Jalwa-latest-run-pretrain-10k / trainer_state.json
fhai50032's picture
Upload folder using huggingface_hub
eb88aa3 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5769674590353104,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0005769674590353104,
"grad_norm": 0.7104284763336182,
"learning_rate": 3.846153846153847e-07,
"loss": 2.6483,
"step": 10
},
{
"epoch": 0.0011539349180706207,
"grad_norm": 0.6569196581840515,
"learning_rate": 7.692307692307694e-07,
"loss": 2.6474,
"step": 20
},
{
"epoch": 0.0017309023771059312,
"grad_norm": 0.5456475615501404,
"learning_rate": 1.153846153846154e-06,
"loss": 2.6342,
"step": 30
},
{
"epoch": 0.0023078698361412415,
"grad_norm": 0.3985019028186798,
"learning_rate": 1.5384615384615387e-06,
"loss": 2.6414,
"step": 40
},
{
"epoch": 0.002884837295176552,
"grad_norm": 0.48592114448547363,
"learning_rate": 1.9230769230769234e-06,
"loss": 2.6257,
"step": 50
},
{
"epoch": 0.0034618047542118624,
"grad_norm": 0.32980403304100037,
"learning_rate": 2.307692307692308e-06,
"loss": 2.5972,
"step": 60
},
{
"epoch": 0.004038772213247173,
"grad_norm": 0.3044310510158539,
"learning_rate": 2.6923076923076928e-06,
"loss": 2.6144,
"step": 70
},
{
"epoch": 0.004615739672282483,
"grad_norm": 0.2920003831386566,
"learning_rate": 3.0769230769230774e-06,
"loss": 2.6037,
"step": 80
},
{
"epoch": 0.005192707131317794,
"grad_norm": 0.28606897592544556,
"learning_rate": 3.4615384615384617e-06,
"loss": 2.5913,
"step": 90
},
{
"epoch": 0.005769674590353104,
"grad_norm": 0.2709102928638458,
"learning_rate": 3.846153846153847e-06,
"loss": 2.5825,
"step": 100
},
{
"epoch": 0.006346642049388415,
"grad_norm": 0.28747376799583435,
"learning_rate": 4.230769230769231e-06,
"loss": 2.5782,
"step": 110
},
{
"epoch": 0.006923609508423725,
"grad_norm": 0.2798466384410858,
"learning_rate": 4.615384615384616e-06,
"loss": 2.5689,
"step": 120
},
{
"epoch": 0.007500576967459035,
"grad_norm": 0.27855777740478516,
"learning_rate": 5e-06,
"loss": 2.5601,
"step": 130
},
{
"epoch": 0.008077544426494346,
"grad_norm": 0.29982247948646545,
"learning_rate": 5.3846153846153855e-06,
"loss": 2.5455,
"step": 140
},
{
"epoch": 0.008654511885529657,
"grad_norm": 0.2967548668384552,
"learning_rate": 5.76923076923077e-06,
"loss": 2.5456,
"step": 150
},
{
"epoch": 0.009231479344564966,
"grad_norm": 0.4355414807796478,
"learning_rate": 6.153846153846155e-06,
"loss": 2.5428,
"step": 160
},
{
"epoch": 0.009808446803600277,
"grad_norm": 0.2910907566547394,
"learning_rate": 6.538461538461539e-06,
"loss": 2.543,
"step": 170
},
{
"epoch": 0.010385414262635588,
"grad_norm": 0.28916531801223755,
"learning_rate": 6.923076923076923e-06,
"loss": 2.5261,
"step": 180
},
{
"epoch": 0.010962381721670899,
"grad_norm": 0.4095863699913025,
"learning_rate": 7.3076923076923085e-06,
"loss": 2.5241,
"step": 190
},
{
"epoch": 0.011539349180706208,
"grad_norm": 0.4236544668674469,
"learning_rate": 7.692307692307694e-06,
"loss": 2.534,
"step": 200
},
{
"epoch": 0.012116316639741519,
"grad_norm": 0.3286638557910919,
"learning_rate": 8.076923076923077e-06,
"loss": 2.5113,
"step": 210
},
{
"epoch": 0.01269328409877683,
"grad_norm": 0.36950576305389404,
"learning_rate": 8.461538461538462e-06,
"loss": 2.5166,
"step": 220
},
{
"epoch": 0.013270251557812139,
"grad_norm": 0.3060932755470276,
"learning_rate": 8.846153846153847e-06,
"loss": 2.4978,
"step": 230
},
{
"epoch": 0.01384721901684745,
"grad_norm": 0.29038047790527344,
"learning_rate": 9.230769230769232e-06,
"loss": 2.5106,
"step": 240
},
{
"epoch": 0.01442418647588276,
"grad_norm": 0.3052986264228821,
"learning_rate": 9.615384615384616e-06,
"loss": 2.4951,
"step": 250
},
{
"epoch": 0.01500115393491807,
"grad_norm": 0.314464271068573,
"learning_rate": 1e-05,
"loss": 2.4922,
"step": 260
},
{
"epoch": 0.01557812139395338,
"grad_norm": 0.3013986051082611,
"learning_rate": 1.0384615384615386e-05,
"loss": 2.4956,
"step": 270
},
{
"epoch": 0.01615508885298869,
"grad_norm": 0.33075615763664246,
"learning_rate": 1.0769230769230771e-05,
"loss": 2.4814,
"step": 280
},
{
"epoch": 0.016732056312024002,
"grad_norm": 0.4657692313194275,
"learning_rate": 1.1153846153846154e-05,
"loss": 2.4951,
"step": 290
},
{
"epoch": 0.017309023771059313,
"grad_norm": 0.40145745873451233,
"learning_rate": 1.153846153846154e-05,
"loss": 2.4692,
"step": 300
},
{
"epoch": 0.017885991230094624,
"grad_norm": 0.3732956349849701,
"learning_rate": 1.1923076923076925e-05,
"loss": 2.4855,
"step": 310
},
{
"epoch": 0.018462958689129932,
"grad_norm": 0.3338468074798584,
"learning_rate": 1.230769230769231e-05,
"loss": 2.4902,
"step": 320
},
{
"epoch": 0.019039926148165243,
"grad_norm": 0.3962608277797699,
"learning_rate": 1.2692307692307691e-05,
"loss": 2.4725,
"step": 330
},
{
"epoch": 0.019616893607200554,
"grad_norm": 0.34549835324287415,
"learning_rate": 1.3076923076923078e-05,
"loss": 2.4665,
"step": 340
},
{
"epoch": 0.020193861066235864,
"grad_norm": 0.5531402826309204,
"learning_rate": 1.3461538461538462e-05,
"loss": 2.4784,
"step": 350
},
{
"epoch": 0.020770828525271175,
"grad_norm": 0.3071824610233307,
"learning_rate": 1.3846153846153847e-05,
"loss": 2.4665,
"step": 360
},
{
"epoch": 0.021347795984306486,
"grad_norm": 0.32911962270736694,
"learning_rate": 1.423076923076923e-05,
"loss": 2.4597,
"step": 370
},
{
"epoch": 0.021924763443341797,
"grad_norm": 0.354469358921051,
"learning_rate": 1.4615384615384617e-05,
"loss": 2.4638,
"step": 380
},
{
"epoch": 0.022501730902377105,
"grad_norm": 0.3209117650985718,
"learning_rate": 1.5e-05,
"loss": 2.4492,
"step": 390
},
{
"epoch": 0.023078698361412416,
"grad_norm": 0.44386786222457886,
"learning_rate": 1.5384615384615387e-05,
"loss": 2.4579,
"step": 400
},
{
"epoch": 0.023655665820447726,
"grad_norm": 0.4315970838069916,
"learning_rate": 1.576923076923077e-05,
"loss": 2.4712,
"step": 410
},
{
"epoch": 0.024232633279483037,
"grad_norm": 0.345197468996048,
"learning_rate": 1.6153846153846154e-05,
"loss": 2.4637,
"step": 420
},
{
"epoch": 0.02480960073851835,
"grad_norm": 0.43101733922958374,
"learning_rate": 1.653846153846154e-05,
"loss": 2.4584,
"step": 430
},
{
"epoch": 0.02538656819755366,
"grad_norm": 0.3445722460746765,
"learning_rate": 1.6923076923076924e-05,
"loss": 2.4607,
"step": 440
},
{
"epoch": 0.02596353565658897,
"grad_norm": 0.38286539912223816,
"learning_rate": 1.730769230769231e-05,
"loss": 2.4451,
"step": 450
},
{
"epoch": 0.026540503115624278,
"grad_norm": 0.7410378456115723,
"learning_rate": 1.7692307692307694e-05,
"loss": 2.4443,
"step": 460
},
{
"epoch": 0.02711747057465959,
"grad_norm": 0.36650899052619934,
"learning_rate": 1.8076923076923076e-05,
"loss": 2.4513,
"step": 470
},
{
"epoch": 0.0276944380336949,
"grad_norm": 0.33747929334640503,
"learning_rate": 1.8461538461538465e-05,
"loss": 2.4286,
"step": 480
},
{
"epoch": 0.02827140549273021,
"grad_norm": 0.3847596347332001,
"learning_rate": 1.8846153846153846e-05,
"loss": 2.4371,
"step": 490
},
{
"epoch": 0.02884837295176552,
"grad_norm": 0.3613141179084778,
"learning_rate": 1.923076923076923e-05,
"loss": 2.4316,
"step": 500
},
{
"epoch": 0.029425340410800832,
"grad_norm": 0.515495240688324,
"learning_rate": 1.9615384615384617e-05,
"loss": 2.4454,
"step": 510
},
{
"epoch": 0.03000230786983614,
"grad_norm": 0.4233556389808655,
"learning_rate": 2e-05,
"loss": 2.4263,
"step": 520
},
{
"epoch": 0.03057927532887145,
"grad_norm": 0.34941181540489197,
"learning_rate": 2.0384615384615387e-05,
"loss": 2.4429,
"step": 530
},
{
"epoch": 0.03115624278790676,
"grad_norm": 0.47630569338798523,
"learning_rate": 2.0769230769230772e-05,
"loss": 2.4181,
"step": 540
},
{
"epoch": 0.031733210246942076,
"grad_norm": 0.6628788709640503,
"learning_rate": 2.1153846153846154e-05,
"loss": 2.4181,
"step": 550
},
{
"epoch": 0.03231017770597738,
"grad_norm": 0.6385440826416016,
"learning_rate": 2.1538461538461542e-05,
"loss": 2.4207,
"step": 560
},
{
"epoch": 0.03288714516501269,
"grad_norm": 0.35360780358314514,
"learning_rate": 2.1923076923076924e-05,
"loss": 2.435,
"step": 570
},
{
"epoch": 0.033464112624048005,
"grad_norm": 0.6603436470031738,
"learning_rate": 2.230769230769231e-05,
"loss": 2.413,
"step": 580
},
{
"epoch": 0.03404108008308331,
"grad_norm": 0.49220070242881775,
"learning_rate": 2.2692307692307694e-05,
"loss": 2.4161,
"step": 590
},
{
"epoch": 0.03461804754211863,
"grad_norm": 0.33461564779281616,
"learning_rate": 2.307692307692308e-05,
"loss": 2.4363,
"step": 600
},
{
"epoch": 0.035195015001153934,
"grad_norm": 0.3938097357749939,
"learning_rate": 2.3461538461538464e-05,
"loss": 2.4059,
"step": 610
},
{
"epoch": 0.03577198246018925,
"grad_norm": 0.559905469417572,
"learning_rate": 2.384615384615385e-05,
"loss": 2.4163,
"step": 620
},
{
"epoch": 0.036348949919224556,
"grad_norm": 0.4643310606479645,
"learning_rate": 2.423076923076923e-05,
"loss": 2.4223,
"step": 630
},
{
"epoch": 0.036925917378259863,
"grad_norm": 0.44358041882514954,
"learning_rate": 2.461538461538462e-05,
"loss": 2.4181,
"step": 640
},
{
"epoch": 0.03750288483729518,
"grad_norm": 0.4786848723888397,
"learning_rate": 2.5e-05,
"loss": 2.419,
"step": 650
},
{
"epoch": 0.038079852296330485,
"grad_norm": 2.538026809692383,
"learning_rate": 2.5384615384615383e-05,
"loss": 2.4174,
"step": 660
},
{
"epoch": 0.0386568197553658,
"grad_norm": 0.8546634912490845,
"learning_rate": 2.5769230769230768e-05,
"loss": 2.4146,
"step": 670
},
{
"epoch": 0.03923378721440111,
"grad_norm": 0.4435669481754303,
"learning_rate": 2.6153846153846157e-05,
"loss": 2.4042,
"step": 680
},
{
"epoch": 0.03981075467343642,
"grad_norm": 0.4790751039981842,
"learning_rate": 2.6538461538461538e-05,
"loss": 2.3906,
"step": 690
},
{
"epoch": 0.04038772213247173,
"grad_norm": 0.40961435437202454,
"learning_rate": 2.6923076923076923e-05,
"loss": 2.4012,
"step": 700
},
{
"epoch": 0.040964689591507036,
"grad_norm": 0.5555410385131836,
"learning_rate": 2.7307692307692305e-05,
"loss": 2.385,
"step": 710
},
{
"epoch": 0.04154165705054235,
"grad_norm": 0.5821479558944702,
"learning_rate": 2.7692307692307694e-05,
"loss": 2.3953,
"step": 720
},
{
"epoch": 0.04211862450957766,
"grad_norm": 0.40736421942710876,
"learning_rate": 2.807692307692308e-05,
"loss": 2.3962,
"step": 730
},
{
"epoch": 0.04269559196861297,
"grad_norm": 0.4075765907764435,
"learning_rate": 2.846153846153846e-05,
"loss": 2.39,
"step": 740
},
{
"epoch": 0.04327255942764828,
"grad_norm": 0.33844342827796936,
"learning_rate": 2.8846153846153845e-05,
"loss": 2.3841,
"step": 750
},
{
"epoch": 0.043849526886683594,
"grad_norm": 0.40746697783470154,
"learning_rate": 2.9230769230769234e-05,
"loss": 2.4011,
"step": 760
},
{
"epoch": 0.0444264943457189,
"grad_norm": 0.5998666882514954,
"learning_rate": 2.9615384615384616e-05,
"loss": 2.3927,
"step": 770
},
{
"epoch": 0.04500346180475421,
"grad_norm": 0.875724196434021,
"learning_rate": 3e-05,
"loss": 2.3916,
"step": 780
},
{
"epoch": 0.045580429263789524,
"grad_norm": 0.9473147392272949,
"learning_rate": 3.0384615384615382e-05,
"loss": 2.3865,
"step": 790
},
{
"epoch": 0.04615739672282483,
"grad_norm": 0.6188512444496155,
"learning_rate": 3.0769230769230774e-05,
"loss": 2.3809,
"step": 800
},
{
"epoch": 0.046734364181860145,
"grad_norm": 0.6592457890510559,
"learning_rate": 3.115384615384615e-05,
"loss": 2.3864,
"step": 810
},
{
"epoch": 0.04731133164089545,
"grad_norm": 0.718359649181366,
"learning_rate": 3.153846153846154e-05,
"loss": 2.3697,
"step": 820
},
{
"epoch": 0.04788829909993077,
"grad_norm": 0.6391853094100952,
"learning_rate": 3.192307692307692e-05,
"loss": 2.3903,
"step": 830
},
{
"epoch": 0.048465266558966075,
"grad_norm": 0.6799693703651428,
"learning_rate": 3.230769230769231e-05,
"loss": 2.3955,
"step": 840
},
{
"epoch": 0.04904223401800138,
"grad_norm": 0.47043943405151367,
"learning_rate": 3.269230769230769e-05,
"loss": 2.376,
"step": 850
},
{
"epoch": 0.0496192014770367,
"grad_norm": 0.45935624837875366,
