ablation-Qwen2.5-7B-ttc / trainer_state.json
wf8888884's picture
Upload fine-tuned model
c16d2e3 verified
Raw
History Blame Contribute Delete
76 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 4347,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006901311249137336,
"grad_norm": 30.456523895263672,
"learning_rate": 2.2988505747126437e-07,
"loss": 1.3071,
"step": 10
},
{
"epoch": 0.013802622498274672,
"grad_norm": 13.81282901763916,
"learning_rate": 4.5977011494252875e-07,
"loss": 1.3012,
"step": 20
},
{
"epoch": 0.020703933747412008,
"grad_norm": 15.553468704223633,
"learning_rate": 6.896551724137931e-07,
"loss": 1.0279,
"step": 30
},
{
"epoch": 0.027605244996549344,
"grad_norm": 12.811663627624512,
"learning_rate": 9.195402298850575e-07,
"loss": 1.1768,
"step": 40
},
{
"epoch": 0.03450655624568668,
"grad_norm": 10.282151222229004,
"learning_rate": 1.1494252873563219e-06,
"loss": 1.0811,
"step": 50
},
{
"epoch": 0.041407867494824016,
"grad_norm": 18.447317123413086,
"learning_rate": 1.3793103448275862e-06,
"loss": 1.0825,
"step": 60
},
{
"epoch": 0.04830917874396135,
"grad_norm": 15.161924362182617,
"learning_rate": 1.6091954022988506e-06,
"loss": 1.0283,
"step": 70
},
{
"epoch": 0.05521048999309869,
"grad_norm": 6.859762191772461,
"learning_rate": 1.839080459770115e-06,
"loss": 0.8963,
"step": 80
},
{
"epoch": 0.062111801242236024,
"grad_norm": 8.195549011230469,
"learning_rate": 2.0689655172413796e-06,
"loss": 0.9299,
"step": 90
},
{
"epoch": 0.06901311249137336,
"grad_norm": 37.93837356567383,
"learning_rate": 2.2988505747126437e-06,
"loss": 0.7865,
"step": 100
},
{
"epoch": 0.0759144237405107,
"grad_norm": 8.226261138916016,
"learning_rate": 2.5287356321839083e-06,
"loss": 0.8928,
"step": 110
},
{
"epoch": 0.08281573498964803,
"grad_norm": 10.093214988708496,
"learning_rate": 2.7586206896551725e-06,
"loss": 0.773,
"step": 120
},
{
"epoch": 0.08971704623878538,
"grad_norm": 7.429821968078613,
"learning_rate": 2.988505747126437e-06,
"loss": 0.8849,
"step": 130
},
{
"epoch": 0.0966183574879227,
"grad_norm": 6.883029937744141,
"learning_rate": 3.2183908045977012e-06,
"loss": 0.7699,
"step": 140
},
{
"epoch": 0.10351966873706005,
"grad_norm": 62.745662689208984,
"learning_rate": 3.448275862068966e-06,
"loss": 0.8079,
"step": 150
},
{
"epoch": 0.11042097998619738,
"grad_norm": 6.498140335083008,
"learning_rate": 3.67816091954023e-06,
"loss": 0.7096,
"step": 160
},
{
"epoch": 0.11732229123533472,
"grad_norm": 25.81262969970703,
"learning_rate": 3.908045977011495e-06,
"loss": 0.7361,
"step": 170
},
{
"epoch": 0.12422360248447205,
"grad_norm": 9.785597801208496,
"learning_rate": 4.137931034482759e-06,
"loss": 0.8101,
"step": 180
},
{
"epoch": 0.13112491373360938,
"grad_norm": 7.6141557693481445,
"learning_rate": 4.367816091954023e-06,
"loss": 0.6649,
"step": 190
},
{
"epoch": 0.13802622498274672,
"grad_norm": 15.329546928405762,
"learning_rate": 4.5977011494252875e-06,
"loss": 0.7395,
"step": 200
},
{
"epoch": 0.14492753623188406,
"grad_norm": 4.4098801612854,
"learning_rate": 4.8275862068965525e-06,
"loss": 0.6573,
"step": 210
},
{
"epoch": 0.1518288474810214,
"grad_norm": 4.633988857269287,
"learning_rate": 5.057471264367817e-06,
"loss": 0.6731,
"step": 220
},
{
"epoch": 0.15873015873015872,
"grad_norm": 6.017467498779297,
"learning_rate": 5.287356321839081e-06,
"loss": 0.6898,
"step": 230
},
{
"epoch": 0.16563146997929606,
"grad_norm": 6.0080885887146,
"learning_rate": 5.517241379310345e-06,
"loss": 0.6938,
"step": 240
},
{
"epoch": 0.1725327812284334,
"grad_norm": 4.427531719207764,
"learning_rate": 5.747126436781609e-06,
"loss": 0.7657,
"step": 250
},
{
"epoch": 0.17943409247757075,
"grad_norm": 4.281004905700684,
"learning_rate": 5.977011494252874e-06,
"loss": 0.5897,
"step": 260
},
{
"epoch": 0.18633540372670807,
"grad_norm": 8.194170951843262,
"learning_rate": 6.206896551724138e-06,
"loss": 0.7891,
"step": 270
},
{
"epoch": 0.1932367149758454,
"grad_norm": 6.162697792053223,
"learning_rate": 6.4367816091954025e-06,
"loss": 0.7238,
"step": 280
},
{
"epoch": 0.20013802622498275,
"grad_norm": 3.846493721008301,
"learning_rate": 6.666666666666667e-06,
"loss": 0.5978,
"step": 290
},
{
"epoch": 0.2070393374741201,
"grad_norm": 3.6054069995880127,
"learning_rate": 6.896551724137932e-06,
"loss": 0.691,
"step": 300
},
{
"epoch": 0.2139406487232574,
"grad_norm": 6.958858013153076,
"learning_rate": 7.126436781609196e-06,
"loss": 0.7185,
"step": 310
},
{
"epoch": 0.22084195997239475,
"grad_norm": 22.483959197998047,
"learning_rate": 7.35632183908046e-06,
"loss": 0.6854,
"step": 320
},
{
"epoch": 0.2277432712215321,
"grad_norm": 7.925286293029785,
"learning_rate": 7.586206896551724e-06,
"loss": 0.7096,
"step": 330
},
{
"epoch": 0.23464458247066944,
"grad_norm": 8.171796798706055,
"learning_rate": 7.81609195402299e-06,
"loss": 0.8267,
"step": 340
},
{
"epoch": 0.24154589371980675,
"grad_norm": 5.633179664611816,
"learning_rate": 8.045977011494253e-06,
"loss": 0.7404,
"step": 350
},
{
"epoch": 0.2484472049689441,
"grad_norm": 4.697131156921387,
"learning_rate": 8.275862068965518e-06,
"loss": 0.6488,
"step": 360
},
{
"epoch": 0.2553485162180814,
"grad_norm": 3.750340700149536,
"learning_rate": 8.505747126436782e-06,
"loss": 0.6117,
"step": 370
},
{
"epoch": 0.26224982746721875,
"grad_norm": 6.92358922958374,
"learning_rate": 8.735632183908047e-06,
"loss": 0.5504,
"step": 380
},
{
"epoch": 0.2691511387163561,
"grad_norm": 4.210056304931641,
"learning_rate": 8.965517241379312e-06,
"loss": 0.5898,
"step": 390
},
{
"epoch": 0.27605244996549344,
"grad_norm": 4.635003566741943,
"learning_rate": 9.195402298850575e-06,
"loss": 0.5893,
"step": 400
},
{
"epoch": 0.2829537612146308,
"grad_norm": 4.109310626983643,
"learning_rate": 9.42528735632184e-06,
"loss": 0.7137,
"step": 410
},
{
"epoch": 0.2898550724637681,
"grad_norm": 6.997936725616455,
"learning_rate": 9.655172413793105e-06,
"loss": 0.6277,
"step": 420
},
{
"epoch": 0.29675638371290547,
"grad_norm": 7.919456481933594,
"learning_rate": 9.885057471264368e-06,
"loss": 0.539,
"step": 430
},
{
"epoch": 0.3036576949620428,
"grad_norm": 5.434922218322754,
"learning_rate": 9.99995969290611e-06,
"loss": 0.6279,
"step": 440
},
{
"epoch": 0.3105590062111801,
"grad_norm": 10.044814109802246,
"learning_rate": 9.999637240054163e-06,
"loss": 0.6317,
"step": 450
},
{
"epoch": 0.31746031746031744,
"grad_norm": 3.6754772663116455,
"learning_rate": 9.998992355145604e-06,
"loss": 0.5687,
"step": 460
},
{
"epoch": 0.3243616287094548,
"grad_norm": 4.302608013153076,
"learning_rate": 9.998025079769766e-06,
"loss": 0.5997,
"step": 470
},
{
"epoch": 0.33126293995859213,
"grad_norm": 5.8444929122924805,
"learning_rate": 9.996735476307292e-06,
"loss": 0.6204,
"step": 480
},
{
"epoch": 0.33816425120772947,
"grad_norm": 5.4927496910095215,
"learning_rate": 9.995123627926113e-06,
"loss": 0.6309,
"step": 490
},
{
"epoch": 0.3450655624568668,
"grad_norm": 5.747093677520752,
"learning_rate": 9.99318963857609e-06,
"loss": 0.6019,
"step": 500
},
{
"epoch": 0.35196687370600416,
"grad_norm": 6.010488033294678,
"learning_rate": 9.990933632982307e-06,
"loss": 0.5639,
"step": 510
},
{
"epoch": 0.3588681849551415,
"grad_norm": 7.603153705596924,
"learning_rate": 9.98835575663702e-06,
"loss": 0.64,
"step": 520
},
{
"epoch": 0.3657694962042788,
"grad_norm": 4.128995895385742,
"learning_rate": 9.98545617579029e-06,
"loss": 0.648,
"step": 530
},
{
