_rate": 0.0001312554872695347,
"loss": 0.5991420149803162,
"step": 2126
},
{
"epoch": 0.8779783393501805,
"grad_norm": 0.329656720161438,
"learning_rate": 0.00013037752414398596,
"loss": 1.1534395217895508,
"step": 2128
},
{
"epoch": 0.8788035069623518,
"grad_norm": 0.41913482546806335,
"learning_rate": 0.00012949956101843723,
"loss": 0.809046745300293,
"step": 2130
},
{
"epoch": 0.879628674574523,
"grad_norm": 0.332945317029953,
"learning_rate": 0.00012862159789288848,
"loss": 1.2948366403579712,
"step": 2132
},
{
"epoch": 0.8804538421866942,
"grad_norm": 0.5400812029838562,
"learning_rate": 0.00012774363476733978,
"loss": 0.5145004987716675,
"step": 2134
},
{
"epoch": 0.8812790097988654,
"grad_norm": 0.34961971640586853,
"learning_rate": 0.00012686567164179105,
"loss": 0.9554323554039001,
"step": 2136
},
{
"epoch": 0.8821041774110366,
"grad_norm": 0.36800262331962585,
"learning_rate": 0.00012598770851624232,
"loss": 1.1778321266174316,
"step": 2138
},
{
"epoch": 0.8829293450232079,
"grad_norm": 0.5734240412712097,
"learning_rate": 0.0001251097453906936,
"loss": 0.7692675590515137,
"step": 2140
},
{
"epoch": 0.8837545126353791,
"grad_norm": 0.45778125524520874,
"learning_rate": 0.00012423178226514487,
"loss": 0.9306511282920837,
"step": 2142
},
{
"epoch": 0.8845796802475503,
"grad_norm": 0.44252243638038635,
"learning_rate": 0.00012335381913959614,
"loss": 1.1382296085357666,
"step": 2144
},
{
"epoch": 0.8854048478597215,
"grad_norm": 0.38627496361732483,
"learning_rate": 0.00012247585601404742,
"loss": 1.0644432306289673,
"step": 2146
},
{
"epoch": 0.8862300154718927,
"grad_norm": 0.3631526529788971,
"learning_rate": 0.00012159789288849867,
"loss": 1.1662378311157227,
"step": 2148
},
{
"epoch": 0.887055183084064,
"grad_norm": 0.3402274549007416,
"learning_rate": 0.00012071992976294996,
"loss": 0.9726006388664246,
"step": 2150
},
{
"epoch": 0.8878803506962352,
"grad_norm": 0.44530025124549866,
"learning_rate": 0.00011984196663740122,
"loss": 0.5405691862106323,
"step": 2152
},
{
"epoch": 0.8887055183084064,
"grad_norm": 0.34933000802993774,
"learning_rate": 0.0001189640035118525,
"loss": 0.8531785607337952,
"step": 2154
},
{
"epoch": 0.8895306859205776,
"grad_norm": 0.3409692049026489,
"learning_rate": 0.00011808604038630378,
"loss": 0.676679790019989,
"step": 2156
},
{
"epoch": 0.8903558535327488,
"grad_norm": 0.3313666582107544,
"learning_rate": 0.00011720807726075505,
"loss": 1.3292533159255981,
"step": 2158
},
{
"epoch": 0.89118102114492,
"grad_norm": 0.3319145441055298,
"learning_rate": 0.00011633011413520632,
"loss": 1.195509672164917,
"step": 2160
},
{
"epoch": 0.8920061887570913,
"grad_norm": 0.37336769700050354,
"learning_rate": 0.0001154521510096576,
"loss": 0.682113528251648,
"step": 2162
},
{
"epoch": 0.8928313563692625,
"grad_norm": 0.42223241925239563,
"learning_rate": 0.00011457418788410887,
"loss": 0.9790430665016174,
"step": 2164
},
{
"epoch": 0.8936565239814337,
"grad_norm": 0.3851698040962219,
"learning_rate": 0.00011369622475856014,
"loss": 1.043063759803772,
"step": 2166
},
{
"epoch": 0.8944816915936049,
"grad_norm": 0.47979748249053955,
"learning_rate": 0.00011281826163301141,
"loss": 0.28356316685676575,
"step": 2168
},
{
"epoch": 0.8953068592057761,