"learning_rate": 3.307692307692308e-05,
"loss": 2.3873,
"step": 860
},
{
"epoch": 0.050196168936072004,
"grad_norm": 0.44479900598526,
"learning_rate": 3.346153846153846e-05,
"loss": 2.3721,
"step": 870
},
{
"epoch": 0.05077313639510732,
"grad_norm": 0.37469252943992615,
"learning_rate": 3.384615384615385e-05,
"loss": 2.3748,
"step": 880
},
{
"epoch": 0.051350103854142626,
"grad_norm": 0.5100722312927246,
"learning_rate": 3.4230769230769234e-05,
"loss": 2.3828,
"step": 890
},
{
"epoch": 0.05192707131317794,
"grad_norm": 0.43965014815330505,
"learning_rate": 3.461538461538462e-05,
"loss": 2.3758,
"step": 900
},
{
"epoch": 0.05250403877221325,
"grad_norm": 0.8576379418373108,
"learning_rate": 3.5e-05,
"loss": 2.3844,
"step": 910
},
{
"epoch": 0.053081006231248555,
"grad_norm": 0.8042026162147522,
"learning_rate": 3.538461538461539e-05,
"loss": 2.3671,
"step": 920
},
{
"epoch": 0.05365797369028387,
"grad_norm": 0.7367364764213562,
"learning_rate": 3.5769230769230774e-05,
"loss": 2.3879,
"step": 930
},
{
"epoch": 0.05423494114931918,
"grad_norm": 0.7298218607902527,
"learning_rate": 3.615384615384615e-05,
"loss": 2.3766,
"step": 940
},
{
"epoch": 0.05481190860835449,
"grad_norm": 0.39043882489204407,
"learning_rate": 3.653846153846154e-05,
"loss": 2.3736,
"step": 950
},
{
"epoch": 0.0553888760673898,
"grad_norm": 0.6030943989753723,
"learning_rate": 3.692307692307693e-05,
"loss": 2.3579,
"step": 960
},
{
"epoch": 0.05596584352642511,
"grad_norm": 0.6056037545204163,
"learning_rate": 3.730769230769231e-05,
"loss": 2.3579,
"step": 970
},
{
"epoch": 0.05654281098546042,
"grad_norm": 0.37389466166496277,
"learning_rate": 3.769230769230769e-05,
"loss": 2.3746,
"step": 980
},
{
"epoch": 0.05711977844449573,
"grad_norm": 0.3487268388271332,
"learning_rate": 3.807692307692308e-05,
"loss": 2.3633,
"step": 990
},
{
"epoch": 0.05769674590353104,
"grad_norm": 0.4989362955093384,
"learning_rate": 3.846153846153846e-05,
"loss": 2.3639,
"step": 1000
},
{
"epoch": 0.05827371336256635,
"grad_norm": 0.3485865890979767,
"learning_rate": 3.884615384615385e-05,
"loss": 2.3655,
"step": 1010
},
{
"epoch": 0.058850680821601664,
"grad_norm": 1.3665262460708618,
"learning_rate": 3.923076923076923e-05,
"loss": 2.3559,
"step": 1020
},
{
"epoch": 0.05942764828063697,
"grad_norm": 1.7110989093780518,
"learning_rate": 3.961538461538462e-05,
"loss": 2.372,
"step": 1030
},
{
"epoch": 0.06000461573967228,
"grad_norm": 1.2965339422225952,
"learning_rate": 4e-05,
"loss": 2.3641,
"step": 1040
},
{
"epoch": 0.06058158319870759,
"grad_norm": 1.0707265138626099,
"learning_rate": 4.038461538461539e-05,
"loss": 2.3682,
"step": 1050
},
{
"epoch": 0.0611585506577429,
"grad_norm": 1.106713056564331,
"learning_rate": 4.0769230769230773e-05,
"loss": 2.3559,
"step": 1060
},
{
"epoch": 0.061735518116778215,
"grad_norm": 2.2424588203430176,
"learning_rate": 4.115384615384615e-05,
"loss": 2.3661,
"step": 1070
},
{
"epoch": 0.06231248557581352,
"grad_norm": 0.48183363676071167,
"learning_rate": 4.1538461538461544e-05,
"loss": 2.3442,
"step": 1080
},
{
"epoch": 0.06288945303484883,
"grad_norm": 0.9450508952140808,
"learning_rate": 4.192307692307693e-05,
"loss": 2.3386,
"step": 1090
},
{
"epoch": 0.06346642049388415,
"grad_norm": 1.067024827003479,
"learning_rate": 4.230769230769231e-05,
"loss": 2.3532,
"step": 1100
},
{
"epoch": 0.06404338795291946,
"grad_norm": 0.9552332162857056,
"learning_rate": 4.269230769230769e-05,
"loss": 2.3392,
"step": 1110
},
{
"epoch": 0.06462035541195477,
"grad_norm": 0.8050335645675659,
"learning_rate": 4.3076923076923084e-05,
"loss": 2.3371,
"step": 1120
},
{
"epoch": 0.06519732287099007,
"grad_norm": 0.6774647235870361,
"learning_rate": 4.346153846153846e-05,
"loss": 2.3532,
"step": 1130
},
{
"epoch": 0.06577429033002538,
"grad_norm": 0.4367026388645172,
"learning_rate": 4.384615384615385e-05,
"loss": 2.3487,
"step": 1140
},
{
"epoch": 0.0663512577890607,
"grad_norm": 0.4666714072227478,
"learning_rate": 4.423076923076923e-05,
"loss": 2.3596,
"step": 1150
},
{
"epoch": 0.06692822524809601,
"grad_norm": 0.4387041926383972,
"learning_rate": 4.461538461538462e-05,
"loss": 2.3468,
"step": 1160
},
{
"epoch": 0.06750519270713132,
"grad_norm": 0.4057537913322449,
"learning_rate": 4.5e-05,
"loss": 2.323,
"step": 1170
},
{
"epoch": 0.06808216016616662,
"grad_norm": 0.5073262453079224,
"learning_rate": 4.538461538461539e-05,
"loss": 2.3377,
"step": 1180
},
{
"epoch": 0.06865912762520193,
"grad_norm": 0.40886059403419495,
"learning_rate": 4.576923076923077e-05,
"loss": 2.3369,
"step": 1190
},
{
"epoch": 0.06923609508423725,
"grad_norm": 0.46122193336486816,
"learning_rate": 4.615384615384616e-05,
"loss": 2.3357,
"step": 1200
},
{
"epoch": 0.06981306254327256,
"grad_norm": 0.7845937013626099,
"learning_rate": 4.653846153846154e-05,
"loss": 2.3496,
"step": 1210
},
{
"epoch": 0.07039003000230787,
"grad_norm": 0.4420381188392639,
"learning_rate": 4.692307692307693e-05,
"loss": 2.3529,
"step": 1220
},
{
"epoch": 0.07096699746134318,
"grad_norm": 0.4529377222061157,
"learning_rate": 4.730769230769231e-05,
"loss": 2.3197,
"step": 1230
},
{
"epoch": 0.0715439649203785,
"grad_norm": 0.6386104822158813,
"learning_rate": 4.76923076923077e-05,
"loss": 2.3446,
"step": 1240
},
{
"epoch": 0.0721209323794138,
"grad_norm": 0.7226986289024353,
"learning_rate": 4.8076923076923084e-05,
"loss": 2.3497,
"step": 1250
},
{
"epoch": 0.07269789983844911,
"grad_norm": 1.063812017440796,
"learning_rate": 4.846153846153846e-05,
"loss": 2.3506,
"step": 1260
},
{
"epoch": 0.07327486729748442,
"grad_norm": 0.511073112487793,
"learning_rate": 4.884615384615385e-05,
"loss": 2.3446,
"step": 1270
},
{
"epoch": 0.07385183475651973,
"grad_norm": 0.3652757406234741,
"learning_rate": 4.923076923076924e-05,
"loss": 2.3282,
"step": 1280
},
{
"epoch": 0.07442880221555505,
"grad_norm": 0.4461408853530884,
"learning_rate": 4.961538461538462e-05,
"loss": 2.3417,
"step": 1290
},
{
"epoch": 0.07500576967459036,
"grad_norm": 0.47319433093070984,
"learning_rate": 5e-05,
"loss": 2.3418,
"step": 1300
},
{
"epoch": 0.07558273713362566,
"grad_norm": 0.4424804151058197,
"learning_rate": 5.038461538461539e-05,
"loss": 2.3349,
"step": 1310
},
{
"epoch": 0.07615970459266097,
"grad_norm": 0.509090781211853,
"learning_rate": 5.0769230769230766e-05,
"loss": 2.3236,
"step": 1320
},
{
"epoch": 0.07673667205169628,
"grad_norm": 1.1174790859222412,
"learning_rate": 5.115384615384615e-05,
"loss": 2.3317,
"step": 1330
},
{
"epoch": 0.0773136395107316,
"grad_norm": 0.4461131989955902,
"learning_rate": 5.1538461538461536e-05,
"loss": 2.325,
"step": 1340
},
{
"epoch": 0.0778906069697669,
"grad_norm": 0.40025824308395386,
"learning_rate": 5.192307692307693e-05,
"loss": 2.3391,
"step": 1350
},
{
"epoch": 0.07846757442880221,
"grad_norm": 0.5350430011749268,
"learning_rate": 5.230769230769231e-05,
"loss": 2.3427,
"step": 1360
},
{
"epoch": 0.07904454188783752,
"grad_norm": 0.4814503788948059,
"learning_rate": 5.26923076923077e-05,
"loss": 2.3374,
"step": 1370
},
{
"epoch": 0.07962150934687284,
"grad_norm": 0.8118190765380859,
"learning_rate": 5.3076923076923076e-05,
"loss": 2.3342,
"step": 1380
},
{
"epoch": 0.08019847680590815,
"grad_norm": 0.8076496720314026,
"learning_rate": 5.346153846153846e-05,
"loss": 2.3348,
"step": 1390
},
{
"epoch": 0.08077544426494346,
"grad_norm": 0.43335264921188354,
"learning_rate": 5.384615384615385e-05,
"loss": 2.3369,
"step": 1400
},
{
"epoch": 0.08135241172397877,
"grad_norm": 0.704337477684021,
"learning_rate": 5.423076923076923e-05,
"loss": 2.3227,
"step": 1410
},
{
"epoch": 0.08192937918301407,
"grad_norm": 0.5695681571960449,
"learning_rate": 5.461538461538461e-05,
"loss": 2.3209,
"step": 1420
},
{
"epoch": 0.0825063466420494,
"grad_norm": 0.5841794610023499,
"learning_rate": 5.500000000000001e-05,
"loss": 2.3135,
"step": 1430
},
{
"epoch": 0.0830833141010847,
"grad_norm": 0.5450507402420044,
"learning_rate": 5.538461538461539e-05,
"loss": 2.322,
"step": 1440
},
{
"epoch": 0.08366028156012001,
"grad_norm": 0.5215712785720825,
"learning_rate": 5.576923076923077e-05,
"loss": 2.3356,
"step": 1450
},
{
"epoch": 0.08423724901915532,
"grad_norm": 0.5433002710342407,
"learning_rate": 5.615384615384616e-05,
"loss": 2.3187,
"step": 1460
},
{
"epoch": 0.08481421647819062,
"grad_norm": 0.4980168342590332,
"learning_rate": 5.653846153846154e-05,
"loss": 2.3164,
"step": 1470
},
{
"epoch": 0.08539118393722595,
"grad_norm": 1.0669074058532715,
"learning_rate": 5.692307692307692e-05,
"loss": 2.3226,
"step": 1480
},
{
"epoch": 0.08596815139626125,
"grad_norm": 0.569298505783081,
"learning_rate": 5.7307692307692306e-05,
"loss": 2.3098,
"step": 1490
},
{
"epoch": 0.08654511885529656,
"grad_norm": 0.7760196328163147,
"learning_rate": 5.769230769230769e-05,
"loss": 2.3265,
"step": 1500
},
{
"epoch": 0.08712208631433187,
"grad_norm": 0.5390911102294922,
"learning_rate": 5.807692307692308e-05,
"loss": 2.3122,
"step": 1510
},
{
"epoch": 0.08769905377336719,
"grad_norm": 0.8864639401435852,
"learning_rate": 5.846153846153847e-05,
"loss": 2.3155,
"step": 1520
},
{
"epoch": 0.0882760212324025,
"grad_norm": 0.4188685715198517,
"learning_rate": 5.884615384615385e-05,
"loss": 2.3087,
"step": 1530
},
{
"epoch": 0.0888529886914378,
"grad_norm": 0.5072064399719238,
"learning_rate": 5.923076923076923e-05,
"loss": 2.3091,
"step": 1540
},
{
"epoch": 0.08942995615047311,
"grad_norm": 0.7691831588745117,
"learning_rate": 5.9615384615384616e-05,
"loss": 2.3165,
"step": 1550
},
{
"epoch": 0.09000692360950842,
"grad_norm": 0.3790508806705475,
"learning_rate": 6e-05,
"loss": 2.3162,
"step": 1560
},
{
"epoch": 0.09058389106854374,
"grad_norm": 0.9513830542564392,
"learning_rate": 6.038461538461539e-05,
"loss": 2.3184,
"step": 1570
},
{
"epoch": 0.09116085852757905,
"grad_norm": 0.6464409828186035,
"learning_rate": 6.0769230769230765e-05,
"loss": 2.3297,
"step": 1580
},
{
"epoch": 0.09173782598661435,
"grad_norm": 0.5319147109985352,
"learning_rate": 6.115384615384616e-05,
"loss": 2.2935,
"step": 1590
},
{
"epoch": 0.09231479344564966,
"grad_norm": 0.9692733883857727,
"learning_rate": 6.153846153846155e-05,
"loss": 2.3077,
"step": 1600
},
{
"epoch": 0.09289176090468497,
"grad_norm": 0.448768675327301,
"learning_rate": 6.192307692307693e-05,
"loss": 2.3078,
"step": 1610
},
{
"epoch": 0.09346872836372029,
"grad_norm": 0.6407192945480347,
"learning_rate": 6.23076923076923e-05,
"loss": 2.3069,
"step": 1620
},
{
"epoch": 0.0940456958227556,
"grad_norm": 0.6339554786682129,
"learning_rate": 6.26923076923077e-05,
"loss": 2.31,
"step": 1630
},
{
"epoch": 0.0946226632817909,
"grad_norm": 0.634101152420044,
"learning_rate": 6.307692307692308e-05,
"loss": 2.3068,
"step": 1640
},
{
"epoch": 0.09519963074082621,
"grad_norm": 0.6417617797851562,
"learning_rate": 6.346153846153847e-05,
"loss": 2.297,
"step": 1650
},
{
"epoch": 0.09577659819986153,
"grad_norm": 0.5501047968864441,
"learning_rate": 6.384615384615385e-05,
"loss": 2.2972,
"step": 1660
},
{
"epoch": 0.09635356565889684,
"grad_norm": 1.1125547885894775,
"learning_rate": 6.423076923076924e-05,
"loss": 2.3142,
"step": 1670
},
{
"epoch": 0.09693053311793215,
"grad_norm": 1.900116205215454,
"learning_rate": 6.461538461538462e-05,
"loss": 2.2999,
"step": 1680
},
{
"epoch": 0.09750750057696746,
"grad_norm": 1.4570536613464355,
"learning_rate": 6.500000000000001e-05,
"loss": 2.2935,
"step": 1690
},
{
"epoch": 0.09808446803600276,
"grad_norm": 1.8386913537979126,
"learning_rate": 6.538461538461539e-05,
"loss": 2.2942,
"step": 1700
},
{
"epoch": 0.09866143549503809,
"grad_norm": 1.289682388305664,
"learning_rate": 6.576923076923078e-05,
"loss": 2.306,
"step": 1710
},
{
"epoch": 0.0992384029540734,
"grad_norm": 1.354390025138855,
"learning_rate": 6.615384615384616e-05,
"loss": 2.2973,
"step": 1720
},
{
"epoch": 0.0998153704131087,
"grad_norm": 1.333463430404663,
"learning_rate": 6.653846153846153e-05,
"loss": 2.2863,
"step": 1730