"epoch": 0.37267080745341613,
"grad_norm": 5.207589626312256,
"learning_rate": 9.982235077439242e-06,
"loss": 0.61,
"step": 540
},
{
"epoch": 0.3795721187025535,
"grad_norm": 4.482063293457031,
"learning_rate": 9.978692669316024e-06,
"loss": 0.5378,
"step": 550
},
{
"epoch": 0.3864734299516908,
"grad_norm": 10.252486228942871,
"learning_rate": 9.974829179874398e-06,
"loss": 0.5527,
"step": 560
},
{
"epoch": 0.39337474120082816,
"grad_norm": 4.370613098144531,
"learning_rate": 9.970644858275008e-06,
"loss": 0.603,
"step": 570
},
{
"epoch": 0.4002760524499655,
"grad_norm": 7.507190227508545,
"learning_rate": 9.96613997436932e-06,
"loss": 0.6112,
"step": 580
},
{
"epoch": 0.40717736369910285,
"grad_norm": 4.562262058258057,
"learning_rate": 9.961314818682205e-06,
"loss": 0.5068,
"step": 590
},
{
"epoch": 0.4140786749482402,
"grad_norm": 6.171989917755127,
"learning_rate": 9.956169702393215e-06,
"loss": 0.5664,
"step": 600
},
{
"epoch": 0.4209799861973775,
"grad_norm": 5.033712387084961,
"learning_rate": 9.95070495731651e-06,
"loss": 0.5445,
"step": 610
},
{
"epoch": 0.4278812974465148,
"grad_norm": 8.164334297180176,
"learning_rate": 9.944920935879456e-06,
"loss": 0.7071,
"step": 620
},
{
"epoch": 0.43478260869565216,
"grad_norm": 3.0369226932525635,
"learning_rate": 9.938818011099906e-06,
"loss": 0.512,
"step": 630
},
{
"epoch": 0.4416839199447895,
"grad_norm": 3.959850549697876,
"learning_rate": 9.932396576562128e-06,
"loss": 0.491,
"step": 640
},
{
"epoch": 0.44858523119392685,
"grad_norm": 4.791125774383545,
"learning_rate": 9.925657046391438e-06,
"loss": 0.567,
"step": 650
},
{
"epoch": 0.4554865424430642,
"grad_norm": 4.160429000854492,
"learning_rate": 9.918599855227487e-06,
"loss": 0.5922,
"step": 660
},
{
"epoch": 0.46238785369220153,
"grad_norm": 2.7612223625183105,
"learning_rate": 9.911225458196226e-06,
"loss": 0.4774,
"step": 670
},
{
"epoch": 0.4692891649413389,
"grad_norm": 3.864403486251831,
"learning_rate": 9.903534330880557e-06,
"loss": 0.5926,
"step": 680
},
{
"epoch": 0.47619047619047616,
"grad_norm": 13.458492279052734,
"learning_rate": 9.895526969289671e-06,
"loss": 0.6549,
"step": 690
},
{
"epoch": 0.4830917874396135,
"grad_norm": 6.3047075271606445,
"learning_rate": 9.887203889827045e-06,
"loss": 0.5129,
"step": 700
},
{
"epoch": 0.48999309868875085,
"grad_norm": 5.5510663986206055,
"learning_rate": 9.878565629257146e-06,
"loss": 0.6367,
"step": 710
},
{
"epoch": 0.4968944099378882,
"grad_norm": 5.698916912078857,
"learning_rate": 9.869612744670817e-06,
"loss": 0.6353,
"step": 720
},
{
"epoch": 0.5037957211870255,
"grad_norm": 3.8244333267211914,
"learning_rate": 9.86034581344935e-06,
"loss": 0.63,
"step": 730
},
{
"epoch": 0.5106970324361628,
"grad_norm": 6.79824161529541,
"learning_rate": 9.850765433227239e-06,
"loss": 0.6139,
"step": 740
},
{
"epoch": 0.5175983436853002,
"grad_norm": 7.884110450744629,
"learning_rate": 9.84087222185365e-06,
"loss": 0.5452,
"step": 750
},
{
"epoch": 0.5244996549344375,
"grad_norm": 5.6547722816467285,
"learning_rate": 9.830666817352575e-06,
"loss": 0.561,
"step": 760
},
{
"epoch": 0.5314009661835749,
"grad_norm": 3.915212631225586,
"learning_rate": 9.82014987788167e-06,
"loss": 0.592,
"step": 770
},
{
"epoch": 0.5383022774327122,
"grad_norm": 4.357222557067871,
"learning_rate": 9.809322081689834e-06,
"loss": 0.6482,
"step": 780
},
{
"epoch": 0.5452035886818496,
"grad_norm": 8.227004051208496,
"learning_rate": 9.798184127073446e-06,
"loss": 0.6183,
"step": 790
},
{
"epoch": 0.5521048999309869,
"grad_norm": 4.247949600219727,
"learning_rate": 9.786736732331343e-06,
"loss": 0.6327,
"step": 800
},
{
"epoch": 0.5590062111801242,
"grad_norm": 4.7755866050720215,
"learning_rate": 9.774980635718494e-06,
"loss": 0.5837,
"step": 810
},
{
"epoch": 0.5659075224292616,
"grad_norm": 4.1178460121154785,
"learning_rate": 9.762916595398384e-06,
"loss": 0.4785,
"step": 820
},
{
"epoch": 0.5728088336783989,
"grad_norm": 3.032378911972046,
"learning_rate": 9.75054538939413e-06,
"loss": 0.4775,
"step": 830
},
{
"epoch": 0.5797101449275363,
"grad_norm": 6.076838970184326,
"learning_rate": 9.737867815538292e-06,
"loss": 0.7093,
"step": 840
},
{
"epoch": 0.5866114561766735,
"grad_norm": 4.363618850708008,
"learning_rate": 9.724884691421433e-06,
"loss": 0.501,
"step": 850
},
{
"epoch": 0.5935127674258109,
"grad_norm": 8.495121955871582,
"learning_rate": 9.71159685433938e-06,
"loss": 0.5323,
"step": 860
},
{
"epoch": 0.6004140786749482,
"grad_norm": 3.3251333236694336,
"learning_rate": 9.698005161239237e-06,
"loss": 0.5396,
"step": 870
},
{
"epoch": 0.6073153899240856,
"grad_norm": 23.633102416992188,
"learning_rate": 9.684110488664107e-06,
"loss": 0.5438,
"step": 880
},
{
"epoch": 0.6142167011732229,
"grad_norm": 3.573366641998291,
"learning_rate": 9.669913732696577e-06,
"loss": 0.5228,
"step": 890
},
{
"epoch": 0.6211180124223602,
"grad_norm": 5.600038051605225,
"learning_rate": 9.655415808900915e-06,
"loss": 0.7029,
"step": 900
},
{
"epoch": 0.6280193236714976,
"grad_norm": 5.561460018157959,
"learning_rate": 9.640617652264033e-06,
"loss": 0.5605,
"step": 910
},
{
"epoch": 0.6349206349206349,
"grad_norm": 5.411248207092285,
"learning_rate": 9.62552021713519e-06,
"loss": 0.5564,
"step": 920
},
{
"epoch": 0.6418219461697723,
"grad_norm": 3.9415833950042725,
"learning_rate": 9.610124477164436e-06,
"loss": 0.5577,
"step": 930
},
{
"epoch": 0.6487232574189096,
"grad_norm": 4.239798069000244,
"learning_rate": 9.594431425239825e-06,
"loss": 0.5076,
"step": 940
},
{
"epoch": 0.655624568668047,
"grad_norm": 5.770053386688232,
"learning_rate": 9.57844207342339e-06,
"loss": 0.6494,
"step": 950
},
{
"epoch": 0.6625258799171843,
"grad_norm": 6.171411037445068,
"learning_rate": 9.56215745288586e-06,
"loss": 0.5499,
"step": 960
},
{
"epoch": 0.6694271911663217,
"grad_norm": 3.8769118785858154,
"learning_rate": 9.545578613840168e-06,
"loss": 0.447,
"step": 970
},
{
"epoch": 0.6763285024154589,
"grad_norm": 7.931893825531006,
"learning_rate": 9.52870662547372e-06,
"loss": 0.5629,
"step": 980
},
{
"epoch": 0.6832298136645962,
"grad_norm": 4.503838539123535,
"learning_rate": 9.51154257587944e-06,
"loss": 0.5933,
"step": 990
},
{
"epoch": 0.6901311249137336,
"grad_norm": 6.611269950866699,
"learning_rate": 9.494087571985604e-06,
"loss": 0.4783,
"step": 1000
},
{
"epoch": 0.6970324361628709,
"grad_norm": 13.41486930847168,
"learning_rate": 9.476342739484438e-06,
"loss": 0.4632,
"step": 1010
},
{
"epoch": 0.7039337474120083,
"grad_norm": 8.071541786193848,
"learning_rate": 9.458309222759542e-06,
"loss": 0.5246,
"step": 1020
},
{
"epoch": 0.7108350586611456,
"grad_norm": 2.5699656009674072,
"learning_rate": 9.439988184812065e-06,
"loss": 0.5443,
"step": 1030
},
{
"epoch": 0.717736369910283,
"grad_norm": 3.9053971767425537,
"learning_rate": 9.421380807185727e-06,
"loss": 0.4909,
"step": 1040
},
{
"epoch": 0.7246376811594203,
"grad_norm": 4.790355682373047,
"learning_rate": 9.402488289890593e-06,
"loss": 0.5011,
"step": 1050
},
{
"epoch": 0.7315389924085576,
"grad_norm": 8.724527359008789,
"learning_rate": 9.383311851325702e-06,
"loss": 0.698,
"step": 1060
},
{
"epoch": 0.738440303657695,
"grad_norm": 5.10377836227417,
"learning_rate": 9.363852728200488e-06,
"loss": 0.5925,
"step": 1070
},
{
"epoch": 0.7453416149068323,
"grad_norm": 4.863698482513428,
"learning_rate": 9.344112175455013e-06,
"loss": 0.5543,
"step": 1080
},
{
"epoch": 0.7522429261559697,