"grad_norm": 0.3596416115760803,
"learning_rate": 0.00011194029850746269,
"loss": 0.6587579250335693,
"step": 2170
},
{
"epoch": 0.8961320268179473,
"grad_norm": 0.394227534532547,
"learning_rate": 0.00011106233538191396,
"loss": 1.0800199508666992,
"step": 2172
},
{
"epoch": 0.8969571944301186,
"grad_norm": 0.5065049529075623,
"learning_rate": 0.00011018437225636523,
"loss": 0.2870980501174927,
"step": 2174
},
{
"epoch": 0.8977823620422899,
"grad_norm": 0.3561447858810425,
"learning_rate": 0.0001093064091308165,
"loss": 1.1055365800857544,
"step": 2176
},
{
"epoch": 0.8986075296544611,
"grad_norm": 0.35583215951919556,
"learning_rate": 0.00010842844600526779,
"loss": 1.0420362949371338,
"step": 2178
},
{
"epoch": 0.8994326972666323,
"grad_norm": 0.46746307611465454,
"learning_rate": 0.00010755048287971905,
"loss": 0.2688465714454651,
"step": 2180
},
{
"epoch": 0.9002578648788035,
"grad_norm": 0.3819965720176697,
"learning_rate": 0.00010667251975417034,
"loss": 0.6475980877876282,
"step": 2182
},
{
"epoch": 0.9010830324909748,
"grad_norm": 0.3294242024421692,
"learning_rate": 0.0001057945566286216,
"loss": 0.9739023447036743,
"step": 2184
},
{
"epoch": 0.901908200103146,
"grad_norm": 0.38986486196517944,
"learning_rate": 0.00010491659350307288,
"loss": 0.6386804580688477,
"step": 2186
},
{
"epoch": 0.9027333677153172,
"grad_norm": 0.4394684433937073,
"learning_rate": 0.00010403863037752414,
"loss": 0.6476434469223022,
"step": 2188
},
{
"epoch": 0.9035585353274884,
"grad_norm": 0.4149572551250458,
"learning_rate": 0.00010316066725197541,
"loss": 1.1065031290054321,
"step": 2190
},
{
"epoch": 0.9043837029396596,
"grad_norm": 0.33889737725257874,
"learning_rate": 0.00010228270412642669,
"loss": 0.7109832763671875,
"step": 2192
},
{
"epoch": 0.9052088705518309,
"grad_norm": 0.4915042817592621,
"learning_rate": 0.00010140474100087796,
"loss": 0.6094765663146973,
"step": 2194
},
{
"epoch": 0.9060340381640021,
"grad_norm": 0.34433478116989136,
"learning_rate": 0.00010052677787532923,
"loss": 1.1867214441299438,
"step": 2196
},
{
"epoch": 0.9068592057761733,
"grad_norm": 0.34856536984443665,
"learning_rate": 9.96488147497805e-05,
"loss": 0.6846607327461243,
"step": 2198
},
{
"epoch": 0.9076843733883445,
"grad_norm": 0.5036457777023315,
"learning_rate": 9.877085162423179e-05,
"loss": 0.9738303422927856,
"step": 2200
},
{
"epoch": 0.9085095410005157,
"grad_norm": 0.37732449173927307,
"learning_rate": 9.789288849868305e-05,
"loss": 0.6344292759895325,
"step": 2202
},
{
"epoch": 0.909334708612687,
"grad_norm": 0.47652241587638855,
"learning_rate": 9.701492537313434e-05,
"loss": 1.1111469268798828,
"step": 2204
},
{
"epoch": 0.9101598762248582,
"grad_norm": 0.330166757106781,
"learning_rate": 9.61369622475856e-05,
"loss": 1.0589932203292847,
"step": 2206
},
{
"epoch": 0.9109850438370294,
"grad_norm": 0.36877819895744324,
"learning_rate": 9.525899912203688e-05,
"loss": 1.3222074508666992,
"step": 2208
},
{
"epoch": 0.9118102114492006,
"grad_norm": 0.42300641536712646,
"learning_rate": 9.438103599648814e-05,
"loss": 0.9142203330993652,
"step": 2210
},
{
"epoch": 0.9126353790613718,
"grad_norm": 0.4391476809978485,
"learning_rate": 9.350307287093943e-05,
"loss": 0.2476496547460556,