},
{
"epoch": 0.10039233787214401,
"grad_norm": 1.2394797801971436,
"learning_rate": 6.692307692307693e-05,
"loss": 2.2857,
"step": 1740
},
{
"epoch": 0.10096930533117932,
"grad_norm": 1.4308770895004272,
"learning_rate": 6.730769230769232e-05,
"loss": 2.2901,
"step": 1750
},
{
"epoch": 0.10154627279021464,
"grad_norm": 1.2967873811721802,
"learning_rate": 6.76923076923077e-05,
"loss": 2.2997,
"step": 1760
},
{
"epoch": 0.10212324024924994,
"grad_norm": 1.192817211151123,
"learning_rate": 6.807692307692309e-05,
"loss": 2.2841,
"step": 1770
},
{
"epoch": 0.10270020770828525,
"grad_norm": 1.3187321424484253,
"learning_rate": 6.846153846153847e-05,
"loss": 2.2893,
"step": 1780
},
{
"epoch": 0.10327717516732056,
"grad_norm": 1.267285943031311,
"learning_rate": 6.884615384615385e-05,
"loss": 2.2901,
"step": 1790
},
{
"epoch": 0.10385414262635588,
"grad_norm": 1.3133481740951538,
"learning_rate": 6.923076923076924e-05,
"loss": 2.2961,
"step": 1800
},
{
"epoch": 0.10443111008539119,
"grad_norm": 1.097119927406311,
"learning_rate": 6.961538461538462e-05,
"loss": 2.301,
"step": 1810
},
{
"epoch": 0.1050080775444265,
"grad_norm": 1.2084568738937378,
"learning_rate": 7e-05,
"loss": 2.2936,
"step": 1820
},
{
"epoch": 0.1055850450034618,
"grad_norm": 1.1895954608917236,
"learning_rate": 7.03846153846154e-05,
"loss": 2.2985,
"step": 1830
},
{
"epoch": 0.10616201246249711,
"grad_norm": 0.8945775628089905,
"learning_rate": 7.076923076923078e-05,
"loss": 2.2972,
"step": 1840
},
{
"epoch": 0.10673897992153243,
"grad_norm": 0.39187899231910706,
"learning_rate": 7.115384615384616e-05,
"loss": 2.2916,
"step": 1850
},
{
"epoch": 0.10731594738056774,
"grad_norm": 0.9170978665351868,
"learning_rate": 7.153846153846155e-05,
"loss": 2.2832,
"step": 1860
},
{
"epoch": 0.10789291483960305,
"grad_norm": 0.37489083409309387,
"learning_rate": 7.192307692307693e-05,
"loss": 2.2882,
"step": 1870
},
{
"epoch": 0.10846988229863835,
"grad_norm": 0.3608030080795288,
"learning_rate": 7.23076923076923e-05,
"loss": 2.2875,
"step": 1880
},
{
"epoch": 0.10904684975767366,
"grad_norm": 0.3541902005672455,
"learning_rate": 7.26923076923077e-05,
"loss": 2.2817,
"step": 1890
},
{
"epoch": 0.10962381721670898,
"grad_norm": 0.42174556851387024,
"learning_rate": 7.307692307692307e-05,
"loss": 2.2812,
"step": 1900
},
{
"epoch": 0.11020078467574429,
"grad_norm": 0.3870134651660919,
"learning_rate": 7.346153846153847e-05,
"loss": 2.291,
"step": 1910
},
{
"epoch": 0.1107777521347796,
"grad_norm": 0.5459268689155579,
"learning_rate": 7.384615384615386e-05,
"loss": 2.2766,
"step": 1920
},
{
"epoch": 0.1113547195938149,
"grad_norm": 0.4281494915485382,
"learning_rate": 7.423076923076924e-05,
"loss": 2.2809,
"step": 1930
},
{
"epoch": 0.11193168705285023,
"grad_norm": 0.5660754442214966,
"learning_rate": 7.461538461538462e-05,
"loss": 2.2832,
"step": 1940
},
{
"epoch": 0.11250865451188553,
"grad_norm": 0.3824213445186615,
"learning_rate": 7.500000000000001e-05,
"loss": 2.2833,
"step": 1950
},
{
"epoch": 0.11308562197092084,
"grad_norm": 0.39723706245422363,
"learning_rate": 7.538461538461539e-05,
"loss": 2.2826,
"step": 1960
},
{
"epoch": 0.11366258942995615,
"grad_norm": 0.3764449656009674,
"learning_rate": 7.576923076923076e-05,
"loss": 2.2773,
"step": 1970
},
{
"epoch": 0.11423955688899146,
"grad_norm": 0.4552217423915863,
"learning_rate": 7.615384615384616e-05,
"loss": 2.2741,
"step": 1980
},
{
"epoch": 0.11481652434802678,
"grad_norm": 0.37920665740966797,
"learning_rate": 7.653846153846153e-05,
"loss": 2.2841,
"step": 1990
},
{
"epoch": 0.11539349180706208,
"grad_norm": 0.5208818316459656,
"learning_rate": 7.692307692307693e-05,
"loss": 2.2713,
"step": 2000
},
{
"epoch": 0.11597045926609739,
"grad_norm": 0.40460675954818726,
"learning_rate": 7.730769230769232e-05,
"loss": 2.2902,
"step": 2010
},
{
"epoch": 0.1165474267251327,
"grad_norm": 0.41813430190086365,
"learning_rate": 7.76923076923077e-05,
"loss": 2.2752,
"step": 2020
},
{
"epoch": 0.11712439418416801,
"grad_norm": 0.45470941066741943,
"learning_rate": 7.807692307692307e-05,
"loss": 2.2829,
"step": 2030
},
{
"epoch": 0.11770136164320333,
"grad_norm": 0.5748339891433716,
"learning_rate": 7.846153846153847e-05,
"loss": 2.2865,
"step": 2040
},
{
"epoch": 0.11827832910223864,
"grad_norm": 0.6586531400680542,
"learning_rate": 7.884615384615384e-05,
"loss": 2.2764,
"step": 2050
},
{
"epoch": 0.11885529656127394,
"grad_norm": 0.47713714838027954,
"learning_rate": 7.923076923076924e-05,
"loss": 2.2763,
"step": 2060
},
{
"epoch": 0.11943226402030925,
"grad_norm": 0.5001585483551025,
"learning_rate": 7.961538461538461e-05,
"loss": 2.2637,
"step": 2070
},
{
"epoch": 0.12000923147934456,
"grad_norm": 0.49730342626571655,
"learning_rate": 8e-05,
"loss": 2.2697,
"step": 2080
},
{
"epoch": 0.12058619893837988,
"grad_norm": 0.7335867285728455,
"learning_rate": 8.038461538461538e-05,
"loss": 2.2798,
"step": 2090
},
{
"epoch": 0.12116316639741519,
"grad_norm": 0.7102777361869812,
"learning_rate": 8.076923076923078e-05,
"loss": 2.2659,
"step": 2100
},
{
"epoch": 0.1217401338564505,
"grad_norm": 0.5046675801277161,
"learning_rate": 8.115384615384616e-05,
"loss": 2.2692,
"step": 2110
},
{
"epoch": 0.1223171013154858,
"grad_norm": 0.48471033573150635,
"learning_rate": 8.153846153846155e-05,
"loss": 2.2664,
"step": 2120
},
{
"epoch": 0.12289406877452112,
"grad_norm": 0.5488987565040588,
"learning_rate": 8.192307692307693e-05,
"loss": 2.2862,
"step": 2130
},
{
"epoch": 0.12347103623355643,
"grad_norm": 0.5509325265884399,
"learning_rate": 8.23076923076923e-05,
"loss": 2.2736,
"step": 2140
},
{
"epoch": 0.12404800369259174,
"grad_norm": 0.47769248485565186,
"learning_rate": 8.26923076923077e-05,
"loss": 2.259,
"step": 2150
},
{
"epoch": 0.12462497115162705,
"grad_norm": 0.5516425371170044,
"learning_rate": 8.307692307692309e-05,
"loss": 2.2642,
"step": 2160
},
{
"epoch": 0.12520193861066237,
"grad_norm": 0.3504526615142822,
"learning_rate": 8.346153846153847e-05,
"loss": 2.2652,
"step": 2170
},
{
"epoch": 0.12577890606969766,
"grad_norm": 0.720689058303833,
"learning_rate": 8.384615384615386e-05,
"loss": 2.2765,
"step": 2180
},
{
"epoch": 0.12635587352873298,
"grad_norm": 0.7643449306488037,
"learning_rate": 8.423076923076924e-05,
"loss": 2.27,
"step": 2190
},
{
"epoch": 0.1269328409877683,
"grad_norm": 0.47102758288383484,
"learning_rate": 8.461538461538461e-05,
"loss": 2.2595,
"step": 2200
},
{
"epoch": 0.1275098084468036,
"grad_norm": 0.416532427072525,
"learning_rate": 8.5e-05,
"loss": 2.2669,
"step": 2210
},
{
"epoch": 0.12808677590583892,
"grad_norm": 0.37624692916870117,
"learning_rate": 8.538461538461538e-05,
"loss": 2.2748,
"step": 2220
},
{
"epoch": 0.1286637433648742,
"grad_norm": 0.4788731634616852,
"learning_rate": 8.576923076923076e-05,
"loss": 2.2701,
"step": 2230
},
{
"epoch": 0.12924071082390953,
"grad_norm": 0.34961792826652527,
"learning_rate": 8.615384615384617e-05,
"loss": 2.2658,
"step": 2240
},
{
"epoch": 0.12981767828294485,
"grad_norm": 0.5307486057281494,
"learning_rate": 8.653846153846155e-05,
"loss": 2.2651,
"step": 2250
},
{
"epoch": 0.13039464574198015,
"grad_norm": 0.6055422425270081,
"learning_rate": 8.692307692307692e-05,
"loss": 2.2625,
"step": 2260
},
{
"epoch": 0.13097161320101547,
"grad_norm": 0.5680871605873108,
"learning_rate": 8.730769230769232e-05,
"loss": 2.256,
"step": 2270
},
{
"epoch": 0.13154858066005076,
"grad_norm": 1.0871108770370483,
"learning_rate": 8.76923076923077e-05,
"loss": 2.2568,
"step": 2280
},
{
"epoch": 0.13212554811908608,
"grad_norm": 0.5224435329437256,
"learning_rate": 8.807692307692307e-05,
"loss": 2.2741,
"step": 2290
},
{
"epoch": 0.1327025155781214,
"grad_norm": 0.587762713432312,
"learning_rate": 8.846153846153847e-05,
"loss": 2.2708,
"step": 2300
},
{
"epoch": 0.1332794830371567,
"grad_norm": 0.6314758062362671,
"learning_rate": 8.884615384615384e-05,
"loss": 2.2716,
"step": 2310
},
{
"epoch": 0.13385645049619202,
"grad_norm": 0.4869266748428345,
"learning_rate": 8.923076923076924e-05,
"loss": 2.2511,
"step": 2320
},
{
"epoch": 0.1344334179552273,
"grad_norm": 0.4189651608467102,
"learning_rate": 8.961538461538463e-05,
"loss": 2.2634,
"step": 2330
},
{
"epoch": 0.13501038541426263,
"grad_norm": 0.6171414852142334,
"learning_rate": 9e-05,
"loss": 2.2448,
"step": 2340
},
{
"epoch": 0.13558735287329796,
"grad_norm": 0.8491079807281494,
"learning_rate": 9.038461538461538e-05,
"loss": 2.2592,
"step": 2350
},
{
"epoch": 0.13616432033233325,
"grad_norm": 0.32807987928390503,
"learning_rate": 9.076923076923078e-05,
"loss": 2.2538,
"step": 2360
},
{
"epoch": 0.13674128779136857,
"grad_norm": 0.3959665894508362,
"learning_rate": 9.115384615384615e-05,
"loss": 2.2574,
"step": 2370
},
{
"epoch": 0.13731825525040386,
"grad_norm": 0.8039582967758179,
"learning_rate": 9.153846153846155e-05,
"loss": 2.2581,
"step": 2380
},
{
"epoch": 0.13789522270943919,
"grad_norm": 0.6241025924682617,
"learning_rate": 9.192307692307692e-05,
"loss": 2.2503,
"step": 2390
},
{
"epoch": 0.1384721901684745,
"grad_norm": 0.678353488445282,
"learning_rate": 9.230769230769232e-05,
"loss": 2.2723,
"step": 2400
},
{
"epoch": 0.1390491576275098,
"grad_norm": 0.9290316104888916,
"learning_rate": 9.26923076923077e-05,
"loss": 2.2542,
"step": 2410
},
{
"epoch": 0.13962612508654512,
"grad_norm": 0.3607237935066223,
"learning_rate": 9.307692307692309e-05,
"loss": 2.2639,
"step": 2420
},
{
"epoch": 0.14020309254558042,
"grad_norm": 0.47618988156318665,
"learning_rate": 9.346153846153846e-05,
"loss": 2.2493,
"step": 2430
},
{
"epoch": 0.14078006000461574,
"grad_norm": 0.40734684467315674,
"learning_rate": 9.384615384615386e-05,
"loss": 2.2563,
"step": 2440
},
{
"epoch": 0.14135702746365106,
"grad_norm": 0.3996923267841339,
"learning_rate": 9.423076923076924e-05,
"loss": 2.2489,
"step": 2450
},
{
"epoch": 0.14193399492268635,
"grad_norm": 0.509465754032135,
"learning_rate": 9.461538461538461e-05,
"loss": 2.2477,
"step": 2460
},
{
"epoch": 0.14251096238172167,
"grad_norm": 0.3685857653617859,
"learning_rate": 9.5e-05,
"loss": 2.2667,
"step": 2470
},
{
"epoch": 0.143087929840757,
"grad_norm": 0.5704885125160217,
"learning_rate": 9.53846153846154e-05,
"loss": 2.2584,
"step": 2480
},
{
"epoch": 0.1436648972997923,
"grad_norm": 0.5545893907546997,
"learning_rate": 9.576923076923078e-05,
"loss": 2.2614,
"step": 2490
},
{
"epoch": 0.1442418647588276,
"grad_norm": 0.4671754539012909,
"learning_rate": 9.615384615384617e-05,
"loss": 2.2643,
"step": 2500
},
{
"epoch": 0.1448188322178629,
"grad_norm": 0.4108332097530365,
"learning_rate": 9.653846153846155e-05,
"loss": 2.2519,
"step": 2510
},
{
"epoch": 0.14539579967689822,
"grad_norm": 0.6676947474479675,
"learning_rate": 9.692307692307692e-05,
"loss": 2.2634,
"step": 2520
},
{
"epoch": 0.14597276713593355,
"grad_norm": 0.7467900514602661,
"learning_rate": 9.730769230769232e-05,
"loss": 2.2526,
"step": 2530
},
{
"epoch": 0.14654973459496884,
"grad_norm": 0.7136681079864502,
"learning_rate": 9.76923076923077e-05,
"loss": 2.2539,
"step": 2540
},
{
"epoch": 0.14712670205400416,
"grad_norm": 0.4307969808578491,
"learning_rate": 9.807692307692307e-05,
"loss": 2.2483,
"step": 2550
},
{
"epoch": 0.14770366951303945,
"grad_norm": 0.5553449392318726,
"learning_rate": 9.846153846153848e-05,
"loss": 2.2388,
"step": 2560
},
{
"epoch": 0.14828063697207478,
"grad_norm": 0.8331669569015503,
"learning_rate": 9.884615384615386e-05,
"loss": 2.2571,
"step": 2570
},
{
"epoch": 0.1488576044311101,
"grad_norm": 0.437548965215683,
"learning_rate": 9.923076923076923e-05,
"loss": 2.2604,
"step": 2580
},
{
"epoch": 0.1494345718901454,
"grad_norm": 0.42178213596343994,
"learning_rate": 9.961538461538463e-05,
"loss": 2.2477,
"step": 2590
},
{
"epoch": 0.1500115393491807,
"grad_norm": 0.3375515341758728,
"learning_rate": 0.0001,
"loss": 2.2502,
"step": 2600
},
{
"epoch": 0.150588506808216,
"grad_norm": 0.42404118180274963,