"grad_norm": 5.813719272613525,
"learning_rate": 9.324091466179051e-06,
"loss": 0.5592,
"step": 1090
},
{
"epoch": 0.759144237405107,
"grad_norm": 6.771941661834717,
"learning_rate": 9.303791891529966e-06,
"loss": 0.5901,
"step": 1100
},
{
"epoch": 0.7660455486542443,
"grad_norm": 5.885433673858643,
"learning_rate": 9.283214760649462e-06,
"loss": 0.5491,
"step": 1110
},
{
"epoch": 0.7729468599033816,
"grad_norm": 7.528721332550049,
"learning_rate": 9.262361400579145e-06,
"loss": 0.5494,
"step": 1120
},
{
"epoch": 0.779848171152519,
"grad_norm": 5.483184337615967,
"learning_rate": 9.24123315617494e-06,
"loss": 0.5211,
"step": 1130
},
{
"epoch": 0.7867494824016563,
"grad_norm": 6.512402534484863,
"learning_rate": 9.219831390020367e-06,
"loss": 0.4994,
"step": 1140
},
{
"epoch": 0.7936507936507936,
"grad_norm": 4.124596118927002,
"learning_rate": 9.198157482338655e-06,
"loss": 0.502,
"step": 1150
},
{
"epoch": 0.800552104899931,
"grad_norm": 6.286987781524658,
"learning_rate": 9.176212830903746e-06,
"loss": 0.6133,
"step": 1160
},
{
"epoch": 0.8074534161490683,
"grad_norm": 3.699578046798706,
"learning_rate": 9.153998850950133e-06,
"loss": 0.528,
"step": 1170
},
{
"epoch": 0.8143547273982057,
"grad_norm": 5.51190185546875,
"learning_rate": 9.131516975081604e-06,
"loss": 0.5957,
"step": 1180
},
{
"epoch": 0.821256038647343,
"grad_norm": 6.867079257965088,
"learning_rate": 9.108768653178848e-06,
"loss": 0.51,
"step": 1190
},
{
"epoch": 0.8281573498964804,
"grad_norm": 7.200817108154297,
"learning_rate": 9.08575535230594e-06,
"loss": 0.4962,
"step": 1200
},
{
"epoch": 0.8350586611456177,
"grad_norm": 5.3222246170043945,
"learning_rate": 9.062478556615751e-06,
"loss": 0.6118,
"step": 1210
},
{
"epoch": 0.841959972394755,
"grad_norm": 3.403719902038574,
"learning_rate": 9.03893976725421e-06,
"loss": 0.4587,
"step": 1220
},
{
"epoch": 0.8488612836438924,
"grad_norm": 10.111266136169434,
"learning_rate": 9.015140502263506e-06,
"loss": 0.5364,
"step": 1230
},
{
"epoch": 0.8557625948930296,
"grad_norm": 6.337829113006592,
"learning_rate": 8.991082296484188e-06,
"loss": 0.5346,
"step": 1240
},
{
"epoch": 0.862663906142167,
"grad_norm": 4.5559468269348145,
"learning_rate": 8.966766701456177e-06,
"loss": 0.5322,
"step": 1250
},
{
"epoch": 0.8695652173913043,
"grad_norm": 5.625792026519775,
"learning_rate": 8.942195285318705e-06,
"loss": 0.5908,
"step": 1260
},
{
"epoch": 0.8764665286404417,
"grad_norm": 5.8429460525512695,
"learning_rate": 8.917369632709189e-06,
"loss": 0.5976,
"step": 1270
},
{
"epoch": 0.883367839889579,
"grad_norm": 4.2202887535095215,
"learning_rate": 8.892291344661035e-06,
"loss": 0.5034,
"step": 1280
},
{
"epoch": 0.8902691511387164,
"grad_norm": 3.855746269226074,
"learning_rate": 8.86696203850038e-06,
"loss": 0.5174,
"step": 1290
},
{
"epoch": 0.8971704623878537,
"grad_norm": 4.842246055603027,
"learning_rate": 8.841383347741794e-06,
"loss": 0.5891,
"step": 1300
},
{
"epoch": 0.904071773636991,
"grad_norm": 4.951056957244873,
"learning_rate": 8.81555692198293e-06,
"loss": 0.5214,
"step": 1310
},
{
"epoch": 0.9109730848861284,
"grad_norm": 4.995665073394775,
"learning_rate": 8.789484426798148e-06,
"loss": 0.5777,
"step": 1320
},
{
"epoch": 0.9178743961352657,
"grad_norm": 6.235661506652832,
"learning_rate": 8.763167543631084e-06,
"loss": 0.5184,
"step": 1330
},
{
"epoch": 0.9247757073844031,
"grad_norm": 3.287186622619629,
"learning_rate": 8.736607969686231e-06,
"loss": 0.5442,
"step": 1340
},
{
"epoch": 0.9316770186335404,
"grad_norm": 6.245815753936768,
"learning_rate": 8.709807417819464e-06,
"loss": 0.5082,
"step": 1350
},
{
"epoch": 0.9385783298826778,
"grad_norm": 9.525437355041504,
"learning_rate": 8.682767616427598e-06,
"loss": 0.6263,
"step": 1360
},
{
"epoch": 0.945479641131815,
"grad_norm": 6.914644718170166,
"learning_rate": 8.655490309336905e-06,
"loss": 0.558,
"step": 1370
},
{
"epoch": 0.9523809523809523,
"grad_norm": 3.8063876628875732,
"learning_rate": 8.627977255690658e-06,
"loss": 0.5849,
"step": 1380
},
{
"epoch": 0.9592822636300897,
"grad_norm": 3.1215429306030273,
"learning_rate": 8.600230229835682e-06,
"loss": 0.4418,
"step": 1390
},
{
"epoch": 0.966183574879227,
"grad_norm": 5.066253185272217,
"learning_rate": 8.57225102120793e-06,
"loss": 0.4487,
"step": 1400
},
{
"epoch": 0.9730848861283644,
"grad_norm": 5.166496753692627,
"learning_rate": 8.544041434217067e-06,
"loss": 0.5146,
"step": 1410
},
{
"epoch": 0.9799861973775017,
"grad_norm": 4.333652496337891,
"learning_rate": 8.515603288130119e-06,
"loss": 0.592,
"step": 1420
},
{
"epoch": 0.9868875086266391,
"grad_norm": 4.584316730499268,
"learning_rate": 8.48693841695413e-06,
"loss": 0.5018,
"step": 1430
},
{
"epoch": 0.9937888198757764,
"grad_norm": 3.0310325622558594,
"learning_rate": 8.458048669317897e-06,
"loss": 0.4669,
"step": 1440
},
{
"epoch": 1.0006901311249137,
"grad_norm": 5.318614482879639,
"learning_rate": 8.42893590835274e-06,
"loss": 0.5024,
"step": 1450
},
{
"epoch": 1.007591442374051,
"grad_norm": 3.550750494003296,
"learning_rate": 8.399602011572356e-06,
"loss": 0.4485,
"step": 1460
},
{
"epoch": 1.0144927536231885,
"grad_norm": 3.6674296855926514,
"learning_rate": 8.370048870751732e-06,
"loss": 0.4645,
"step": 1470
},
{
"epoch": 1.0213940648723256,
"grad_norm": 8.107131004333496,
"learning_rate": 8.34027839180514e-06,
"loss": 0.5043,
"step": 1480
},
{
"epoch": 1.028295376121463,
"grad_norm": 3.9312679767608643,
"learning_rate": 8.310292494663222e-06,
"loss": 0.5097,
"step": 1490
},
{
"epoch": 1.0351966873706004,
"grad_norm": 29.156564712524414,
"learning_rate": 8.280093113149182e-06,
"loss": 0.5511,
"step": 1500
},
{
"epoch": 1.0420979986197378,
"grad_norm": 3.583965301513672,
"learning_rate": 8.249682194854059e-06,
"loss": 0.4414,
"step": 1510
},
{
"epoch": 1.048999309868875,
"grad_norm": 4.666857719421387,
"learning_rate": 8.219061701011125e-06,
"loss": 0.5604,
"step": 1520
},
{
"epoch": 1.0559006211180124,
"grad_norm": 2.8852901458740234,
"learning_rate": 8.188233606369422e-06,
"loss": 0.5155,
"step": 1530
},
{
"epoch": 1.0628019323671498,
"grad_norm": 6.806229591369629,
"learning_rate": 8.157199899066379e-06,
"loss": 0.5764,
"step": 1540
},
{
"epoch": 1.069703243616287,
"grad_norm": 3.5518808364868164,
"learning_rate": 8.125962580499616e-06,
"loss": 0.4269,
"step": 1550
},
{
"epoch": 1.0766045548654244,
"grad_norm": 7.634500980377197,
"learning_rate": 8.094523665197862e-06,
"loss": 0.503,
"step": 1560
},
{
"epoch": 1.0835058661145618,
"grad_norm": 4.027690887451172,
"learning_rate": 8.062885180691046e-06,
"loss": 0.5463,
"step": 1570
},
{
"epoch": 1.0904071773636992,
"grad_norm": 3.588782787322998,
"learning_rate": 8.03104916737953e-06,
"loss": 0.4588,
"step": 1580
},
{
"epoch": 1.0973084886128364,
"grad_norm": 4.588827133178711,
"learning_rate": 7.999017678402518e-06,
"loss": 0.4267,
"step": 1590
},
{
"epoch": 1.1042097998619738,
"grad_norm": 5.544867515563965,
"learning_rate": 7.96679277950566e-06,
"loss": 0.4986,
"step": 1600
},
{
"epoch": 1.1111111111111112,
"grad_norm": 3.672931671142578,
"learning_rate": 7.93437654890782e-06,
"loss": 0.5319,
"step": 1610
},
{
"epoch": 1.1180124223602483,
"grad_norm": 5.405056476593018,
"learning_rate": 7.901771077167051e-06,
"loss": 0.444,
"step": 1620
},
{
"epoch": 1.1249137336093857,
"grad_norm": 4.525319576263428,
"learning_rate": 7.868978467045777e-06,
"loss": 0.5528,
"step": 1630
},
{
