bge-m3-more-epoch / trainer_state.json
Mithilss's picture
Upload folder using huggingface_hub
53495a5 verified
{
"best_global_step": 3616,
"best_metric": 0.08387347,
"best_model_checkpoint": "/home/mithil/PycharmProjects/HebrewSearch/output/BAAI/bge-m3/v0-20250928-034732/checkpoint-3616",
"epoch": 4.0,
"eval_steps": 500,
"global_step": 3616,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0011061946902654867,
"grad_norm": 5.15625,
"learning_rate": 5.999998867771367e-06,
"loss": 0.20637531578540802,
"step": 1
},
{
"epoch": 0.0055309734513274336,
"grad_norm": 5.65625,
"learning_rate": 5.999971694326911e-06,
"loss": 0.2206825315952301,
"step": 5
},
{
"epoch": 0.011061946902654867,
"grad_norm": 6.8125,
"learning_rate": 5.9998867778417844e-06,
"loss": 0.19484057426452636,
"step": 10
},
{
"epoch": 0.016592920353982302,
"grad_norm": 5.15625,
"learning_rate": 5.999745252147032e-06,
"loss": 0.16861529350280763,
"step": 15
},
{
"epoch": 0.022123893805309734,
"grad_norm": 6.4375,
"learning_rate": 5.999547119913308e-06,
"loss": 0.16001431941986083,
"step": 20
},
{
"epoch": 0.02765486725663717,
"grad_norm": 4.5,
"learning_rate": 5.999292384879455e-06,
"loss": 0.13545579910278321,
"step": 25
},
{
"epoch": 0.033185840707964605,
"grad_norm": 5.9375,
"learning_rate": 5.998981051852438e-06,
"loss": 0.12482795715332032,
"step": 30
},
{
"epoch": 0.03871681415929203,
"grad_norm": 9.4375,
"learning_rate": 5.998613126707254e-06,
"loss": 0.10733649730682374,
"step": 35
},
{
"epoch": 0.04424778761061947,
"grad_norm": 4.59375,
"learning_rate": 5.998188616386811e-06,
"loss": 0.12228715419769287,
"step": 40
},
{
"epoch": 0.049778761061946904,
"grad_norm": 8.875,
"learning_rate": 5.997707528901811e-06,
"loss": 0.09464394450187683,
"step": 45
},
{
"epoch": 0.05530973451327434,
"grad_norm": 5.40625,
"learning_rate": 5.997169873330592e-06,
"loss": 0.11600016355514527,
"step": 50
},
{
"epoch": 0.06084070796460177,
"grad_norm": 4.53125,
"learning_rate": 5.996575659818955e-06,
"loss": 0.11194223165512085,
"step": 55
},
{
"epoch": 0.06637168141592921,
"grad_norm": 3.375,
"learning_rate": 5.995924899579974e-06,
"loss": 0.0987135112285614,
"step": 60
},
{
"epoch": 0.07190265486725664,
"grad_norm": 2.765625,
"learning_rate": 5.995217604893789e-06,
"loss": 0.09049043655395508,
"step": 65
},
{
"epoch": 0.07743362831858407,
"grad_norm": 3.828125,
"learning_rate": 5.994453789107367e-06,
"loss": 0.10438909530639648,
"step": 70
},
{
"epoch": 0.08296460176991151,
"grad_norm": 4.90625,
"learning_rate": 5.9936334666342555e-06,
"loss": 0.09688463807106018,
"step": 75
},
{
"epoch": 0.08849557522123894,
"grad_norm": 2.953125,
"learning_rate": 5.992756652954306e-06,
"loss": 0.08381547331809998,
"step": 80
},
{
"epoch": 0.09402654867256637,
"grad_norm": 3.21875,
"learning_rate": 5.991823364613388e-06,
"loss": 0.09717974662780762,
"step": 85
},
{
"epoch": 0.09955752212389381,
"grad_norm": 4.03125,
"learning_rate": 5.990833619223069e-06,
"loss": 0.08909911513328553,
"step": 90
},
{
"epoch": 0.10508849557522124,
"grad_norm": 3.359375,
"learning_rate": 5.989787435460292e-06,
"loss": 0.09559545516967774,
"step": 95
},
{
"epoch": 0.11061946902654868,
"grad_norm": 4.90625,
"learning_rate": 5.988684833067012e-06,
"loss": 0.12518904209136963,
"step": 100
},
{
"epoch": 0.1161504424778761,
"grad_norm": 3.359375,
"learning_rate": 5.987525832849829e-06,
"loss": 0.10833564996719361,
"step": 105
},
{
"epoch": 0.12168141592920353,
"grad_norm": 3.390625,
"learning_rate": 5.9863104566796e-06,
"loss": 0.0930842101573944,
"step": 110
},
{
"epoch": 0.12721238938053098,
"grad_norm": 2.828125,
"learning_rate": 5.985038727491019e-06,
"loss": 0.08122435212135315,
"step": 115
},
{
"epoch": 0.13274336283185842,
"grad_norm": 1.8984375,
"learning_rate": 5.983710669282184e-06,
"loss": 0.10641155242919922,
"step": 120
},
{
"epoch": 0.13827433628318583,
"grad_norm": 2.484375,
"learning_rate": 5.982326307114152e-06,
"loss": 0.10035157203674316,
"step": 125
},
{
"epoch": 0.14380530973451328,
"grad_norm": 4.25,
"learning_rate": 5.980885667110457e-06,
"loss": 0.11728062629699706,
"step": 130
},
{
"epoch": 0.14933628318584072,
"grad_norm": 4.53125,
"learning_rate": 5.979388776456621e-06,
"loss": 0.10820536613464356,
"step": 135
},
{
"epoch": 0.15486725663716813,
"grad_norm": 2.421875,
"learning_rate": 5.977835663399645e-06,
"loss": 0.10037193298339844,
"step": 140
},
{
"epoch": 0.16039823008849557,
"grad_norm": 9.8125,
"learning_rate": 5.976226357247468e-06,
"loss": 0.1056329607963562,
"step": 145
},
{
"epoch": 0.16592920353982302,
"grad_norm": 2.671875,
"learning_rate": 5.974560888368418e-06,
"loss": 0.10252338647842407,
"step": 150
},
{
"epoch": 0.17146017699115043,
"grad_norm": 4.875,
"learning_rate": 5.972839288190641e-06,
"loss": 0.10308029651641845,
"step": 155
},
{
"epoch": 0.17699115044247787,
"grad_norm": 2.96875,
"learning_rate": 5.971061589201506e-06,
"loss": 0.1045421838760376,
"step": 160
},
{
"epoch": 0.18252212389380532,
"grad_norm": 2.9375,
"learning_rate": 5.96922782494699e-06,
"loss": 0.10497348308563233,
"step": 165
},
{
"epoch": 0.18805309734513273,
"grad_norm": 3.53125,
"learning_rate": 5.967338030031048e-06,
"loss": 0.07789823412895203,
"step": 170
},
{
"epoch": 0.19358407079646017,
"grad_norm": 3.0,
"learning_rate": 5.965392240114955e-06,
"loss": 0.09011430144309998,
"step": 175
},
{
"epoch": 0.19911504424778761,
"grad_norm": 4.65625,
"learning_rate": 5.963390491916644e-06,
"loss": 0.1126249074935913,
"step": 180
},
{
"epoch": 0.20464601769911506,
"grad_norm": 3.25,
"learning_rate": 5.961332823209998e-06,
"loss": 0.08271955251693726,
"step": 185
},
{
"epoch": 0.21017699115044247,
"grad_norm": 4.4375,
"learning_rate": 5.959219272824152e-06,
"loss": 0.08680923581123352,
"step": 190
},
{
"epoch": 0.2157079646017699,
"grad_norm": 4.0,
"learning_rate": 5.957049880642748e-06,
"loss": 0.0809370756149292,
"step": 195
},
{
"epoch": 0.22123893805309736,
"grad_norm": 6.0625,
"learning_rate": 5.954824687603191e-06,
"loss": 0.08687911629676819,
"step": 200
},
{
"epoch": 0.22676991150442477,
"grad_norm": 4.15625,
"learning_rate": 5.952543735695872e-06,
"loss": 0.09522160291671752,
"step": 205
},
{
"epoch": 0.2323008849557522,
"grad_norm": 15.75,
"learning_rate": 5.950207067963377e-06,
"loss": 0.11227219104766846,
"step": 210
},
{
"epoch": 0.23783185840707965,
"grad_norm": 4.71875,
"learning_rate": 5.947814728499674e-06,
"loss": 0.09369527101516724,
"step": 215
},
{
"epoch": 0.24336283185840707,
"grad_norm": 3.140625,
"learning_rate": 5.945366762449283e-06,
"loss": 0.09389888048171997,
"step": 220
},
{
"epoch": 0.2488938053097345,
"grad_norm": 3.84375,
"learning_rate": 5.942863216006421e-06,
"loss": 0.08466490507125854,
"step": 225
},
{
"epoch": 0.25442477876106195,
"grad_norm": 6.0625,
"learning_rate": 5.940304136414135e-06,
"loss": 0.0902035117149353,
"step": 230
},
{
"epoch": 0.25995575221238937,
"grad_norm": 3.71875,
"learning_rate": 5.937689571963402e-06,
"loss": 0.09655627012252807,
"step": 235
},
{
"epoch": 0.26548672566371684,
"grad_norm": 3.046875,
"learning_rate": 5.935019571992228e-06,
"loss": 0.09126712083816528,
"step": 240
},
{
"epoch": 0.27101769911504425,
"grad_norm": 2.953125,
"learning_rate": 5.932294186884711e-06,
"loss": 0.10732594728469849,
"step": 245
},
{
"epoch": 0.27654867256637167,
"grad_norm": 1.9296875,
"learning_rate": 5.92951346807009e-06,
"loss": 0.08174673318862916,
"step": 250
},
{
"epoch": 0.28207964601769914,
"grad_norm": 2.609375,
"learning_rate": 5.9266774680217784e-06,
"loss": 0.08370364904403686,
"step": 255
},
{
"epoch": 0.28761061946902655,
"grad_norm": 3.21875,
"learning_rate": 5.923786240256368e-06,
"loss": 0.09700153470039367,
"step": 260
},
{
"epoch": 0.29314159292035397,
"grad_norm": 3.390625,
"learning_rate": 5.920839839332624e-06,
"loss": 0.11723319292068482,
"step": 265
},
{
"epoch": 0.29867256637168144,
"grad_norm": 3.28125,
"learning_rate": 5.917838320850455e-06,
"loss": 0.1025324821472168,
"step": 270
},
{
"epoch": 0.30420353982300885,
"grad_norm": 3.4375,
"learning_rate": 5.914781741449862e-06,
"loss": 0.08462166786193848,
"step": 275
},
{
"epoch": 0.30973451327433627,
"grad_norm": 3.390625,
"learning_rate": 5.911670158809868e-06,
"loss": 0.08254794478416443,
"step": 280
},
{
"epoch": 0.31526548672566373,
"grad_norm": 2.46875,
"learning_rate": 5.908503631647435e-06,
"loss": 0.09051342010498047,
"step": 285
},
{
"epoch": 0.32079646017699115,
"grad_norm": 4.28125,
"learning_rate": 5.905282219716351e-06,
"loss": 0.09051858186721802,
"step": 290
},
{
"epoch": 0.32632743362831856,
"grad_norm": 3.34375,
"learning_rate": 5.902005983806106e-06,
"loss": 0.11182348728179932,
"step": 295
},
{
"epoch": 0.33185840707964603,
"grad_norm": 1.6953125,
"learning_rate": 5.89867498574074e-06,
"loss": 0.10160108804702758,
"step": 300
},
{
"epoch": 0.33738938053097345,
"grad_norm": 1.875,
"learning_rate": 5.895289288377682e-06,
"loss": 0.09192607998847961,
"step": 305
},
{
"epoch": 0.34292035398230086,
"grad_norm": 2.65625,
"learning_rate": 5.891848955606559e-06,
"loss": 0.08775659203529358,
"step": 310
},
{
"epoch": 0.34845132743362833,
"grad_norm": 3.4375,
"learning_rate": 5.8883540523479976e-06,
"loss": 0.08484236001968384,
"step": 315
},
{
"epoch": 0.35398230088495575,
"grad_norm": 3.59375,
"learning_rate": 5.884804644552388e-06,
"loss": 0.08966814875602722,
"step": 320
},
{
"epoch": 0.35951327433628316,
"grad_norm": 2.09375,
"learning_rate": 5.881200799198647e-06,
"loss": 0.0682072401046753,
"step": 325
},
{
"epoch": 0.36504424778761063,
"grad_norm": 10.875,
"learning_rate": 5.877542584292956e-06,
"loss": 0.09088088274002075,
"step": 330
},
{
"epoch": 0.37057522123893805,
"grad_norm": 3.390625,
"learning_rate": 5.873830068867472e-06,
"loss": 0.08733047842979431,
"step": 335
},
{
"epoch": 0.37610619469026546,
"grad_norm": 2.65625,
"learning_rate": 5.870063322979024e-06,
"loss": 0.09741016626358032,
"step": 340
},
{
"epoch": 0.38163716814159293,
"grad_norm": 3.046875,
"learning_rate": 5.866242417707799e-06,
"loss": 0.10568550825119019,
"step": 345
},
{
"epoch": 0.38716814159292035,
"grad_norm": 5.5,
"learning_rate": 5.8623674251559945e-06,
"loss": 0.1037368655204773,
"step": 350
},
{
"epoch": 0.3926991150442478,
"grad_norm": 4.0,
"learning_rate": 5.858438418446458e-06,
"loss": 0.10279269218444824,
"step": 355
},
{
"epoch": 0.39823008849557523,
"grad_norm": 4.65625,
"learning_rate": 5.85445547172131e-06,
"loss": 0.07905886173248292,
"step": 360
},
{
"epoch": 0.40376106194690264,
"grad_norm": 2.859375,
"learning_rate": 5.850418660140542e-06,
"loss": 0.08219751119613647,
"step": 365
},
{
"epoch": 0.4092920353982301,
"grad_norm": 9.125,
"learning_rate": 5.8463280598806e-06,
"loss": 0.0956901729106903,
"step": 370
},
{
"epoch": 0.41482300884955753,
"grad_norm": 6.3125,
"learning_rate": 5.842183748132946e-06,
"loss": 0.0784976065158844,
"step": 375
},
{
"epoch": 0.42035398230088494,
"grad_norm": 3.0,
"learning_rate": 5.837985803102602e-06,
"loss": 0.08359727263450623,
"step": 380
},
{
"epoch": 0.4258849557522124,
"grad_norm": 2.1875,
"learning_rate": 5.833734304006674e-06,
"loss": 0.0771904468536377,
"step": 385
},
{
"epoch": 0.4314159292035398,
"grad_norm": 3.5,
"learning_rate": 5.82942933107286e-06,
"loss": 0.080874103307724,
"step": 390
},
{
"epoch": 0.43694690265486724,
"grad_norm": 3.75,
"learning_rate": 5.825070965537931e-06,
"loss": 0.09273157119750977,
"step": 395
},
{
"epoch": 0.4424778761061947,
"grad_norm": 3.265625,
"learning_rate": 5.820659289646197e-06,
"loss": 0.10167332887649536,
"step": 400
},
{
"epoch": 0.4480088495575221,
"grad_norm": 4.25,
"learning_rate": 5.816194386647965e-06,
"loss": 0.0901688277721405,
"step": 405
},
{
"epoch": 0.45353982300884954,
"grad_norm": 3.234375,
"learning_rate": 5.811676340797957e-06,
"loss": 0.08500126600265503,
"step": 410
},
{
"epoch": 0.459070796460177,
"grad_norm": 18.125,
"learning_rate": 5.8071052373537245e-06,
"loss": 0.07273505330085754,
"step": 415
},
{
"epoch": 0.4646017699115044,
"grad_norm": 4.03125,
"learning_rate": 5.802481162574043e-06,
"loss": 0.07213820815086365,
"step": 420
},
{
"epoch": 0.47013274336283184,
"grad_norm": 3.96875,
"learning_rate": 5.797804203717277e-06,
"loss": 0.09481836557388305,
"step": 425
},
{
"epoch": 0.4756637168141593,
"grad_norm": 3.0625,
"learning_rate": 5.793074449039738e-06,
"loss": 0.07019275426864624,
"step": 430
},
{
"epoch": 0.4811946902654867,
"grad_norm": 2.015625,
"learning_rate": 5.7882919877940205e-06,
"loss": 0.10080122947692871,
"step": 435
},
{
"epoch": 0.48672566371681414,
"grad_norm": 2.328125,
"learning_rate": 5.783456910227314e-06,
"loss": 0.09479623436927795,
"step": 440
},
{
"epoch": 0.4922566371681416,
"grad_norm": 3.78125,
"learning_rate": 5.778569307579702e-06,
"loss": 0.10009855031967163,
"step": 445
},
{
"epoch": 0.497787610619469,
"grad_norm": 2.78125,
"learning_rate": 5.773629272082439e-06,
"loss": 0.0831712543964386,
"step": 450
},
{