"learning_rate": 9.999988631159617e-05,
"loss": 2.2535,
"step": 2610
},
{
"epoch": 0.15116547426725133,
"grad_norm": 0.7891207933425903,
"learning_rate": 9.999954524690165e-05,
"loss": 2.2498,
"step": 2620
},
{
"epoch": 0.15174244172628665,
"grad_norm": 1.1495307683944702,
"learning_rate": 9.999897680746749e-05,
"loss": 2.2489,
"step": 2630
},
{
"epoch": 0.15231940918532194,
"grad_norm": 1.0135457515716553,
"learning_rate": 9.999818099587866e-05,
"loss": 2.2668,
"step": 2640
},
{
"epoch": 0.15289637664435726,
"grad_norm": 0.676613450050354,
"learning_rate": 9.999715781575412e-05,
"loss": 2.2425,
"step": 2650
},
{
"epoch": 0.15347334410339256,
"grad_norm": 0.4525866210460663,
"learning_rate": 9.999590727174686e-05,
"loss": 2.23,
"step": 2660
},
{
"epoch": 0.15405031156242788,
"grad_norm": 0.35913950204849243,
"learning_rate": 9.999442936954375e-05,
"loss": 2.2487,
"step": 2670
},
{
"epoch": 0.1546272790214632,
"grad_norm": 0.5417553186416626,
"learning_rate": 9.999272411586562e-05,
"loss": 2.2526,
"step": 2680
},
{
"epoch": 0.1552042464804985,
"grad_norm": 0.3553147315979004,
"learning_rate": 9.999079151846715e-05,
"loss": 2.2419,
"step": 2690
},
{
"epoch": 0.1557812139395338,
"grad_norm": 0.4257926642894745,
"learning_rate": 9.998863158613692e-05,
"loss": 2.2468,
"step": 2700
},
{
"epoch": 0.1563581813985691,
"grad_norm": 0.3629404306411743,
"learning_rate": 9.99862443286973e-05,
"loss": 2.2514,
"step": 2710
},
{
"epoch": 0.15693514885760443,
"grad_norm": 0.3703051805496216,
"learning_rate": 9.998362975700442e-05,
"loss": 2.2419,
"step": 2720
},
{
"epoch": 0.15751211631663975,
"grad_norm": 0.5970242619514465,
"learning_rate": 9.998078788294811e-05,
"loss": 2.2395,
"step": 2730
},
{
"epoch": 0.15808908377567504,
"grad_norm": 0.5654362440109253,
"learning_rate": 9.997771871945195e-05,
"loss": 2.2408,
"step": 2740
},
{
"epoch": 0.15866605123471036,
"grad_norm": 0.3666626811027527,
"learning_rate": 9.997442228047304e-05,
"loss": 2.231,
"step": 2750
},
{
"epoch": 0.15924301869374569,
"grad_norm": 0.5000890493392944,
"learning_rate": 9.997089858100205e-05,
"loss": 2.2351,
"step": 2760
},
{
"epoch": 0.15981998615278098,
"grad_norm": 0.35561060905456543,
"learning_rate": 9.996714763706318e-05,
"loss": 2.2401,
"step": 2770
},
{
"epoch": 0.1603969536118163,
"grad_norm": 0.803989589214325,
"learning_rate": 9.996316946571391e-05,
"loss": 2.2356,
"step": 2780
},
{
"epoch": 0.1609739210708516,
"grad_norm": 0.7431883215904236,
"learning_rate": 9.995896408504518e-05,
"loss": 2.2353,
"step": 2790
},
{
"epoch": 0.16155088852988692,
"grad_norm": 0.3578578233718872,
"learning_rate": 9.995453151418106e-05,
"loss": 2.2419,
"step": 2800
},
{
"epoch": 0.16212785598892224,
"grad_norm": 0.3431973159313202,
"learning_rate": 9.994987177327885e-05,
"loss": 2.2316,
"step": 2810
},
{
"epoch": 0.16270482344795753,
"grad_norm": 0.3207797706127167,
"learning_rate": 9.99449848835289e-05,
"loss": 2.2381,
"step": 2820
},
{
"epoch": 0.16328179090699285,
"grad_norm": 0.5509916543960571,
"learning_rate": 9.99398708671545e-05,
"loss": 2.2297,
"step": 2830
},
{
"epoch": 0.16385875836602815,
"grad_norm": 0.3692900538444519,
"learning_rate": 9.993452974741185e-05,
"loss": 2.2148,
"step": 2840
},
{
"epoch": 0.16443572582506347,
"grad_norm": 0.9321035146713257,
"learning_rate": 9.992896154858987e-05,
"loss": 2.2323,
"step": 2850
},
{
"epoch": 0.1650126932840988,
"grad_norm": 0.5195390582084656,
"learning_rate": 9.992316629601012e-05,
"loss": 2.2435,
"step": 2860
},
{
"epoch": 0.16558966074313408,
"grad_norm": 0.407969206571579,
"learning_rate": 9.991714401602676e-05,
"loss": 2.2384,
"step": 2870
},
{
"epoch": 0.1661666282021694,
"grad_norm": 0.3711942136287689,
"learning_rate": 9.991089473602631e-05,
"loss": 2.2422,
"step": 2880
},
{
"epoch": 0.1667435956612047,
"grad_norm": 0.3882133364677429,
"learning_rate": 9.99044184844276e-05,
"loss": 2.2175,
"step": 2890
},
{
"epoch": 0.16732056312024002,
"grad_norm": 0.6136539578437805,
"learning_rate": 9.989771529068162e-05,
"loss": 2.2226,
"step": 2900
},
{
"epoch": 0.16789753057927534,
"grad_norm": 0.48447325825691223,
"learning_rate": 9.989078518527137e-05,
"loss": 2.2111,
"step": 2910
},
{
"epoch": 0.16847449803831063,
"grad_norm": 0.47318318486213684,
"learning_rate": 9.988362819971175e-05,
"loss": 2.2225,
"step": 2920
},
{
"epoch": 0.16905146549734595,
"grad_norm": 0.5245835185050964,
"learning_rate": 9.987624436654946e-05,
"loss": 2.2251,
"step": 2930
},
{
"epoch": 0.16962843295638125,
"grad_norm": 0.3767394721508026,
"learning_rate": 9.98686337193627e-05,
"loss": 2.2276,
"step": 2940
},
{
"epoch": 0.17020540041541657,
"grad_norm": 0.5148694515228271,
"learning_rate": 9.986079629276118e-05,
"loss": 2.2154,
"step": 2950
},
{
"epoch": 0.1707823678744519,
"grad_norm": 0.7400733232498169,
"learning_rate": 9.985273212238586e-05,
"loss": 2.2322,
"step": 2960
},
{
"epoch": 0.17135933533348718,
"grad_norm": 0.5233339071273804,
"learning_rate": 9.984444124490888e-05,
"loss": 2.2257,
"step": 2970
},
{
"epoch": 0.1719363027925225,
"grad_norm": 0.3448404371738434,
"learning_rate": 9.98359236980333e-05,
"loss": 2.2264,
"step": 2980
},
{
"epoch": 0.1725132702515578,
"grad_norm": 0.38899293541908264,
"learning_rate": 9.982717952049296e-05,
"loss": 2.2255,
"step": 2990
},
{
"epoch": 0.17309023771059312,
"grad_norm": 0.38548049330711365,
"learning_rate": 9.98182087520523e-05,
"loss": 2.2258,
"step": 3000
},
{
"epoch": 0.17366720516962844,
"grad_norm": 0.4106454849243164,
"learning_rate": 9.980901143350627e-05,
"loss": 2.2252,
"step": 3010
},
{
"epoch": 0.17424417262866373,
"grad_norm": 0.31857094168663025,
"learning_rate": 9.979958760667997e-05,
"loss": 2.2277,
"step": 3020
},
{
"epoch": 0.17482114008769906,
"grad_norm": 0.41078636050224304,
"learning_rate": 9.97899373144286e-05,
"loss": 2.2281,
"step": 3030
},
{
"epoch": 0.17539810754673438,
"grad_norm": 0.411409467458725,
"learning_rate": 9.978006060063723e-05,
"loss": 2.2331,
"step": 3040
},
{
"epoch": 0.17597507500576967,
"grad_norm": 0.9199445247650146,
"learning_rate": 9.976995751022054e-05,
"loss": 2.2305,
"step": 3050
},
{
"epoch": 0.176552042464805,
"grad_norm": 0.5076198577880859,
"learning_rate": 9.975962808912271e-05,
"loss": 2.2105,
"step": 3060
},
{
"epoch": 0.17712900992384029,
"grad_norm": 0.5077535510063171,
"learning_rate": 9.974907238431716e-05,
"loss": 2.2286,
"step": 3070
},
{
"epoch": 0.1777059773828756,
"grad_norm": 0.398752897977829,
"learning_rate": 9.973829044380637e-05,
"loss": 2.2134,
"step": 3080
},
{
"epoch": 0.17828294484191093,
"grad_norm": 0.32480180263519287,
"learning_rate": 9.972728231662157e-05,
"loss": 2.2225,
"step": 3090
},
{
"epoch": 0.17885991230094622,
"grad_norm": 0.49645814299583435,
"learning_rate": 9.971604805282259e-05,
"loss": 2.2179,
"step": 3100
},
{
"epoch": 0.17943687975998154,
"grad_norm": 0.6805341243743896,
"learning_rate": 9.97045877034977e-05,
"loss": 2.2093,
"step": 3110
},
{
"epoch": 0.18001384721901684,
"grad_norm": 0.3878099024295807,
"learning_rate": 9.969290132076326e-05,
"loss": 2.2088,
"step": 3120
},
{
"epoch": 0.18059081467805216,
"grad_norm": 0.34581634402275085,
"learning_rate": 9.968098895776348e-05,
"loss": 2.232,
"step": 3130
},
{
"epoch": 0.18116778213708748,
"grad_norm": 0.36607232689857483,
"learning_rate": 9.966885066867027e-05,
"loss": 2.2202,
"step": 3140
},
{
"epoch": 0.18174474959612277,
"grad_norm": 0.3712445795536041,
"learning_rate": 9.965648650868293e-05,
"loss": 2.203,
"step": 3150
},
{
"epoch": 0.1823217170551581,
"grad_norm": 0.3221428096294403,
"learning_rate": 9.964389653402797e-05,
"loss": 2.2184,
"step": 3160
},
{
"epoch": 0.1828986845141934,
"grad_norm": 0.4847376048564911,
"learning_rate": 9.963108080195869e-05,
"loss": 2.2127,
"step": 3170
},
{
"epoch": 0.1834756519732287,
"grad_norm": 0.5290243625640869,
"learning_rate": 9.961803937075516e-05,
"loss": 2.2163,
"step": 3180
},
{
"epoch": 0.18405261943226403,
"grad_norm": 0.3758445084095001,
"learning_rate": 9.960477229972372e-05,
"loss": 2.2056,
"step": 3190
},
{
"epoch": 0.18462958689129932,
"grad_norm": 0.3582039773464203,
"learning_rate": 9.959127964919684e-05,
"loss": 2.214,
"step": 3200
},
{
"epoch": 0.18520655435033465,
"grad_norm": 0.4270158112049103,
"learning_rate": 9.957756148053288e-05,
"loss": 2.2075,
"step": 3210
},
{
"epoch": 0.18578352180936994,
"grad_norm": 0.5898721218109131,
"learning_rate": 9.956361785611568e-05,
"loss": 2.2099,
"step": 3220
},
{
"epoch": 0.18636048926840526,
"grad_norm": 0.3439638912677765,
"learning_rate": 9.95494488393544e-05,
"loss": 2.2077,
"step": 3230
},
{
"epoch": 0.18693745672744058,
"grad_norm": 0.37806159257888794,
"learning_rate": 9.953505449468313e-05,
"loss": 2.2226,
"step": 3240
},
{
"epoch": 0.18751442418647588,
"grad_norm": 0.4258742332458496,
"learning_rate": 9.952043488756066e-05,
"loss": 2.2137,
"step": 3250
},
{
"epoch": 0.1880913916455112,
"grad_norm": 0.4259757697582245,
"learning_rate": 9.950559008447024e-05,
"loss": 2.21,
"step": 3260
},
{
"epoch": 0.1886683591045465,
"grad_norm": 0.3395422399044037,
"learning_rate": 9.94905201529191e-05,
"loss": 2.211,
"step": 3270
},
{
"epoch": 0.1892453265635818,
"grad_norm": 0.407440721988678,
"learning_rate": 9.94752251614383e-05,
"loss": 2.215,
"step": 3280
},
{
"epoch": 0.18982229402261713,
"grad_norm": 0.3814033269882202,
"learning_rate": 9.945970517958238e-05,
"loss": 2.2249,
"step": 3290
},
{
"epoch": 0.19039926148165243,
"grad_norm": 0.5478818416595459,
"learning_rate": 9.944396027792904e-05,
"loss": 2.2032,
"step": 3300
},
{
"epoch": 0.19097622894068775,
"grad_norm": 0.5092093348503113,
"learning_rate": 9.942799052807874e-05,
"loss": 2.2157,
"step": 3310
},
{
"epoch": 0.19155319639972307,
"grad_norm": 0.4795791506767273,
"learning_rate": 9.941179600265454e-05,
"loss": 2.2013,
"step": 3320
},
{
"epoch": 0.19213016385875836,
"grad_norm": 0.6791263222694397,
"learning_rate": 9.93953767753016e-05,
"loss": 2.1997,
"step": 3330
},
{
"epoch": 0.19270713131779368,
"grad_norm": 0.48331305384635925,
"learning_rate": 9.937873292068696e-05,
"loss": 2.1939,
"step": 3340
},
{
"epoch": 0.19328409877682898,
"grad_norm": 0.41813355684280396,
"learning_rate": 9.936186451449916e-05,
"loss": 2.2098,
"step": 3350
},
{
"epoch": 0.1938610662358643,
"grad_norm": 0.3425246775150299,
"learning_rate": 9.934477163344787e-05,
"loss": 2.2023,
"step": 3360
},
{
"epoch": 0.19443803369489962,
"grad_norm": 0.5277093648910522,
"learning_rate": 9.93274543552636e-05,
"loss": 2.199,
"step": 3370
},
{
"epoch": 0.1950150011539349,
"grad_norm": 0.4232868552207947,
"learning_rate": 9.930991275869727e-05,
"loss": 2.1923,
"step": 3380
},
{
"epoch": 0.19559196861297024,
"grad_norm": 0.6466622352600098,
"learning_rate": 9.929214692351998e-05,
"loss": 2.2013,
"step": 3390
},
{
"epoch": 0.19616893607200553,
"grad_norm": 0.48593297600746155,
"learning_rate": 9.927415693052245e-05,
"loss": 2.2042,
"step": 3400
},
{
"epoch": 0.19674590353104085,
"grad_norm": 0.6403309106826782,
"learning_rate": 9.925594286151487e-05,
"loss": 2.188,
"step": 3410
},
{
"epoch": 0.19732287099007617,
"grad_norm": 0.4060690402984619,
"learning_rate": 9.923750479932635e-05,
"loss": 2.2146,
"step": 3420
},
{
"epoch": 0.19789983844911146,
"grad_norm": 0.3018631339073181,
"learning_rate": 9.921884282780463e-05,
"loss": 2.2013,
"step": 3430
},
{
"epoch": 0.1984768059081468,
"grad_norm": 0.3768801689147949,
"learning_rate": 9.919995703181573e-05,
"loss": 2.1963,
"step": 3440
},
{
"epoch": 0.19905377336718208,
"grad_norm": 0.5511378645896912,
"learning_rate": 9.918084749724348e-05,
"loss": 2.1979,
"step": 3450
},
{
"epoch": 0.1996307408262174,
"grad_norm": 0.5576174855232239,
"learning_rate": 9.916151431098918e-05,
"loss": 2.1961,
"step": 3460
},
{
"epoch": 0.20020770828525272,
"grad_norm": 0.4171042740345001,
"learning_rate": 9.91419575609712e-05,
"loss": 2.2061,
"step": 3470
},
{
"epoch": 0.20078467574428802,
"grad_norm": 0.3103582561016083,