"epoch": 1.1318150448585231,
"grad_norm": 2.5616278648376465,
"learning_rate": 7.836000833375174e-06,
"loss": 0.451,
"step": 1640
},
{
"epoch": 1.1387163561076605,
"grad_norm": 4.437591552734375,
"learning_rate": 7.802840302918794e-06,
"loss": 0.4389,
"step": 1650
},
{
"epoch": 1.1456176673567977,
"grad_norm": 4.839269161224365,
"learning_rate": 7.769499014235401e-06,
"loss": 0.4973,
"step": 1660
},
{
"epoch": 1.152518978605935,
"grad_norm": 3.4790279865264893,
"learning_rate": 7.735979117541054e-06,
"loss": 0.5495,
"step": 1670
},
{
"epoch": 1.1594202898550725,
"grad_norm": 4.62631893157959,
"learning_rate": 7.702282774570434e-06,
"loss": 0.4872,
"step": 1680
},
{
"epoch": 1.16632160110421,
"grad_norm": 10.294990539550781,
"learning_rate": 7.668412158437442e-06,
"loss": 0.4533,
"step": 1690
},
{
"epoch": 1.173222912353347,
"grad_norm": 6.42893648147583,
"learning_rate": 7.634369453495036e-06,
"loss": 0.4866,
"step": 1700
},
{
"epoch": 1.1801242236024845,
"grad_norm": 6.9018988609313965,
"learning_rate": 7.600156855194376e-06,
"loss": 0.4324,
"step": 1710
},
{
"epoch": 1.1870255348516219,
"grad_norm": 5.8724164962768555,
"learning_rate": 7.565776569943231e-06,
"loss": 0.4312,
"step": 1720
},
{
"epoch": 1.193926846100759,
"grad_norm": 5.410865306854248,
"learning_rate": 7.531230814963675e-06,
"loss": 0.4914,
"step": 1730
},
{
"epoch": 1.2008281573498965,
"grad_norm": 6.02437162399292,
"learning_rate": 7.496521818149114e-06,
"loss": 0.4368,
"step": 1740
},
{
"epoch": 1.2077294685990339,
"grad_norm": 5.565640926361084,
"learning_rate": 7.461651817920595e-06,
"loss": 0.6265,
"step": 1750
},
{
"epoch": 1.2146307798481712,
"grad_norm": 5.0998125076293945,
"learning_rate": 7.426623063082452e-06,
"loss": 0.4067,
"step": 1760
},
{
"epoch": 1.2215320910973084,
"grad_norm": 8.882064819335938,
"learning_rate": 7.391437812677274e-06,
"loss": 0.4788,
"step": 1770
},
{
"epoch": 1.2284334023464458,
"grad_norm": 7.057101726531982,
"learning_rate": 7.356098335840221e-06,
"loss": 0.4375,
"step": 1780
},
{
"epoch": 1.2353347135955832,
"grad_norm": 3.037950277328491,
"learning_rate": 7.3206069116526855e-06,
"loss": 0.5243,
"step": 1790
},
{
"epoch": 1.2422360248447206,
"grad_norm": 5.529163360595703,
"learning_rate": 7.2849658289953084e-06,
"loss": 0.5253,
"step": 1800
},
{
"epoch": 1.2491373360938578,
"grad_norm": 3.62673020362854,
"learning_rate": 7.249177386400369e-06,
"loss": 0.4805,
"step": 1810
},
{
"epoch": 1.2560386473429952,
"grad_norm": 5.269548416137695,
"learning_rate": 7.213243891903549e-06,
"loss": 0.4707,
"step": 1820
},
{
"epoch": 1.2629399585921326,
"grad_norm": 6.054306983947754,
"learning_rate": 7.177167662895087e-06,
"loss": 0.5421,
"step": 1830
},
{
"epoch": 1.2698412698412698,
"grad_norm": 12.915227890014648,
"learning_rate": 7.140951025970324e-06,
"loss": 0.4806,
"step": 1840
},
{
"epoch": 1.2767425810904072,
"grad_norm": 3.1157984733581543,
"learning_rate": 7.104596316779663e-06,
"loss": 0.524,
"step": 1850
},
{
"epoch": 1.2836438923395446,
"grad_norm": 3.958435297012329,
"learning_rate": 7.0681058798779335e-06,
"loss": 0.4982,
"step": 1860
},
{
"epoch": 1.2905452035886817,
"grad_norm": 3.1013059616088867,
"learning_rate": 7.031482068573201e-06,
"loss": 0.4501,
"step": 1870
},
{
"epoch": 1.2974465148378191,
"grad_norm": 3.420816659927368,
"learning_rate": 6.994727244774981e-06,
"loss": 0.4503,
"step": 1880
},
{
"epoch": 1.3043478260869565,
"grad_norm": 7.524533271789551,
"learning_rate": 6.957843778841937e-06,
"loss": 0.5178,
"step": 1890
},
{
"epoch": 1.311249137336094,
"grad_norm": 7.762824535369873,
"learning_rate": 6.920834049429002e-06,
"loss": 0.5071,
"step": 1900
},
{
"epoch": 1.3181504485852311,
"grad_norm": 3.4833388328552246,
"learning_rate": 6.88370044333397e-06,
"loss": 0.4669,
"step": 1910
},
{
"epoch": 1.3250517598343685,
"grad_norm": 4.939413547515869,
"learning_rate": 6.846445355343591e-06,
"loss": 0.4622,
"step": 1920
},
{
"epoch": 1.331953071083506,
"grad_norm": 3.8528079986572266,
"learning_rate": 6.8090711880791025e-06,
"loss": 0.45,
"step": 1930
},
{
"epoch": 1.3388543823326433,
"grad_norm": 5.8818793296813965,
"learning_rate": 6.771580351841304e-06,
"loss": 0.485,
"step": 1940
},
{
"epoch": 1.3457556935817805,
"grad_norm": 11.815837860107422,
"learning_rate": 6.733975264455097e-06,
"loss": 0.5142,
"step": 1950
},
{
"epoch": 1.3526570048309179,
"grad_norm": 6.753859519958496,
"learning_rate": 6.696258351113569e-06,
"loss": 0.5225,
"step": 1960
},
{
"epoch": 1.3595583160800553,
"grad_norm": 6.15470027923584,
"learning_rate": 6.658432044221584e-06,
"loss": 0.5448,
"step": 1970
},
{
"epoch": 1.3664596273291925,
"grad_norm": 6.573286533355713,
"learning_rate": 6.6204987832389115e-06,
"loss": 0.5493,
"step": 1980
},
{
"epoch": 1.3733609385783299,
"grad_norm": 5.581584453582764,
"learning_rate": 6.582461014522913e-06,
"loss": 0.486,
"step": 1990
},
{
"epoch": 1.3802622498274673,
"grad_norm": 7.525827407836914,
"learning_rate": 6.544321191170764e-06,
"loss": 0.4931,
"step": 2000
},
{
"epoch": 1.3871635610766044,
"grad_norm": 4.291807651519775,
"learning_rate": 6.5060817728612544e-06,
"loss": 0.4259,
"step": 2010
},
{
"epoch": 1.3940648723257418,
"grad_norm": 3.788956642150879,
"learning_rate": 6.467745225696164e-06,
"loss": 0.5021,
"step": 2020
},
{
"epoch": 1.4009661835748792,
"grad_norm": 7.374408721923828,
"learning_rate": 6.429314022041216e-06,
"loss": 0.4508,
"step": 2030
},
{
"epoch": 1.4078674948240166,
"grad_norm": 4.020993232727051,
"learning_rate": 6.3907906403666355e-06,
"loss": 0.4709,
"step": 2040
},
{
"epoch": 1.414768806073154,
"grad_norm": 4.260829925537109,
"learning_rate": 6.352177565087308e-06,
"loss": 0.4233,
"step": 2050
},
{
"epoch": 1.4216701173222912,
"grad_norm": 4.548175811767578,
"learning_rate": 6.313477286402557e-06,
"loss": 0.5209,
"step": 2060
},
{
"epoch": 1.4285714285714286,
"grad_norm": 3.74542498588562,
"learning_rate": 6.27469230013555e-06,
"loss": 0.4751,
"step": 2070
},
{
"epoch": 1.435472739820566,
"grad_norm": 5.57232141494751,
"learning_rate": 6.235825107572335e-06,
"loss": 0.4697,
"step": 2080
},
{
"epoch": 1.4423740510697032,
"grad_norm": 4.981284141540527,
"learning_rate": 6.196878215300538e-06,
"loss": 0.3933,
"step": 2090
},
{
"epoch": 1.4492753623188406,
"grad_norm": 3.621471643447876,
"learning_rate": 6.1578541350477076e-06,
"loss": 0.4174,
"step": 2100
},
{
"epoch": 1.456176673567978,
"grad_norm": 3.4967782497406006,
"learning_rate": 6.118755383519323e-06,
"loss": 0.5396,
"step": 2110
},
{
"epoch": 1.4630779848171152,
"grad_norm": 2.8297739028930664,
"learning_rate": 6.079584482236499e-06,
"loss": 0.431,
"step": 2120
},
{
"epoch": 1.4699792960662525,
"grad_norm": 4.409465789794922,
"learning_rate": 6.040343957373367e-06,
"loss": 0.4249,
"step": 2130
},
{
"epoch": 1.47688060731539,
"grad_norm": 6.355079174041748,
"learning_rate": 6.0010363395941585e-06,
"loss": 0.5098,
"step": 2140
},
{
"epoch": 1.4837819185645271,
"grad_norm": 7.394165515899658,
"learning_rate": 5.961664163890002e-06,
"loss": 0.5587,
"step": 2150
},
{
"epoch": 1.4906832298136645,
"grad_norm": 5.7639970779418945,
"learning_rate": 5.922229969415432e-06,
"loss": 0.4362,
"step": 2160
},
{
"epoch": 1.497584541062802,
"grad_norm": 5.568869590759277,
"learning_rate": 5.882736299324648e-06,
"loss": 0.4344,
"step": 2170
},
{
"epoch": 1.5044858523119393,
"grad_norm": 4.6455278396606445,
"learning_rate": 5.84318570060749e-06,