"epoch": 0.5033185840707964,
"grad_norm": 6.90625,
"learning_rate": 5.768636896956212e-06,
"loss": 0.09191598296165467,
"step": 455
},
{
"epoch": 0.5088495575221239,
"grad_norm": 6.40625,
"learning_rate": 5.763592276409379e-06,
"loss": 0.10570597648620605,
"step": 460
},
{
"epoch": 0.5143805309734514,
"grad_norm": 1.7578125,
"learning_rate": 5.758495505636195e-06,
"loss": 0.07560634016990661,
"step": 465
},
{
"epoch": 0.5199115044247787,
"grad_norm": 1.953125,
"learning_rate": 5.75334668081501e-06,
"loss": 0.09575658440589904,
"step": 470
},
{
"epoch": 0.5254424778761062,
"grad_norm": 2.609375,
"learning_rate": 5.74814589910646e-06,
"loss": 0.07793788313865661,
"step": 475
},
{
"epoch": 0.5309734513274337,
"grad_norm": 5.625,
"learning_rate": 5.742893258651628e-06,
"loss": 0.08670571446418762,
"step": 480
},
{
"epoch": 0.536504424778761,
"grad_norm": 5.375,
"learning_rate": 5.737588858570198e-06,
"loss": 0.10170416831970215,
"step": 485
},
{
"epoch": 0.5420353982300885,
"grad_norm": 1.921875,
"learning_rate": 5.732232798958579e-06,
"loss": 0.11373239755630493,
"step": 490
},
{
"epoch": 0.547566371681416,
"grad_norm": 3.453125,
"learning_rate": 5.726825180888019e-06,
"loss": 0.10954059362411499,
"step": 495
},
{
"epoch": 0.5530973451327433,
"grad_norm": 2.578125,
"learning_rate": 5.721366106402699e-06,
"loss": 0.09733225703239441,
"step": 500
},
{
"epoch": 0.5586283185840708,
"grad_norm": 4.21875,
"learning_rate": 5.715855678517802e-06,
"loss": 0.0946130096912384,
"step": 505
},
{
"epoch": 0.5641592920353983,
"grad_norm": 1.796875,
"learning_rate": 5.710294001217577e-06,
"loss": 0.058262968063354494,
"step": 510
},
{
"epoch": 0.5696902654867256,
"grad_norm": 1.84375,
"learning_rate": 5.7046811794533695e-06,
"loss": 0.07477419972419738,
"step": 515
},
{
"epoch": 0.5752212389380531,
"grad_norm": 4.78125,
"learning_rate": 5.699017319141643e-06,
"loss": 0.09675314426422119,
"step": 520
},
{
"epoch": 0.5807522123893806,
"grad_norm": 2.90625,
"learning_rate": 5.693302527161986e-06,
"loss": 0.08058277368545533,
"step": 525
},
{
"epoch": 0.5862831858407079,
"grad_norm": 4.34375,
"learning_rate": 5.6875369113550855e-06,
"loss": 0.0709432303905487,
"step": 530
},
{
"epoch": 0.5918141592920354,
"grad_norm": 7.78125,
"learning_rate": 5.6817205805207e-06,
"loss": 0.0839331328868866,
"step": 535
},
{
"epoch": 0.5973451327433629,
"grad_norm": 2.59375,
"learning_rate": 5.675853644415601e-06,
"loss": 0.0789577066898346,
"step": 540
},
{
"epoch": 0.6028761061946902,
"grad_norm": 8.5625,
"learning_rate": 5.6699362137515085e-06,
"loss": 0.08718615770339966,
"step": 545
},
{
"epoch": 0.6084070796460177,
"grad_norm": 1.9296875,
"learning_rate": 5.6639684001929906e-06,
"loss": 0.09491783976554871,
"step": 550
},
{
"epoch": 0.6139380530973452,
"grad_norm": 2.0625,
"learning_rate": 5.65795031635537e-06,
"loss": 0.08632619976997376,
"step": 555
},
{
"epoch": 0.6194690265486725,
"grad_norm": 3.59375,
"learning_rate": 5.651882075802588e-06,
"loss": 0.07699216604232788,
"step": 560
},
{
"epoch": 0.625,
"grad_norm": 3.015625,
"learning_rate": 5.645763793045065e-06,
"loss": 0.0748705267906189,
"step": 565
},
{
"epoch": 0.6305309734513275,
"grad_norm": 1.6796875,
"learning_rate": 5.639595583537546e-06,
"loss": 0.07814546823501586,
"step": 570
},
{
"epoch": 0.6360619469026548,
"grad_norm": 2.34375,
"learning_rate": 5.633377563676908e-06,
"loss": 0.08367200493812561,
"step": 575
},
{
"epoch": 0.6415929203539823,
"grad_norm": 2.1875,
"learning_rate": 5.627109850799979e-06,
"loss": 0.07812417149543763,
"step": 580
},
{
"epoch": 0.6471238938053098,
"grad_norm": 1.4609375,
"learning_rate": 5.620792563181313e-06,
"loss": 0.07464907169342042,
"step": 585
},
{
"epoch": 0.6526548672566371,
"grad_norm": 5.3125,
"learning_rate": 5.6144258200309625e-06,
"loss": 0.07074556350708008,
"step": 590
},
{
"epoch": 0.6581858407079646,
"grad_norm": 1.8671875,
"learning_rate": 5.608009741492225e-06,
"loss": 0.09674428105354309,
"step": 595
},
{
"epoch": 0.6637168141592921,
"grad_norm": 5.59375,
"learning_rate": 5.6015444486393845e-06,
"loss": 0.07741823196411132,
"step": 600
},
{
"epoch": 0.6692477876106194,
"grad_norm": 2.15625,
"learning_rate": 5.595030063475417e-06,
"loss": 0.09045513868331909,
"step": 605
},
{
"epoch": 0.6747787610619469,
"grad_norm": 1.5859375,
"learning_rate": 5.588466708929694e-06,
"loss": 0.11130952835083008,
"step": 610
},
{
"epoch": 0.6803097345132744,
"grad_norm": 3.765625,
"learning_rate": 5.58185450885566e-06,
"loss": 0.09267461895942689,
"step": 615
},
{
"epoch": 0.6858407079646017,
"grad_norm": 2.28125,
"learning_rate": 5.5751935880285e-06,
"loss": 0.08096293210983277,
"step": 620
},
{
"epoch": 0.6913716814159292,
"grad_norm": 3.265625,
"learning_rate": 5.568484072142776e-06,
"loss": 0.09142077565193177,
"step": 625
},
{
"epoch": 0.6969026548672567,
"grad_norm": 3.734375,
"learning_rate": 5.561726087810065e-06,
"loss": 0.09943547248840331,
"step": 630
},
{
"epoch": 0.702433628318584,
"grad_norm": 1.9765625,
"learning_rate": 5.554919762556564e-06,
"loss": 0.07108739614486695,
"step": 635
},
{
"epoch": 0.7079646017699115,
"grad_norm": 3.265625,
"learning_rate": 5.548065224820683e-06,
"loss": 0.09564738273620606,
"step": 640
},
{
"epoch": 0.713495575221239,
"grad_norm": 1.796875,
"learning_rate": 5.541162603950627e-06,
"loss": 0.08729345798492431,
"step": 645
},
{
"epoch": 0.7190265486725663,
"grad_norm": 3.28125,
"learning_rate": 5.534212030201949e-06,
"loss": 0.07513746619224548,
"step": 650
},
{
"epoch": 0.7245575221238938,
"grad_norm": 2.484375,
"learning_rate": 5.527213634735094e-06,
"loss": 0.09536111354827881,
"step": 655
},
{
"epoch": 0.7300884955752213,
"grad_norm": 1.625,
"learning_rate": 5.520167549612923e-06,
"loss": 0.08774694204330444,
"step": 660
},
{
"epoch": 0.7356194690265486,
"grad_norm": 1.8359375,
"learning_rate": 5.513073907798227e-06,
"loss": 0.10526577234268189,
"step": 665
},
{
"epoch": 0.7411504424778761,
"grad_norm": 2.453125,
"learning_rate": 5.505932843151209e-06,
"loss": 0.08371620178222657,
"step": 670
},
{
"epoch": 0.7466814159292036,
"grad_norm": 3.28125,
"learning_rate": 5.498744490426963e-06,
"loss": 0.10913375616073609,
"step": 675
},
{
"epoch": 0.7522123893805309,
"grad_norm": 4.0,
"learning_rate": 5.49150898527293e-06,
"loss": 0.08618167638778687,
"step": 680
},
{
"epoch": 0.7577433628318584,
"grad_norm": 11.5625,
"learning_rate": 5.484226464226341e-06,
"loss": 0.09385718703269959,
"step": 685
},
{
"epoch": 0.7632743362831859,
"grad_norm": 7.25,
"learning_rate": 5.476897064711633e-06,
"loss": 0.07986197471618653,
"step": 690
},
{
"epoch": 0.7688053097345132,
"grad_norm": 3.046875,
"learning_rate": 5.469520925037867e-06,
"loss": 0.10280755758285523,
"step": 695
},
{
"epoch": 0.7743362831858407,
"grad_norm": 1.75,
"learning_rate": 5.4620981843961064e-06,
"loss": 0.06520473361015319,
"step": 700
},
{
"epoch": 0.7798672566371682,
"grad_norm": 3.453125,
"learning_rate": 5.454628982856798e-06,
"loss": 0.10451754331588745,
"step": 705
},
{
"epoch": 0.7853982300884956,
"grad_norm": 4.40625,
"learning_rate": 5.447113461367127e-06,
"loss": 0.07575499415397643,
"step": 710
},
{
"epoch": 0.790929203539823,
"grad_norm": 2.828125,
"learning_rate": 5.439551761748356e-06,
"loss": 0.06732524037361146,
"step": 715
},
{
"epoch": 0.7964601769911505,
"grad_norm": 2.171875,
"learning_rate": 5.431944026693152e-06,
"loss": 0.06861334443092346,
"step": 720
},
{
"epoch": 0.8019911504424779,
"grad_norm": 3.359375,
"learning_rate": 5.4242903997628865e-06,
"loss": 0.09167773723602295,
"step": 725
},
{
"epoch": 0.8075221238938053,
"grad_norm": 4.09375,
"learning_rate": 5.416591025384935e-06,
"loss": 0.10097070932388305,
"step": 730
},
{
"epoch": 0.8130530973451328,
"grad_norm": 1.5546875,
"learning_rate": 5.408846048849947e-06,
"loss": 0.09188099503517151,
"step": 735
},
{
"epoch": 0.8185840707964602,
"grad_norm": 2.125,
"learning_rate": 5.401055616309104e-06,
"loss": 0.08375394344329834,
"step": 740
},
{
"epoch": 0.8241150442477876,
"grad_norm": 2.375,
"learning_rate": 5.393219874771366e-06,
"loss": 0.11473640203475952,
"step": 745
},
{
"epoch": 0.8296460176991151,
"grad_norm": 4.53125,
"learning_rate": 5.38533897210069e-06,
"loss": 0.09071962237358093,
"step": 750
},
{
"epoch": 0.8351769911504425,
"grad_norm": 1.875,
"learning_rate": 5.3774130570132465e-06,
"loss": 0.09773556590080261,
"step": 755
},
{
"epoch": 0.8407079646017699,
"grad_norm": 1.53125,
"learning_rate": 5.369442279074611e-06,
"loss": 0.0812099814414978,
"step": 760
},
{
"epoch": 0.8462389380530974,
"grad_norm": 4.65625,
"learning_rate": 5.361426788696937e-06,
"loss": 0.09079021811485291,
"step": 765
},
{
"epoch": 0.8517699115044248,
"grad_norm": 3.078125,
"learning_rate": 5.353366737136128e-06,
"loss": 0.08313687443733216,
"step": 770
},
{
"epoch": 0.8573008849557522,
"grad_norm": 2.03125,
"learning_rate": 5.34526227648897e-06,
"loss": 0.07656715512275696,
"step": 775
},
{
"epoch": 0.8628318584070797,
"grad_norm": 2.140625,
"learning_rate": 5.337113559690277e-06,
"loss": 0.06714207530021668,
"step": 780
},
{
"epoch": 0.8683628318584071,
"grad_norm": 2.390625,
"learning_rate": 5.328920740509987e-06,
"loss": 0.08704954385757446,
"step": 785
},
{
"epoch": 0.8738938053097345,
"grad_norm": 3.4375,
"learning_rate": 5.320683973550276e-06,
"loss": 0.09085990190505981,
"step": 790
},
{
"epoch": 0.879424778761062,
"grad_norm": 2.453125,
"learning_rate": 5.312403414242633e-06,
"loss": 0.08155676126480102,
"step": 795
},
{
"epoch": 0.8849557522123894,
"grad_norm": 3.015625,
"learning_rate": 5.304079218844926e-06,
"loss": 0.09094826579093933,
"step": 800
},
{
"epoch": 0.8904867256637168,
"grad_norm": 2.5625,
"learning_rate": 5.295711544438461e-06,
"loss": 0.078577721118927,
"step": 805
},
{
"epoch": 0.8960176991150443,
"grad_norm": 1.8203125,
"learning_rate": 5.287300548925005e-06,
"loss": 0.08717013597488403,
"step": 810
},
{
"epoch": 0.9015486725663717,
"grad_norm": 2.875,
"learning_rate": 5.27884639102382e-06,
"loss": 0.10738222599029541,
"step": 815
},
{
"epoch": 0.9070796460176991,
"grad_norm": 3.15625,
"learning_rate": 5.270349230268659e-06,
"loss": 0.09782763719558715,
"step": 820
},
{
"epoch": 0.9126106194690266,
"grad_norm": 2.703125,
"learning_rate": 5.261809227004757e-06,
"loss": 0.09719208478927613,
"step": 825
},
{
"epoch": 0.918141592920354,
"grad_norm": 2.4375,
"learning_rate": 5.253226542385808e-06,
"loss": 0.07652133107185363,
"step": 830
},
{
"epoch": 0.9236725663716814,
"grad_norm": 2.734375,
"learning_rate": 5.2446013383709235e-06,
"loss": 0.08185278177261353,
"step": 835
},
{
"epoch": 0.9292035398230089,
"grad_norm": 2.0,
"learning_rate": 5.235933777721573e-06,
"loss": 0.11720459461212158,
"step": 840
},
{
"epoch": 0.9347345132743363,
"grad_norm": 3.640625,
"learning_rate": 5.227224023998513e-06,
"loss": 0.08421225547790527,
"step": 845
},
{
"epoch": 0.9402654867256637,
"grad_norm": 5.4375,
"learning_rate": 5.2184722415587085e-06,
"loss": 0.10485668182373047,
"step": 850
},
{
"epoch": 0.9457964601769911,
"grad_norm": 1.9609375,
"learning_rate": 5.209678595552219e-06,
"loss": 0.07587134838104248,
"step": 855
},
{
"epoch": 0.9513274336283186,
"grad_norm": 1.8671875,
"learning_rate": 5.200843251919091e-06,
"loss": 0.10049251317977906,
"step": 860
},
{
"epoch": 0.956858407079646,
"grad_norm": 1.7578125,
"learning_rate": 5.191966377386222e-06,
"loss": 0.0784164011478424,
"step": 865
},
{
"epoch": 0.9623893805309734,
"grad_norm": 34.0,
"learning_rate": 5.18304813946422e-06,
"loss": 0.08536301255226135,
"step": 870
},
{
"epoch": 0.9679203539823009,
"grad_norm": 1.5625,
"learning_rate": 5.174088706444236e-06,
"loss": 0.08391154408454896,
"step": 875
},
{
"epoch": 0.9734513274336283,
"grad_norm": 3.0625,
"learning_rate": 5.16508824739479e-06,
"loss": 0.09063887000083923,
"step": 880
},
{
"epoch": 0.9789823008849557,
"grad_norm": 2.171875,
"learning_rate": 5.156046932158586e-06,
"loss": 0.09883906245231629,
"step": 885
},
{
"epoch": 0.9845132743362832,
"grad_norm": 1.8125,
"learning_rate": 5.146964931349295e-06,
"loss": 0.07933374643325805,
"step": 890
},
{
"epoch": 0.9900442477876106,
"grad_norm": 69.0,
"learning_rate": 5.137842416348352e-06,
"loss": 0.07782045602798462,
"step": 895
},
{
"epoch": 0.995575221238938,
"grad_norm": 4.9375,
"learning_rate": 5.128679559301706e-06,
"loss": 0.09287483692169189,
"step": 900
},
{
"epoch": 1.0,
"eval_loss": 0.08557523787021637,
"eval_pearson_cosine": 0.24448971450328827,
"eval_pearson_dot_product": 0.2444896697998047,
"eval_pearson_euclidean": 0.24357418715953827,
"eval_pearson_manhattan": 0.23164989054203033,
"eval_runtime": 75.0319,
"eval_samples_per_second": 108.434,
"eval_spearman_cosine": 0.2719746111830238,
"eval_spearman_dot_product": 0.27197414741000026,
"eval_spearman_euclidean": 0.27197469004649544,
"eval_spearman_manhattan": 0.26044463868326206,
"eval_steps_per_second": 18.072,
"step": 904
},
{
"epoch": 1.0011061946902655,