"step": 2212
},
{
"epoch": 0.913460546673543,
"grad_norm": 0.40787941217422485,
"learning_rate": 9.262510974539069e-05,
"loss": 1.0889965295791626,
"step": 2214
},
{
"epoch": 0.9142857142857143,
"grad_norm": 0.3469390571117401,
"learning_rate": 9.174714661984197e-05,
"loss": 1.1763725280761719,
"step": 2216
},
{
"epoch": 0.9151108818978855,
"grad_norm": 0.42488953471183777,
"learning_rate": 9.086918349429323e-05,
"loss": 1.0028226375579834,
"step": 2218
},
{
"epoch": 0.9159360495100567,
"grad_norm": 0.39252761006355286,
"learning_rate": 8.999122036874452e-05,
"loss": 0.6802842020988464,
"step": 2220
},
{
"epoch": 0.9167612171222279,
"grad_norm": 0.37534213066101074,
"learning_rate": 8.911325724319579e-05,
"loss": 0.7141227722167969,
"step": 2222
},
{
"epoch": 0.9175863847343991,
"grad_norm": 0.49157610535621643,
"learning_rate": 8.823529411764706e-05,
"loss": 0.9489649534225464,
"step": 2224
},
{
"epoch": 0.9184115523465703,
"grad_norm": 0.3925117254257202,
"learning_rate": 8.735733099209834e-05,
"loss": 1.0078821182250977,
"step": 2226
},
{
"epoch": 0.9192367199587416,
"grad_norm": 0.3824593722820282,
"learning_rate": 8.647936786654961e-05,
"loss": 0.573438286781311,
"step": 2228
},
{
"epoch": 0.9200618875709129,
"grad_norm": 0.3944372236728668,
"learning_rate": 8.560140474100088e-05,
"loss": 0.6466689705848694,
"step": 2230
},
{
"epoch": 0.9208870551830841,
"grad_norm": 0.3959895670413971,
"learning_rate": 8.472344161545214e-05,
"loss": 1.0455079078674316,
"step": 2232
},
{
"epoch": 0.9217122227952553,
"grad_norm": 0.3652520477771759,
"learning_rate": 8.384547848990343e-05,
"loss": 0.9111723303794861,
"step": 2234
},
{
"epoch": 0.9225373904074265,
"grad_norm": 0.4143148958683014,
"learning_rate": 8.296751536435469e-05,
"loss": 0.6016215085983276,
"step": 2236
},
{
"epoch": 0.9233625580195978,
"grad_norm": 0.3278208076953888,
"learning_rate": 8.208955223880597e-05,
"loss": 1.371229648590088,
"step": 2238
},
{
"epoch": 0.924187725631769,
"grad_norm": 0.331640362739563,
"learning_rate": 8.121158911325723e-05,
"loss": 0.8705099821090698,
"step": 2240
},
{
"epoch": 0.9250128932439402,
"grad_norm": 0.4457402527332306,
"learning_rate": 8.033362598770852e-05,
"loss": 0.5557809472084045,
"step": 2242
},
{
"epoch": 0.9258380608561114,
"grad_norm": 0.42742833495140076,
"learning_rate": 7.945566286215979e-05,
"loss": 0.847883939743042,
"step": 2244
},
{
"epoch": 0.9266632284682826,
"grad_norm": 0.35928651690483093,
"learning_rate": 7.857769973661106e-05,
"loss": 1.1678866147994995,
"step": 2246
},
{
"epoch": 0.9274883960804539,
"grad_norm": 0.48982396721839905,
"learning_rate": 7.769973661106234e-05,
"loss": 1.015211820602417,
"step": 2248
},
{
"epoch": 0.9283135636926251,
"grad_norm": 0.3965141475200653,
"learning_rate": 7.682177348551361e-05,
"loss": 0.9306076169013977,
"step": 2250
},
{
"epoch": 0.9291387313047963,
"grad_norm": 0.518174946308136,
"learning_rate": 7.594381035996488e-05,
"loss": 0.5669870376586914,
"step": 2252
},
{
"epoch": 0.9299638989169675,
"grad_norm": 0.5797333121299744,
"learning_rate": 7.506584723441616e-05,
"loss": 0.31021666526794434,
"step": 2254
},
{
"epoch": 0.9307890665291387,
"grad_norm": 0.3715957999229431,
"learning_rate": 7.418788410886743e-05,