"loss": 0.4703,
"step": 2180
},
{
"epoch": 1.5113871635610767,
"grad_norm": 6.105690956115723,
"learning_rate": 5.803580723925193e-06,
"loss": 0.4452,
"step": 2190
},
{
"epoch": 1.518288474810214,
"grad_norm": 4.214017868041992,
"learning_rate": 5.7639239234458846e-06,
"loss": 0.4021,
"step": 2200
},
{
"epoch": 1.5251897860593513,
"grad_norm": 6.385672092437744,
"learning_rate": 5.724217856679859e-06,
"loss": 0.474,
"step": 2210
},
{
"epoch": 1.5320910973084887,
"grad_norm": 3.936816453933716,
"learning_rate": 5.6844650843146595e-06,
"loss": 0.4866,
"step": 2220
},
{
"epoch": 1.5389924085576259,
"grad_norm": 7.008480548858643,
"learning_rate": 5.644668170049913e-06,
"loss": 0.542,
"step": 2230
},
{
"epoch": 1.5458937198067633,
"grad_norm": 5.321537494659424,
"learning_rate": 5.604829680432012e-06,
"loss": 0.4752,
"step": 2240
},
{
"epoch": 1.5527950310559007,
"grad_norm": 3.9321799278259277,
"learning_rate": 5.564952184688588e-06,
"loss": 0.4947,
"step": 2250
},
{
"epoch": 1.5596963423050378,
"grad_norm": 4.093950271606445,
"learning_rate": 5.525038254562817e-06,
"loss": 0.5209,
"step": 2260
},
{
"epoch": 1.5665976535541752,
"grad_norm": 5.226855278015137,
"learning_rate": 5.485090464147571e-06,
"loss": 0.5588,
"step": 2270
},
{
"epoch": 1.5734989648033126,
"grad_norm": 3.595454216003418,
"learning_rate": 5.445111389719408e-06,
"loss": 0.431,
"step": 2280
},
{
"epoch": 1.5804002760524498,
"grad_norm": 5.264852523803711,
"learning_rate": 5.405103609572427e-06,
"loss": 0.49,
"step": 2290
},
{
"epoch": 1.5873015873015874,
"grad_norm": 6.781251430511475,
"learning_rate": 5.3650697038519885e-06,
"loss": 0.5021,
"step": 2300
},
{
"epoch": 1.5942028985507246,
"grad_norm": 5.205682277679443,
"learning_rate": 5.32501225438832e-06,
"loss": 0.4516,
"step": 2310
},
{
"epoch": 1.601104209799862,
"grad_norm": 3.0770986080169678,
"learning_rate": 5.284933844530013e-06,
"loss": 0.4717,
"step": 2320
},
{
"epoch": 1.6080055210489994,
"grad_norm": 9.906235694885254,
"learning_rate": 5.244837058977413e-06,
"loss": 0.4778,
"step": 2330
},
{
"epoch": 1.6149068322981366,
"grad_norm": 4.150860786437988,
"learning_rate": 5.204724483615941e-06,
"loss": 0.4752,
"step": 2340
},
{
"epoch": 1.621808143547274,
"grad_norm": 4.955432891845703,
"learning_rate": 5.164598705349313e-06,
"loss": 0.4561,
"step": 2350
},
{
"epoch": 1.6287094547964114,
"grad_norm": 5.098195552825928,
"learning_rate": 5.124462311932719e-06,
"loss": 0.4592,
"step": 2360
},
{
"epoch": 1.6356107660455486,
"grad_norm": 7.2271952629089355,
"learning_rate": 5.084317891805928e-06,
"loss": 0.4418,
"step": 2370
},
{
"epoch": 1.642512077294686,
"grad_norm": 3.4083011150360107,
"learning_rate": 5.0441680339263655e-06,
"loss": 0.5026,
"step": 2380
},
{
"epoch": 1.6494133885438234,
"grad_norm": 5.3739914894104,
"learning_rate": 5.004015327602139e-06,
"loss": 0.4333,
"step": 2390
},
{
"epoch": 1.6563146997929605,
"grad_norm": 8.130451202392578,
"learning_rate": 4.963862362325058e-06,
"loss": 0.4922,
"step": 2400
},
{
"epoch": 1.6632160110420982,
"grad_norm": 11.71609115600586,
"learning_rate": 4.923711727603632e-06,
"loss": 0.4559,
"step": 2410
},
{
"epoch": 1.6701173222912353,
"grad_norm": 5.356748104095459,
"learning_rate": 4.883566012796068e-06,
"loss": 0.4993,
"step": 2420
},
{
"epoch": 1.6770186335403725,
"grad_norm": 6.024771690368652,
"learning_rate": 4.843427806943283e-06,
"loss": 0.4538,
"step": 2430
},
{
"epoch": 1.6839199447895101,
"grad_norm": 3.9793906211853027,
"learning_rate": 4.803299698601935e-06,
"loss": 0.5609,
"step": 2440
},
{
"epoch": 1.6908212560386473,
"grad_norm": 4.044511795043945,
"learning_rate": 4.763184275677484e-06,
"loss": 0.5378,
"step": 2450
},
{
"epoch": 1.6977225672877847,
"grad_norm": 3.551792860031128,
"learning_rate": 4.723084125257291e-06,
"loss": 0.4701,
"step": 2460
},
{
"epoch": 1.704623878536922,
"grad_norm": 5.491049766540527,
"learning_rate": 4.683001833443774e-06,
"loss": 0.436,
"step": 2470
},
{
"epoch": 1.7115251897860593,
"grad_norm": 3.685840368270874,
"learning_rate": 4.642939985187634e-06,
"loss": 0.478,
"step": 2480
},
{
"epoch": 1.7184265010351967,
"grad_norm": 6.465378284454346,
"learning_rate": 4.602901164121145e-06,
"loss": 0.479,
"step": 2490
},
{
"epoch": 1.725327812284334,
"grad_norm": 5.6965556144714355,
"learning_rate": 4.562887952391531e-06,
"loss": 0.5081,
"step": 2500
},
{
"epoch": 1.7322291235334712,
"grad_norm": 2.0033233165740967,
"learning_rate": 4.522902930494442e-06,
"loss": 0.4958,
"step": 2510
},
{
"epoch": 1.7391304347826086,
"grad_norm": 4.223212242126465,
"learning_rate": 4.482948677107533e-06,
"loss": 0.4717,
"step": 2520
},
{
"epoch": 1.746031746031746,
"grad_norm": 4.229642391204834,
"learning_rate": 4.443027768924168e-06,
"loss": 0.4457,
"step": 2530
},
{
"epoch": 1.7529330572808832,
"grad_norm": 5.072473049163818,
"learning_rate": 4.40314278048724e-06,
"loss": 0.4221,
"step": 2540
},
{
"epoch": 1.7598343685300208,
"grad_norm": 6.257089614868164,
"learning_rate": 4.363296284023139e-06,
"loss": 0.4509,
"step": 2550
},
{
"epoch": 1.766735679779158,
"grad_norm": 7.792009353637695,
"learning_rate": 4.323490849275868e-06,
"loss": 0.4681,
"step": 2560
},
{
"epoch": 1.7736369910282954,
"grad_norm": 8.221329689025879,
"learning_rate": 4.283729043341308e-06,
"loss": 0.493,
"step": 2570
},
{
"epoch": 1.7805383022774328,
"grad_norm": 7.550902366638184,
"learning_rate": 4.244013430501681e-06,
"loss": 0.4632,
"step": 2580
},
{
"epoch": 1.78743961352657,
"grad_norm": 6.375574588775635,
"learning_rate": 4.204346572060156e-06,
"loss": 0.4653,
"step": 2590
},
{
"epoch": 1.7943409247757074,
"grad_norm": 2.9537665843963623,
"learning_rate": 4.164731026175682e-06,
"loss": 0.3974,
"step": 2600
},
{
"epoch": 1.8012422360248448,
"grad_norm": 5.192391395568848,
"learning_rate": 4.125169347698009e-06,
"loss": 0.4684,
"step": 2610
},
{
"epoch": 1.808143547273982,
"grad_norm": 5.661397457122803,
"learning_rate": 4.085664088002911e-06,
"loss": 0.4299,
"step": 2620
},
{
"epoch": 1.8150448585231194,
"grad_norm": 3.6837475299835205,
"learning_rate": 4.046217794827659e-06,
"loss": 0.4464,
"step": 2630
},
{
"epoch": 1.8219461697722568,
"grad_norm": 4.128551483154297,
"learning_rate": 4.006833012106707e-06,
"loss": 0.463,
"step": 2640
},
{
"epoch": 1.828847481021394,
"grad_norm": 5.615044593811035,
"learning_rate": 3.967512279807633e-06,
"loss": 0.4455,
"step": 2650
},
{
"epoch": 1.8357487922705316,
"grad_norm": 11.22446060180664,
"learning_rate": 3.928258133767336e-06,
"loss": 0.5461,
"step": 2660
},
{
"epoch": 1.8426501035196687,
"grad_norm": 3.3922340869903564,
"learning_rate": 3.889073105528489e-06,
"loss": 0.5,
"step": 2670
},
{
"epoch": 1.849551414768806,
"grad_norm": 8.71829605102539,
"learning_rate": 3.849959722176287e-06,
"loss": 0.5477,
"step": 2680
},
{
"epoch": 1.8564527260179435,
"grad_norm": 16.89905548095703,
"learning_rate": 3.8109205061754657e-06,
"loss": 0.4661,
"step": 2690
},
{
"epoch": 1.8633540372670807,
"grad_norm": 5.0564165115356445,
"learning_rate": 3.77195797520763e-06,
"loss": 0.4295,
"step": 2700
},
{
"epoch": 1.870255348516218,
"grad_norm": 6.244871616363525,
"learning_rate": 3.7330746420088837e-06,
"loss": 0.485,
"step": 2710
},
{
"epoch": 1.8771566597653555,
"grad_norm": 3.3127167224884033,
"learning_rate": 3.6942730142077756e-06,
"loss": 0.481,
"step": 2720
},
{
"epoch": 1.8840579710144927,
"grad_norm": 4.086329460144043,
"learning_rate": 3.655555594163587e-06,