"grad_norm": 10.8125,
"learning_rate": 5.1194765331165816e-06,
"loss": 0.09788246154785156,
"step": 905
},
{
"epoch": 1.0066371681415929,
"grad_norm": 1.9453125,
"learning_rate": 5.110233511458213e-06,
"loss": 0.08525151014328003,
"step": 910
},
{
"epoch": 1.0121681415929205,
"grad_norm": 2.359375,
"learning_rate": 5.1009506687465665e-06,
"loss": 0.07281992435455323,
"step": 915
},
{
"epoch": 1.0176991150442478,
"grad_norm": 3.28125,
"learning_rate": 5.09162818015305e-06,
"loss": 0.08622983694076539,
"step": 920
},
{
"epoch": 1.0232300884955752,
"grad_norm": 2.359375,
"learning_rate": 5.082266221597205e-06,
"loss": 0.08948267102241517,
"step": 925
},
{
"epoch": 1.0287610619469028,
"grad_norm": 14.125,
"learning_rate": 5.0728649697433905e-06,
"loss": 0.09953801035881042,
"step": 930
},
{
"epoch": 1.0342920353982301,
"grad_norm": 1.46875,
"learning_rate": 5.063424601997449e-06,
"loss": 0.08026633858680725,
"step": 935
},
{
"epoch": 1.0398230088495575,
"grad_norm": 1.6953125,
"learning_rate": 5.053945296503355e-06,
"loss": 0.07641162276268006,
"step": 940
},
{
"epoch": 1.045353982300885,
"grad_norm": 3.890625,
"learning_rate": 5.044427232139856e-06,
"loss": 0.08721265196800232,
"step": 945
},
{
"epoch": 1.0508849557522124,
"grad_norm": 1.6796875,
"learning_rate": 5.0348705885171e-06,
"loss": 0.0862955629825592,
"step": 950
},
{
"epoch": 1.0564159292035398,
"grad_norm": 1.34375,
"learning_rate": 5.02527554597324e-06,
"loss": 0.09808052778244018,
"step": 955
},
{
"epoch": 1.0619469026548674,
"grad_norm": 2.578125,
"learning_rate": 5.015642285571032e-06,
"loss": 0.07419036030769348,
"step": 960
},
{
"epoch": 1.0674778761061947,
"grad_norm": 6.65625,
"learning_rate": 5.005970989094424e-06,
"loss": 0.09272552132606507,
"step": 965
},
{
"epoch": 1.073008849557522,
"grad_norm": 1.8359375,
"learning_rate": 4.996261839045121e-06,
"loss": 0.07260675430297851,
"step": 970
},
{
"epoch": 1.0785398230088497,
"grad_norm": 2.28125,
"learning_rate": 4.986515018639139e-06,
"loss": 0.07248570919036865,
"step": 975
},
{
"epoch": 1.084070796460177,
"grad_norm": 1.953125,
"learning_rate": 4.976730711803355e-06,
"loss": 0.08079921007156372,
"step": 980
},
{
"epoch": 1.0896017699115044,
"grad_norm": 1.875,
"learning_rate": 4.9669091031720285e-06,
"loss": 0.07663200497627258,
"step": 985
},
{
"epoch": 1.095132743362832,
"grad_norm": 2.046875,
"learning_rate": 4.9570503780833215e-06,
"loss": 0.07535027861595153,
"step": 990
},
{
"epoch": 1.1006637168141593,
"grad_norm": 9.6875,
"learning_rate": 4.9471547225758e-06,
"loss": 0.08428272008895873,
"step": 995
},
{
"epoch": 1.1061946902654867,
"grad_norm": 5.25,
"learning_rate": 4.937222323384925e-06,
"loss": 0.07841547727584838,
"step": 1000
},
{
"epoch": 1.1117256637168142,
"grad_norm": 6.96875,
"learning_rate": 4.927253367939525e-06,
"loss": 0.09149413704872131,
"step": 1005
},
{
"epoch": 1.1172566371681416,
"grad_norm": 4.0625,
"learning_rate": 4.917248044358263e-06,
"loss": 0.09244868755340577,
"step": 1010
},
{
"epoch": 1.122787610619469,
"grad_norm": 2.3125,
"learning_rate": 4.907206541446086e-06,
"loss": 0.0952062726020813,
"step": 1015
},
{
"epoch": 1.1283185840707965,
"grad_norm": 2.3125,
"learning_rate": 4.897129048690657e-06,
"loss": 0.07673588991165162,
"step": 1020
},
{
"epoch": 1.133849557522124,
"grad_norm": 2.859375,
"learning_rate": 4.887015756258787e-06,
"loss": 0.09415974617004394,
"step": 1025
},
{
"epoch": 1.1393805309734513,
"grad_norm": 10.875,
"learning_rate": 4.876866854992845e-06,
"loss": 0.07608025670051574,
"step": 1030
},
{
"epoch": 1.1449115044247788,
"grad_norm": 2.25,
"learning_rate": 4.866682536407149e-06,
"loss": 0.07764704823493958,
"step": 1035
},
{
"epoch": 1.1504424778761062,
"grad_norm": 2.5,
"learning_rate": 4.856462992684362e-06,
"loss": 0.07415613532066345,
"step": 1040
},
{
"epoch": 1.1559734513274336,
"grad_norm": 3.421875,
"learning_rate": 4.84620841667186e-06,
"loss": 0.09022656083106995,
"step": 1045
},
{
"epoch": 1.1615044247787611,
"grad_norm": 3.203125,
"learning_rate": 4.835919001878093e-06,
"loss": 0.09267069101333618,
"step": 1050
},
{
"epoch": 1.1670353982300885,
"grad_norm": 1.84375,
"learning_rate": 4.825594942468936e-06,
"loss": 0.0662878394126892,
"step": 1055
},
{
"epoch": 1.1725663716814159,
"grad_norm": 1.5859375,
"learning_rate": 4.8152364332640214e-06,
"loss": 0.07941198348999023,
"step": 1060
},
{
"epoch": 1.1780973451327434,
"grad_norm": 3.078125,
"learning_rate": 4.804843669733068e-06,
"loss": 0.07384738922119141,
"step": 1065
},
{
"epoch": 1.1836283185840708,
"grad_norm": 1.8046875,
"learning_rate": 4.7944168479921865e-06,
"loss": 0.08366084098815918,
"step": 1070
},
{
"epoch": 1.1891592920353982,
"grad_norm": 3.171875,
"learning_rate": 4.783956164800181e-06,
"loss": 0.0913965106010437,
"step": 1075
},
{
"epoch": 1.1946902654867257,
"grad_norm": 1.90625,
"learning_rate": 4.773461817554839e-06,
"loss": 0.07782629728317261,
"step": 1080
},
{
"epoch": 1.200221238938053,
"grad_norm": 2.28125,
"learning_rate": 4.762934004289201e-06,
"loss": 0.0773772954940796,
"step": 1085
},
{
"epoch": 1.2057522123893805,
"grad_norm": 1.6484375,
"learning_rate": 4.752372923667827e-06,
"loss": 0.05787185430526733,
"step": 1090
},
{
"epoch": 1.211283185840708,
"grad_norm": 1.515625,
"learning_rate": 4.7417787749830465e-06,
"loss": 0.06832894086837768,
"step": 1095
},
{
"epoch": 1.2168141592920354,
"grad_norm": 2.59375,
"learning_rate": 4.7311517581512016e-06,
"loss": 0.10163642168045044,
"step": 1100
},
{
"epoch": 1.2223451327433628,
"grad_norm": 1.7890625,
"learning_rate": 4.720492073708868e-06,
"loss": 0.09122365713119507,
"step": 1105
},
{
"epoch": 1.2278761061946903,
"grad_norm": 3.40625,
"learning_rate": 4.709799922809073e-06,
"loss": 0.09118576645851136,
"step": 1110
},
{
"epoch": 1.2334070796460177,
"grad_norm": 2.09375,
"learning_rate": 4.699075507217503e-06,
"loss": 0.10236841440200806,
"step": 1115
},
{
"epoch": 1.238938053097345,
"grad_norm": 3.765625,
"learning_rate": 4.688319029308692e-06,
"loss": 0.09580781459808349,
"step": 1120
},
{
"epoch": 1.2444690265486726,
"grad_norm": 3.15625,
"learning_rate": 4.677530692062205e-06,
"loss": 0.10360836982727051,
"step": 1125
},
{
"epoch": 1.25,
"grad_norm": 1.8359375,
"learning_rate": 4.666710699058807e-06,
"loss": 0.07111564874649048,
"step": 1130
},
{
"epoch": 1.2555309734513274,
"grad_norm": 1.3125,
"learning_rate": 4.655859254476622e-06,
"loss": 0.07404866814613342,
"step": 1135
},
{
"epoch": 1.261061946902655,
"grad_norm": 7.25,
"learning_rate": 4.644976563087276e-06,
"loss": 0.07595281600952149,
"step": 1140
},
{
"epoch": 1.2665929203539823,
"grad_norm": 3.96875,
"learning_rate": 4.634062830252041e-06,
"loss": 0.09457840323448181,
"step": 1145
},
{
"epoch": 1.2721238938053097,
"grad_norm": 3.0625,
"learning_rate": 4.623118261917953e-06,
"loss": 0.07928698658943176,
"step": 1150
},
{
"epoch": 1.2776548672566372,
"grad_norm": 2.234375,
"learning_rate": 4.612143064613928e-06,
"loss": 0.09228884577751159,
"step": 1155
},
{
"epoch": 1.2831858407079646,
"grad_norm": 2.765625,
"learning_rate": 4.601137445446861e-06,
"loss": 0.08573116660118103,
"step": 1160
},
{
"epoch": 1.288716814159292,
"grad_norm": 3.21875,
"learning_rate": 4.590101612097727e-06,
"loss": 0.1049655795097351,
"step": 1165
},
{
"epoch": 1.2942477876106195,
"grad_norm": 1.75,
"learning_rate": 4.579035772817653e-06,
"loss": 0.06504960060119629,
"step": 1170
},
{
"epoch": 1.299778761061947,
"grad_norm": 2.125,
"learning_rate": 4.567940136423991e-06,
"loss": 0.07238067984580994,
"step": 1175
},
{
"epoch": 1.3053097345132743,
"grad_norm": 3.09375,
"learning_rate": 4.5568149122963786e-06,
"loss": 0.07839985489845276,
"step": 1180
},
{
"epoch": 1.3108407079646018,
"grad_norm": 2.921875,
"learning_rate": 4.545660310372788e-06,
"loss": 0.1008988857269287,
"step": 1185
},
{
"epoch": 1.3163716814159292,
"grad_norm": 1.6640625,
"learning_rate": 4.534476541145563e-06,
"loss": 0.06239774823188782,
"step": 1190
},
{
"epoch": 1.3219026548672566,
"grad_norm": 1.1875,
"learning_rate": 4.523263815657447e-06,
"loss": 0.12135466337203979,
"step": 1195
},
{
"epoch": 1.3274336283185841,
"grad_norm": 3.8125,
"learning_rate": 4.5120223454976e-06,
"loss": 0.07085340023040772,
"step": 1200
},
{
"epoch": 1.3329646017699115,
"grad_norm": 3.984375,
"learning_rate": 4.50075234279761e-06,
"loss": 0.06139802932739258,
"step": 1205
},
{
"epoch": 1.338495575221239,
"grad_norm": 4.34375,
"learning_rate": 4.489454020227485e-06,
"loss": 0.10039302110671997,
"step": 1210
},
{
"epoch": 1.3440265486725664,
"grad_norm": 2.328125,
"learning_rate": 4.47812759099164e-06,
"loss": 0.08999085426330566,
"step": 1215
},
{
"epoch": 1.3495575221238938,
"grad_norm": 2.484375,
"learning_rate": 4.466773268824881e-06,
"loss": 0.07911369800567628,
"step": 1220
},
{
"epoch": 1.3550884955752212,
"grad_norm": 1.8828125,
"learning_rate": 4.455391267988358e-06,
"loss": 0.07740116119384766,
"step": 1225
},
{
"epoch": 1.3606194690265487,
"grad_norm": 1.9140625,
"learning_rate": 4.443981803265536e-06,
"loss": 0.09790969491004944,
"step": 1230
},
{
"epoch": 1.366150442477876,
"grad_norm": 1.9375,
"learning_rate": 4.432545089958135e-06,
"loss": 0.08635541796684265,
"step": 1235
},
{
"epoch": 1.3716814159292037,
"grad_norm": 4.25,
"learning_rate": 4.421081343882065e-06,
"loss": 0.09469048976898194,
"step": 1240
},
{
"epoch": 1.377212389380531,
"grad_norm": 2.328125,
"learning_rate": 4.40959078136336e-06,
"loss": 0.07265591025352477,
"step": 1245
},
{
"epoch": 1.3827433628318584,
"grad_norm": 2.8125,
"learning_rate": 4.39807361923409e-06,
"loss": 0.0741220772266388,
"step": 1250
},
{
"epoch": 1.3882743362831858,
"grad_norm": 2.96875,
"learning_rate": 4.3865300748282744e-06,
"loss": 0.09189764857292175,
"step": 1255
},
{
"epoch": 1.3938053097345133,
"grad_norm": 2.53125,
"learning_rate": 4.374960365977772e-06,
"loss": 0.09559110403060914,
"step": 1260
},
{
"epoch": 1.3993362831858407,
"grad_norm": 1.4453125,
"learning_rate": 4.363364711008185e-06,
"loss": 0.10397497415542603,
"step": 1265
},
{
"epoch": 1.4048672566371683,
"grad_norm": 1.9375,
"learning_rate": 4.351743328734723e-06,
"loss": 0.07507475614547729,
"step": 1270
},
{
"epoch": 1.4103982300884956,
"grad_norm": 2.59375,
"learning_rate": 4.340096438458086e-06,
"loss": 0.08261100053787232,
"step": 1275
},
{
"epoch": 1.415929203539823,
"grad_norm": 3.0,
"learning_rate": 4.328424259960319e-06,
"loss": 0.0959708571434021,
"step": 1280
},
{
"epoch": 1.4214601769911503,
"grad_norm": 2.328125,
"learning_rate": 4.316727013500667e-06,
"loss": 0.0735737144947052,
"step": 1285
},
{
"epoch": 1.426991150442478,
"grad_norm": 3.265625,
"learning_rate": 4.305004919811421e-06,
"loss": 0.10711610317230225,
"step": 1290
},
{
"epoch": 1.4325221238938053,
"grad_norm": 1.734375,
"learning_rate": 4.293258200093748e-06,
"loss": 0.06000299453735351,
"step": 1295
},
{
"epoch": 1.4380530973451329,
"grad_norm": 1.5625,
"learning_rate": 4.281487076013521e-06,
"loss": 0.08671942353248596,
"step": 1300
},
{
"epoch": 1.4435840707964602,
"grad_norm": 1.6796875,
"learning_rate": 4.269691769697132e-06,
"loss": 0.06764653325080872,
"step": 1305
},
{
"epoch": 1.4491150442477876,
"grad_norm": 4.0,
"learning_rate": 4.257872503727306e-06,
"loss": 0.06291778087615967,
"step": 1310
},
{
"epoch": 1.454646017699115,
"grad_norm": 4.21875,
"learning_rate": 4.246029501138894e-06,
"loss": 0.09451824426651001,
"step": 1315
},
{
"epoch": 1.4601769911504425,
"grad_norm": 2.0,
"learning_rate": 4.234162985414668e-06,
"loss": 0.09219925999641418,
"step": 1320
},
{
"epoch": 1.4657079646017699,
"grad_norm": 14.0,
"learning_rate": 4.222273180481109e-06,
"loss": 0.09066170454025269,
"step": 1325
},
{
"epoch": 1.4712389380530975,
"grad_norm": 5.9375,
"learning_rate": 4.210360310704166e-06,
"loss": 0.08797928094863891,
"step": 1330
},
{
"epoch": 1.4767699115044248,
"grad_norm": 2.46875,
"learning_rate": 4.198424600885041e-06,
"loss": 0.08425722122192383,
"step": 1335
},
{
"epoch": 1.4823008849557522,
"grad_norm": 2.265625,
"learning_rate": 4.186466276255932e-06,
"loss": 0.08120394945144653,
"step": 1340
},
{
"epoch": 1.4878318584070795,
"grad_norm": 1.7109375,
"learning_rate": 4.174485562475792e-06,
"loss": 0.09548501372337341,
"step": 1345
},
{
"epoch": 1.4933628318584071,
"grad_norm": 1.7109375,
"learning_rate": 4.162482685626066e-06,
"loss": 0.11302947998046875,
"step": 1350
},
{
"epoch": 1.4988938053097345,
"grad_norm": 1.6953125,
"learning_rate": 4.150457872206425e-06,
"loss": 0.08306962847709656,
"step": 1355
},
{
"epoch": 1.504424778761062,
"grad_norm": 2.1875,
"learning_rate": 4.138411349130497e-06,
"loss": 0.08865585327148437,
"step": 1360
},
{
"epoch": 1.5099557522123894,
"grad_norm": 2.796875,
"learning_rate": 4.126343343721574e-06,
"loss": 0.08652371764183045,
"step": 1365
},
{
"epoch": 1.5154867256637168,
"grad_norm": 3.546875,
"learning_rate": 4.114254083708336e-06,
"loss": 0.11913001537322998,