"loss": 0.9423503279685974,
"step": 2256
},
{
"epoch": 0.93161423414131,
"grad_norm": 0.3591626286506653,
"learning_rate": 7.33099209833187e-05,
"loss": 1.0626194477081299,
"step": 2258
},
{
"epoch": 0.9324394017534812,
"grad_norm": 0.3182604908943176,
"learning_rate": 7.243195785776997e-05,
"loss": 1.2764723300933838,
"step": 2260
},
{
"epoch": 0.9332645693656524,
"grad_norm": 0.4209969639778137,
"learning_rate": 7.155399473222126e-05,
"loss": 0.7196886539459229,
"step": 2262
},
{
"epoch": 0.9340897369778236,
"grad_norm": 0.35605111718177795,
"learning_rate": 7.067603160667252e-05,
"loss": 1.0196263790130615,
"step": 2264
},
{
"epoch": 0.9349149045899948,
"grad_norm": 0.447392076253891,
"learning_rate": 6.97980684811238e-05,
"loss": 0.9616338014602661,
"step": 2266
},
{
"epoch": 0.935740072202166,
"grad_norm": 0.3329978883266449,
"learning_rate": 6.892010535557506e-05,
"loss": 0.9454256892204285,
"step": 2268
},
{
"epoch": 0.9365652398143373,
"grad_norm": 0.5372968912124634,
"learning_rate": 6.804214223002635e-05,
"loss": 0.8368072509765625,
"step": 2270
},
{
"epoch": 0.9373904074265085,
"grad_norm": 0.41135427355766296,
"learning_rate": 6.716417910447761e-05,
"loss": 0.5012705326080322,
"step": 2272
},
{
"epoch": 0.9382155750386797,
"grad_norm": 0.412280410528183,
"learning_rate": 6.62862159789289e-05,
"loss": 0.6955962181091309,
"step": 2274
},
{
"epoch": 0.9390407426508509,
"grad_norm": 0.5137490034103394,
"learning_rate": 6.540825285338016e-05,
"loss": 0.5611425638198853,
"step": 2276
},
{
"epoch": 0.9398659102630221,
"grad_norm": 0.42874646186828613,
"learning_rate": 6.453028972783143e-05,
"loss": 0.8422022461891174,
"step": 2278
},
{
"epoch": 0.9406910778751933,
"grad_norm": 0.3852092921733856,
"learning_rate": 6.36523266022827e-05,
"loss": 0.9753254652023315,
"step": 2280
},
{
"epoch": 0.9415162454873646,
"grad_norm": 0.330733060836792,
"learning_rate": 6.277436347673397e-05,
"loss": 0.7355703711509705,
"step": 2282
},
{
"epoch": 0.9423414130995359,
"grad_norm": 0.3486326336860657,
"learning_rate": 6.189640035118526e-05,
"loss": 0.9386199712753296,
"step": 2284
},
{
"epoch": 0.9431665807117071,
"grad_norm": 0.42362621426582336,
"learning_rate": 6.1018437225636526e-05,
"loss": 1.003790020942688,
"step": 2286
},
{
"epoch": 0.9439917483238783,
"grad_norm": 0.5310701727867126,
"learning_rate": 6.01404741000878e-05,
"loss": 0.4871599078178406,
"step": 2288
},
{
"epoch": 0.9448169159360496,
"grad_norm": 0.5021318197250366,
"learning_rate": 5.926251097453907e-05,
"loss": 0.6965718269348145,
"step": 2290
},
{
"epoch": 0.9456420835482208,
"grad_norm": 0.44528308510780334,
"learning_rate": 5.8384547848990344e-05,
"loss": 0.8355302810668945,
"step": 2292
},
{
"epoch": 0.946467251160392,
"grad_norm": 0.5687183737754822,
"learning_rate": 5.7506584723441617e-05,
"loss": 0.2601590156555176,
"step": 2294
},
{
"epoch": 0.9472924187725632,
"grad_norm": 0.34947848320007324,
"learning_rate": 5.6628621597892896e-05,
"loss": 0.7886998057365417,
"step": 2296
},
{
"epoch": 0.9481175863847344,
"grad_norm": 0.3575672507286072,
"learning_rate": 5.575065847234416e-05,
"loss": 1.0595288276672363,
"step": 2298
},
{
"epoch": 0.9489427539969056,
"grad_norm": 0.522102415561676,