"loss": 0.4406,
"step": 2730
},
{
"epoch": 1.89095928226363,
"grad_norm": 6.198220252990723,
"learning_rate": 3.61692487880495e-06,
"loss": 0.4664,
"step": 2740
},
{
"epoch": 1.8978605935127675,
"grad_norm": 6.655552864074707,
"learning_rate": 3.5783833594688177e-06,
"loss": 0.4329,
"step": 2750
},
{
"epoch": 1.9047619047619047,
"grad_norm": 7.991542339324951,
"learning_rate": 3.5399335217397936e-06,
"loss": 0.46,
"step": 2760
},
{
"epoch": 1.911663216011042,
"grad_norm": 5.558722496032715,
"learning_rate": 3.5015778452898306e-06,
"loss": 0.4601,
"step": 2770
},
{
"epoch": 1.9185645272601795,
"grad_norm": 4.689853668212891,
"learning_rate": 3.463318803718323e-06,
"loss": 0.4177,
"step": 2780
},
{
"epoch": 1.9254658385093166,
"grad_norm": 3.181793451309204,
"learning_rate": 3.4251588643925757e-06,
"loss": 0.4102,
"step": 2790
},
{
"epoch": 1.9323671497584543,
"grad_norm": 3.8942883014678955,
"learning_rate": 3.3871004882886805e-06,
"loss": 0.4638,
"step": 2800
},
{
"epoch": 1.9392684610075914,
"grad_norm": 2.4674880504608154,
"learning_rate": 3.3491461298328064e-06,
"loss": 0.4704,
"step": 2810
},
{
"epoch": 1.9461697722567288,
"grad_norm": 2.6942079067230225,
"learning_rate": 3.3112982367429105e-06,
"loss": 0.468,
"step": 2820
},
{
"epoch": 1.9530710835058662,
"grad_norm": 5.350464344024658,
"learning_rate": 3.2735592498708818e-06,
"loss": 0.4399,
"step": 2830
},
{
"epoch": 1.9599723947550034,
"grad_norm": 3.550755739212036,
"learning_rate": 3.2359316030451327e-06,
"loss": 0.4392,
"step": 2840
},
{
"epoch": 1.9668737060041408,
"grad_norm": 3.441821575164795,
"learning_rate": 3.1984177229136287e-06,
"loss": 0.4772,
"step": 2850
},
{
"epoch": 1.9737750172532782,
"grad_norm": 4.649478435516357,
"learning_rate": 3.161020028787405e-06,
"loss": 0.5217,
"step": 2860
},
{
"epoch": 1.9806763285024154,
"grad_norm": 4.30385684967041,
"learning_rate": 3.1237409324845224e-06,
"loss": 0.4551,
"step": 2870
},
{
"epoch": 1.9875776397515528,
"grad_norm": 3.710928440093994,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.4438,
"step": 2880
},
{
"epoch": 1.9944789510006902,
"grad_norm": 8.352008819580078,
"learning_rate": 3.049548142223506e-06,
"loss": 0.5109,
"step": 2890
},
{
"epoch": 2.0013802622498273,
"grad_norm": 7.670598030090332,
"learning_rate": 3.0126392330393033e-06,
"loss": 0.4793,
"step": 2900
},
{
"epoch": 2.008281573498965,
"grad_norm": 5.645792484283447,
"learning_rate": 2.9758584909177434e-06,
"loss": 0.467,
"step": 2910
},
{
"epoch": 2.015182884748102,
"grad_norm": 5.914472579956055,
"learning_rate": 2.9392082878889804e-06,
"loss": 0.4541,
"step": 2920
},
{
"epoch": 2.0220841959972393,
"grad_norm": 3.6343367099761963,
"learning_rate": 2.9026909875645693e-06,
"loss": 0.3755,
"step": 2930
},
{
"epoch": 2.028985507246377,
"grad_norm": 4.04212760925293,
"learning_rate": 2.8663089449850246e-06,
"loss": 0.3841,
"step": 2940
},
{
"epoch": 2.035886818495514,
"grad_norm": 5.742480278015137,
"learning_rate": 2.8300645064679345e-06,
"loss": 0.3529,
"step": 2950
},
{
"epoch": 2.0427881297446513,
"grad_norm": 3.632901906967163,
"learning_rate": 2.79396000945666e-06,
"loss": 0.4137,
"step": 2960
},
{
"epoch": 2.049689440993789,
"grad_norm": 5.725683212280273,
"learning_rate": 2.7579977823695727e-06,
"loss": 0.4445,
"step": 2970
},
{
"epoch": 2.056590752242926,
"grad_norm": 3.75038480758667,
"learning_rate": 2.7221801444499104e-06,
"loss": 0.4638,
"step": 2980
},
{
"epoch": 2.0634920634920633,
"grad_norm": 4.58835506439209,
"learning_rate": 2.68650940561619e-06,
"loss": 0.3893,
"step": 2990
},
{
"epoch": 2.070393374741201,
"grad_norm": 3.112630605697632,
"learning_rate": 2.6509878663132514e-06,
"loss": 0.4373,
"step": 3000
},
{
"epoch": 2.077294685990338,
"grad_norm": 4.088521957397461,
"learning_rate": 2.6156178173638978e-06,
"loss": 0.4711,
"step": 3010
},
{
"epoch": 2.0841959972394757,
"grad_norm": 8.322639465332031,
"learning_rate": 2.58040153982114e-06,
"loss": 0.3957,
"step": 3020
},
{
"epoch": 2.091097308488613,
"grad_norm": 21.649370193481445,
"learning_rate": 2.5453413048211175e-06,
"loss": 0.4576,
"step": 3030
},
{
"epoch": 2.09799861973775,
"grad_norm": 5.0623016357421875,
"learning_rate": 2.5104393734366113e-06,
"loss": 0.4235,
"step": 3040
},
{
"epoch": 2.1048999309868877,
"grad_norm": 11.47243595123291,
"learning_rate": 2.4756979965312262e-06,
"loss": 0.5128,
"step": 3050
},
{
"epoch": 2.111801242236025,
"grad_norm": 4.394364356994629,
"learning_rate": 2.441119414614241e-06,
"loss": 0.4797,
"step": 3060
},
{
"epoch": 2.118702553485162,
"grad_norm": 5.27672815322876,
"learning_rate": 2.406705857696099e-06,
"loss": 0.5014,
"step": 3070
},
{
"epoch": 2.1256038647342996,
"grad_norm": 9.366743087768555,
"learning_rate": 2.3724595451446104e-06,
"loss": 0.4434,
"step": 3080
},
{
"epoch": 2.132505175983437,
"grad_norm": 4.564538478851318,
"learning_rate": 2.3383826855418064e-06,
"loss": 0.4565,
"step": 3090
},
{
"epoch": 2.139406487232574,
"grad_norm": 3.357450246810913,
"learning_rate": 2.3044774765415172e-06,
"loss": 0.4243,
"step": 3100
},
{
"epoch": 2.1463077984817116,
"grad_norm": 2.929481267929077,
"learning_rate": 2.2707461047276395e-06,
"loss": 0.4194,
"step": 3110
},
{
"epoch": 2.153209109730849,
"grad_norm": 4.902336597442627,
"learning_rate": 2.2371907454731168e-06,
"loss": 0.4489,
"step": 3120
},
{
"epoch": 2.1601104209799864,
"grad_norm": 8.501842498779297,
"learning_rate": 2.2038135627996498e-06,
"loss": 0.4407,
"step": 3130
},
{
"epoch": 2.1670117322291236,
"grad_norm": 4.641388416290283,
"learning_rate": 2.1706167092381416e-06,
"loss": 0.4422,
"step": 3140
},
{
"epoch": 2.1739130434782608,
"grad_norm": 4.227949142456055,
"learning_rate": 2.137602325689873e-06,
"loss": 0.4713,
"step": 3150
},
{
"epoch": 2.1808143547273984,
"grad_norm": 4.0596771240234375,
"learning_rate": 2.104772541288434e-06,
"loss": 0.4662,
"step": 3160
},
{
"epoch": 2.1877156659765356,
"grad_norm": 2.5764706134796143,
"learning_rate": 2.0721294732624113e-06,
"loss": 0.4678,
"step": 3170
},
{
"epoch": 2.1946169772256727,
"grad_norm": 4.290266513824463,
"learning_rate": 2.039675226798854e-06,
"loss": 0.4344,
"step": 3180
},
{
"epoch": 2.2015182884748103,
"grad_norm": 4.121200084686279,
"learning_rate": 2.0074118949075024e-06,
"loss": 0.3841,
"step": 3190
},
{
"epoch": 2.2084195997239475,
"grad_norm": 6.931027412414551,
"learning_rate": 1.9753415582858026e-06,
"loss": 0.4366,
"step": 3200
},
{
"epoch": 2.2153209109730847,
"grad_norm": 4.345970630645752,
"learning_rate": 1.9434662851847325e-06,
"loss": 0.4317,
"step": 3210
},
{
"epoch": 2.2222222222222223,
"grad_norm": 3.9762792587280273,
"learning_rate": 1.911788131275406e-06,
"loss": 0.4997,
"step": 3220
},
{
"epoch": 2.2291235334713595,
"grad_norm": 3.652177572250366,
"learning_rate": 1.8803091395165046e-06,
"loss": 0.4822,
"step": 3230
},
{
"epoch": 2.2360248447204967,
"grad_norm": 3.2718632221221924,
"learning_rate": 1.8490313400225296e-06,
"loss": 0.3906,
"step": 3240
},
{
"epoch": 2.2429261559696343,
"grad_norm": 5.452423572540283,
"learning_rate": 1.817956749932873e-06,
"loss": 0.5147,
"step": 3250
},
{
"epoch": 2.2498274672187715,
"grad_norm": 7.504180908203125,
"learning_rate": 1.787087373281729e-06,
"loss": 0.5492,
"step": 3260
},
{
"epoch": 2.256728778467909,
"grad_norm": 4.878859043121338,
"learning_rate": 1.7564252008688514e-06,
"loss": 0.4429,
"step": 3270
},
{
"epoch": 2.2636300897170463,