"step": 1370
},
{
"epoch": 1.5210176991150441,
"grad_norm": 3.15625,
"learning_rate": 4.102143797220543e-06,
"loss": 0.07272455096244812,
"step": 1375
},
{
"epoch": 1.5265486725663717,
"grad_norm": 3.125,
"learning_rate": 4.090012712784736e-06,
"loss": 0.08868218064308167,
"step": 1380
},
{
"epoch": 1.532079646017699,
"grad_norm": 2.125,
"learning_rate": 4.077861059319922e-06,
"loss": 0.09112980961799622,
"step": 1385
},
{
"epoch": 1.5376106194690267,
"grad_norm": 2.390625,
"learning_rate": 4.065689066133254e-06,
"loss": 0.0885285496711731,
"step": 1390
},
{
"epoch": 1.543141592920354,
"grad_norm": 7.9375,
"learning_rate": 4.0534969629157045e-06,
"loss": 0.07703336477279663,
"step": 1395
},
{
"epoch": 1.5486725663716814,
"grad_norm": 29.0,
"learning_rate": 4.041284979737735e-06,
"loss": 0.07731183171272278,
"step": 1400
},
{
"epoch": 1.5542035398230087,
"grad_norm": 2.265625,
"learning_rate": 4.029053347044944e-06,
"loss": 0.0940183937549591,
"step": 1405
},
{
"epoch": 1.5597345132743363,
"grad_norm": 2.390625,
"learning_rate": 4.016802295653732e-06,
"loss": 0.09459161758422852,
"step": 1410
},
{
"epoch": 1.5652654867256637,
"grad_norm": 2.703125,
"learning_rate": 4.004532056746936e-06,
"loss": 0.095030677318573,
"step": 1415
},
{
"epoch": 1.5707964601769913,
"grad_norm": 2.546875,
"learning_rate": 3.992242861869468e-06,
"loss": 0.09716604948043824,
"step": 1420
},
{
"epoch": 1.5763274336283186,
"grad_norm": 2.328125,
"learning_rate": 3.979934942923953e-06,
"loss": 0.07542461156845093,
"step": 1425
},
{
"epoch": 1.581858407079646,
"grad_norm": 2.78125,
"learning_rate": 3.967608532166342e-06,
"loss": 0.10648454427719116,
"step": 1430
},
{
"epoch": 1.5873893805309733,
"grad_norm": 1.71875,
"learning_rate": 3.955263862201537e-06,
"loss": 0.09427541494369507,
"step": 1435
},
{
"epoch": 1.592920353982301,
"grad_norm": 4.625,
"learning_rate": 3.9429011659790005e-06,
"loss": 0.09536692500114441,
"step": 1440
},
{
"epoch": 1.5984513274336283,
"grad_norm": 2.234375,
"learning_rate": 3.930520676788359e-06,
"loss": 0.08253237009048461,
"step": 1445
},
{
"epoch": 1.6039823008849559,
"grad_norm": 1.671875,
"learning_rate": 3.918122628254996e-06,
"loss": 0.09138402342796326,
"step": 1450
},
{
"epoch": 1.6095132743362832,
"grad_norm": 3.21875,
"learning_rate": 3.905707254335653e-06,
"loss": 0.09183993339538574,
"step": 1455
},
{
"epoch": 1.6150442477876106,
"grad_norm": 1.78125,
"learning_rate": 3.893274789314006e-06,
"loss": 0.07995411157608032,
"step": 1460
},
{
"epoch": 1.620575221238938,
"grad_norm": 1.6015625,
"learning_rate": 3.880825467796249e-06,
"loss": 0.08449186086654663,
"step": 1465
},
{
"epoch": 1.6261061946902655,
"grad_norm": 2.546875,
"learning_rate": 3.8683595247066646e-06,
"loss": 0.08018342852592468,
"step": 1470
},
{
"epoch": 1.6316371681415929,
"grad_norm": 1.4375,
"learning_rate": 3.855877195283194e-06,
"loss": 0.06539214253425599,
"step": 1475
},
{
"epoch": 1.6371681415929205,
"grad_norm": 1.75,
"learning_rate": 3.843378715072993e-06,
"loss": 0.09079237580299378,
"step": 1480
},
{
"epoch": 1.6426991150442478,
"grad_norm": 1.6171875,
"learning_rate": 3.830864319927994e-06,
"loss": 0.10145295858383178,
"step": 1485
},
{
"epoch": 1.6482300884955752,
"grad_norm": 2.21875,
"learning_rate": 3.8183342460004475e-06,
"loss": 0.06505280137062072,
"step": 1490
},
{
"epoch": 1.6537610619469025,
"grad_norm": 1.7578125,
"learning_rate": 3.8057887297384685e-06,
"loss": 0.1106723666191101,
"step": 1495
},
{
"epoch": 1.6592920353982301,
"grad_norm": 1.6328125,
"learning_rate": 3.7932280078815834e-06,
"loss": 0.07789478302001954,
"step": 1500
},
{
"epoch": 1.6648230088495575,
"grad_norm": 2.078125,
"learning_rate": 3.7806523174562465e-06,
"loss": 0.0815811276435852,
"step": 1505
},
{
"epoch": 1.670353982300885,
"grad_norm": 3.109375,
"learning_rate": 3.7680618957713805e-06,
"loss": 0.08219192028045655,
"step": 1510
},
{
"epoch": 1.6758849557522124,
"grad_norm": 1.765625,
"learning_rate": 3.755456980413892e-06,
"loss": 0.10590378046035767,
"step": 1515
},
{
"epoch": 1.6814159292035398,
"grad_norm": 3.9375,
"learning_rate": 3.7428378092441892e-06,
"loss": 0.07235798835754395,
"step": 1520
},
{
"epoch": 1.6869469026548671,
"grad_norm": 1.953125,
"learning_rate": 3.7302046203916957e-06,
"loss": 0.07492388486862182,
"step": 1525
},
{
"epoch": 1.6924778761061947,
"grad_norm": 1.59375,
"learning_rate": 3.7175576522503528e-06,
"loss": 0.06490190625190735,
"step": 1530
},
{
"epoch": 1.698008849557522,
"grad_norm": 1.9921875,
"learning_rate": 3.704897143474126e-06,
"loss": 0.08304203748703003,
"step": 1535
},
{
"epoch": 1.7035398230088497,
"grad_norm": 3.3125,
"learning_rate": 3.6922233329724945e-06,
"loss": 0.07527768611907959,
"step": 1540
},
{
"epoch": 1.709070796460177,
"grad_norm": 3.71875,
"learning_rate": 3.679536459905953e-06,
"loss": 0.06223577260971069,
"step": 1545
},
{
"epoch": 1.7146017699115044,
"grad_norm": 2.3125,
"learning_rate": 3.6668367636814865e-06,
"loss": 0.07909644842147827,
"step": 1550
},
{
"epoch": 1.7201327433628317,
"grad_norm": 2.625,
"learning_rate": 3.654124483948062e-06,
"loss": 0.0856441855430603,
"step": 1555
},
{
"epoch": 1.7256637168141593,
"grad_norm": 2.5,
"learning_rate": 3.641399860592103e-06,
"loss": 0.07336902618408203,
"step": 1560
},
{
"epoch": 1.731194690265487,
"grad_norm": 2.0,
"learning_rate": 3.628663133732962e-06,
"loss": 0.09839301705360412,
"step": 1565
},
{
"epoch": 1.7367256637168142,
"grad_norm": 2.359375,
"learning_rate": 3.6159145437183893e-06,
"loss": 0.1075641393661499,
"step": 1570
},
{
"epoch": 1.7422566371681416,
"grad_norm": 1.875,
"learning_rate": 3.6031543311199993e-06,
"loss": 0.10360686779022217,
"step": 1575
},
{
"epoch": 1.747787610619469,
"grad_norm": 15.75,
"learning_rate": 3.5903827367287313e-06,
"loss": 0.09614421725273133,
"step": 1580
},
{
"epoch": 1.7533185840707963,
"grad_norm": 1.515625,
"learning_rate": 3.5776000015502997e-06,
"loss": 0.08801310062408448,
"step": 1585
},
{
"epoch": 1.758849557522124,
"grad_norm": 3.6875,
"learning_rate": 3.564806366800655e-06,
"loss": 0.11702193021774292,
"step": 1590
},
{
"epoch": 1.7643805309734515,
"grad_norm": 2.625,
"learning_rate": 3.552002073901426e-06,
"loss": 0.07679486870765687,
"step": 1595
},
{
"epoch": 1.7699115044247788,
"grad_norm": 4.8125,
"learning_rate": 3.539187364475364e-06,
"loss": 0.09703764915466309,
"step": 1600
},
{
"epoch": 1.7754424778761062,
"grad_norm": 2.421875,
"learning_rate": 3.5263624803417876e-06,
"loss": 0.08743624687194824,
"step": 1605
},
{
"epoch": 1.7809734513274336,
"grad_norm": 1.9375,
"learning_rate": 3.5135276635120145e-06,
"loss": 0.08286827802658081,
"step": 1610
},
{
"epoch": 1.786504424778761,
"grad_norm": 2.078125,
"learning_rate": 3.500683156184797e-06,
"loss": 0.07697615027427673,
"step": 1615
},
{
"epoch": 1.7920353982300885,
"grad_norm": 1.2890625,
"learning_rate": 3.487829200741753e-06,
"loss": 0.08791252970695496,
"step": 1620
},
{
"epoch": 1.797566371681416,
"grad_norm": 1.9375,
"learning_rate": 3.4749660397427902e-06,
"loss": 0.08169299364089966,
"step": 1625
},
{
"epoch": 1.8030973451327434,
"grad_norm": 1.640625,
"learning_rate": 3.4620939159215256e-06,
"loss": 0.09056628346443177,
"step": 1630
},
{
"epoch": 1.8086283185840708,
"grad_norm": 2.140625,
"learning_rate": 3.4492130721807157e-06,
"loss": 0.0846575915813446,
"step": 1635
},
{
"epoch": 1.8141592920353982,
"grad_norm": 1.96875,
"learning_rate": 3.43632375158766e-06,
"loss": 0.11277735233306885,
"step": 1640
},
{
"epoch": 1.8196902654867255,
"grad_norm": 1.609375,
"learning_rate": 3.4234261973696212e-06,
"loss": 0.08123862147331237,
"step": 1645
},
{
"epoch": 1.825221238938053,
"grad_norm": 1.7265625,
"learning_rate": 3.4105206529092362e-06,
"loss": 0.09810963869094849,
"step": 1650
},
{
"epoch": 1.8307522123893807,
"grad_norm": 1.6484375,
"learning_rate": 3.397607361739919e-06,
"loss": 0.09722994565963745,
"step": 1655
},
{
"epoch": 1.836283185840708,
"grad_norm": 2.484375,
"learning_rate": 3.3846865675412698e-06,
"loss": 0.08509445190429688,
"step": 1660
},
{
"epoch": 1.8418141592920354,
"grad_norm": 2.03125,
"learning_rate": 3.371758514134471e-06,
"loss": 0.0829291045665741,
"step": 1665
},
{
"epoch": 1.8473451327433628,
"grad_norm": 3.953125,
"learning_rate": 3.3588234454776937e-06,
"loss": 0.09088820815086365,
"step": 1670
},
{
"epoch": 1.8528761061946901,
"grad_norm": 1.8046875,
"learning_rate": 3.3458816056614862e-06,
"loss": 0.08787710070610047,
"step": 1675
},
{
"epoch": 1.8584070796460177,
"grad_norm": 1.234375,
"learning_rate": 3.3329332389041746e-06,
"loss": 0.07930462956428527,
"step": 1680
},
{
"epoch": 1.8639380530973453,
"grad_norm": 2.71875,
"learning_rate": 3.319978589547248e-06,
"loss": 0.05953507423400879,
"step": 1685
},
{
"epoch": 1.8694690265486726,
"grad_norm": 1.2578125,
"learning_rate": 3.307017902050754e-06,
"loss": 0.08887494206428528,
"step": 1690
},
{
"epoch": 1.875,
"grad_norm": 7.0,
"learning_rate": 3.2940514209886826e-06,
"loss": 0.10111449956893921,
"step": 1695
},
{
"epoch": 1.8805309734513274,
"grad_norm": 1.171875,
"learning_rate": 3.281079391044347e-06,
"loss": 0.07487451434135436,
"step": 1700
},
{
"epoch": 1.8860619469026547,
"grad_norm": 2.28125,
"learning_rate": 3.2681020570057758e-06,
"loss": 0.09109893441200256,
"step": 1705
},
{
"epoch": 1.8915929203539823,
"grad_norm": 3.234375,
"learning_rate": 3.2551196637610835e-06,
"loss": 0.09872841835021973,
"step": 1710
},
{
"epoch": 1.8971238938053099,
"grad_norm": 1.9453125,
"learning_rate": 3.2421324562938584e-06,
"loss": 0.08543952703475952,
"step": 1715
},
{
"epoch": 1.9026548672566372,
"grad_norm": 2.125,
"learning_rate": 3.2291406796785306e-06,
"loss": 0.07763241529464722,
"step": 1720
},
{
"epoch": 1.9081858407079646,
"grad_norm": 2.234375,
"learning_rate": 3.216144579075757e-06,
"loss": 0.08440614938735962,
"step": 1725
},
{
"epoch": 1.913716814159292,
"grad_norm": 1.6640625,
"learning_rate": 3.203144399727785e-06,
"loss": 0.07370088696479797,
"step": 1730
},
{
"epoch": 1.9192477876106193,
"grad_norm": 2.265625,
"learning_rate": 3.1901403869538357e-06,
"loss": 0.07980594038963318,
"step": 1735
},
{
"epoch": 1.924778761061947,
"grad_norm": 1.3828125,
"learning_rate": 3.1771327861454624e-06,
"loss": 0.08534276485443115,
"step": 1740
},
{
"epoch": 1.9303097345132745,
"grad_norm": 2.390625,
"learning_rate": 3.1641218427619313e-06,
"loss": 0.11289265155792236,
"step": 1745
},
{
"epoch": 1.9358407079646018,
"grad_norm": 5.59375,
"learning_rate": 3.151107802325581e-06,
"loss": 0.07448163628578186,
"step": 1750
},
{
"epoch": 1.9413716814159292,
"grad_norm": 2.265625,
"learning_rate": 3.138090910417195e-06,
"loss": 0.10556936264038086,
"step": 1755
},
{
"epoch": 1.9469026548672566,
"grad_norm": 2.609375,
"learning_rate": 3.1250714126713647e-06,
"loss": 0.11516400575637817,
"step": 1760
},
{
"epoch": 1.952433628318584,
"grad_norm": 1.5234375,
"learning_rate": 3.112049554771853e-06,
"loss": 0.06777188181877136,
"step": 1765
},
{
"epoch": 1.9579646017699115,
"grad_norm": 1.59375,
"learning_rate": 3.0990255824469655e-06,
"loss": 0.07335808277130126,
"step": 1770
},
{
"epoch": 1.963495575221239,
"grad_norm": 2.140625,
"learning_rate": 3.0859997414649e-06,
"loss": 0.0712091326713562,
"step": 1775
},
{
"epoch": 1.9690265486725664,
"grad_norm": 2.71875,
"learning_rate": 3.0729722776291233e-06,
"loss": 0.07971543669700623,
"step": 1780
},
{
"epoch": 1.9745575221238938,
"grad_norm": 1.9140625,
"learning_rate": 3.0599434367737232e-06,
"loss": 0.08747852444648743,
"step": 1785
},
{
"epoch": 1.9800884955752212,
"grad_norm": 2.296875,
"learning_rate": 3.0469134647587735e-06,
"loss": 0.08784466385841369,
"step": 1790
},
{
"epoch": 1.9856194690265485,
"grad_norm": 1.359375,
"learning_rate": 3.0338826074656915e-06,
"loss": 0.09164035916328431,
"step": 1795
},
{
"epoch": 1.991150442477876,
"grad_norm": 1.8671875,
"learning_rate": 3.0208511107926035e-06,
"loss": 0.07686671018600463,
"step": 1800
},
{
"epoch": 1.9966814159292037,
"grad_norm": 2.46875,
"learning_rate": 3.0078192206496985e-06,
"loss": 0.08835670351982117,
"step": 1805
},
{
"epoch": 2.0,
"eval_loss": 0.08414819836616516,
"eval_pearson_cosine": 0.2709871828556061,
"eval_pearson_dot_product": 0.2709871828556061,
"eval_pearson_euclidean": 0.2703634202480316,
"eval_pearson_manhattan": 0.25963324308395386,
"eval_runtime": 74.5667,
"eval_samples_per_second": 109.11,
"eval_spearman_cosine": 0.29511210445833813,
"eval_spearman_dot_product": 0.29511191815615745,
"eval_spearman_euclidean": 0.29511186776411674,
"eval_spearman_manhattan": 0.28502995074975684,
"eval_steps_per_second": 18.185,
"step": 1808
},
{
"epoch": 2.002212389380531,
"grad_norm": 7.4375,
"learning_rate": 2.9947871829545896e-06,
"loss": 0.09820975661277771,
"step": 1810
},
{
"epoch": 2.0077433628318584,
"grad_norm": 85.5,
"learning_rate": 2.9817552436276793e-06,
"loss": 0.07302388548851013,
"step": 1815