"learning_rate": 5.4872695346795435e-05,
"loss": 0.2766351103782654,
"step": 2300
},
{
"epoch": 0.9497679216090769,
"grad_norm": 0.38553956151008606,
"learning_rate": 5.399473222124671e-05,
"loss": 1.1284608840942383,
"step": 2302
},
{
"epoch": 0.9505930892212481,
"grad_norm": 0.4197658598423004,
"learning_rate": 5.311676909569798e-05,
"loss": 1.0164026021957397,
"step": 2304
},
{
"epoch": 0.9514182568334193,
"grad_norm": 0.3319571912288666,
"learning_rate": 5.223880597014925e-05,
"loss": 0.7956835627555847,
"step": 2306
},
{
"epoch": 0.9522434244455905,
"grad_norm": 0.41206222772598267,
"learning_rate": 5.1360842844600526e-05,
"loss": 0.8848831653594971,
"step": 2308
},
{
"epoch": 0.9530685920577617,
"grad_norm": 0.3841288089752197,
"learning_rate": 5.04828797190518e-05,
"loss": 0.667270839214325,
"step": 2310
},
{
"epoch": 0.953893759669933,
"grad_norm": 0.3630666136741638,
"learning_rate": 4.960491659350307e-05,
"loss": 0.9738955497741699,
"step": 2312
},
{
"epoch": 0.9547189272821042,
"grad_norm": 0.3791244924068451,
"learning_rate": 4.8726953467954344e-05,
"loss": 0.771942675113678,
"step": 2314
},
{
"epoch": 0.9555440948942754,
"grad_norm": 0.5045363306999207,
"learning_rate": 4.7848990342405616e-05,
"loss": 0.8737837672233582,
"step": 2316
},
{
"epoch": 0.9563692625064466,
"grad_norm": 0.6962354779243469,
"learning_rate": 4.6971027216856896e-05,
"loss": 0.7971687912940979,
"step": 2318
},
{
"epoch": 0.9571944301186178,
"grad_norm": 0.30969011783599854,
"learning_rate": 4.609306409130817e-05,
"loss": 1.3567135334014893,
"step": 2320
},
{
"epoch": 0.958019597730789,
"grad_norm": 0.3647533357143402,
"learning_rate": 4.521510096575944e-05,
"loss": 0.7342512011528015,
"step": 2322
},
{
"epoch": 0.9588447653429603,
"grad_norm": 0.42492303252220154,
"learning_rate": 4.4337137840210714e-05,
"loss": 0.6293901205062866,
"step": 2324
},
{
"epoch": 0.9596699329551315,
"grad_norm": 0.3570888638496399,
"learning_rate": 4.345917471466199e-05,
"loss": 0.7224306464195251,
"step": 2326
},
{
"epoch": 0.9604951005673027,
"grad_norm": 0.33239689469337463,
"learning_rate": 4.258121158911326e-05,
"loss": 1.0472383499145508,
"step": 2328
},
{
"epoch": 0.9613202681794739,
"grad_norm": 0.3778957724571228,
"learning_rate": 4.170324846356453e-05,
"loss": 0.7374131679534912,
"step": 2330
},
{
"epoch": 0.9621454357916451,
"grad_norm": 0.47549328207969666,
"learning_rate": 4.0825285338015805e-05,
"loss": 0.4838019609451294,
"step": 2332
},
{
"epoch": 0.9629706034038163,
"grad_norm": 0.4702673554420471,
"learning_rate": 3.994732221246708e-05,
"loss": 0.5558277368545532,
"step": 2334
},
{
"epoch": 0.9637957710159876,
"grad_norm": 0.42936426401138306,
"learning_rate": 3.906935908691835e-05,
"loss": 0.9682615399360657,
"step": 2336
},
{
"epoch": 0.9646209386281589,
"grad_norm": 0.33468520641326904,
"learning_rate": 3.819139596136963e-05,
"loss": 0.9221891164779663,
"step": 2338
},
{
"epoch": 0.9654461062403301,
"grad_norm": 0.416604608297348,
"learning_rate": 3.7313432835820896e-05,
"loss": 0.7495571374893188,
"step": 2340
},
{
"epoch": 0.9662712738525013,
"grad_norm": 0.33602046966552734,
"learning_rate": 3.643546971027217e-05,
"loss": 1.231461763381958,
"step": 2342
},
{