"grad_norm": 10.33424186706543,
"learning_rate": 1.725972210131172e-06,
"loss": 0.4669,
"step": 3280
},
{
"epoch": 2.2705314009661834,
"grad_norm": 5.427865982055664,
"learning_rate": 1.6957303650152674e-06,
"loss": 0.439,
"step": 3290
},
{
"epoch": 2.277432712215321,
"grad_norm": 3.2975456714630127,
"learning_rate": 1.6657016158506966e-06,
"loss": 0.4173,
"step": 3300
},
{
"epoch": 2.2843340234644582,
"grad_norm": 5.5312652587890625,
"learning_rate": 1.6358878992242377e-06,
"loss": 0.4418,
"step": 3310
},
{
"epoch": 2.2912353347135954,
"grad_norm": 8.227662086486816,
"learning_rate": 1.6062911378549768e-06,
"loss": 0.5331,
"step": 3320
},
{
"epoch": 2.298136645962733,
"grad_norm": 12.097111701965332,
"learning_rate": 1.5769132404703274e-06,
"loss": 0.4287,
"step": 3330
},
{
"epoch": 2.30503795721187,
"grad_norm": 14.756723403930664,
"learning_rate": 1.5477561016829185e-06,
"loss": 0.4447,
"step": 3340
},
{
"epoch": 2.311939268461008,
"grad_norm": 3.6619608402252197,
"learning_rate": 1.518821601868421e-06,
"loss": 0.4515,
"step": 3350
},
{
"epoch": 2.318840579710145,
"grad_norm": 4.210367679595947,
"learning_rate": 1.4901116070442772e-06,
"loss": 0.4171,
"step": 3360
},
{
"epoch": 2.325741890959282,
"grad_norm": 6.9380903244018555,
"learning_rate": 1.4616279687493528e-06,
"loss": 0.4402,
"step": 3370
},
{
"epoch": 2.33264320220842,
"grad_norm": 5.626720428466797,
"learning_rate": 1.4333725239245355e-06,
"loss": 0.3809,
"step": 3380
},
{
"epoch": 2.339544513457557,
"grad_norm": 3.1848628520965576,
"learning_rate": 1.4053470947942694e-06,
"loss": 0.4753,
"step": 3390
},
{
"epoch": 2.346445824706694,
"grad_norm": 5.034173965454102,
"learning_rate": 1.3775534887490365e-06,
"loss": 0.5457,
"step": 3400
},
{
"epoch": 2.353347135955832,
"grad_norm": 3.3620548248291016,
"learning_rate": 1.3499934982287922e-06,
"loss": 0.3679,
"step": 3410
},
{
"epoch": 2.360248447204969,
"grad_norm": 3.6021339893341064,
"learning_rate": 1.3226689006073712e-06,
"loss": 0.4447,
"step": 3420
},
{
"epoch": 2.367149758454106,
"grad_norm": 8.112746238708496,
"learning_rate": 1.2955814580778686e-06,
"loss": 0.4838,
"step": 3430
},
{
"epoch": 2.3740510697032438,
"grad_norm": 4.7525248527526855,
"learning_rate": 1.2687329175389835e-06,
"loss": 0.3782,
"step": 3440
},
{
"epoch": 2.380952380952381,
"grad_norm": 3.802643299102783,
"learning_rate": 1.2421250104823683e-06,
"loss": 0.4681,
"step": 3450
},
{
"epoch": 2.387853692201518,
"grad_norm": 4.141634464263916,
"learning_rate": 1.2157594528809619e-06,
"loss": 0.4475,
"step": 3460
},
{
"epoch": 2.3947550034506557,
"grad_norm": 3.0190589427948,
"learning_rate": 1.1896379450783174e-06,
"loss": 0.3805,
"step": 3470
},
{
"epoch": 2.401656314699793,
"grad_norm": 3.8419723510742188,
"learning_rate": 1.163762171678951e-06,
"loss": 0.4576,
"step": 3480
},
{
"epoch": 2.40855762594893,
"grad_norm": 3.847914218902588,
"learning_rate": 1.1381338014397035e-06,
"loss": 0.4412,
"step": 3490
},
{
"epoch": 2.4154589371980677,
"grad_norm": 4.5137481689453125,
"learning_rate": 1.1127544871621144e-06,
"loss": 0.45,
"step": 3500
},
{
"epoch": 2.422360248447205,
"grad_norm": 2.8401100635528564,
"learning_rate": 1.0876258655858307e-06,
"loss": 0.5054,
"step": 3510
},
{
"epoch": 2.4292615596963425,
"grad_norm": 4.151172637939453,
"learning_rate": 1.0627495572830537e-06,
"loss": 0.4,
"step": 3520
},
{
"epoch": 2.4361628709454797,
"grad_norm": 6.4405837059021,
"learning_rate": 1.0381271665540294e-06,
"loss": 0.4465,
"step": 3530
},
{
"epoch": 2.443064182194617,
"grad_norm": 4.475886821746826,
"learning_rate": 1.0137602813235824e-06,
"loss": 0.4136,
"step": 3540
},
{
"epoch": 2.4499654934437545,
"grad_norm": 4.4384355545043945,
"learning_rate": 9.896504730387059e-07,
"loss": 0.4242,
"step": 3550
},
{
"epoch": 2.4568668046928916,
"grad_norm": 4.164675712585449,
"learning_rate": 9.657992965672242e-07,
"loss": 0.4119,
"step": 3560
},
{
"epoch": 2.463768115942029,
"grad_norm": 4.215827465057373,
"learning_rate": 9.422082900975105e-07,
"loss": 0.5137,
"step": 3570
},
{
"epoch": 2.4706694271911664,
"grad_norm": 4.377105236053467,
"learning_rate": 9.188789750392951e-07,
"loss": 0.4474,
"step": 3580
},
{
"epoch": 2.4775707384403036,
"grad_norm": 7.331305503845215,
"learning_rate": 8.958128559255397e-07,
"loss": 0.4678,
"step": 3590
},
{
"epoch": 2.4844720496894412,
"grad_norm": 2.0584566593170166,
"learning_rate": 8.730114203154144e-07,
"loss": 0.3597,
"step": 3600
},
{
"epoch": 2.4913733609385784,
"grad_norm": 7.080709934234619,
"learning_rate": 8.504761386983651e-07,
"loss": 0.5171,
"step": 3610
},
{
"epoch": 2.4982746721877156,
"grad_norm": 3.9697988033294678,
"learning_rate": 8.282084643992677e-07,
"loss": 0.3977,
"step": 3620
},
{
"epoch": 2.505175983436853,
"grad_norm": 4.740238666534424,
"learning_rate": 8.062098334847185e-07,
"loss": 0.4319,
"step": 3630
},
{
"epoch": 2.5120772946859904,
"grad_norm": 3.5427558422088623,
"learning_rate": 7.844816646704096e-07,
"loss": 0.3772,
"step": 3640
},
{
"epoch": 2.5189786059351276,
"grad_norm": 3.859440803527832,
"learning_rate": 7.630253592296388e-07,
"loss": 0.5061,
"step": 3650
},
{
"epoch": 2.525879917184265,
"grad_norm": 6.630671501159668,
"learning_rate": 7.418423009029363e-07,
"loss": 0.3935,
"step": 3660
},
{
"epoch": 2.5327812284334024,
"grad_norm": 5.899334907531738,
"learning_rate": 7.209338558088275e-07,
"loss": 0.4222,
"step": 3670
},
{
"epoch": 2.5396825396825395,
"grad_norm": 3.2894980907440186,
"learning_rate": 7.003013723557356e-07,
"loss": 0.3958,
"step": 3680
},
{
"epoch": 2.546583850931677,
"grad_norm": 6.762352466583252,
"learning_rate": 6.799461811550101e-07,
"loss": 0.4644,
"step": 3690
},
{
"epoch": 2.5534851621808143,
"grad_norm": 2.811450481414795,
"learning_rate": 6.59869594935127e-07,
"loss": 0.5311,
"step": 3700
},
{
"epoch": 2.5603864734299515,
"grad_norm": 4.0911664962768555,
"learning_rate": 6.400729084570212e-07,
"loss": 0.499,
"step": 3710
},
{
"epoch": 2.567287784679089,
"grad_norm": 2.0677435398101807,
"learning_rate": 6.205573984305857e-07,
"loss": 0.4363,
"step": 3720
},
{
"epoch": 2.5741890959282263,
"grad_norm": 6.652555465698242,
"learning_rate": 6.013243234323396e-07,
"loss": 0.4329,
"step": 3730
},
{
"epoch": 2.5810904071773635,
"grad_norm": 5.920029163360596,
"learning_rate": 5.823749238242593e-07,
"loss": 0.4332,
"step": 3740
},
{
"epoch": 2.587991718426501,
"grad_norm": 4.426562309265137,
"learning_rate": 5.637104216737871e-07,
"loss": 0.3951,
"step": 3750
},
{
"epoch": 2.5948930296756383,
"grad_norm": 4.995473384857178,
"learning_rate": 5.453320206750152e-07,
"loss": 0.379,
"step": 3760
},
{
"epoch": 2.6017943409247755,
"grad_norm": 10.261146545410156,
"learning_rate": 5.272409060710598e-07,
"loss": 0.4505,
"step": 3770
},
{
"epoch": 2.608695652173913,
"grad_norm": 7.172719478607178,
"learning_rate": 5.094382445776281e-07,
"loss": 0.5054,
"step": 3780
},
{
"epoch": 2.6155969634230503,
"grad_norm": 8.207676887512207,
"learning_rate": 4.919251843077699e-07,
"loss": 0.4809,
"step": 3790
},
{
"epoch": 2.622498274672188,
"grad_norm": 2.518497943878174,
"learning_rate": 4.7470285469783493e-07,
"loss": 0.4353,
"step": 3800
},
{
"epoch": 2.629399585921325,
"grad_norm": 4.08033561706543,
"learning_rate": 4.5777236643463787e-07,
"loss": 0.4148,
"step": 3810
},
{
"epoch": 2.6363008971704622,
"grad_norm": 5.848834037780762,
"learning_rate": 4.411348113838243e-07,
"loss": 0.4687,
"step": 3820