"learning_rate": 9.912217733612455e-05,
"loss": 2.1972,
"step": 3480
},
{
"epoch": 0.20136164320332334,
"grad_norm": 0.39360588788986206,
"learning_rate": 9.910217372640052e-05,
"loss": 2.2038,
"step": 3490
},
{
"epoch": 0.20193861066235863,
"grad_norm": 0.3420889377593994,
"learning_rate": 9.908194682276626e-05,
"loss": 2.2043,
"step": 3500
},
{
"epoch": 0.20251557812139395,
"grad_norm": 0.30040401220321655,
"learning_rate": 9.906149671720436e-05,
"loss": 2.2147,
"step": 3510
},
{
"epoch": 0.20309254558042927,
"grad_norm": 0.3862975835800171,
"learning_rate": 9.904082350271238e-05,
"loss": 2.216,
"step": 3520
},
{
"epoch": 0.20366951303946457,
"grad_norm": 0.5109495520591736,
"learning_rate": 9.901992727330254e-05,
"loss": 2.194,
"step": 3530
},
{
"epoch": 0.2042464804984999,
"grad_norm": 0.5943346619606018,
"learning_rate": 9.899880812400118e-05,
"loss": 2.1829,
"step": 3540
},
{
"epoch": 0.20482344795753518,
"grad_norm": 0.38951772451400757,
"learning_rate": 9.89774661508484e-05,
"loss": 2.2003,
"step": 3550
},
{
"epoch": 0.2054004154165705,
"grad_norm": 0.4462457001209259,
"learning_rate": 9.895590145089759e-05,
"loss": 2.2059,
"step": 3560
},
{
"epoch": 0.20597738287560582,
"grad_norm": 0.3021964132785797,
"learning_rate": 9.8934114122215e-05,
"loss": 2.1936,
"step": 3570
},
{
"epoch": 0.20655435033464112,
"grad_norm": 0.3611934781074524,
"learning_rate": 9.891210426387931e-05,
"loss": 2.2004,
"step": 3580
},
{
"epoch": 0.20713131779367644,
"grad_norm": 0.6618946194648743,
"learning_rate": 9.888987197598115e-05,
"loss": 2.1984,
"step": 3590
},
{
"epoch": 0.20770828525271176,
"grad_norm": 0.6646985411643982,
"learning_rate": 9.886741735962261e-05,
"loss": 2.1952,
"step": 3600
},
{
"epoch": 0.20828525271174705,
"grad_norm": 0.7527153491973877,
"learning_rate": 9.884474051691693e-05,
"loss": 2.1803,
"step": 3610
},
{
"epoch": 0.20886222017078238,
"grad_norm": 0.5738996267318726,
"learning_rate": 9.882184155098784e-05,
"loss": 2.2139,
"step": 3620
},
{
"epoch": 0.20943918762981767,
"grad_norm": 0.3935083746910095,
"learning_rate": 9.879872056596922e-05,
"loss": 2.189,
"step": 3630
},
{
"epoch": 0.210016155088853,
"grad_norm": 0.4685090482234955,
"learning_rate": 9.877537766700458e-05,
"loss": 2.1919,
"step": 3640
},
{
"epoch": 0.2105931225478883,
"grad_norm": 0.4964298903942108,
"learning_rate": 9.87518129602466e-05,
"loss": 2.1843,
"step": 3650
},
{
"epoch": 0.2111700900069236,
"grad_norm": 0.33545660972595215,
"learning_rate": 9.872802655285663e-05,
"loss": 2.1826,
"step": 3660
},
{
"epoch": 0.21174705746595893,
"grad_norm": 0.3587926924228668,
"learning_rate": 9.870401855300423e-05,
"loss": 2.2016,
"step": 3670
},
{
"epoch": 0.21232402492499422,
"grad_norm": 0.6967931985855103,
"learning_rate": 9.867978906986665e-05,
"loss": 2.1923,
"step": 3680
},
{
"epoch": 0.21290099238402954,
"grad_norm": 0.5460529923439026,
"learning_rate": 9.865533821362832e-05,
"loss": 2.1825,
"step": 3690
},
{
"epoch": 0.21347795984306486,
"grad_norm": 0.32380005717277527,
"learning_rate": 9.863066609548042e-05,
"loss": 2.194,
"step": 3700
},
{
"epoch": 0.21405492730210016,
"grad_norm": 0.4450247287750244,
"learning_rate": 9.860577282762027e-05,
"loss": 2.2023,
"step": 3710
},
{
"epoch": 0.21463189476113548,
"grad_norm": 0.3208085298538208,
"learning_rate": 9.858065852325094e-05,
"loss": 2.1782,
"step": 3720
},
{
"epoch": 0.21520886222017077,
"grad_norm": 0.34282881021499634,
"learning_rate": 9.85553232965806e-05,
"loss": 2.186,
"step": 3730
},
{
"epoch": 0.2157858296792061,
"grad_norm": 0.4419040381908417,
"learning_rate": 9.852976726282213e-05,
"loss": 2.1831,
"step": 3740
},
{
"epoch": 0.21636279713824141,
"grad_norm": 0.706082820892334,
"learning_rate": 9.85039905381925e-05,
"loss": 2.1929,
"step": 3750
},
{
"epoch": 0.2169397645972767,
"grad_norm": 0.2975011467933655,
"learning_rate": 9.847799323991234e-05,
"loss": 2.1937,
"step": 3760
},
{
"epoch": 0.21751673205631203,
"grad_norm": 0.36291176080703735,
"learning_rate": 9.845177548620525e-05,
"loss": 2.1949,
"step": 3770
},
{
"epoch": 0.21809369951534732,
"grad_norm": 0.31104329228401184,
"learning_rate": 9.842533739629744e-05,
"loss": 2.1829,
"step": 3780
},
{
"epoch": 0.21867066697438264,
"grad_norm": 0.3598311245441437,
"learning_rate": 9.83986790904171e-05,
"loss": 2.1724,
"step": 3790
},
{
"epoch": 0.21924763443341796,
"grad_norm": 0.717742383480072,
"learning_rate": 9.837180068979379e-05,
"loss": 2.1984,
"step": 3800
},
{
"epoch": 0.21982460189245326,
"grad_norm": 0.5237902402877808,
"learning_rate": 9.834470231665806e-05,
"loss": 2.1791,
"step": 3810
},
{
"epoch": 0.22040156935148858,
"grad_norm": 0.633453905582428,
"learning_rate": 9.83173840942407e-05,
"loss": 2.2034,
"step": 3820
},
{
"epoch": 0.22097853681052387,
"grad_norm": 0.5341862440109253,
"learning_rate": 9.828984614677234e-05,
"loss": 2.1894,
"step": 3830
},
{
"epoch": 0.2215555042695592,
"grad_norm": 0.38566407561302185,
"learning_rate": 9.826208859948279e-05,
"loss": 2.1807,
"step": 3840
},
{
"epoch": 0.22213247172859452,
"grad_norm": 0.4219278395175934,
"learning_rate": 9.823411157860048e-05,
"loss": 2.1827,
"step": 3850
},
{
"epoch": 0.2227094391876298,
"grad_norm": 0.3236340582370758,
"learning_rate": 9.820591521135193e-05,
"loss": 2.1748,
"step": 3860
},
{
"epoch": 0.22328640664666513,
"grad_norm": 0.335493803024292,
"learning_rate": 9.817749962596115e-05,
"loss": 2.1778,
"step": 3870
},
{
"epoch": 0.22386337410570045,
"grad_norm": 0.3819381594657898,
"learning_rate": 9.814886495164905e-05,
"loss": 2.1836,
"step": 3880
},
{
"epoch": 0.22444034156473575,
"grad_norm": 0.370803564786911,
"learning_rate": 9.812001131863281e-05,
"loss": 2.1896,
"step": 3890
},
{
"epoch": 0.22501730902377107,
"grad_norm": 0.406332790851593,
"learning_rate": 9.809093885812543e-05,
"loss": 2.187,
"step": 3900
},
{
"epoch": 0.22559427648280636,
"grad_norm": 0.6314796209335327,
"learning_rate": 9.806164770233493e-05,
"loss": 2.196,
"step": 3910
},
{
"epoch": 0.22617124394184168,
"grad_norm": 0.6008214950561523,
"learning_rate": 9.803213798446389e-05,
"loss": 2.1867,
"step": 3920
},
{
"epoch": 0.226748211400877,
"grad_norm": 0.41306042671203613,
"learning_rate": 9.800240983870886e-05,
"loss": 2.1802,
"step": 3930
},
{
"epoch": 0.2273251788599123,
"grad_norm": 0.3542007803916931,
"learning_rate": 9.797246340025962e-05,
"loss": 2.1872,
"step": 3940
},
{
"epoch": 0.22790214631894762,
"grad_norm": 0.35550656914711,
"learning_rate": 9.79422988052987e-05,
"loss": 2.1811,
"step": 3950
},
{
"epoch": 0.2284791137779829,
"grad_norm": 0.4810178875923157,
"learning_rate": 9.791191619100069e-05,
"loss": 2.1716,
"step": 3960
},
{
"epoch": 0.22905608123701823,
"grad_norm": 0.4199106693267822,
"learning_rate": 9.78813156955316e-05,
"loss": 2.1816,
"step": 3970
},
{
"epoch": 0.22963304869605355,
"grad_norm": 0.5093641877174377,
"learning_rate": 9.785049745804833e-05,
"loss": 2.1931,
"step": 3980
},
{
"epoch": 0.23021001615508885,
"grad_norm": 0.32428303360939026,
"learning_rate": 9.78194616186979e-05,
"loss": 2.1839,
"step": 3990
},
{
"epoch": 0.23078698361412417,
"grad_norm": 0.4841248691082001,
"learning_rate": 9.778820831861693e-05,
"loss": 2.1695,
"step": 4000
},
{
"epoch": 0.23136395107315946,
"grad_norm": 0.4829302430152893,
"learning_rate": 9.775673769993092e-05,
"loss": 2.1786,
"step": 4010
},
{
"epoch": 0.23194091853219478,
"grad_norm": 0.6401785612106323,
"learning_rate": 9.772504990575365e-05,
"loss": 2.1814,
"step": 4020
},
{
"epoch": 0.2325178859912301,
"grad_norm": 0.3681984543800354,
"learning_rate": 9.76931450801865e-05,
"loss": 2.1719,
"step": 4030
},
{
"epoch": 0.2330948534502654,
"grad_norm": 0.35280972719192505,
"learning_rate": 9.766102336831785e-05,
"loss": 2.1695,
"step": 4040
},
{
"epoch": 0.23367182090930072,
"grad_norm": 0.3713974058628082,
"learning_rate": 9.76286849162223e-05,
"loss": 2.1657,
"step": 4050
},
{
"epoch": 0.23424878836833601,
"grad_norm": 0.7190024852752686,
"learning_rate": 9.759612987096016e-05,
"loss": 2.182,
"step": 4060
},
{
"epoch": 0.23482575582737134,
"grad_norm": 0.6214960217475891,
"learning_rate": 9.756335838057665e-05,
"loss": 2.1656,
"step": 4070
},
{
"epoch": 0.23540272328640666,
"grad_norm": 0.28883612155914307,
"learning_rate": 9.753037059410134e-05,
"loss": 2.1723,
"step": 4080
},
{
"epoch": 0.23597969074544195,
"grad_norm": 0.4463951289653778,
"learning_rate": 9.749716666154737e-05,
"loss": 2.1799,
"step": 4090
},
{
"epoch": 0.23655665820447727,
"grad_norm": 0.6259954571723938,
"learning_rate": 9.746374673391082e-05,
"loss": 2.173,
"step": 4100
},
{
"epoch": 0.23713362566351256,
"grad_norm": 0.35418662428855896,
"learning_rate": 9.743011096317e-05,
"loss": 2.1809,
"step": 4110
},
{
"epoch": 0.2377105931225479,
"grad_norm": 0.5010941028594971,
"learning_rate": 9.739625950228484e-05,
"loss": 2.179,
"step": 4120
},
{
"epoch": 0.2382875605815832,
"grad_norm": 0.37506598234176636,
"learning_rate": 9.736219250519604e-05,
"loss": 2.1754,
"step": 4130
},
{
"epoch": 0.2388645280406185,
"grad_norm": 0.5773160457611084,
"learning_rate": 9.732791012682452e-05,
"loss": 2.1694,
"step": 4140
},
{
"epoch": 0.23944149549965382,
"grad_norm": 0.3286305069923401,
"learning_rate": 9.729341252307063e-05,
"loss": 2.1704,
"step": 4150
},
{
"epoch": 0.24001846295868912,
"grad_norm": 0.40018871426582336,
"learning_rate": 9.725869985081346e-05,
"loss": 2.1802,
"step": 4160
},
{
"epoch": 0.24059543041772444,
"grad_norm": 0.46409788727760315,
"learning_rate": 9.722377226791017e-05,
"loss": 2.1773,
"step": 4170
},
{
"epoch": 0.24117239787675976,
"grad_norm": 0.4142248332500458,
"learning_rate": 9.718862993319518e-05,
"loss": 2.1747,
"step": 4180
},
{
"epoch": 0.24174936533579505,
"grad_norm": 0.5945191979408264,
"learning_rate": 9.715327300647953e-05,
"loss": 2.1689,
"step": 4190
},
{
"epoch": 0.24232633279483037,
"grad_norm": 0.44592514634132385,
"learning_rate": 9.711770164855015e-05,
"loss": 2.1694,
"step": 4200
},
{
"epoch": 0.2429033002538657,
"grad_norm": 0.2986921966075897,
"learning_rate": 9.708191602116905e-05,
"loss": 2.1637,
"step": 4210
},
{
"epoch": 0.243480267712901,
"grad_norm": 0.4355255663394928,
"learning_rate": 9.704591628707267e-05,
"loss": 2.152,
"step": 4220
},
{
"epoch": 0.2440572351719363,
"grad_norm": 0.46000930666923523,
"learning_rate": 9.70097026099711e-05,
"loss": 2.1636,
"step": 4230
},
{
"epoch": 0.2446342026309716,
"grad_norm": 0.3301403820514679,
"learning_rate": 9.697327515454735e-05,
"loss": 2.1596,
"step": 4240
},
{
"epoch": 0.24521117009000692,
"grad_norm": 0.4323584735393524,
"learning_rate": 9.693663408645658e-05,
"loss": 2.1723,
"step": 4250
},
{
"epoch": 0.24578813754904225,
"grad_norm": 0.28445109724998474,
"learning_rate": 9.689977957232539e-05,
"loss": 2.1728,
"step": 4260
},
{
"epoch": 0.24636510500807754,
"grad_norm": 0.3563917279243469,
"learning_rate": 9.686271177975099e-05,
"loss": 2.162,
"step": 4270
},
{
"epoch": 0.24694207246711286,
"grad_norm": 0.29485827684402466,
"learning_rate": 9.682543087730054e-05,
"loss": 2.1781,
"step": 4280
},
{
"epoch": 0.24751903992614815,
"grad_norm": 0.4161454141139984,
"learning_rate": 9.678793703451028e-05,
"loss": 2.17,
"step": 4290
},
{
"epoch": 0.24809600738518348,
"grad_norm": 0.3382590413093567,
"learning_rate": 9.67502304218848e-05,
"loss": 2.1748,
"step": 4300
},
{
"epoch": 0.2486729748442188,
"grad_norm": 0.3781547248363495,
"learning_rate": 9.671231121089629e-05,
"loss": 2.152,
"step": 4310
},
{
"epoch": 0.2492499423032541,
"grad_norm": 0.2784830331802368,
"learning_rate": 9.667417957398375e-05,
"loss": 2.1656,
"step": 4320
},
{
"epoch": 0.2498269097622894,
"grad_norm": 0.3041353225708008,
"learning_rate": 9.663583568455214e-05,
"loss": 2.1665,
"step": 4330
},
{
"epoch": 0.25040387722132473,
"grad_norm": 0.30996307730674744,
"learning_rate": 9.659727971697174e-05,
"loss": 2.1701,
"step": 4340
},
{
"epoch": 0.25098084468036,
"grad_norm": 0.42379140853881836,