},
{
"epoch": 2.0132743362831858,
"grad_norm": 2.0,
"learning_rate": 2.968723648587508e-06,
"loss": 0.10155719518661499,
"step": 1820
},
{
"epoch": 2.018805309734513,
"grad_norm": 2.015625,
"learning_rate": 2.9556926437461225e-06,
"loss": 0.08605079650878907,
"step": 1825
},
{
"epoch": 2.024336283185841,
"grad_norm": 15.875,
"learning_rate": 2.9426624750044306e-06,
"loss": 0.08645280003547669,
"step": 1830
},
{
"epoch": 2.0298672566371683,
"grad_norm": 1.7890625,
"learning_rate": 2.9296333882475647e-06,
"loss": 0.09440279006958008,
"step": 1835
},
{
"epoch": 2.0353982300884956,
"grad_norm": 1.953125,
"learning_rate": 2.916605629340238e-06,
"loss": 0.06409942507743835,
"step": 1840
},
{
"epoch": 2.040929203539823,
"grad_norm": 3.140625,
"learning_rate": 2.903579444122107e-06,
"loss": 0.06556311249732971,
"step": 1845
},
{
"epoch": 2.0464601769911503,
"grad_norm": 1.5546875,
"learning_rate": 2.890555078403132e-06,
"loss": 0.09075504541397095,
"step": 1850
},
{
"epoch": 2.0519911504424777,
"grad_norm": 3.828125,
"learning_rate": 2.877532777958938e-06,
"loss": 0.08203197717666626,
"step": 1855
},
{
"epoch": 2.0575221238938055,
"grad_norm": 1.4140625,
"learning_rate": 2.8645127885261778e-06,
"loss": 0.07762320041656494,
"step": 1860
},
{
"epoch": 2.063053097345133,
"grad_norm": 2.03125,
"learning_rate": 2.8514953557978948e-06,
"loss": 0.07188876271247864,
"step": 1865
},
{
"epoch": 2.0685840707964602,
"grad_norm": 1.8828125,
"learning_rate": 2.838480725418887e-06,
"loss": 0.07778414487838745,
"step": 1870
},
{
"epoch": 2.0741150442477876,
"grad_norm": 3.71875,
"learning_rate": 2.8254691429810674e-06,
"loss": 0.09088488817214965,
"step": 1875
},
{
"epoch": 2.079646017699115,
"grad_norm": 1.7421875,
"learning_rate": 2.8124608540188366e-06,
"loss": 0.08848857879638672,
"step": 1880
},
{
"epoch": 2.0851769911504423,
"grad_norm": 2.796875,
"learning_rate": 2.799456104004444e-06,
"loss": 0.0977685272693634,
"step": 1885
},
{
"epoch": 2.09070796460177,
"grad_norm": 57.75,
"learning_rate": 2.7864551383433575e-06,
"loss": 0.11080644130706788,
"step": 1890
},
{
"epoch": 2.0962389380530975,
"grad_norm": 2.46875,
"learning_rate": 2.773458202369635e-06,
"loss": 0.08939802050590515,
"step": 1895
},
{
"epoch": 2.101769911504425,
"grad_norm": 1.9296875,
"learning_rate": 2.7604655413412885e-06,
"loss": 0.10864754915237426,
"step": 1900
},
{
"epoch": 2.107300884955752,
"grad_norm": 1.6796875,
"learning_rate": 2.7474774004356622e-06,
"loss": 0.08222600817680359,
"step": 1905
},
{
"epoch": 2.1128318584070795,
"grad_norm": 2.71875,
"learning_rate": 2.734494024744803e-06,
"loss": 0.08235861659049988,
"step": 1910
},
{
"epoch": 2.118362831858407,
"grad_norm": 1.796875,
"learning_rate": 2.721515659270837e-06,
"loss": 0.06869395971298217,
"step": 1915
},
{
"epoch": 2.1238938053097347,
"grad_norm": 2.078125,
"learning_rate": 2.7085425489213425e-06,
"loss": 0.07857872843742371,
"step": 1920
},
{
"epoch": 2.129424778761062,
"grad_norm": 1.84375,
"learning_rate": 2.695574938504735e-06,
"loss": 0.08845473527908325,
"step": 1925
},
{
"epoch": 2.1349557522123894,
"grad_norm": 1.65625,
"learning_rate": 2.682613072725642e-06,
"loss": 0.08582295775413513,
"step": 1930
},
{
"epoch": 2.140486725663717,
"grad_norm": 1.34375,
"learning_rate": 2.6696571961802857e-06,
"loss": 0.0818107008934021,
"step": 1935
},
{
"epoch": 2.146017699115044,
"grad_norm": 2.34375,
"learning_rate": 2.656707553351871e-06,
"loss": 0.07559136748313904,
"step": 1940
},
{
"epoch": 2.1515486725663715,
"grad_norm": 1.6484375,
"learning_rate": 2.64376438860597e-06,
"loss": 0.09777607321739197,
"step": 1945
},
{
"epoch": 2.1570796460176993,
"grad_norm": 2.3125,
"learning_rate": 2.6308279461859066e-06,
"loss": 0.09236737489700317,
"step": 1950
},
{
"epoch": 2.1626106194690267,
"grad_norm": 1.6953125,
"learning_rate": 2.6178984702081564e-06,
"loss": 0.07170077562332153,
"step": 1955
},
{
"epoch": 2.168141592920354,
"grad_norm": 1.84375,
"learning_rate": 2.6049762046577326e-06,
"loss": 0.09276016950607299,
"step": 1960
},
{
"epoch": 2.1736725663716814,
"grad_norm": 2.59375,
"learning_rate": 2.592061393383584e-06,
"loss": 0.09366627931594848,
"step": 1965
},
{
"epoch": 2.1792035398230087,
"grad_norm": 1.5625,
"learning_rate": 2.579154280093996e-06,
"loss": 0.06509131789207459,
"step": 1970
},
{
"epoch": 2.184734513274336,
"grad_norm": 2.046875,
"learning_rate": 2.5662551083519863e-06,
"loss": 0.08264510035514831,
"step": 1975
},
{
"epoch": 2.190265486725664,
"grad_norm": 1.75,
"learning_rate": 2.5533641215707136e-06,
"loss": 0.08778979182243347,
"step": 1980
},
{
"epoch": 2.1957964601769913,
"grad_norm": 1.8203125,
"learning_rate": 2.540481563008886e-06,
"loss": 0.11057205200195312,
"step": 1985
},
{
"epoch": 2.2013274336283186,
"grad_norm": 2.40625,
"learning_rate": 2.527607675766161e-06,
"loss": 0.0900986671447754,
"step": 1990
},
{
"epoch": 2.206858407079646,
"grad_norm": 2.59375,
"learning_rate": 2.51474270277857e-06,
"loss": 0.09554226398468017,
"step": 1995
},
{
"epoch": 2.2123893805309733,
"grad_norm": 12.0,
"learning_rate": 2.5018868868139248e-06,
"loss": 0.08063278794288635,
"step": 2000
},
{
"epoch": 2.2179203539823007,
"grad_norm": 3.578125,
"learning_rate": 2.4890404704672423e-06,
"loss": 0.0709685206413269,
"step": 2005
},
{
"epoch": 2.2234513274336285,
"grad_norm": 1.53125,
"learning_rate": 2.4762036961561615e-06,
"loss": 0.0909338116645813,
"step": 2010
},
{
"epoch": 2.228982300884956,
"grad_norm": 1.421875,
"learning_rate": 2.4633768061163763e-06,
"loss": 0.09112616777420043,
"step": 2015
},
{
"epoch": 2.234513274336283,
"grad_norm": 1.9296875,
"learning_rate": 2.4505600423970564e-06,
"loss": 0.07998245358467101,
"step": 2020
},
{
"epoch": 2.2400442477876106,
"grad_norm": 2.859375,
"learning_rate": 2.437753646856285e-06,
"loss": 0.09550694227218628,
"step": 2025
},
{
"epoch": 2.245575221238938,
"grad_norm": 3.4375,
"learning_rate": 2.4249578611564915e-06,
"loss": 0.06952092647552491,
"step": 2030
},
{
"epoch": 2.2511061946902653,
"grad_norm": 1.5703125,
"learning_rate": 2.412172926759895e-06,
"loss": 0.0683376669883728,
"step": 2035
},
{
"epoch": 2.256637168141593,
"grad_norm": 2.328125,
"learning_rate": 2.399399084923943e-06,
"loss": 0.08672719597816467,
"step": 2040
},
{
"epoch": 2.2621681415929205,
"grad_norm": 1.9375,
"learning_rate": 2.386636576696764e-06,
"loss": 0.0991967499256134,
"step": 2045
},
{
"epoch": 2.267699115044248,
"grad_norm": 1.453125,
"learning_rate": 2.373885642912616e-06,
"loss": 0.09153903722763061,
"step": 2050
},
{
"epoch": 2.273230088495575,
"grad_norm": 2.578125,
"learning_rate": 2.361146524187339e-06,
"loss": 0.09814254045486451,
"step": 2055
},
{
"epoch": 2.2787610619469025,
"grad_norm": 2.71875,
"learning_rate": 2.3484194609138216e-06,
"loss": 0.0924601435661316,
"step": 2060
},
{
"epoch": 2.28429203539823,
"grad_norm": 2.453125,
"learning_rate": 2.3357046932574577e-06,
"loss": 0.08476918339729309,
"step": 2065
},
{
"epoch": 2.2898230088495577,
"grad_norm": 3.359375,
"learning_rate": 2.323002461151618e-06,
"loss": 0.10835702419281006,
"step": 2070
},
{
"epoch": 2.295353982300885,
"grad_norm": 2.71875,
"learning_rate": 2.3103130042931237e-06,
"loss": 0.08235512971878052,
"step": 2075
},
{
"epoch": 2.3008849557522124,
"grad_norm": 2.171875,
"learning_rate": 2.2976365621377188e-06,
"loss": 0.10071982145309448,
"step": 2080
},
{
"epoch": 2.3064159292035398,
"grad_norm": 2.421875,
"learning_rate": 2.2849733738955552e-06,
"loss": 0.08264999389648438,
"step": 2085
},
{
"epoch": 2.311946902654867,
"grad_norm": 3.296875,
"learning_rate": 2.272323678526677e-06,
"loss": 0.09692053198814392,
"step": 2090
},
{
"epoch": 2.317477876106195,
"grad_norm": 3.59375,
"learning_rate": 2.259687714736513e-06,
"loss": 0.0863722801208496,
"step": 2095
},
{
"epoch": 2.3230088495575223,
"grad_norm": 1.8671875,
"learning_rate": 2.247065720971368e-06,
"loss": 0.10569133758544921,
"step": 2100
},
{
"epoch": 2.3285398230088497,
"grad_norm": 1.578125,
"learning_rate": 2.2344579354139305e-06,
"loss": 0.08622775673866272,
"step": 2105
},
{
"epoch": 2.334070796460177,
"grad_norm": 2.71875,
"learning_rate": 2.2218645959787706e-06,
"loss": 0.11562827825546265,
"step": 2110
},
{
"epoch": 2.3396017699115044,
"grad_norm": 1.4765625,
"learning_rate": 2.2092859403078545e-06,
"loss": 0.08747315406799316,
"step": 2115
},
{
"epoch": 2.3451327433628317,
"grad_norm": 2.328125,
"learning_rate": 2.1967222057660583e-06,
"loss": 0.08471767902374268,
"step": 2120
},
{
"epoch": 2.350663716814159,
"grad_norm": 1.90625,
"learning_rate": 2.1841736294366922e-06,
"loss": 0.08561186790466309,
"step": 2125
},
{
"epoch": 2.356194690265487,
"grad_norm": 2.359375,
"learning_rate": 2.1716404481170196e-06,
"loss": 0.08371530175209045,
"step": 2130
},
{
"epoch": 2.3617256637168142,
"grad_norm": 2.15625,
"learning_rate": 2.159122898313798e-06,
"loss": 0.07868375182151795,
"step": 2135
},
{
"epoch": 2.3672566371681416,
"grad_norm": 4.5,
"learning_rate": 2.1466212162388097e-06,
"loss": 0.08004286885261536,
"step": 2140
},
{
"epoch": 2.372787610619469,
"grad_norm": 2.234375,
"learning_rate": 2.1341356378044035e-06,
"loss": 0.07826697826385498,
"step": 2145
},
{
"epoch": 2.3783185840707963,
"grad_norm": 2.46875,
"learning_rate": 2.1216663986190483e-06,
"loss": 0.08701767921447753,
"step": 2150
},
{
"epoch": 2.383849557522124,
"grad_norm": 4.0,
"learning_rate": 2.1092137339828824e-06,
"loss": 0.08059791326522828,
"step": 2155
},
{
"epoch": 2.3893805309734515,
"grad_norm": 1.5546875,
"learning_rate": 2.0967778788832745e-06,
"loss": 0.06480356454849243,
"step": 2160
},
{
"epoch": 2.394911504424779,
"grad_norm": 1.640625,
"learning_rate": 2.0843590679903926e-06,
"loss": 0.07256920337677002,
"step": 2165
},
{
"epoch": 2.400442477876106,
"grad_norm": 1.46875,
"learning_rate": 2.071957535652769e-06,
"loss": 0.10539222955703735,
"step": 2170
},
{
"epoch": 2.4059734513274336,
"grad_norm": 1.296875,
"learning_rate": 2.059573515892885e-06,
"loss": 0.07799110412597657,
"step": 2175
},
{
"epoch": 2.411504424778761,
"grad_norm": 1.5546875,
"learning_rate": 2.04720724240275e-06,
"loss": 0.08215337395668029,
"step": 2180
},
{
"epoch": 2.4170353982300883,
"grad_norm": 2.09375,
"learning_rate": 2.034858948539493e-06,
"loss": 0.08607609272003174,
"step": 2185
},
{
"epoch": 2.422566371681416,
"grad_norm": 2.75,
"learning_rate": 2.0225288673209614e-06,
"loss": 0.07781745195388794,
"step": 2190
},
{
"epoch": 2.4280973451327434,
"grad_norm": 3.46875,
"learning_rate": 2.010217231421321e-06,
"loss": 0.08161718845367431,
"step": 2195
},
{
"epoch": 2.433628318584071,
"grad_norm": 1.78125,
"learning_rate": 1.9979242731666646e-06,
"loss": 0.09122164845466614,
"step": 2200
},
{
"epoch": 2.439159292035398,
"grad_norm": 3.890625,
"learning_rate": 1.9856502245306316e-06,
"loss": 0.0937601387500763,
"step": 2205
},
{
"epoch": 2.4446902654867255,
"grad_norm": 1.921875,
"learning_rate": 1.973395317130026e-06,
"loss": 0.10289661884307862,
"step": 2210
},
{
"epoch": 2.4502212389380533,
"grad_norm": 3.5,
"learning_rate": 1.9611597822204517e-06,
"loss": 0.06480934619903564,
"step": 2215
},
{
"epoch": 2.4557522123893807,
"grad_norm": 1.4765625,
"learning_rate": 1.948943850691941e-06,
"loss": 0.0627164900302887,
"step": 2220
},
{
"epoch": 2.461283185840708,
"grad_norm": 5.6875,
"learning_rate": 1.9367477530646036e-06,
"loss": 0.0907912790775299,
"step": 2225
},
{
"epoch": 2.4668141592920354,
"grad_norm": 1.875,
"learning_rate": 1.924571719484276e-06,
"loss": 0.08383381366729736,
"step": 2230
},
{
"epoch": 2.4723451327433628,
"grad_norm": 1.9609375,
"learning_rate": 1.912415979718174e-06,
"loss": 0.06725156307220459,
"step": 2235
},
{
"epoch": 2.47787610619469,
"grad_norm": 2.3125,
"learning_rate": 1.9002807631505625e-06,
"loss": 0.0827639639377594,
"step": 2240
},
{
"epoch": 2.4834070796460175,
"grad_norm": 3.984375,
"learning_rate": 1.8881662987784226e-06,
"loss": 0.09498741030693054,
"step": 2245
},
{
"epoch": 2.4889380530973453,
"grad_norm": 1.4296875,
"learning_rate": 1.8760728152071325e-06,
"loss": 0.07206577062606812,
"step": 2250
},
{
"epoch": 2.4944690265486726,
"grad_norm": 1.640625,
"learning_rate": 1.8640005406461569e-06,
"loss": 0.09282957911491393,
"step": 2255
},
{
"epoch": 2.5,
"grad_norm": 1.609375,
"learning_rate": 1.8519497029047308e-06,
"loss": 0.06762008666992188,
"step": 2260
},
{
"epoch": 2.5055309734513274,
"grad_norm": 3.71875,
"learning_rate": 1.8399205293875716e-06,
"loss": 0.08229448795318603,
"step": 2265
},
{
"epoch": 2.5110619469026547,
"grad_norm": 1.4140625,
"learning_rate": 1.827913247090581e-06,
"loss": 0.08683115839958191,
"step": 2270
},
{
"epoch": 2.5165929203539825,
"grad_norm": 1.4140625,
"learning_rate": 1.8159280825965628e-06,
"loss": 0.07722365856170654,
"step": 2275
},
{
"epoch": 2.52212389380531,
"grad_norm": 2.140625,
"learning_rate": 1.8039652620709506e-06,
"loss": 0.09522076845169067,
"step": 2280
},
{