"epoch": 0.9670964414646726,
"grad_norm": 0.4742812216281891,
"learning_rate": 3.555750658472344e-05,
"loss": 0.6368871927261353,
"step": 2344
},
{
"epoch": 0.9679216090768438,
"grad_norm": 0.32410314679145813,
"learning_rate": 3.4679543459174714e-05,
"loss": 1.0154671669006348,
"step": 2346
},
{
"epoch": 0.968746776689015,
"grad_norm": 0.3856157958507538,
"learning_rate": 3.380158033362599e-05,
"loss": 0.9918252229690552,
"step": 2348
},
{
"epoch": 0.9695719443011862,
"grad_norm": 0.33930912613868713,
"learning_rate": 3.292361720807726e-05,
"loss": 0.8040989637374878,
"step": 2350
},
{
"epoch": 0.9703971119133574,
"grad_norm": 0.33802446722984314,
"learning_rate": 3.204565408252853e-05,
"loss": 1.1971032619476318,
"step": 2352
},
{
"epoch": 0.9712222795255286,
"grad_norm": 0.39261794090270996,
"learning_rate": 3.1167690956979805e-05,
"loss": 1.0805972814559937,
"step": 2354
},
{
"epoch": 0.9720474471376999,
"grad_norm": 0.37178656458854675,
"learning_rate": 3.028972783143108e-05,
"loss": 1.1777889728546143,
"step": 2356
},
{
"epoch": 0.9728726147498711,
"grad_norm": 0.3386428654193878,
"learning_rate": 2.9411764705882354e-05,
"loss": 0.7068308591842651,
"step": 2358
},
{
"epoch": 0.9736977823620423,
"grad_norm": 0.32404541969299316,
"learning_rate": 2.8533801580333626e-05,
"loss": 1.1133564710617065,
"step": 2360
},
{
"epoch": 0.9745229499742135,
"grad_norm": 0.5029610395431519,
"learning_rate": 2.76558384547849e-05,
"loss": 0.6080154776573181,
"step": 2362
},
{
"epoch": 0.9753481175863847,
"grad_norm": 0.5243930816650391,
"learning_rate": 2.6777875329236172e-05,
"loss": 0.6097514033317566,
"step": 2364
},
{
"epoch": 0.976173285198556,
"grad_norm": 0.35282057523727417,
"learning_rate": 2.5899912203687448e-05,
"loss": 1.0285130739212036,
"step": 2366
},
{
"epoch": 0.9769984528107272,
"grad_norm": 0.3112531304359436,
"learning_rate": 2.502194907813872e-05,
"loss": 0.7054395079612732,
"step": 2368
},
{
"epoch": 0.9778236204228984,
"grad_norm": 0.4182526767253876,
"learning_rate": 2.4143985952589993e-05,
"loss": 1.0121715068817139,
"step": 2370
},
{
"epoch": 0.9786487880350696,
"grad_norm": 0.507233738899231,
"learning_rate": 2.3266022827041263e-05,
"loss": 0.766700029373169,
"step": 2372
},
{
"epoch": 0.9794739556472408,
"grad_norm": 0.3593057096004486,
"learning_rate": 2.2388059701492536e-05,
"loss": 0.9971503019332886,
"step": 2374
},
{
"epoch": 0.980299123259412,
"grad_norm": 0.4490787982940674,
"learning_rate": 2.1510096575943808e-05,
"loss": 0.8696000576019287,
"step": 2376
},
{
"epoch": 0.9811242908715833,
"grad_norm": 0.4025909900665283,
"learning_rate": 2.0632133450395084e-05,
"loss": 1.014345407485962,
"step": 2378
},
{
"epoch": 0.9819494584837545,
"grad_norm": 0.43438106775283813,
"learning_rate": 1.9754170324846357e-05,
"loss": 0.988254189491272,
"step": 2380
},
{
"epoch": 0.9827746260959257,
"grad_norm": 0.40626466274261475,
"learning_rate": 1.887620719929763e-05,
"loss": 1.177801251411438,
"step": 2382
},
{
"epoch": 0.9835997937080969,
"grad_norm": 0.40217188000679016,
"learning_rate": 1.7998244073748902e-05,
"loss": 0.9934738874435425,
"step": 2384
},
{
"epoch": 0.9844249613202681,
"grad_norm": 0.35306984186172485,
"learning_rate": 1.7120280948200175e-05,