"learning_rate": 9.655851184657716e-05,
"loss": 2.158,
"step": 4350
},
{
"epoch": 0.2515578121393953,
"grad_norm": 0.5364662408828735,
"learning_rate": 9.65195322496667e-05,
"loss": 2.1658,
"step": 4360
},
{
"epoch": 0.25213477959843067,
"grad_norm": 0.447578102350235,
"learning_rate": 9.648034110350151e-05,
"loss": 2.1626,
"step": 4370
},
{
"epoch": 0.25271174705746596,
"grad_norm": 0.4996080696582794,
"learning_rate": 9.644093858630471e-05,
"loss": 2.1659,
"step": 4380
},
{
"epoch": 0.25328871451650126,
"grad_norm": 0.2662523090839386,
"learning_rate": 9.640132487726069e-05,
"loss": 2.1824,
"step": 4390
},
{
"epoch": 0.2538656819755366,
"grad_norm": 0.2966640293598175,
"learning_rate": 9.636150015651422e-05,
"loss": 2.1665,
"step": 4400
},
{
"epoch": 0.2544426494345719,
"grad_norm": 0.3168790638446808,
"learning_rate": 9.632146460516967e-05,
"loss": 2.1814,
"step": 4410
},
{
"epoch": 0.2550196168936072,
"grad_norm": 0.7422902584075928,
"learning_rate": 9.628121840529015e-05,
"loss": 2.1674,
"step": 4420
},
{
"epoch": 0.2555965843526425,
"grad_norm": 0.447091668844223,
"learning_rate": 9.624076173989672e-05,
"loss": 2.1606,
"step": 4430
},
{
"epoch": 0.25617355181167784,
"grad_norm": 0.37554842233657837,
"learning_rate": 9.620009479296748e-05,
"loss": 2.158,
"step": 4440
},
{
"epoch": 0.25675051927071313,
"grad_norm": 0.353153795003891,
"learning_rate": 9.61592177494369e-05,
"loss": 2.1627,
"step": 4450
},
{
"epoch": 0.2573274867297484,
"grad_norm": 0.4235435724258423,
"learning_rate": 9.611813079519479e-05,
"loss": 2.1617,
"step": 4460
},
{
"epoch": 0.25790445418878377,
"grad_norm": 0.34072136878967285,
"learning_rate": 9.607683411708555e-05,
"loss": 2.1651,
"step": 4470
},
{
"epoch": 0.25848142164781907,
"grad_norm": 0.33193790912628174,
"learning_rate": 9.603532790290733e-05,
"loss": 2.1592,
"step": 4480
},
{
"epoch": 0.25905838910685436,
"grad_norm": 0.3969743549823761,
"learning_rate": 9.599361234141113e-05,
"loss": 2.1709,
"step": 4490
},
{
"epoch": 0.2596353565658897,
"grad_norm": 0.5880393981933594,
"learning_rate": 9.595168762229998e-05,
"loss": 2.1589,
"step": 4500
},
{
"epoch": 0.260212324024925,
"grad_norm": 0.4169381856918335,
"learning_rate": 9.590955393622804e-05,
"loss": 2.166,
"step": 4510
},
{
"epoch": 0.2607892914839603,
"grad_norm": 0.5424039959907532,
"learning_rate": 9.586721147479983e-05,
"loss": 2.1592,
"step": 4520
},
{
"epoch": 0.2613662589429956,
"grad_norm": 0.6034404039382935,
"learning_rate": 9.582466043056914e-05,
"loss": 2.1546,
"step": 4530
},
{
"epoch": 0.26194322640203094,
"grad_norm": 0.34911131858825684,
"learning_rate": 9.578190099703845e-05,
"loss": 2.1667,
"step": 4540
},
{
"epoch": 0.26252019386106623,
"grad_norm": 0.40795624256134033,
"learning_rate": 9.573893336865778e-05,
"loss": 2.1544,
"step": 4550
},
{
"epoch": 0.2630971613201015,
"grad_norm": 0.36005672812461853,
"learning_rate": 9.569575774082401e-05,
"loss": 2.1547,
"step": 4560
},
{
"epoch": 0.2636741287791369,
"grad_norm": 0.3152981102466583,
"learning_rate": 9.565237430987986e-05,
"loss": 2.1557,
"step": 4570
},
{
"epoch": 0.26425109623817217,
"grad_norm": 0.33613887429237366,
"learning_rate": 9.560878327311302e-05,
"loss": 2.1617,
"step": 4580
},
{
"epoch": 0.26482806369720746,
"grad_norm": 0.4851570427417755,
"learning_rate": 9.556498482875535e-05,
"loss": 2.1462,
"step": 4590
},
{
"epoch": 0.2654050311562428,
"grad_norm": 0.2919342517852783,
"learning_rate": 9.552097917598184e-05,
"loss": 2.1683,
"step": 4600
},
{
"epoch": 0.2659819986152781,
"grad_norm": 0.31903278827667236,
"learning_rate": 9.547676651490978e-05,
"loss": 2.1608,
"step": 4610
},
{
"epoch": 0.2665589660743134,
"grad_norm": 1.5958030223846436,
"learning_rate": 9.543234704659784e-05,
"loss": 2.1548,
"step": 4620
},
{
"epoch": 0.26713593353334875,
"grad_norm": 0.26471462845802307,
"learning_rate": 9.538772097304521e-05,
"loss": 2.1566,
"step": 4630
},
{
"epoch": 0.26771290099238404,
"grad_norm": 0.45223453640937805,
"learning_rate": 9.534288849719049e-05,
"loss": 2.1653,
"step": 4640
},
{
"epoch": 0.26828986845141933,
"grad_norm": 0.33436188101768494,
"learning_rate": 9.529784982291105e-05,
"loss": 2.1547,
"step": 4650
},
{
"epoch": 0.2688668359104546,
"grad_norm": 0.34297022223472595,
"learning_rate": 9.525260515502185e-05,
"loss": 2.1604,
"step": 4660
},
{
"epoch": 0.26944380336949,
"grad_norm": 0.572670578956604,
"learning_rate": 9.520715469927468e-05,
"loss": 2.155,
"step": 4670
},
{
"epoch": 0.27002077082852527,
"grad_norm": 0.28033772110939026,
"learning_rate": 9.516149866235713e-05,
"loss": 2.1731,
"step": 4680
},
{
"epoch": 0.27059773828756056,
"grad_norm": 0.460891455411911,
"learning_rate": 9.511563725189165e-05,
"loss": 2.153,
"step": 4690
},
{
"epoch": 0.2711747057465959,
"grad_norm": 0.34910324215888977,
"learning_rate": 9.506957067643469e-05,
"loss": 2.1617,
"step": 4700
},
{
"epoch": 0.2717516732056312,
"grad_norm": 0.4306391775608063,
"learning_rate": 9.502329914547567e-05,
"loss": 2.1362,
"step": 4710
},
{
"epoch": 0.2723286406646665,
"grad_norm": 0.39422643184661865,
"learning_rate": 9.497682286943604e-05,
"loss": 2.1513,
"step": 4720
},
{
"epoch": 0.27290560812370185,
"grad_norm": 0.3839055001735687,
"learning_rate": 9.493014205966832e-05,
"loss": 2.1553,
"step": 4730
},
{
"epoch": 0.27348257558273714,
"grad_norm": 0.31915026903152466,
"learning_rate": 9.488325692845522e-05,
"loss": 2.155,
"step": 4740
},
{
"epoch": 0.27405954304177244,
"grad_norm": 0.29322880506515503,
"learning_rate": 9.483616768900854e-05,
"loss": 2.1543,
"step": 4750
},
{
"epoch": 0.27463651050080773,
"grad_norm": 0.4080626666545868,
"learning_rate": 9.478887455546832e-05,
"loss": 2.1478,
"step": 4760
},
{
"epoch": 0.2752134779598431,
"grad_norm": 0.42490825057029724,
"learning_rate": 9.47413777429018e-05,
"loss": 2.1449,
"step": 4770
},
{
"epoch": 0.27579044541887837,
"grad_norm": 0.29601389169692993,
"learning_rate": 9.469367746730243e-05,
"loss": 2.1489,
"step": 4780
},
{
"epoch": 0.27636741287791367,
"grad_norm": 0.28761008381843567,
"learning_rate": 9.464577394558893e-05,
"loss": 2.1542,
"step": 4790
},
{
"epoch": 0.276944380336949,
"grad_norm": 0.35729655623435974,
"learning_rate": 9.459766739560433e-05,
"loss": 2.146,
"step": 4800
},
{
"epoch": 0.2775213477959843,
"grad_norm": 0.5332178473472595,
"learning_rate": 9.45493580361149e-05,
"loss": 2.1498,
"step": 4810
},
{
"epoch": 0.2780983152550196,
"grad_norm": 0.3445925712585449,
"learning_rate": 9.450084608680915e-05,
"loss": 2.1369,
"step": 4820
},
{
"epoch": 0.27867528271405495,
"grad_norm": 0.37524721026420593,
"learning_rate": 9.4452131768297e-05,
"loss": 2.1465,
"step": 4830
},
{
"epoch": 0.27925225017309024,
"grad_norm": 0.35284727811813354,
"learning_rate": 9.440321530210852e-05,
"loss": 2.145,
"step": 4840
},
{
"epoch": 0.27982921763212554,
"grad_norm": 0.387416273355484,
"learning_rate": 9.435409691069312e-05,
"loss": 2.147,
"step": 4850
},
{
"epoch": 0.28040618509116083,
"grad_norm": 0.33263909816741943,
"learning_rate": 9.430477681741847e-05,
"loss": 2.1449,
"step": 4860
},
{
"epoch": 0.2809831525501962,
"grad_norm": 0.31773439049720764,
"learning_rate": 9.425525524656948e-05,
"loss": 2.1528,
"step": 4870
},
{
"epoch": 0.2815601200092315,
"grad_norm": 0.3170188367366791,
"learning_rate": 9.420553242334727e-05,
"loss": 2.1564,
"step": 4880
},
{
"epoch": 0.28213708746826677,
"grad_norm": 0.4686289429664612,
"learning_rate": 9.415560857386817e-05,
"loss": 2.1557,
"step": 4890
},
{
"epoch": 0.2827140549273021,
"grad_norm": 0.2960648834705353,
"learning_rate": 9.410548392516271e-05,
"loss": 2.1348,
"step": 4900
},
{
"epoch": 0.2832910223863374,
"grad_norm": 0.6153275370597839,
"learning_rate": 9.405515870517455e-05,
"loss": 2.1498,
"step": 4910
},
{
"epoch": 0.2838679898453727,
"grad_norm": 0.549048900604248,
"learning_rate": 9.400463314275943e-05,
"loss": 2.1461,
"step": 4920
},
{
"epoch": 0.28444495730440805,
"grad_norm": 0.2764069736003876,
"learning_rate": 9.395390746768415e-05,
"loss": 2.1421,
"step": 4930
},
{
"epoch": 0.28502192476344335,
"grad_norm": 0.5139395594596863,
"learning_rate": 9.39029819106256e-05,
"loss": 2.1489,
"step": 4940
},
{
"epoch": 0.28559889222247864,
"grad_norm": 0.30540773272514343,
"learning_rate": 9.385185670316957e-05,
"loss": 2.1571,
"step": 4950
},
{
"epoch": 0.286175859681514,
"grad_norm": 0.3731764853000641,
"learning_rate": 9.380053207780977e-05,
"loss": 2.146,
"step": 4960
},
{
"epoch": 0.2867528271405493,
"grad_norm": 0.3358662724494934,
"learning_rate": 9.37490082679468e-05,
"loss": 2.1426,
"step": 4970
},
{
"epoch": 0.2873297945995846,
"grad_norm": 0.37629401683807373,
"learning_rate": 9.369728550788706e-05,
"loss": 2.1478,
"step": 4980
},
{
"epoch": 0.28790676205861987,
"grad_norm": 0.2608135938644409,
"learning_rate": 9.364536403284167e-05,
"loss": 2.147,
"step": 4990
},
{
"epoch": 0.2884837295176552,
"grad_norm": 0.3051717281341553,
"learning_rate": 9.359324407892542e-05,
"loss": 2.1509,
"step": 5000
},
{
"epoch": 0.2890606969766905,
"grad_norm": 0.2828206717967987,
"learning_rate": 9.354092588315566e-05,
"loss": 2.1467,
"step": 5010
},
{
"epoch": 0.2896376644357258,
"grad_norm": 0.3248600959777832,
"learning_rate": 9.348840968345129e-05,
"loss": 2.1441,
"step": 5020
},
{
"epoch": 0.29021463189476115,
"grad_norm": 0.322581022977829,
"learning_rate": 9.343569571863165e-05,
"loss": 2.1526,
"step": 5030
},
{
"epoch": 0.29079159935379645,
"grad_norm": 0.5681893825531006,
"learning_rate": 9.338278422841536e-05,
"loss": 2.1577,
"step": 5040
},
{
"epoch": 0.29136856681283174,
"grad_norm": 0.27868878841400146,
"learning_rate": 9.332967545341935e-05,
"loss": 2.1542,
"step": 5050
},
{
"epoch": 0.2919455342718671,
"grad_norm": 0.26688241958618164,
"learning_rate": 9.327636963515771e-05,
"loss": 2.1497,
"step": 5060
},
{
"epoch": 0.2925225017309024,
"grad_norm": 0.4590870440006256,
"learning_rate": 9.322286701604056e-05,
"loss": 2.1481,
"step": 5070
},
{
"epoch": 0.2930994691899377,
"grad_norm": 0.44435915350914,
"learning_rate": 9.316916783937299e-05,
"loss": 2.1551,
"step": 5080
},
{
"epoch": 0.29367643664897297,
"grad_norm": 0.5026682019233704,
"learning_rate": 9.311527234935397e-05,
"loss": 2.1594,
"step": 5090
},
{
"epoch": 0.2942534041080083,
"grad_norm": 0.28176602721214294,
"learning_rate": 9.306118079107518e-05,
"loss": 2.1427,
"step": 5100
},
{
"epoch": 0.2948303715670436,
"grad_norm": 0.3687533140182495,
"learning_rate": 9.300689341051992e-05,
"loss": 2.136,
"step": 5110
},
{
"epoch": 0.2954073390260789,
"grad_norm": 0.35213086009025574,
"learning_rate": 9.295241045456204e-05,
"loss": 2.1411,
"step": 5120
},
{
"epoch": 0.29598430648511426,
"grad_norm": 0.26359936594963074,
"learning_rate": 9.289773217096473e-05,
"loss": 2.1484,
"step": 5130
},
{
"epoch": 0.29656127394414955,
"grad_norm": 0.37725865840911865,
"learning_rate": 9.284285880837946e-05,
"loss": 2.1444,
"step": 5140
},
{
"epoch": 0.29713824140318484,
"grad_norm": 0.5063810348510742,
"learning_rate": 9.278779061634486e-05,
"loss": 2.1468,
"step": 5150
},
{
"epoch": 0.2977152088622202,
"grad_norm": 0.40431004762649536,
"learning_rate": 9.27325278452855e-05,
"loss": 2.1479,
"step": 5160
},
{
"epoch": 0.2982921763212555,
"grad_norm": 0.47989556193351746,
"learning_rate": 9.267707074651084e-05,
"loss": 2.1465,
"step": 5170
},
{
"epoch": 0.2988691437802908,
"grad_norm": 0.5401825308799744,
"learning_rate": 9.262141957221403e-05,
"loss": 2.1406,
"step": 5180
},
{
"epoch": 0.29944611123932613,
"grad_norm": 0.4957210123538971,
"learning_rate": 9.256557457547081e-05,
"loss": 2.1657,
"step": 5190
},
{
"epoch": 0.3000230786983614,
"grad_norm": 0.46101513504981995,
"learning_rate": 9.25095360102383e-05,
"loss": 2.1387,
"step": 5200
},
{
"epoch": 0.3006000461573967,
"grad_norm": 0.5260946154594421,
"learning_rate": 9.245330413135395e-05,
"loss": 2.1417,
"step": 5210
},
{
"epoch": 0.301177013616432,