"epoch": 2.5276548672566372,
"grad_norm": 1.734375,
"learning_rate": 1.7920250112575357e-06,
"loss": 0.08804202079772949,
"step": 2285
},
{
"epoch": 2.5331858407079646,
"grad_norm": 1.703125,
"learning_rate": 1.7801075554742075e-06,
"loss": 0.08685529828071595,
"step": 2290
},
{
"epoch": 2.538716814159292,
"grad_norm": 1.6875,
"learning_rate": 1.768213119608706e-06,
"loss": 0.0919269859790802,
"step": 2295
},
{
"epoch": 2.5442477876106193,
"grad_norm": 1.8046875,
"learning_rate": 1.7563419281143715e-06,
"loss": 0.07678677439689637,
"step": 2300
},
{
"epoch": 2.5497787610619467,
"grad_norm": 1.3671875,
"learning_rate": 1.744494205005916e-06,
"loss": 0.085833078622818,
"step": 2305
},
{
"epoch": 2.5553097345132745,
"grad_norm": 1.625,
"learning_rate": 1.7326701738551897e-06,
"loss": 0.07251712083816528,
"step": 2310
},
{
"epoch": 2.560840707964602,
"grad_norm": 2.46875,
"learning_rate": 1.7208700577869667e-06,
"loss": 0.10770748853683472,
"step": 2315
},
{
"epoch": 2.566371681415929,
"grad_norm": 1.7265625,
"learning_rate": 1.709094079474733e-06,
"loss": 0.08924276828765869,
"step": 2320
},
{
"epoch": 2.5719026548672566,
"grad_norm": 2.28125,
"learning_rate": 1.697342461136483e-06,
"loss": 0.09545642137527466,
"step": 2325
},
{
"epoch": 2.577433628318584,
"grad_norm": 4.71875,
"learning_rate": 1.6856154245305277e-06,
"loss": 0.11620496511459351,
"step": 2330
},
{
"epoch": 2.5829646017699117,
"grad_norm": 2.234375,
"learning_rate": 1.6739131909513084e-06,
"loss": 0.09573878049850464,
"step": 2335
},
{
"epoch": 2.588495575221239,
"grad_norm": 2.828125,
"learning_rate": 1.662235981225227e-06,
"loss": 0.0887173593044281,
"step": 2340
},
{
"epoch": 2.5940265486725664,
"grad_norm": 2.4375,
"learning_rate": 1.6505840157064706e-06,
"loss": 0.07949482798576354,
"step": 2345
},
{
"epoch": 2.599557522123894,
"grad_norm": 1.609375,
"learning_rate": 1.6389575142728553e-06,
"loss": 0.09370492696762085,
"step": 2350
},
{
"epoch": 2.605088495575221,
"grad_norm": 2.765625,
"learning_rate": 1.6273566963216808e-06,
"loss": 0.06538494825363159,
"step": 2355
},
{
"epoch": 2.6106194690265485,
"grad_norm": 2.640625,
"learning_rate": 1.6157817807655874e-06,
"loss": 0.06570954322814941,
"step": 2360
},
{
"epoch": 2.616150442477876,
"grad_norm": 1.8203125,
"learning_rate": 1.604232986028426e-06,
"loss": 0.07775924801826477,
"step": 2365
},
{
"epoch": 2.6216814159292037,
"grad_norm": 2.078125,
"learning_rate": 1.5927105300411356e-06,
"loss": 0.07373055219650268,
"step": 2370
},
{
"epoch": 2.627212389380531,
"grad_norm": 4.0625,
"learning_rate": 1.5812146302376306e-06,
"loss": 0.08471120595932007,
"step": 2375
},
{
"epoch": 2.6327433628318584,
"grad_norm": 1.7265625,
"learning_rate": 1.5697455035507001e-06,
"loss": 0.06944184303283692,
"step": 2380
},
{
"epoch": 2.6382743362831858,
"grad_norm": 1.625,
"learning_rate": 1.5583033664079094e-06,
"loss": 0.07597512006759644,
"step": 2385
},
{
"epoch": 2.643805309734513,
"grad_norm": 2.5,
"learning_rate": 1.5468884347275212e-06,
"loss": 0.09081778526306153,
"step": 2390
},
{
"epoch": 2.649336283185841,
"grad_norm": 1.828125,
"learning_rate": 1.535500923914419e-06,
"loss": 0.081801438331604,
"step": 2395
},
{
"epoch": 2.6548672566371683,
"grad_norm": 4.0625,
"learning_rate": 1.5241410488560418e-06,
"loss": 0.08338025808334351,
"step": 2400
},
{
"epoch": 2.6603982300884956,
"grad_norm": 1.3828125,
"learning_rate": 1.5128090239183285e-06,
"loss": 0.07939456701278687,
"step": 2405
},
{
"epoch": 2.665929203539823,
"grad_norm": 1.7890625,
"learning_rate": 1.5015050629416758e-06,
"loss": 0.07578387260437011,
"step": 2410
},
{
"epoch": 2.6714601769911503,
"grad_norm": 7.25,
"learning_rate": 1.4902293792369004e-06,
"loss": 0.10105445384979247,
"step": 2415
},
{
"epoch": 2.676991150442478,
"grad_norm": 2.890625,
"learning_rate": 1.4789821855812111e-06,
"loss": 0.0861863374710083,
"step": 2420
},
{
"epoch": 2.682522123893805,
"grad_norm": 2.1875,
"learning_rate": 1.4677636942141991e-06,
"loss": 0.09112591147422791,
"step": 2425
},
{
"epoch": 2.688053097345133,
"grad_norm": 1.859375,
"learning_rate": 1.456574116833833e-06,
"loss": 0.08214497566223145,
"step": 2430
},
{
"epoch": 2.6935840707964602,
"grad_norm": 1.609375,
"learning_rate": 1.4454136645924589e-06,
"loss": 0.07330217361450195,
"step": 2435
},
{
"epoch": 2.6991150442477876,
"grad_norm": 14.0625,
"learning_rate": 1.4342825480928162e-06,
"loss": 0.10092228651046753,
"step": 2440
},
{
"epoch": 2.704646017699115,
"grad_norm": 1.3515625,
"learning_rate": 1.4231809773840694e-06,
"loss": 0.076936274766922,
"step": 2445
},
{
"epoch": 2.7101769911504423,
"grad_norm": 5.1875,
"learning_rate": 1.4121091619578396e-06,
"loss": 0.0856077253818512,
"step": 2450
},
{
"epoch": 2.71570796460177,
"grad_norm": 1.5703125,
"learning_rate": 1.4010673107442512e-06,
"loss": 0.08901992440223694,
"step": 2455
},
{
"epoch": 2.7212389380530975,
"grad_norm": 1.8203125,
"learning_rate": 1.390055632107992e-06,
"loss": 0.09287630319595337,
"step": 2460
},
{
"epoch": 2.726769911504425,
"grad_norm": 4.875,
"learning_rate": 1.3790743338443794e-06,
"loss": 0.08547717332839966,
"step": 2465
},
{
"epoch": 2.732300884955752,
"grad_norm": 1.2890625,
"learning_rate": 1.3681236231754397e-06,
"loss": 0.08154829144477845,
"step": 2470
},
{
"epoch": 2.7378318584070795,
"grad_norm": 1.5,
"learning_rate": 1.3572037067459953e-06,
"loss": 0.08663605451583863,
"step": 2475
},
{
"epoch": 2.7433628318584073,
"grad_norm": 2.03125,
"learning_rate": 1.3463147906197702e-06,
"loss": 0.09524447321891785,
"step": 2480
},
{
"epoch": 2.7488938053097343,
"grad_norm": 1.8203125,
"learning_rate": 1.335457080275498e-06,
"loss": 0.08607526421546936,
"step": 2485
},
{
"epoch": 2.754424778761062,
"grad_norm": 1.8515625,
"learning_rate": 1.3246307806030447e-06,
"loss": 0.09461776614189148,
"step": 2490
},
{
"epoch": 2.7599557522123894,
"grad_norm": 2.015625,
"learning_rate": 1.3138360958995435e-06,
"loss": 0.09269558787345886,
"step": 2495
},
{
"epoch": 2.765486725663717,
"grad_norm": 2.953125,
"learning_rate": 1.3030732298655381e-06,
"loss": 0.07900597453117371,
"step": 2500
},
{
"epoch": 2.771017699115044,
"grad_norm": 2.78125,
"learning_rate": 1.2923423856011412e-06,
"loss": 0.07077383995056152,
"step": 2505
},
{
"epoch": 2.7765486725663715,
"grad_norm": 2.171875,
"learning_rate": 1.2816437656021993e-06,
"loss": 0.0925867199897766,
"step": 2510
},
{
"epoch": 2.7820796460176993,
"grad_norm": 2.34375,
"learning_rate": 1.2709775717564695e-06,
"loss": 0.08598564267158508,
"step": 2515
},
{
"epoch": 2.7876106194690267,
"grad_norm": 2.1875,
"learning_rate": 1.2603440053398196e-06,
"loss": 0.1018025279045105,
"step": 2520
},
{
"epoch": 2.793141592920354,
"grad_norm": 2.0625,
"learning_rate": 1.2497432670124192e-06,
"loss": 0.08638688325881957,
"step": 2525
},
{
"epoch": 2.7986725663716814,
"grad_norm": 3.390625,
"learning_rate": 1.2391755568149547e-06,
"loss": 0.08196425437927246,
"step": 2530
},
{
"epoch": 2.8042035398230087,
"grad_norm": 2.5,
"learning_rate": 1.2286410741648615e-06,
"loss": 0.08200865387916564,
"step": 2535
},
{
"epoch": 2.8097345132743365,
"grad_norm": 1.8671875,
"learning_rate": 1.2181400178525537e-06,
"loss": 0.09552202224731446,
"step": 2540
},
{
"epoch": 2.8152654867256635,
"grad_norm": 3.921875,
"learning_rate": 1.207672586037676e-06,
"loss": 0.07617225050926209,
"step": 2545
},
{
"epoch": 2.8207964601769913,
"grad_norm": 2.25,
"learning_rate": 1.1972389762453644e-06,
"loss": 0.09064263105392456,
"step": 2550
},
{
"epoch": 2.8263274336283186,
"grad_norm": 1.4453125,
"learning_rate": 1.1868393853625166e-06,
"loss": 0.0835217833518982,
"step": 2555
},
{
"epoch": 2.831858407079646,
"grad_norm": 3.453125,
"learning_rate": 1.1764740096340805e-06,
"loss": 0.09065532088279724,
"step": 2560
},
{
"epoch": 2.8373893805309733,
"grad_norm": 1.5,
"learning_rate": 1.1661430446593455e-06,
"loss": 0.07405591607093812,
"step": 2565
},
{
"epoch": 2.8429203539823007,
"grad_norm": 1.53125,
"learning_rate": 1.1558466853882573e-06,
"loss": 0.06981947422027587,
"step": 2570
},
{
"epoch": 2.8484513274336285,
"grad_norm": 2.125,
"learning_rate": 1.1455851261177356e-06,
"loss": 0.07571223974227906,
"step": 2575
},
{
"epoch": 2.853982300884956,
"grad_norm": 3.71875,
"learning_rate": 1.1353585604880087e-06,
"loss": 0.08819946646690369,
"step": 2580
},
{
"epoch": 2.859513274336283,
"grad_norm": 1.421875,
"learning_rate": 1.1251671814789585e-06,
"loss": 0.08156806826591492,
"step": 2585
},
{
"epoch": 2.8650442477876106,
"grad_norm": 1.84375,
"learning_rate": 1.1150111814064802e-06,
"loss": 0.07055151462554932,
"step": 2590
},
{
"epoch": 2.870575221238938,
"grad_norm": 2.703125,
"learning_rate": 1.1048907519188527e-06,
"loss": 0.0768340766429901,
"step": 2595
},
{
"epoch": 2.8761061946902657,
"grad_norm": 19.0,
"learning_rate": 1.0948060839931227e-06,
"loss": 0.07272812128067016,
"step": 2600
},
{
"epoch": 2.8816371681415927,
"grad_norm": 2.03125,
"learning_rate": 1.0847573679314957e-06,
"loss": 0.12251821756362916,
"step": 2605
},
{
"epoch": 2.8871681415929205,
"grad_norm": 12.5625,
"learning_rate": 1.0747447933577562e-06,
"loss": 0.08684195876121521,
"step": 2610
},
{
"epoch": 2.892699115044248,
"grad_norm": 2.390625,
"learning_rate": 1.0647685492136796e-06,
"loss": 0.08701669573783874,
"step": 2615
},
{
"epoch": 2.898230088495575,
"grad_norm": 3.765625,
"learning_rate": 1.054828823755467e-06,
"loss": 0.08582164645195008,
"step": 2620
},
{
"epoch": 2.9037610619469025,
"grad_norm": 1.3203125,
"learning_rate": 1.0449258045501997e-06,
"loss": 0.0718487560749054,
"step": 2625
},
{
"epoch": 2.90929203539823,
"grad_norm": 2.234375,
"learning_rate": 1.0350596784722932e-06,
"loss": 0.09879735112190247,
"step": 2630
},
{
"epoch": 2.9148230088495577,
"grad_norm": 2.15625,
"learning_rate": 1.0252306316999743e-06,
"loss": 0.07352356910705567,
"step": 2635
},
{
"epoch": 2.920353982300885,
"grad_norm": 2.234375,
"learning_rate": 1.0154388497117654e-06,
"loss": 0.08437286615371704,
"step": 2640
},
{
"epoch": 2.9258849557522124,
"grad_norm": 1.6953125,
"learning_rate": 1.005684517282987e-06,
"loss": 0.09863988757133484,
"step": 2645
},
{
"epoch": 2.9314159292035398,
"grad_norm": 1.7734375,
"learning_rate": 9.959678184822699e-07,
"loss": 0.08480924367904663,
"step": 2650
},
{
"epoch": 2.936946902654867,
"grad_norm": 2.90625,
"learning_rate": 9.862889366680787e-07,
"loss": 0.07831119298934937,
"step": 2655
},
{
"epoch": 2.942477876106195,
"grad_norm": 1.8203125,
"learning_rate": 9.766480544852575e-07,
"loss": 0.08476048111915588,
"step": 2660
},
{
"epoch": 2.948008849557522,
"grad_norm": 1.46875,
"learning_rate": 9.67045353861578e-07,
"loss": 0.0749458134174347,
"step": 2665
},
{
"epoch": 2.9535398230088497,
"grad_norm": 2.921875,
"learning_rate": 9.57481016004313e-07,
"loss": 0.09438548088073731,
"step": 2670
},
{
"epoch": 2.959070796460177,
"grad_norm": 2.265625,
"learning_rate": 9.479552213968065e-07,
"loss": 0.07784347534179688,
"step": 2675
},
{
"epoch": 2.9646017699115044,
"grad_norm": 3.3125,
"learning_rate": 9.384681497950785e-07,
"loss": 0.0916098952293396,
"step": 2680
},
{
"epoch": 2.9701327433628317,
"grad_norm": 1.515625,
"learning_rate": 9.290199802244273e-07,
"loss": 0.06830742359161376,
"step": 2685
},
{
"epoch": 2.975663716814159,
"grad_norm": 1.1953125,
"learning_rate": 9.196108909760526e-07,
"loss": 0.06312470436096192,
"step": 2690
},
{
"epoch": 2.981194690265487,
"grad_norm": 2.5,
"learning_rate": 9.102410596036884e-07,
"loss": 0.08321941494941712,
"step": 2695
},
{
"epoch": 2.9867256637168142,
"grad_norm": 1.3984375,
"learning_rate": 9.009106629202602e-07,
"loss": 0.06696463227272034,
"step": 2700
},
{
"epoch": 2.9922566371681416,
"grad_norm": 8.5,
"learning_rate": 8.916198769945394e-07,
"loss": 0.09367920160293579,
"step": 2705
},
{
"epoch": 2.997787610619469,
"grad_norm": 1.5859375,
"learning_rate": 8.823688771478243e-07,
"loss": 0.07851234078407288,
"step": 2710
},
{
"epoch": 3.0,
"eval_loss": 0.083900585770607,
"eval_pearson_cosine": 0.2747773826122284,
"eval_pearson_dot_product": 0.2747773826122284,
"eval_pearson_euclidean": 0.27428966760635376,
"eval_pearson_manhattan": 0.26390936970710754,
"eval_runtime": 74.238,
"eval_samples_per_second": 109.593,
"eval_spearman_cosine": 0.298245669322679,
"eval_spearman_dot_product": 0.298245075064119,
"eval_spearman_euclidean": 0.2982453028501419,
"eval_spearman_manhattan": 0.28816982726622004,
"eval_steps_per_second": 18.266,
"step": 2712
},
{
"epoch": 3.0033185840707963,
"grad_norm": 1.6015625,
"learning_rate": 8.73157837950634e-07,
"loss": 0.08410517573356628,
"step": 2715
},
{
"epoch": 3.0088495575221237,
"grad_norm": 2.3125,
"learning_rate": 8.639869332194109e-07,
"loss": 0.09520081281661988,
"step": 2720
},
{
"epoch": 3.0143805309734515,
"grad_norm": 1.7890625,
"learning_rate": 8.548563360132432e-07,
"loss": 0.07079311609268188,
"step": 2725
},
{
"epoch": 3.019911504424779,