"loss": 1.014739990234375,
"step": 2386
},
{
"epoch": 0.9852501289324394,
"grad_norm": 0.5966408252716064,
"learning_rate": 1.624231782265145e-05,
"loss": 0.5051560401916504,
"step": 2388
},
{
"epoch": 0.9860752965446106,
"grad_norm": 0.3856194317340851,
"learning_rate": 1.536435469710272e-05,
"loss": 0.7751734256744385,
"step": 2390
},
{
"epoch": 0.9869004641567819,
"grad_norm": 0.41418737173080444,
"learning_rate": 1.4486391571553995e-05,
"loss": 0.8432037234306335,
"step": 2392
},
{
"epoch": 0.9877256317689531,
"grad_norm": 0.31449630856513977,
"learning_rate": 1.3608428446005268e-05,
"loss": 0.8296864032745361,
"step": 2394
},
{
"epoch": 0.9885507993811243,
"grad_norm": 0.3470984995365143,
"learning_rate": 1.2730465320456542e-05,
"loss": 0.5408712029457092,
"step": 2396
},
{
"epoch": 0.9893759669932956,
"grad_norm": 0.3735104203224182,
"learning_rate": 1.1852502194907815e-05,
"loss": 0.8418866395950317,
"step": 2398
},
{
"epoch": 0.9902011346054668,
"grad_norm": 0.3870505690574646,
"learning_rate": 1.0974539069359086e-05,
"loss": 1.2456138134002686,
"step": 2400
},
{
"epoch": 0.991026302217638,
"grad_norm": 0.4642637372016907,
"learning_rate": 1.009657594381036e-05,
"loss": 0.38519394397735596,
"step": 2402
},
{
"epoch": 0.9918514698298092,
"grad_norm": 0.4713614284992218,
"learning_rate": 9.218612818261633e-06,
"loss": 1.1183679103851318,
"step": 2404
},
{
"epoch": 0.9926766374419804,
"grad_norm": 0.42575523257255554,
"learning_rate": 8.340649692712907e-06,
"loss": 0.5299312472343445,
"step": 2406
},
{
"epoch": 0.9935018050541516,
"grad_norm": 0.5539902448654175,
"learning_rate": 7.462686567164179e-06,
"loss": 0.4136897921562195,
"step": 2408
},
{
"epoch": 0.9943269726663229,
"grad_norm": 0.4781537652015686,
"learning_rate": 6.584723441615452e-06,
"loss": 0.9022220969200134,
"step": 2410
},
{
"epoch": 0.9951521402784941,
"grad_norm": 0.4406108558177948,
"learning_rate": 5.706760316066725e-06,
"loss": 0.7472458481788635,
"step": 2412
},
{
"epoch": 0.9959773078906653,
"grad_norm": 0.37820929288864136,
"learning_rate": 4.828797190517998e-06,
"loss": 1.0683876276016235,
"step": 2414
},
{
"epoch": 0.9968024755028365,
"grad_norm": 0.3749732971191406,
"learning_rate": 3.950834064969272e-06,
"loss": 0.9929365515708923,
"step": 2416
},
{
"epoch": 0.9976276431150077,
"grad_norm": 0.4763321876525879,
"learning_rate": 3.0728709394205446e-06,
"loss": 0.8095572590827942,
"step": 2418
},
{
"epoch": 0.998452810727179,
"grad_norm": 0.5003067255020142,
"learning_rate": 2.1949078138718174e-06,
"loss": 0.8071651458740234,
"step": 2420
},
{
"epoch": 0.9992779783393502,
"grad_norm": 0.3505983054637909,
"learning_rate": 1.3169446883230905e-06,
"loss": 1.1196868419647217,
"step": 2422
},
{
"epoch": 1.0,
"grad_norm": 0.603049635887146,
"learning_rate": 4.389815627743635e-07,
"loss": 0.34408286213874817,
"step": 2424
},
{
"epoch": 1.0,
"eval_loss": 0.6458116769790649,
"eval_runtime": 7.3655,
"eval_samples_per_second": 106.442,
"eval_steps_per_second": 13.305,
"step": 2424
}
],
"logging_steps": 2,
"max_steps": 2424,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 243,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.339120124357376e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}