"grad_norm": 0.36784660816192627,
"learning_rate": 9.239687919453421e-05,
"loss": 2.1502,
"step": 5220
},
{
"epoch": 0.30175398107546736,
"grad_norm": 0.3011978566646576,
"learning_rate": 9.234026145637353e-05,
"loss": 2.1464,
"step": 5230
},
{
"epoch": 0.30233094853450265,
"grad_norm": 0.31565114855766296,
"learning_rate": 9.228345117434314e-05,
"loss": 2.1237,
"step": 5240
},
{
"epoch": 0.30290791599353795,
"grad_norm": 0.2920317053794861,
"learning_rate": 9.222644860678984e-05,
"loss": 2.1287,
"step": 5250
},
{
"epoch": 0.3034848834525733,
"grad_norm": 0.31231454014778137,
"learning_rate": 9.216925401293488e-05,
"loss": 2.1552,
"step": 5260
},
{
"epoch": 0.3040618509116086,
"grad_norm": 0.361036479473114,
"learning_rate": 9.211186765287271e-05,
"loss": 2.1483,
"step": 5270
},
{
"epoch": 0.3046388183706439,
"grad_norm": 0.29103997349739075,
"learning_rate": 9.205428978756991e-05,
"loss": 2.1469,
"step": 5280
},
{
"epoch": 0.30521578582967923,
"grad_norm": 0.38782352209091187,
"learning_rate": 9.199652067886391e-05,
"loss": 2.1371,
"step": 5290
},
{
"epoch": 0.3057927532887145,
"grad_norm": 0.3866995871067047,
"learning_rate": 9.193856058946179e-05,
"loss": 2.1287,
"step": 5300
},
{
"epoch": 0.3063697207477498,
"grad_norm": 0.3839791417121887,
"learning_rate": 9.188040978293916e-05,
"loss": 2.1424,
"step": 5310
},
{
"epoch": 0.3069466882067851,
"grad_norm": 0.6226606369018555,
"learning_rate": 9.182206852373892e-05,
"loss": 2.1447,
"step": 5320
},
{
"epoch": 0.30752365566582046,
"grad_norm": 0.6132786870002747,
"learning_rate": 9.176353707717007e-05,
"loss": 2.1326,
"step": 5330
},
{
"epoch": 0.30810062312485575,
"grad_norm": 0.4488280117511749,
"learning_rate": 9.170481570940645e-05,
"loss": 2.1291,
"step": 5340
},
{
"epoch": 0.30867759058389105,
"grad_norm": 0.4035676419734955,
"learning_rate": 9.164590468748563e-05,
"loss": 2.1426,
"step": 5350
},
{
"epoch": 0.3092545580429264,
"grad_norm": 0.403726190328598,
"learning_rate": 9.15868042793076e-05,
"loss": 2.1416,
"step": 5360
},
{
"epoch": 0.3098315255019617,
"grad_norm": 0.3956368565559387,
"learning_rate": 9.152751475363359e-05,
"loss": 2.1317,
"step": 5370
},
{
"epoch": 0.310408492960997,
"grad_norm": 0.3403710722923279,
"learning_rate": 9.14680363800849e-05,
"loss": 2.1368,
"step": 5380
},
{
"epoch": 0.31098546042003233,
"grad_norm": 0.28579944372177124,
"learning_rate": 9.140836942914154e-05,
"loss": 2.1484,
"step": 5390
},
{
"epoch": 0.3115624278790676,
"grad_norm": 0.3212498724460602,
"learning_rate": 9.134851417214114e-05,
"loss": 2.1337,
"step": 5400
},
{
"epoch": 0.3121393953381029,
"grad_norm": 0.2878890335559845,
"learning_rate": 9.128847088127768e-05,
"loss": 2.1359,
"step": 5410
},
{
"epoch": 0.3127163627971382,
"grad_norm": 0.47753390669822693,
"learning_rate": 9.122823982960014e-05,
"loss": 2.1397,
"step": 5420
},
{
"epoch": 0.31329333025617356,
"grad_norm": 0.4015023410320282,
"learning_rate": 9.116782129101144e-05,
"loss": 2.1283,
"step": 5430
},
{
"epoch": 0.31387029771520886,
"grad_norm": 0.4427224099636078,
"learning_rate": 9.110721554026705e-05,
"loss": 2.1355,
"step": 5440
},
{
"epoch": 0.31444726517424415,
"grad_norm": 0.2725686728954315,
"learning_rate": 9.104642285297385e-05,
"loss": 2.1425,
"step": 5450
},
{
"epoch": 0.3150242326332795,
"grad_norm": 0.31836792826652527,
"learning_rate": 9.098544350558872e-05,
"loss": 2.128,
"step": 5460
},
{
"epoch": 0.3156012000923148,
"grad_norm": 0.2776530981063843,
"learning_rate": 9.092427777541751e-05,
"loss": 2.1353,
"step": 5470
},
{
"epoch": 0.3161781675513501,
"grad_norm": 0.28507447242736816,
"learning_rate": 9.086292594061356e-05,
"loss": 2.1378,
"step": 5480
},
{
"epoch": 0.31675513501038544,
"grad_norm": 0.30143943428993225,
"learning_rate": 9.080138828017655e-05,
"loss": 2.1194,
"step": 5490
},
{
"epoch": 0.31733210246942073,
"grad_norm": 0.36624717712402344,
"learning_rate": 9.073966507395122e-05,
"loss": 2.1274,
"step": 5500
},
{
"epoch": 0.317909069928456,
"grad_norm": 0.3170357346534729,
"learning_rate": 9.067775660262609e-05,
"loss": 2.1253,
"step": 5510
},
{
"epoch": 0.31848603738749137,
"grad_norm": 0.3736599087715149,
"learning_rate": 9.061566314773217e-05,
"loss": 2.1236,
"step": 5520
},
{
"epoch": 0.31906300484652667,
"grad_norm": 0.2892272472381592,
"learning_rate": 9.055338499164171e-05,
"loss": 2.1235,
"step": 5530
},
{
"epoch": 0.31963997230556196,
"grad_norm": 0.5071966052055359,
"learning_rate": 9.049092241756683e-05,
"loss": 2.1211,
"step": 5540
},
{
"epoch": 0.32021693976459725,
"grad_norm": 0.5542040467262268,
"learning_rate": 9.042827570955838e-05,
"loss": 2.1413,
"step": 5550
},
{
"epoch": 0.3207939072236326,
"grad_norm": 0.39468914270401,
"learning_rate": 9.036544515250452e-05,
"loss": 2.1484,
"step": 5560
},
{
"epoch": 0.3213708746826679,
"grad_norm": 0.45103156566619873,
"learning_rate": 9.030243103212948e-05,
"loss": 2.1349,
"step": 5570
},
{
"epoch": 0.3219478421417032,
"grad_norm": 0.41866421699523926,
"learning_rate": 9.023923363499225e-05,
"loss": 2.1318,
"step": 5580
},
{
"epoch": 0.32252480960073854,
"grad_norm": 0.3836067020893097,
"learning_rate": 9.017585324848528e-05,
"loss": 2.1367,
"step": 5590
},
{
"epoch": 0.32310177705977383,
"grad_norm": 0.3563435971736908,
"learning_rate": 9.011229016083315e-05,
"loss": 2.1396,
"step": 5600
},
{
"epoch": 0.3236787445188091,
"grad_norm": 0.38590380549430847,
"learning_rate": 9.004854466109134e-05,
"loss": 2.1345,
"step": 5610
},
{
"epoch": 0.3242557119778445,
"grad_norm": 0.3448795676231384,
"learning_rate": 8.998461703914477e-05,
"loss": 2.1233,
"step": 5620
},
{
"epoch": 0.32483267943687977,
"grad_norm": 0.34533822536468506,
"learning_rate": 8.992050758570665e-05,
"loss": 2.1474,
"step": 5630
},
{
"epoch": 0.32540964689591506,
"grad_norm": 0.3775763511657715,
"learning_rate": 8.9856216592317e-05,
"loss": 2.1239,
"step": 5640
},
{
"epoch": 0.32598661435495035,
"grad_norm": 0.5391921401023865,
"learning_rate": 8.979174435134146e-05,
"loss": 2.1213,
"step": 5650
},
{
"epoch": 0.3265635818139857,
"grad_norm": 0.4273969233036041,
"learning_rate": 8.972709115596989e-05,
"loss": 2.1356,
"step": 5660
},
{
"epoch": 0.327140549273021,
"grad_norm": 0.4936189353466034,
"learning_rate": 8.9662257300215e-05,
"loss": 2.1347,
"step": 5670
},
{
"epoch": 0.3277175167320563,
"grad_norm": 0.39583414793014526,
"learning_rate": 8.959724307891113e-05,
"loss": 2.1279,
"step": 5680
},
{
"epoch": 0.32829448419109164,
"grad_norm": 0.27078336477279663,
"learning_rate": 8.953204878771276e-05,
"loss": 2.1386,
"step": 5690
},
{
"epoch": 0.32887145165012693,
"grad_norm": 0.2853717803955078,
"learning_rate": 8.946667472309332e-05,
"loss": 2.1206,
"step": 5700
},
{
"epoch": 0.3294484191091622,
"grad_norm": 0.2861252725124359,
"learning_rate": 8.940112118234372e-05,
"loss": 2.1211,
"step": 5710
},
{
"epoch": 0.3300253865681976,
"grad_norm": 0.5010199546813965,
"learning_rate": 8.933538846357105e-05,
"loss": 2.1313,
"step": 5720
},
{
"epoch": 0.33060235402723287,
"grad_norm": 0.3199447989463806,
"learning_rate": 8.926947686569724e-05,
"loss": 2.1281,
"step": 5730
},
{
"epoch": 0.33117932148626816,
"grad_norm": 0.3005349040031433,
"learning_rate": 8.920338668845764e-05,
"loss": 2.1341,
"step": 5740
},
{
"epoch": 0.3317562889453035,
"grad_norm": 0.2689361274242401,
"learning_rate": 8.913711823239977e-05,
"loss": 2.1323,
"step": 5750
},
{
"epoch": 0.3323332564043388,
"grad_norm": 0.538209855556488,
"learning_rate": 8.907067179888177e-05,
"loss": 2.1425,
"step": 5760
},
{
"epoch": 0.3329102238633741,
"grad_norm": 0.45525121688842773,
"learning_rate": 8.900404769007123e-05,
"loss": 2.1286,
"step": 5770
},
{
"epoch": 0.3334871913224094,
"grad_norm": 0.379524290561676,
"learning_rate": 8.893724620894369e-05,
"loss": 2.1284,
"step": 5780
},
{
"epoch": 0.33406415878144474,
"grad_norm": 0.31159836053848267,
"learning_rate": 8.887026765928129e-05,
"loss": 2.1309,
"step": 5790
},
{
"epoch": 0.33464112624048004,
"grad_norm": 0.31788134574890137,
"learning_rate": 8.880311234567142e-05,
"loss": 2.1339,
"step": 5800
},
{
"epoch": 0.33521809369951533,
"grad_norm": 0.2826105058193207,
"learning_rate": 8.873578057350531e-05,
"loss": 2.1195,
"step": 5810
},
{
"epoch": 0.3357950611585507,
"grad_norm": 0.3322320282459259,
"learning_rate": 8.86682726489766e-05,
"loss": 2.1323,
"step": 5820
},
{
"epoch": 0.33637202861758597,
"grad_norm": 0.35983988642692566,
"learning_rate": 8.860058887908e-05,
"loss": 2.1319,
"step": 5830
},
{
"epoch": 0.33694899607662127,
"grad_norm": 0.27813997864723206,
"learning_rate": 8.853272957160997e-05,
"loss": 2.1251,
"step": 5840
},
{
"epoch": 0.3375259635356566,
"grad_norm": 0.3000926971435547,
"learning_rate": 8.846469503515909e-05,
"loss": 2.1238,
"step": 5850
},
{
"epoch": 0.3381029309946919,
"grad_norm": 0.2732325494289398,
"learning_rate": 8.839648557911691e-05,
"loss": 2.1302,
"step": 5860
},
{
"epoch": 0.3386798984537272,
"grad_norm": 0.3154725134372711,
"learning_rate": 8.832810151366841e-05,
"loss": 2.1185,
"step": 5870
},
{
"epoch": 0.3392568659127625,
"grad_norm": 0.3048752546310425,
"learning_rate": 8.825954314979258e-05,
"loss": 2.1053,
"step": 5880
},
{
"epoch": 0.33983383337179784,
"grad_norm": 0.6547893285751343,
"learning_rate": 8.819081079926106e-05,
"loss": 2.1069,
"step": 5890
},
{
"epoch": 0.34041080083083314,
"grad_norm": 0.27823716402053833,
"learning_rate": 8.812190477463669e-05,
"loss": 2.1345,
"step": 5900
},
{
"epoch": 0.34098776828986843,
"grad_norm": 0.2662949860095978,
"learning_rate": 8.805282538927214e-05,
"loss": 2.1311,
"step": 5910
},
{
"epoch": 0.3415647357489038,
"grad_norm": 0.3390512466430664,
"learning_rate": 8.798357295730836e-05,
"loss": 2.108,
"step": 5920
},
{
"epoch": 0.3421417032079391,
"grad_norm": 0.4059408903121948,
"learning_rate": 8.791414779367335e-05,
"loss": 2.1185,
"step": 5930
},
{
"epoch": 0.34271867066697437,
"grad_norm": 0.5488471388816833,
"learning_rate": 8.784455021408051e-05,
"loss": 2.1193,
"step": 5940
},
{
"epoch": 0.3432956381260097,
"grad_norm": 0.29119235277175903,
"learning_rate": 8.777478053502736e-05,
"loss": 2.1249,
"step": 5950
},
{
"epoch": 0.343872605585045,
"grad_norm": 0.268136203289032,
"learning_rate": 8.770483907379403e-05,
"loss": 2.1307,
"step": 5960
},
{
"epoch": 0.3444495730440803,
"grad_norm": 0.2661965787410736,
"learning_rate": 8.763472614844185e-05,
"loss": 2.1146,
"step": 5970
},
{
"epoch": 0.3450265405031156,
"grad_norm": 0.4353127181529999,
"learning_rate": 8.756444207781192e-05,
"loss": 2.1215,
"step": 5980
},
{
"epoch": 0.34560350796215095,
"grad_norm": 0.5630121827125549,
"learning_rate": 8.749398718152352e-05,
"loss": 2.1231,
"step": 5990
},
{
"epoch": 0.34618047542118624,
"grad_norm": 0.369564026594162,
"learning_rate": 8.742336177997289e-05,
"loss": 2.1185,
"step": 6000
},
{
"epoch": 0.34675744288022153,
"grad_norm": 0.3091670572757721,
"learning_rate": 8.735256619433157e-05,
"loss": 2.1088,
"step": 6010
},
{
"epoch": 0.3473344103392569,
"grad_norm": 0.27325648069381714,
"learning_rate": 8.728160074654507e-05,
"loss": 2.1254,
"step": 6020
},
{
"epoch": 0.3479113777982922,
"grad_norm": 0.266166090965271,
"learning_rate": 8.721046575933132e-05,
"loss": 2.112,
"step": 6030
},
{
"epoch": 0.34848834525732747,
"grad_norm": 0.3261997699737549,
"learning_rate": 8.713916155617922e-05,
"loss": 2.1253,
"step": 6040
},
{
"epoch": 0.3490653127163628,
"grad_norm": 0.34902575612068176,
"learning_rate": 8.706768846134727e-05,
"loss": 2.1253,
"step": 6050
},
{
"epoch": 0.3496422801753981,
"grad_norm": 0.2903362810611725,
"learning_rate": 8.699604679986188e-05,
"loss": 2.1239,
"step": 6060
},
{
"epoch": 0.3502192476344334,
"grad_norm": 0.2810317575931549,
"learning_rate": 8.692423689751617e-05,
"loss": 2.1183,
"step": 6070
},
{
"epoch": 0.35079621509346876,
"grad_norm": 0.3162137269973755,
"learning_rate": 8.68522590808682e-05,
"loss": 2.1171,
"step": 6080
},
{
"epoch": 0.35137318255250405,