"grad_norm": 5.625,
"learning_rate": 8.457662186305966e-07,
"loss": 0.08944098949432373,
"step": 2730
},
{
"epoch": 3.025442477876106,
"grad_norm": 9.3125,
"learning_rate": 8.367167526060655e-07,
"loss": 0.08370385169982911,
"step": 2735
},
{
"epoch": 3.0309734513274336,
"grad_norm": 2.90625,
"learning_rate": 8.277081087071355e-07,
"loss": 0.08712309002876281,
"step": 2740
},
{
"epoch": 3.036504424778761,
"grad_norm": 2.4375,
"learning_rate": 8.187404569309575e-07,
"loss": 0.08104004263877869,
"step": 2745
},
{
"epoch": 3.0420353982300883,
"grad_norm": 2.375,
"learning_rate": 8.09813966501145e-07,
"loss": 0.07224444150924683,
"step": 2750
},
{
"epoch": 3.047566371681416,
"grad_norm": 1.90625,
"learning_rate": 8.009288058645778e-07,
"loss": 0.10097153186798095,
"step": 2755
},
{
"epoch": 3.0530973451327434,
"grad_norm": 2.78125,
"learning_rate": 7.920851426882262e-07,
"loss": 0.08011520504951478,
"step": 2760
},
{
"epoch": 3.058628318584071,
"grad_norm": 2.0625,
"learning_rate": 7.832831438559804e-07,
"loss": 0.08335314989089966,
"step": 2765
},
{
"epoch": 3.064159292035398,
"grad_norm": 1.375,
"learning_rate": 7.745229754655091e-07,
"loss": 0.06681510806083679,
"step": 2770
},
{
"epoch": 3.0696902654867255,
"grad_norm": 2.328125,
"learning_rate": 7.658048028251199e-07,
"loss": 0.09199514985084534,
"step": 2775
},
{
"epoch": 3.0752212389380533,
"grad_norm": 3.84375,
"learning_rate": 7.571287904506445e-07,
"loss": 0.09151683449745178,
"step": 2780
},
{
"epoch": 3.0807522123893807,
"grad_norm": 1.46875,
"learning_rate": 7.484951020623264e-07,
"loss": 0.08503244519233703,
"step": 2785
},
{
"epoch": 3.086283185840708,
"grad_norm": 2.71875,
"learning_rate": 7.399039005817421e-07,
"loss": 0.08643928170204163,
"step": 2790
},
{
"epoch": 3.0918141592920354,
"grad_norm": 2.375,
"learning_rate": 7.31355348128719e-07,
"loss": 0.08839238882064819,
"step": 2795
},
{
"epoch": 3.0973451327433628,
"grad_norm": 2.6875,
"learning_rate": 7.22849606018276e-07,
"loss": 0.09040845036506653,
"step": 2800
},
{
"epoch": 3.10287610619469,
"grad_norm": 1.578125,
"learning_rate": 7.143868347575843e-07,
"loss": 0.09118080139160156,
"step": 2805
},
{
"epoch": 3.1084070796460175,
"grad_norm": 1.953125,
"learning_rate": 7.059671940429353e-07,
"loss": 0.08608399033546447,
"step": 2810
},
{
"epoch": 3.1139380530973453,
"grad_norm": 2.34375,
"learning_rate": 6.975908427567268e-07,
"loss": 0.06726066470146179,
"step": 2815
},
{
"epoch": 3.1194690265486726,
"grad_norm": 2.203125,
"learning_rate": 6.892579389644669e-07,
"loss": 0.059871864318847653,
"step": 2820
},
{
"epoch": 3.125,
"grad_norm": 2.03125,
"learning_rate": 6.80968639911789e-07,
"loss": 0.07590010166168212,
"step": 2825
},
{
"epoch": 3.1305309734513274,
"grad_norm": 3.75,
"learning_rate": 6.72723102021486e-07,
"loss": 0.08746974468231201,
"step": 2830
},
{
"epoch": 3.1360619469026547,
"grad_norm": 2.328125,
"learning_rate": 6.645214808905587e-07,
"loss": 0.07423999309539794,
"step": 2835
},
{
"epoch": 3.1415929203539825,
"grad_norm": 2.90625,
"learning_rate": 6.563639312872755e-07,
"loss": 0.07480744123458863,
"step": 2840
},
{
"epoch": 3.14712389380531,
"grad_norm": 1.1796875,
"learning_rate": 6.482506071482606e-07,
"loss": 0.07017472386360168,
"step": 2845
},
{
"epoch": 3.1526548672566372,
"grad_norm": 2.53125,
"learning_rate": 6.401816615755808e-07,
"loss": 0.0768163561820984,
"step": 2850
},
{
"epoch": 3.1581858407079646,
"grad_norm": 2.640625,
"learning_rate": 6.321572468338587e-07,
"loss": 0.08797813653945923,
"step": 2855
},
{
"epoch": 3.163716814159292,
"grad_norm": 1.375,
"learning_rate": 6.241775143474016e-07,
"loss": 0.07705342173576354,
"step": 2860
},
{
"epoch": 3.1692477876106193,
"grad_norm": 2.5625,
"learning_rate": 6.162426146973423e-07,
"loss": 0.09341556429862977,
"step": 2865
},
{
"epoch": 3.1747787610619467,
"grad_norm": 2.0,
"learning_rate": 6.08352697618798e-07,
"loss": 0.07034317255020142,
"step": 2870
},
{
"epoch": 3.1803097345132745,
"grad_norm": 2.328125,
"learning_rate": 6.005079119980445e-07,
"loss": 0.09640387892723083,
"step": 2875
},
{
"epoch": 3.185840707964602,
"grad_norm": 2.328125,
"learning_rate": 5.92708405869706e-07,
"loss": 0.09538073539733886,
"step": 2880
},
{
"epoch": 3.191371681415929,
"grad_norm": 1.2890625,
"learning_rate": 5.849543264139646e-07,
"loss": 0.0773756444454193,
"step": 2885
},
{
"epoch": 3.1969026548672566,
"grad_norm": 1.6328125,
"learning_rate": 5.772458199537769e-07,
"loss": 0.1019517183303833,
"step": 2890
},
{
"epoch": 3.202433628318584,
"grad_norm": 2.703125,
"learning_rate": 5.695830319521198e-07,
"loss": 0.08872216343879699,
"step": 2895
},
{
"epoch": 3.2079646017699117,
"grad_norm": 1.4921875,
"learning_rate": 5.619661070092411e-07,
"loss": 0.08448187112808228,
"step": 2900
},
{
"epoch": 3.213495575221239,
"grad_norm": 2.4375,
"learning_rate": 5.543951888599329e-07,
"loss": 0.07406718730926513,
"step": 2905
},
{
"epoch": 3.2190265486725664,
"grad_norm": 2.078125,
"learning_rate": 5.468704203708176e-07,
"loss": 0.09174955487251282,
"step": 2910
},
{
"epoch": 3.224557522123894,
"grad_norm": 1.609375,
"learning_rate": 5.393919435376536e-07,
"loss": 0.0792352795600891,
"step": 2915
},
{
"epoch": 3.230088495575221,
"grad_norm": 2.9375,
"learning_rate": 5.31959899482654e-07,
"loss": 0.08766459822654724,
"step": 2920
},
{
"epoch": 3.2356194690265485,
"grad_norm": 3.09375,
"learning_rate": 5.245744284518259e-07,
"loss": 0.0936079204082489,
"step": 2925
},
{
"epoch": 3.241150442477876,
"grad_norm": 1.46875,
"learning_rate": 5.172356698123195e-07,
"loss": 0.07304841876029969,
"step": 2930
},
{
"epoch": 3.2466814159292037,
"grad_norm": 1.6484375,
"learning_rate": 5.099437620498061e-07,
"loss": 0.08829051852226258,
"step": 2935
},
{
"epoch": 3.252212389380531,
"grad_norm": 1.6484375,
"learning_rate": 5.026988427658564e-07,
"loss": 0.09452310800552369,
"step": 2940
},
{
"epoch": 3.2577433628318584,
"grad_norm": 1.2734375,
"learning_rate": 4.955010486753476e-07,
"loss": 0.08840855360031127,
"step": 2945
},
{
"epoch": 3.2632743362831858,
"grad_norm": 8.9375,
"learning_rate": 4.883505156038841e-07,
"loss": 0.0822034239768982,
"step": 2950
},
{
"epoch": 3.268805309734513,
"grad_norm": 2.1875,
"learning_rate": 4.812473784852343e-07,
"loss": 0.10266009569168091,
"step": 2955
},
{
"epoch": 3.274336283185841,
"grad_norm": 5.09375,
"learning_rate": 4.7419177135878236e-07,
"loss": 0.07481509447097778,
"step": 2960
},
{
"epoch": 3.2798672566371683,
"grad_norm": 2.953125,
"learning_rate": 4.6718382736700105e-07,
"loss": 0.1020311951637268,
"step": 2965
},
{
"epoch": 3.2853982300884956,
"grad_norm": 3.546875,
"learning_rate": 4.6022367875293836e-07,
"loss": 0.06875736117362977,
"step": 2970
},
{
"epoch": 3.290929203539823,
"grad_norm": 1.703125,
"learning_rate": 4.5331145685772225e-07,
"loss": 0.07798204421997071,
"step": 2975
},
{
"epoch": 3.2964601769911503,
"grad_norm": 2.5,
"learning_rate": 4.464472921180801e-07,
"loss": 0.11223137378692627,
"step": 2980
},
{
"epoch": 3.3019911504424777,
"grad_norm": 2.0625,
"learning_rate": 4.3963131406388145e-07,
"loss": 0.08697097897529601,
"step": 2985
},
{
"epoch": 3.307522123893805,
"grad_norm": 2.984375,
"learning_rate": 4.3286365131569017e-07,
"loss": 0.10638842582702637,
"step": 2990
},
{
"epoch": 3.313053097345133,
"grad_norm": 1.484375,
"learning_rate": 4.261444315823411e-07,
"loss": 0.08118297457695008,
"step": 2995
},
{
"epoch": 3.3185840707964602,
"grad_norm": 1.28125,
"learning_rate": 4.194737816585236e-07,
"loss": 0.07100878953933716,
"step": 3000
},
{
"epoch": 3.3241150442477876,
"grad_norm": 5.15625,
"learning_rate": 4.128518274223959e-07,
"loss": 0.08557186722755432,
"step": 3005
},
{
"epoch": 3.329646017699115,
"grad_norm": 16.25,
"learning_rate": 4.062786938332057e-07,
"loss": 0.07297680377960206,
"step": 3010
},
{
"epoch": 3.3351769911504423,
"grad_norm": 2.484375,
"learning_rate": 3.9975450492893383e-07,
"loss": 0.12103928327560425,
"step": 3015
},
{
"epoch": 3.34070796460177,
"grad_norm": 1.5390625,
"learning_rate": 3.93279383823951e-07,
"loss": 0.05248732566833496,
"step": 3020
},
{
"epoch": 3.3462389380530975,
"grad_norm": 2.4375,
"learning_rate": 3.868534527066996e-07,
"loss": 0.07409105896949768,
"step": 3025
},
{
"epoch": 3.351769911504425,
"grad_norm": 1.7109375,
"learning_rate": 3.804768328373834e-07,
"loss": 0.09107747077941894,
"step": 3030
},
{
"epoch": 3.357300884955752,
"grad_norm": 2.78125,
"learning_rate": 3.7414964454567946e-07,
"loss": 0.0890788495540619,
"step": 3035
},
{
"epoch": 3.3628318584070795,
"grad_norm": 5.84375,
"learning_rate": 3.6787200722847067e-07,
"loss": 0.0948181390762329,
"step": 3040
},
{
"epoch": 3.368362831858407,
"grad_norm": 2.046875,
"learning_rate": 3.616440393475902e-07,
"loss": 0.09428614377975464,
"step": 3045
},
{
"epoch": 3.3738938053097347,
"grad_norm": 1.15625,
"learning_rate": 3.5546585842758626e-07,
"loss": 0.08605985045433044,
"step": 3050
},
{
"epoch": 3.379424778761062,
"grad_norm": 3.875,
"learning_rate": 3.493375810535057e-07,
"loss": 0.09287367463111877,
"step": 3055
},
{
"epoch": 3.3849557522123894,
"grad_norm": 3.890625,
"learning_rate": 3.432593228686919e-07,
"loss": 0.0755927562713623,
"step": 3060
},
{
"epoch": 3.390486725663717,
"grad_norm": 2.0625,
"learning_rate": 3.372311985726052e-07,
"loss": 0.0675960659980774,
"step": 3065
},
{
"epoch": 3.396017699115044,
"grad_norm": 3.515625,
"learning_rate": 3.3125332191865465e-07,
"loss": 0.08956061601638794,
"step": 3070
},
{
"epoch": 3.4015486725663715,
"grad_norm": 2.984375,
"learning_rate": 3.2532580571205607e-07,
"loss": 0.09910589456558228,
"step": 3075
},
{
"epoch": 3.4070796460176993,
"grad_norm": 2.578125,
"learning_rate": 3.1944876180769946e-07,
"loss": 0.09331427216529846,
"step": 3080
},
{
"epoch": 3.4126106194690267,
"grad_norm": 21.125,
"learning_rate": 3.136223011080417e-07,
"loss": 0.08216321468353271,
"step": 3085
},
{
"epoch": 3.418141592920354,
"grad_norm": 1.34375,
"learning_rate": 3.078465335610096e-07,
"loss": 0.09412244558334351,
"step": 3090
},
{
"epoch": 3.4236725663716814,
"grad_norm": 3.28125,
"learning_rate": 3.021215681579287e-07,
"loss": 0.082271808385849,
"step": 3095
},
{
"epoch": 3.4292035398230087,
"grad_norm": 1.828125,
"learning_rate": 2.964475129314652e-07,
"loss": 0.08677974343299866,
"step": 3100
},
{
"epoch": 3.434734513274336,
"grad_norm": 1.3125,
"learning_rate": 2.908244749535881e-07,
"loss": 0.07961620688438416,
"step": 3105
},
{
"epoch": 3.440265486725664,
"grad_norm": 2.53125,
"learning_rate": 2.852525603335454e-07,
"loss": 0.06239288449287415,
"step": 3110
},
{
"epoch": 3.4457964601769913,
"grad_norm": 3.140625,
"learning_rate": 2.7973187421586756e-07,
"loss": 0.08878275156021118,
"step": 3115
},
{
"epoch": 3.4513274336283186,
"grad_norm": 2.75,
"learning_rate": 2.7426252077837956e-07,
"loss": 0.10823150873184204,
"step": 3120
},
{
"epoch": 3.456858407079646,
"grad_norm": 2.703125,
"learning_rate": 2.688446032302333e-07,
"loss": 0.07904008626937867,
"step": 3125
},
{
"epoch": 3.4623893805309733,
"grad_norm": 2.203125,
"learning_rate": 2.6347822380996513e-07,
"loss": 0.09215710759162903,
"step": 3130
},
{
"epoch": 3.4679203539823007,
"grad_norm": 3.328125,
"learning_rate": 2.5816348378356224e-07,
"loss": 0.08517601490020751,
"step": 3135
},
{
"epoch": 3.4734513274336285,
"grad_norm": 2.5,
"learning_rate": 2.529004834425541e-07,
"loss": 0.09502893686294556,
"step": 3140
},
{
"epoch": 3.478982300884956,
"grad_norm": 1.65625,
"learning_rate": 2.476893221021188e-07,
"loss": 0.079125314950943,
"step": 3145
},
{
"epoch": 3.484513274336283,
"grad_norm": 2.3125,
"learning_rate": 2.4253009809920924e-07,
"loss": 0.07005802989006042,
"step": 3150
},
{
"epoch": 3.4900442477876106,
"grad_norm": 1.375,
"learning_rate": 2.3742290879069727e-07,
"loss": 0.07264252305030823,
"step": 3155
},
{
"epoch": 3.495575221238938,
"grad_norm": 1.6796875,
"learning_rate": 2.3236785055153732e-07,
"loss": 0.08641891479492188,
"step": 3160
},
{
"epoch": 3.5011061946902657,
"grad_norm": 4.0625,
"learning_rate": 2.2736501877294624e-07,
"loss": 0.10106908082962036,
"step": 3165
},
{
"epoch": 3.5066371681415927,
"grad_norm": 1.53125,
"learning_rate": 2.224145078606037e-07,
"loss": 0.09118204712867736,
"step": 3170
},
{
"epoch": 3.5121681415929205,
"grad_norm": 2.390625,
"learning_rate": 2.1751641123287435e-07,
"loss": 0.06414726376533508,
"step": 3175
},
{
"epoch": 3.517699115044248,
"grad_norm": 2.28125,
"learning_rate": 2.1267082131903747e-07,
"loss": 0.09005916714668274,
"step": 3180
},
{
"epoch": 3.523230088495575,
"grad_norm": 1.3046875,
"learning_rate": 2.0787782955754985e-07,
"loss": 0.08173223733901977,
"step": 3185
},
{
"epoch": 3.5287610619469025,
"grad_norm": 2.125,
"learning_rate": 2.0313752639431692e-07,
"loss": 0.07737751603126526,
"step": 3190
},
{
"epoch": 3.53429203539823,
"grad_norm": 3.96875,
"learning_rate": 1.9845000128098644e-07,
"loss": 0.09748212695121765,