},
{
"epoch": 2.6432022084196,
"grad_norm": 3.7531790733337402,
"learning_rate": 4.247912625194611e-07,
"loss": 0.4627,
"step": 3830
},
{
"epoch": 2.650103519668737,
"grad_norm": 9.258443832397461,
"learning_rate": 4.087427738548322e-07,
"loss": 0.4366,
"step": 3840
},
{
"epoch": 2.6570048309178746,
"grad_norm": 8.345389366149902,
"learning_rate": 3.9299038037446837e-07,
"loss": 0.4869,
"step": 3850
},
{
"epoch": 2.663906142167012,
"grad_norm": 3.5337610244750977,
"learning_rate": 3.775350979674025e-07,
"loss": 0.4491,
"step": 3860
},
{
"epoch": 2.670807453416149,
"grad_norm": 7.077737808227539,
"learning_rate": 3.6237792336164424e-07,
"loss": 0.4337,
"step": 3870
},
{
"epoch": 2.6777087646652866,
"grad_norm": 3.8632216453552246,
"learning_rate": 3.4751983405991007e-07,
"loss": 0.4162,
"step": 3880
},
{
"epoch": 2.684610075914424,
"grad_norm": 3.5121543407440186,
"learning_rate": 3.329617882765807e-07,
"loss": 0.3828,
"step": 3890
},
{
"epoch": 2.691511387163561,
"grad_norm": 5.687404155731201,
"learning_rate": 3.1870472487589845e-07,
"loss": 0.4537,
"step": 3900
},
{
"epoch": 2.6984126984126986,
"grad_norm": 5.270540237426758,
"learning_rate": 3.0474956331142837e-07,
"loss": 0.3759,
"step": 3910
},
{
"epoch": 2.7053140096618358,
"grad_norm": 3.591524600982666,
"learning_rate": 2.910972035667531e-07,
"loss": 0.3694,
"step": 3920
},
{
"epoch": 2.712215320910973,
"grad_norm": 7.048372268676758,
"learning_rate": 2.7774852609743887e-07,
"loss": 0.4195,
"step": 3930
},
{
"epoch": 2.7191166321601106,
"grad_norm": 4.01481294631958,
"learning_rate": 2.6470439177424754e-07,
"loss": 0.3936,
"step": 3940
},
{
"epoch": 2.7260179434092477,
"grad_norm": 3.495352268218994,
"learning_rate": 2.5196564182762506e-07,
"loss": 0.4135,
"step": 3950
},
{
"epoch": 2.732919254658385,
"grad_norm": 4.340942859649658,
"learning_rate": 2.39533097793444e-07,
"loss": 0.3933,
"step": 3960
},
{
"epoch": 2.7398205659075225,
"grad_norm": 4.2435455322265625,
"learning_rate": 2.2740756146002409e-07,
"loss": 0.4473,
"step": 3970
},
{
"epoch": 2.7467218771566597,
"grad_norm": 4.400971412658691,
"learning_rate": 2.155898148164226e-07,
"loss": 0.5041,
"step": 3980
},
{
"epoch": 2.753623188405797,
"grad_norm": 7.734391689300537,
"learning_rate": 2.0408062000200623e-07,
"loss": 0.5217,
"step": 3990
},
{
"epoch": 2.7605244996549345,
"grad_norm": 9.11950397491455,
"learning_rate": 1.9288071925729735e-07,
"loss": 0.4603,
"step": 4000
},
{
"epoch": 2.7674258109040717,
"grad_norm": 3.3895680904388428,
"learning_rate": 1.8199083487610514e-07,
"loss": 0.4142,
"step": 4010
},
{
"epoch": 2.774327122153209,
"grad_norm": 5.121877670288086,
"learning_rate": 1.714116691589457e-07,
"loss": 0.4673,
"step": 4020
},
{
"epoch": 2.7812284334023465,
"grad_norm": 5.084831237792969,
"learning_rate": 1.6114390436775095e-07,
"loss": 0.3851,
"step": 4030
},
{
"epoch": 2.7881297446514837,
"grad_norm": 5.831473350524902,
"learning_rate": 1.511882026818684e-07,
"loss": 0.4661,
"step": 4040
},
{
"epoch": 2.795031055900621,
"grad_norm": 6.206417083740234,
"learning_rate": 1.4154520615535185e-07,
"loss": 0.3826,
"step": 4050
},
{
"epoch": 2.8019323671497585,
"grad_norm": 7.832612991333008,
"learning_rate": 1.3221553667556342e-07,
"loss": 0.4691,
"step": 4060
},
{
"epoch": 2.8088336783988956,
"grad_norm": 4.907148361206055,
"learning_rate": 1.2319979592305798e-07,
"loss": 0.4069,
"step": 4070
},
{
"epoch": 2.8157349896480333,
"grad_norm": 4.754528999328613,
"learning_rate": 1.1449856533278624e-07,
"loss": 0.4042,
"step": 4080
},
{
"epoch": 2.8226363008971704,
"grad_norm": 4.072829723358154,
"learning_rate": 1.0611240605659601e-07,
"loss": 0.4023,
"step": 4090
},
{
"epoch": 2.829537612146308,
"grad_norm": 13.20451831817627,
"learning_rate": 9.804185892704376e-08,
"loss": 0.4824,
"step": 4100
},
{
"epoch": 2.8364389233954452,
"grad_norm": 6.202005863189697,
"learning_rate": 9.028744442251103e-08,
"loss": 0.3975,
"step": 4110
},
{
"epoch": 2.8433402346445824,
"grad_norm": 3.3089945316314697,
"learning_rate": 8.284966263364447e-08,
"loss": 0.4158,
"step": 4120
},
{
"epoch": 2.85024154589372,
"grad_norm": 5.233531951904297,
"learning_rate": 7.572899323110116e-08,
"loss": 0.4356,
"step": 4130
},
{
"epoch": 2.857142857142857,
"grad_norm": 4.851590156555176,
"learning_rate": 6.892589543461392e-08,
"loss": 0.4523,
"step": 4140
},
{
"epoch": 2.8640441683919944,
"grad_norm": 5.297482490539551,
"learning_rate": 6.244080798337715e-08,
"loss": 0.4441,
"step": 4150
},
{
"epoch": 2.870945479641132,
"grad_norm": 4.253460884094238,
"learning_rate": 5.627414910775175e-08,
"loss": 0.3956,
"step": 4160
},
{
"epoch": 2.877846790890269,
"grad_norm": 4.016265869140625,
"learning_rate": 5.042631650229057e-08,
"loss": 0.4225,
"step": 4170
},
{
"epoch": 2.8847481021394064,
"grad_norm": 3.6404125690460205,
"learning_rate": 4.489768730009558e-08,
"loss": 0.4694,
"step": 4180
},
{
"epoch": 2.891649413388544,
"grad_norm": 4.740210056304932,
"learning_rate": 3.968861804849067e-08,
"loss": 0.4486,
"step": 4190
},
{
"epoch": 2.898550724637681,
"grad_norm": 3.9483048915863037,
"learning_rate": 3.479944468603169e-08,
"loss": 0.4479,
"step": 4200
},
{
"epoch": 2.9054520358868183,
"grad_norm": 7.645808219909668,
"learning_rate": 3.0230482520839376e-08,
"loss": 0.4272,
"step": 4210
},
{
"epoch": 2.912353347135956,
"grad_norm": 4.717611312866211,
"learning_rate": 2.5982026210265575e-08,
"loss": 0.4509,
"step": 4220
},
{
"epoch": 2.919254658385093,
"grad_norm": 5.330791473388672,
"learning_rate": 2.205434974188847e-08,
"loss": 0.4703,
"step": 4230
},
{
"epoch": 2.9261559696342303,
"grad_norm": 5.309491157531738,
"learning_rate": 1.8447706415847255e-08,
"loss": 0.4515,
"step": 4240
},
{
"epoch": 2.933057280883368,
"grad_norm": 3.1294057369232178,
"learning_rate": 1.5162328828501328e-08,
"loss": 0.4799,
"step": 4250
},
{
"epoch": 2.939958592132505,
"grad_norm": 8.952521324157715,
"learning_rate": 1.2198428857433941e-08,
"loss": 0.448,
"step": 4260
},
{
"epoch": 2.9468599033816423,
"grad_norm": 4.1140570640563965,
"learning_rate": 9.556197647784814e-09,
"loss": 0.479,
"step": 4270
},
{
"epoch": 2.95376121463078,
"grad_norm": 3.1543099880218506,
"learning_rate": 7.235805599926093e-09,
"loss": 0.411,
"step": 4280
},
{
"epoch": 2.960662525879917,
"grad_norm": 7.077431678771973,
"learning_rate": 5.237402358471144e-09,
"loss": 0.4504,
"step": 4290
},
{
"epoch": 2.9675638371290542,
"grad_norm": 5.550919055938721,
"learning_rate": 3.5611168026228326e-09,
"loss": 0.4167,
"step": 4300
},
{
"epoch": 2.974465148378192,
"grad_norm": 5.99105978012085,
"learning_rate": 2.207057037865168e-09,
"loss": 0.4036,
"step": 4310
},
{
"epoch": 2.981366459627329,
"grad_norm": 6.394698619842529,
"learning_rate": 1.1753103889883267e-09,
"loss": 0.3838,
"step": 4320
},
{
"epoch": 2.9882677708764667,
"grad_norm": 5.257230281829834,
"learning_rate": 4.659433944587122e-10,
"loss": 0.4032,
"step": 4330
},
{
"epoch": 2.995169082125604,
"grad_norm": 6.086167812347412,
"learning_rate": 7.90018021268324e-11,
"loss": 0.4149,
"step": 4340
},
{
"epoch": 3.0,
"step": 4347,
"total_flos": 5.696946108136161e+17,
"train_loss": 0.5156504885202061,
"train_runtime": 3167.6896,
"train_samples_per_second": 10.976,
"train_steps_per_second": 1.372
}
],
"logging_steps": 10,
"max_steps": 4347,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.696946108136161e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}