"step": 3195
},
{
"epoch": 3.5398230088495577,
"grad_norm": 1.8984375,
"learning_rate": 1.9381534267325952e-07,
"loss": 0.08687747716903686,
"step": 3200
},
{
"epoch": 3.545353982300885,
"grad_norm": 1.4140625,
"learning_rate": 1.8923363802922523e-07,
"loss": 0.06420722603797913,
"step": 3205
},
{
"epoch": 3.5508849557522124,
"grad_norm": 3.25,
"learning_rate": 1.8470497380770602e-07,
"loss": 0.10305624008178711,
"step": 3210
},
{
"epoch": 3.5564159292035398,
"grad_norm": 25.375,
"learning_rate": 1.8022943546662696e-07,
"loss": 0.08411546349525452,
"step": 3215
},
{
"epoch": 3.561946902654867,
"grad_norm": 3.640625,
"learning_rate": 1.758071074614057e-07,
"loss": 0.06632069945335388,
"step": 3220
},
{
"epoch": 3.567477876106195,
"grad_norm": 1.90625,
"learning_rate": 1.714380732433558e-07,
"loss": 0.09679591655731201,
"step": 3225
},
{
"epoch": 3.573008849557522,
"grad_norm": 3.5625,
"learning_rate": 1.6712241525811323e-07,
"loss": 0.07133694887161254,
"step": 3230
},
{
"epoch": 3.5785398230088497,
"grad_norm": 3.265625,
"learning_rate": 1.6286021494408098e-07,
"loss": 0.07739973068237305,
"step": 3235
},
{
"epoch": 3.584070796460177,
"grad_norm": 1.453125,
"learning_rate": 1.586515527308917e-07,
"loss": 0.08276796936988831,
"step": 3240
},
{
"epoch": 3.5896017699115044,
"grad_norm": 1.390625,
"learning_rate": 1.544965080378895e-07,
"loss": 0.087041175365448,
"step": 3245
},
{
"epoch": 3.5951327433628317,
"grad_norm": 1.4453125,
"learning_rate": 1.503951592726328e-07,
"loss": 0.0790724515914917,
"step": 3250
},
{
"epoch": 3.600663716814159,
"grad_norm": 2.59375,
"learning_rate": 1.463475838294126e-07,
"loss": 0.09119505882263183,
"step": 3255
},
{
"epoch": 3.606194690265487,
"grad_norm": 1.9609375,
"learning_rate": 1.423538580877932e-07,
"loss": 0.0784076988697052,
"step": 3260
},
{
"epoch": 3.6117256637168142,
"grad_norm": 2.71875,
"learning_rate": 1.384140574111732e-07,
"loss": 0.08988794088363647,
"step": 3265
},
{
"epoch": 3.6172566371681416,
"grad_norm": 1.8359375,
"learning_rate": 1.3452825614535803e-07,
"loss": 0.09346536993980407,
"step": 3270
},
{
"epoch": 3.622787610619469,
"grad_norm": 306.0,
"learning_rate": 1.3069652761716133e-07,
"loss": 0.08357428908348083,
"step": 3275
},
{
"epoch": 3.6283185840707963,
"grad_norm": 1.71875,
"learning_rate": 1.2691894413301962e-07,
"loss": 0.09032775163650512,
"step": 3280
},
{
"epoch": 3.633849557522124,
"grad_norm": 2.90625,
"learning_rate": 1.231955769776294e-07,
"loss": 0.0825690507888794,
"step": 3285
},
{
"epoch": 3.6393805309734515,
"grad_norm": 1.3359375,
"learning_rate": 1.1952649641259817e-07,
"loss": 0.07909038662910461,
"step": 3290
},
{
"epoch": 3.644911504424779,
"grad_norm": 1.9765625,
"learning_rate": 1.1591177167512356e-07,
"loss": 0.06863436102867126,
"step": 3295
},
{
"epoch": 3.650442477876106,
"grad_norm": 1.4609375,
"learning_rate": 1.1235147097668375e-07,
"loss": 0.09359605312347412,
"step": 3300
},
{
"epoch": 3.6559734513274336,
"grad_norm": 3.3125,
"learning_rate": 1.0884566150174969e-07,
"loss": 0.08553218245506286,
"step": 3305
},
{
"epoch": 3.661504424778761,
"grad_norm": 1.953125,
"learning_rate": 1.0539440940651923e-07,
"loss": 0.09473593235015869,
"step": 3310
},
{
"epoch": 3.6670353982300883,
"grad_norm": 9.0625,
"learning_rate": 1.0199777981766878e-07,
"loss": 0.08314796686172485,
"step": 3315
},
{
"epoch": 3.672566371681416,
"grad_norm": 1.5625,
"learning_rate": 9.865583683112255e-08,
"loss": 0.06281660795211792,
"step": 3320
},
{
"epoch": 3.6780973451327434,
"grad_norm": 2.90625,
"learning_rate": 9.536864351084396e-08,
"loss": 0.09631728529930114,
"step": 3325
},
{
"epoch": 3.683628318584071,
"grad_norm": 1.5390625,
"learning_rate": 9.213626188764646e-08,
"loss": 0.0964404821395874,
"step": 3330
},
{
"epoch": 3.689159292035398,
"grad_norm": 1.796875,
"learning_rate": 8.895875295802158e-08,
"loss": 0.06045501828193665,
"step": 3335
},
{
"epoch": 3.6946902654867255,
"grad_norm": 2.21875,
"learning_rate": 8.583617668298915e-08,
"loss": 0.08764418363571166,
"step": 3340
},
{
"epoch": 3.7002212389380533,
"grad_norm": 2.71875,
"learning_rate": 8.276859198696385e-08,
"loss": 0.09030008316040039,
"step": 3345
},
{
"epoch": 3.7057522123893807,
"grad_norm": 1.25,
"learning_rate": 7.97560567566451e-08,
"loss": 0.08346371650695801,
"step": 3350
},
{
"epoch": 3.711283185840708,
"grad_norm": 4.25,
"learning_rate": 7.679862783992598e-08,
"loss": 0.08700665831565857,
"step": 3355
},
{
"epoch": 3.7168141592920354,
"grad_norm": 1.8203125,
"learning_rate": 7.389636104481601e-08,
"loss": 0.09551226496696472,
"step": 3360
},
{
"epoch": 3.7223451327433628,
"grad_norm": 1.78125,
"learning_rate": 7.104931113839242e-08,
"loss": 0.07144197225570678,
"step": 3365
},
{
"epoch": 3.72787610619469,
"grad_norm": 1.625,
"learning_rate": 6.825753184576422e-08,
"loss": 0.10204342603683472,
"step": 3370
},
{
"epoch": 3.7334070796460175,
"grad_norm": 2.765625,
"learning_rate": 6.552107584905975e-08,
"loss": 0.0788453459739685,
"step": 3375
},
{
"epoch": 3.7389380530973453,
"grad_norm": 1.9609375,
"learning_rate": 6.28399947864311e-08,
"loss": 0.06642647385597229,
"step": 3380
},
{
"epoch": 3.7444690265486726,
"grad_norm": 2.703125,
"learning_rate": 6.021433925108122e-08,
"loss": 0.07201675772666931,
"step": 3385
},
{
"epoch": 3.75,
"grad_norm": 2.9375,
"learning_rate": 5.7644158790308706e-08,
"loss": 0.08554626703262329,
"step": 3390
},
{
"epoch": 3.7555309734513274,
"grad_norm": 2.671875,
"learning_rate": 5.512950190457189e-08,
"loss": 0.08283672332763672,
"step": 3395
},
{
"epoch": 3.7610619469026547,
"grad_norm": 4.875,
"learning_rate": 5.267041604657419e-08,
"loss": 0.07992309331893921,
"step": 3400
},
{
"epoch": 3.7665929203539825,
"grad_norm": 3.890625,
"learning_rate": 5.026694762036921e-08,
"loss": 0.08749452233314514,
"step": 3405
},
{
"epoch": 3.77212389380531,
"grad_norm": 4.3125,
"learning_rate": 4.791914198048541e-08,
"loss": 0.061494797468185425,
"step": 3410
},
{
"epoch": 3.7776548672566372,
"grad_norm": 2.015625,
"learning_rate": 4.562704343106816e-08,
"loss": 0.09088380932807923,
"step": 3415
},
{
"epoch": 3.7831858407079646,
"grad_norm": 2.71875,
"learning_rate": 4.339069522504568e-08,
"loss": 0.09628398418426513,
"step": 3420
},
{
"epoch": 3.788716814159292,
"grad_norm": 1.8984375,
"learning_rate": 4.12101395633121e-08,
"loss": 0.07528159618377686,
"step": 3425
},
{
"epoch": 3.7942477876106193,
"grad_norm": 1.9921875,
"learning_rate": 3.908541759393136e-08,
"loss": 0.08525064587593079,
"step": 3430
},
{
"epoch": 3.7997787610619467,
"grad_norm": 1.3203125,
"learning_rate": 3.7016569411360225e-08,
"loss": 0.0689652144908905,
"step": 3435
},
{
"epoch": 3.8053097345132745,
"grad_norm": 5.5625,
"learning_rate": 3.500363405569251e-08,
"loss": 0.08421769142150878,
"step": 3440
},
{
"epoch": 3.810840707964602,
"grad_norm": 1.625,
"learning_rate": 3.3046649511921715e-08,
"loss": 0.07707933187484742,
"step": 3445
},
{
"epoch": 3.816371681415929,
"grad_norm": 5.28125,
"learning_rate": 3.114565270922387e-08,
"loss": 0.083420330286026,
"step": 3450
},
{
"epoch": 3.8219026548672566,
"grad_norm": 1.609375,
"learning_rate": 2.9300679520262164e-08,
"loss": 0.08925502896308898,
"step": 3455
},
{
"epoch": 3.827433628318584,
"grad_norm": 4.375,
"learning_rate": 2.7511764760508094e-08,
"loss": 0.10175607204437256,
"step": 3460
},
{
"epoch": 3.8329646017699117,
"grad_norm": 3.1875,
"learning_rate": 2.5778942187586362e-08,
"loss": 0.08114023208618164,
"step": 3465
},
{
"epoch": 3.838495575221239,
"grad_norm": 1.03125,
"learning_rate": 2.410224450063603e-08,
"loss": 0.07333937883377076,
"step": 3470
},
{
"epoch": 3.8440265486725664,
"grad_norm": 1.9609375,
"learning_rate": 2.2481703339695348e-08,
"loss": 0.08941631317138672,
"step": 3475
},
{
"epoch": 3.849557522123894,
"grad_norm": 1.3515625,
"learning_rate": 2.091734928510258e-08,
"loss": 0.08364614844322205,
"step": 3480
},
{
"epoch": 3.855088495575221,
"grad_norm": 1.9609375,
"learning_rate": 1.940921185692146e-08,
"loss": 0.11341748237609864,
"step": 3485
},
{
"epoch": 3.8606194690265485,
"grad_norm": 2.46875,
"learning_rate": 1.7957319514381287e-08,
"loss": 0.0984826147556305,
"step": 3490
},
{
"epoch": 3.866150442477876,
"grad_norm": 17.5,
"learning_rate": 1.6561699655342376e-08,
"loss": 0.10930407047271729,
"step": 3495
},
{
"epoch": 3.8716814159292037,
"grad_norm": 3.34375,
"learning_rate": 1.52223786157778e-08,
"loss": 0.060564398765563965,
"step": 3500
},
{
"epoch": 3.877212389380531,
"grad_norm": 3.109375,
"learning_rate": 1.3939381669276442e-08,
"loss": 0.08095338940620422,
"step": 3505
},
{
"epoch": 3.8827433628318584,
"grad_norm": 2.921875,
"learning_rate": 1.2712733026566391e-08,
"loss": 0.08435689210891724,
"step": 3510
},
{
"epoch": 3.8882743362831858,
"grad_norm": 2.84375,
"learning_rate": 1.1542455835057974e-08,
"loss": 0.0947385549545288,
"step": 3515
},
{
"epoch": 3.893805309734513,
"grad_norm": 3.234375,
"learning_rate": 1.0428572178407092e-08,
"loss": 0.07517052292823792,
"step": 3520
},
{
"epoch": 3.899336283185841,
"grad_norm": 2.34375,
"learning_rate": 9.371103076097898e-09,
"loss": 0.08906136155128479,
"step": 3525
},
{
"epoch": 3.9048672566371683,
"grad_norm": 41.5,
"learning_rate": 8.37006848304711e-09,
"loss": 0.09657859206199645,
"step": 3530
},
{
"epoch": 3.9103982300884956,
"grad_norm": 2.390625,
"learning_rate": 7.4254872892263184e-09,
"loss": 0.08448508381843567,
"step": 3535
},
{
"epoch": 3.915929203539823,
"grad_norm": 3.703125,
"learning_rate": 6.537377319307259e-09,
"loss": 0.07274735569953919,
"step": 3540
},
{
"epoch": 3.9214601769911503,
"grad_norm": 1.8828125,
"learning_rate": 5.70575533232276e-09,
"loss": 0.08858110904693603,
"step": 3545
},
{
"epoch": 3.926991150442478,
"grad_norm": 8.5,
"learning_rate": 4.930637021352991e-09,
"loss": 0.08049632906913758,
"step": 3550
},
{
"epoch": 3.932522123893805,
"grad_norm": 1.796875,
"learning_rate": 4.212037013228698e-09,
"loss": 0.09107894897460937,
"step": 3555
},
{
"epoch": 3.938053097345133,
"grad_norm": 1.8125,
"learning_rate": 3.5499688682544273e-09,
"loss": 0.08556202054023743,
"step": 3560
},
{
"epoch": 3.9435840707964602,
"grad_norm": 1.6640625,
"learning_rate": 2.944445079953062e-09,
"loss": 0.0678092360496521,
"step": 3565
},
{
"epoch": 3.9491150442477876,
"grad_norm": 1.3515625,
"learning_rate": 2.3954770748303436e-09,
"loss": 0.0814163863658905,
"step": 3570
},
{
"epoch": 3.954646017699115,
"grad_norm": 1.8984375,
"learning_rate": 1.9030752121583785e-09,
"loss": 0.09616196751594544,
"step": 3575
},
{
"epoch": 3.9601769911504423,
"grad_norm": 1.9609375,
"learning_rate": 1.4672487837817939e-09,
"loss": 0.08779360055923462,
"step": 3580
},
{
"epoch": 3.96570796460177,
"grad_norm": 3.921875,
"learning_rate": 1.0880060139405456e-09,
"loss": 0.09187158346176147,
"step": 3585
},
{
"epoch": 3.9712389380530975,
"grad_norm": 21.625,
"learning_rate": 7.653540591160412e-10,
"loss": 0.08648747205734253,
"step": 3590
},
{
"epoch": 3.976769911504425,
"grad_norm": 1.6796875,
"learning_rate": 4.992990078955817e-10,
"loss": 0.08265175223350525,
"step": 3595
},
{
"epoch": 3.982300884955752,
"grad_norm": 7.59375,
"learning_rate": 2.898458808567872e-10,
"loss": 0.07560694813728333,
"step": 3600
},
{
"epoch": 3.9878318584070795,
"grad_norm": 3.671875,
"learning_rate": 1.3699863047500484e-10,
"loss": 0.09129037261009217,
"step": 3605
},
{
"epoch": 3.9933628318584073,
"grad_norm": 1.828125,
"learning_rate": 4.0760141045703694e-11,
"loss": 0.0957610547542572,
"step": 3610
},
{
"epoch": 3.9988938053097343,
"grad_norm": 5.5,
"learning_rate": 1.1322286328496745e-12,
"loss": 0.0876231849193573,
"step": 3615
},
{
"epoch": 4.0,
"eval_loss": 0.08387346565723419,
"eval_pearson_cosine": 0.27503812313079834,
"eval_pearson_dot_product": 0.27503806352615356,
"eval_pearson_euclidean": 0.27455732226371765,
"eval_pearson_manhattan": 0.26413509249687195,
"eval_runtime": 75.5447,
"eval_samples_per_second": 107.698,
"eval_spearman_cosine": 0.29903502399664816,
"eval_spearman_dot_product": 0.29903523651008945,
"eval_spearman_euclidean": 0.29903508306832594,
"eval_spearman_manhattan": 0.2888251791647824,
"eval_steps_per_second": 17.95,
"step": 3616
},
{
"epoch": 4.0,
"eval_loss": 0.08387346565723419,
"eval_pearson_cosine": 0.27503812313079834,
"eval_pearson_dot_product": 0.27503806352615356,
"eval_pearson_euclidean": 0.27455732226371765,
"eval_pearson_manhattan": 0.26413509249687195,
"eval_runtime": 75.5973,
"eval_samples_per_second": 107.623,
"eval_spearman_cosine": 0.29903502399664816,
"eval_spearman_dot_product": 0.29903523651008945,
"eval_spearman_euclidean": 0.29903508306832594,
"eval_spearman_manhattan": 0.2888251791647824,
"eval_steps_per_second": 17.937,
"step": 3616
}
],
"logging_steps": 5,
"max_steps": 3616,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.5965355694292992e+17,
"train_batch_size": 6,
"trial_name": null,
"trial_params": null
}