"grad_norm": 0.3248831331729889,
"learning_rate": 8.678011367723973e-05,
"loss": 2.1262,
"step": 6090
},
{
"epoch": 0.35195015001153934,
"grad_norm": 0.35751304030418396,
"learning_rate": 8.670780101471458e-05,
"loss": 2.1107,
"step": 6100
},
{
"epoch": 0.35252711747057464,
"grad_norm": 0.30080780386924744,
"learning_rate": 8.663532142213721e-05,
"loss": 2.1197,
"step": 6110
},
{
"epoch": 0.35310408492961,
"grad_norm": 0.24629926681518555,
"learning_rate": 8.656267522911119e-05,
"loss": 2.137,
"step": 6120
},
{
"epoch": 0.3536810523886453,
"grad_norm": 0.24670863151550293,
"learning_rate": 8.648986276599769e-05,
"loss": 2.1134,
"step": 6130
},
{
"epoch": 0.35425801984768057,
"grad_norm": 0.5493544340133667,
"learning_rate": 8.641688436391402e-05,
"loss": 2.1163,
"step": 6140
},
{
"epoch": 0.3548349873067159,
"grad_norm": 0.34867388010025024,
"learning_rate": 8.634374035473212e-05,
"loss": 2.1076,
"step": 6150
},
{
"epoch": 0.3554119547657512,
"grad_norm": 0.38117578625679016,
"learning_rate": 8.6270431071077e-05,
"loss": 2.1163,
"step": 6160
},
{
"epoch": 0.3559889222247865,
"grad_norm": 0.3320949673652649,
"learning_rate": 8.619695684632529e-05,
"loss": 2.1048,
"step": 6170
},
{
"epoch": 0.35656588968382186,
"grad_norm": 0.36235710978507996,
"learning_rate": 8.612331801460367e-05,
"loss": 2.1019,
"step": 6180
},
{
"epoch": 0.35714285714285715,
"grad_norm": 0.28610482811927795,
"learning_rate": 8.604951491078739e-05,
"loss": 2.1179,
"step": 6190
},
{
"epoch": 0.35771982460189244,
"grad_norm": 0.4073322117328644,
"learning_rate": 8.597554787049875e-05,
"loss": 2.1277,
"step": 6200
},
{
"epoch": 0.35829679206092774,
"grad_norm": 0.2900354862213135,
"learning_rate": 8.590141723010552e-05,
"loss": 2.1257,
"step": 6210
},
{
"epoch": 0.3588737595199631,
"grad_norm": 0.3767155408859253,
"learning_rate": 8.582712332671947e-05,
"loss": 2.1198,
"step": 6220
},
{
"epoch": 0.3594507269789984,
"grad_norm": 0.2501465976238251,
"learning_rate": 8.575266649819484e-05,
"loss": 2.1187,
"step": 6230
},
{
"epoch": 0.3600276944380337,
"grad_norm": 0.30487820506095886,
"learning_rate": 8.56780470831267e-05,
"loss": 2.1127,
"step": 6240
},
{
"epoch": 0.360604661897069,
"grad_norm": 0.33418163657188416,
"learning_rate": 8.560326542084959e-05,
"loss": 2.1206,
"step": 6250
},
{
"epoch": 0.3611816293561043,
"grad_norm": 0.39887627959251404,
"learning_rate": 8.552832185143577e-05,
"loss": 2.109,
"step": 6260
},
{
"epoch": 0.3617585968151396,
"grad_norm": 0.3675427734851837,
"learning_rate": 8.545321671569389e-05,
"loss": 2.1245,
"step": 6270
},
{
"epoch": 0.36233556427417496,
"grad_norm": 0.2894626259803772,
"learning_rate": 8.537795035516722e-05,
"loss": 2.1208,
"step": 6280
},
{
"epoch": 0.36291253173321025,
"grad_norm": 0.3343421518802643,
"learning_rate": 8.53025231121323e-05,
"loss": 2.1077,
"step": 6290
},
{
"epoch": 0.36348949919224555,
"grad_norm": 0.4454777240753174,
"learning_rate": 8.522693532959721e-05,
"loss": 2.1155,
"step": 6300
},
{
"epoch": 0.36406646665128084,
"grad_norm": 0.37064430117607117,
"learning_rate": 8.515118735130011e-05,
"loss": 2.0999,
"step": 6310
},
{
"epoch": 0.3646434341103162,
"grad_norm": 0.429360568523407,
"learning_rate": 8.507527952170771e-05,
"loss": 2.1192,
"step": 6320
},
{
"epoch": 0.3652204015693515,
"grad_norm": 0.4052377939224243,
"learning_rate": 8.499921218601358e-05,
"loss": 2.1094,
"step": 6330
},
{
"epoch": 0.3657973690283868,
"grad_norm": 0.2919637858867645,
"learning_rate": 8.49229856901367e-05,
"loss": 2.1169,
"step": 6340
},
{
"epoch": 0.3663743364874221,
"grad_norm": 0.44187116622924805,
"learning_rate": 8.484660038071981e-05,
"loss": 2.101,
"step": 6350
},
{
"epoch": 0.3669513039464574,
"grad_norm": 0.4569217264652252,
"learning_rate": 8.477005660512787e-05,
"loss": 2.1241,
"step": 6360
},
{
"epoch": 0.3675282714054927,
"grad_norm": 0.40709832310676575,
"learning_rate": 8.469335471144646e-05,
"loss": 2.1134,
"step": 6370
},
{
"epoch": 0.36810523886452806,
"grad_norm": 0.289991170167923,
"learning_rate": 8.46164950484802e-05,
"loss": 2.1119,
"step": 6380
},
{
"epoch": 0.36868220632356336,
"grad_norm": 0.30788782238960266,
"learning_rate": 8.453947796575122e-05,
"loss": 2.1171,
"step": 6390
},
{
"epoch": 0.36925917378259865,
"grad_norm": 0.26853644847869873,
"learning_rate": 8.446230381349745e-05,
"loss": 2.1118,
"step": 6400
},
{
"epoch": 0.369836141241634,
"grad_norm": 0.3687596023082733,
"learning_rate": 8.438497294267117e-05,
"loss": 2.1341,
"step": 6410
},
{
"epoch": 0.3704131087006693,
"grad_norm": 0.4256872534751892,
"learning_rate": 8.430748570493729e-05,
"loss": 2.1198,
"step": 6420
},
{
"epoch": 0.3709900761597046,
"grad_norm": 0.27723926305770874,
"learning_rate": 8.422984245267184e-05,
"loss": 2.1193,
"step": 6430
},
{
"epoch": 0.3715670436187399,
"grad_norm": 0.3537181317806244,
"learning_rate": 8.415204353896031e-05,
"loss": 2.1111,
"step": 6440
},
{
"epoch": 0.3721440110777752,
"grad_norm": 0.30080363154411316,
"learning_rate": 8.407408931759607e-05,
"loss": 2.0906,
"step": 6450
},
{
"epoch": 0.3727209785368105,
"grad_norm": 0.3747290074825287,
"learning_rate": 8.399598014307877e-05,
"loss": 2.1012,
"step": 6460
},
{
"epoch": 0.3732979459958458,
"grad_norm": 0.5772930383682251,
"learning_rate": 8.391771637061268e-05,
"loss": 2.1125,
"step": 6470
},
{
"epoch": 0.37387491345488116,
"grad_norm": 0.5180211067199707,
"learning_rate": 8.383929835610516e-05,
"loss": 2.102,
"step": 6480
},
{
"epoch": 0.37445188091391646,
"grad_norm": 0.5194865465164185,
"learning_rate": 8.376072645616496e-05,
"loss": 2.1105,
"step": 6490
},
{
"epoch": 0.37502884837295175,
"grad_norm": 0.2418607920408249,
"learning_rate": 8.368200102810065e-05,
"loss": 2.1099,
"step": 6500
},
{
"epoch": 0.3756058158319871,
"grad_norm": 0.29503995180130005,
"learning_rate": 8.360312242991892e-05,
"loss": 2.1044,
"step": 6510
},
{
"epoch": 0.3761827832910224,
"grad_norm": 0.3843310475349426,
"learning_rate": 8.352409102032308e-05,
"loss": 2.1014,
"step": 6520
},
{
"epoch": 0.3767597507500577,
"grad_norm": 0.2821659743785858,
"learning_rate": 8.344490715871132e-05,
"loss": 2.1056,
"step": 6530
},
{
"epoch": 0.377336718209093,
"grad_norm": 0.28183305263519287,
"learning_rate": 8.336557120517509e-05,
"loss": 2.0959,
"step": 6540
},
{
"epoch": 0.37791368566812833,
"grad_norm": 0.27406924962997437,
"learning_rate": 8.328608352049754e-05,
"loss": 2.1111,
"step": 6550
},
{
"epoch": 0.3784906531271636,
"grad_norm": 0.41145142912864685,
"learning_rate": 8.320644446615175e-05,
"loss": 2.1011,
"step": 6560
},
{
"epoch": 0.3790676205861989,
"grad_norm": 0.368637353181839,
"learning_rate": 8.312665440429925e-05,
"loss": 2.1185,
"step": 6570
},
{
"epoch": 0.37964458804523427,
"grad_norm": 0.42195016145706177,
"learning_rate": 8.304671369778821e-05,
"loss": 2.1148,
"step": 6580
},
{
"epoch": 0.38022155550426956,
"grad_norm": 0.3770730793476105,
"learning_rate": 8.296662271015185e-05,
"loss": 2.1128,
"step": 6590
},
{
"epoch": 0.38079852296330485,
"grad_norm": 0.31544458866119385,
"learning_rate": 8.288638180560686e-05,
"loss": 2.1162,
"step": 6600
},
{
"epoch": 0.3813754904223402,
"grad_norm": 0.3706103265285492,
"learning_rate": 8.280599134905167e-05,
"loss": 2.1006,
"step": 6610
},
{
"epoch": 0.3819524578813755,
"grad_norm": 0.4784095585346222,
"learning_rate": 8.272545170606476e-05,
"loss": 2.1131,
"step": 6620
},
{
"epoch": 0.3825294253404108,
"grad_norm": 0.5787613987922668,
"learning_rate": 8.264476324290309e-05,
"loss": 2.1004,
"step": 6630
},
{
"epoch": 0.38310639279944614,
"grad_norm": 0.5983478426933289,
"learning_rate": 8.256392632650034e-05,
"loss": 2.1085,
"step": 6640
},
{
"epoch": 0.38368336025848143,
"grad_norm": 0.302979439496994,
"learning_rate": 8.248294132446533e-05,
"loss": 2.1015,
"step": 6650
},
{
"epoch": 0.3842603277175167,
"grad_norm": 0.3463706970214844,
"learning_rate": 8.240180860508027e-05,
"loss": 2.101,
"step": 6660
},
{
"epoch": 0.384837295176552,
"grad_norm": 0.32216259837150574,
"learning_rate": 8.232052853729913e-05,
"loss": 2.1197,
"step": 6670
},
{
"epoch": 0.38541426263558737,
"grad_norm": 0.30237728357315063,
"learning_rate": 8.223910149074598e-05,
"loss": 2.098,
"step": 6680
},
{
"epoch": 0.38599123009462266,
"grad_norm": 0.4659302532672882,
"learning_rate": 8.215752783571325e-05,
"loss": 2.1121,
"step": 6690
},
{
"epoch": 0.38656819755365796,
"grad_norm": 0.314649373292923,
"learning_rate": 8.207580794316009e-05,
"loss": 2.0945,
"step": 6700
},
{
"epoch": 0.3871451650126933,
"grad_norm": 0.29076725244522095,
"learning_rate": 8.199394218471065e-05,
"loss": 2.1033,
"step": 6710
},
{
"epoch": 0.3877221324717286,
"grad_norm": 0.3016183376312256,
"learning_rate": 8.191193093265241e-05,
"loss": 2.0942,
"step": 6720
},
{
"epoch": 0.3882990999307639,
"grad_norm": 0.29181423783302307,
"learning_rate": 8.182977455993456e-05,
"loss": 2.1083,
"step": 6730
},
{
"epoch": 0.38887606738979924,
"grad_norm": 0.3460990786552429,
"learning_rate": 8.174747344016613e-05,
"loss": 2.0995,
"step": 6740
},
{
"epoch": 0.38945303484883453,
"grad_norm": 0.2913837134838104,
"learning_rate": 8.166502794761445e-05,
"loss": 2.111,
"step": 6750
},
{
"epoch": 0.3900300023078698,
"grad_norm": 0.3307998478412628,
"learning_rate": 8.158243845720338e-05,
"loss": 2.1161,
"step": 6760
},
{
"epoch": 0.3906069697669051,
"grad_norm": 0.2763582766056061,
"learning_rate": 8.149970534451161e-05,
"loss": 2.1044,
"step": 6770
},
{
"epoch": 0.39118393722594047,
"grad_norm": 0.2793087065219879,
"learning_rate": 8.141682898577096e-05,
"loss": 2.1016,
"step": 6780
},
{
"epoch": 0.39176090468497576,
"grad_norm": 0.3368696868419647,
"learning_rate": 8.133380975786469e-05,
"loss": 2.1022,
"step": 6790
},
{
"epoch": 0.39233787214401106,
"grad_norm": 0.3350260257720947,
"learning_rate": 8.125064803832569e-05,
"loss": 2.0958,
"step": 6800
},
{
"epoch": 0.3929148396030464,
"grad_norm": 0.2954477369785309,
"learning_rate": 8.116734420533493e-05,
"loss": 2.1082,
"step": 6810
},
{
"epoch": 0.3934918070620817,
"grad_norm": 0.302860289812088,
"learning_rate": 8.108389863771959e-05,
"loss": 2.0965,
"step": 6820
},
{
"epoch": 0.394068774521117,
"grad_norm": 0.34222808480262756,
"learning_rate": 8.100031171495142e-05,
"loss": 2.1034,
"step": 6830
},
{
"epoch": 0.39464574198015234,
"grad_norm": 0.27451932430267334,
"learning_rate": 8.091658381714493e-05,
"loss": 2.1104,
"step": 6840
},
{
"epoch": 0.39522270943918764,
"grad_norm": 0.40477854013442993,
"learning_rate": 8.083271532505581e-05,
"loss": 2.0989,
"step": 6850
},
{
"epoch": 0.39579967689822293,
"grad_norm": 0.3276130259037018,
"learning_rate": 8.074870662007903e-05,
"loss": 2.1041,
"step": 6860
},
{
"epoch": 0.3963766443572582,
"grad_norm": 0.3322683572769165,
"learning_rate": 8.066455808424722e-05,
"loss": 2.1114,
"step": 6870
},
{
"epoch": 0.3969536118162936,
"grad_norm": 0.29346489906311035,
"learning_rate": 8.05802701002289e-05,
"loss": 2.1009,
"step": 6880
},
{
"epoch": 0.39753057927532887,
"grad_norm": 0.275181382894516,
"learning_rate": 8.049584305132673e-05,
"loss": 2.0983,
"step": 6890
},
{
"epoch": 0.39810754673436416,
"grad_norm": 0.2939784526824951,
"learning_rate": 8.041127732147573e-05,
"loss": 2.1173,
"step": 6900
},
{
"epoch": 0.3986845141933995,
"grad_norm": 0.3622361719608307,
"learning_rate": 8.032657329524165e-05,
"loss": 2.092,
"step": 6910
},
{
"epoch": 0.3992614816524348,
"grad_norm": 0.4047354459762573,
"learning_rate": 8.02417313578191e-05,
"loss": 2.1023,
"step": 6920
},
{
"epoch": 0.3998384491114701,
"grad_norm": 0.301595538854599,
"learning_rate": 8.015675189502986e-05,
"loss": 2.105,
"step": 6930
},
{
"epoch": 0.40041541657050544,
"grad_norm": 0.39583104848861694,
"learning_rate": 8.007163529332111e-05,
"loss": 2.107,
"step": 6940
},
{
"epoch": 0.40099238402954074,
"grad_norm": 0.26828697323799133,
"learning_rate": 7.998638193976366e-05,
"loss": 2.1087,
"step": 6950
},
{
"epoch": 0.40156935148857603,
"grad_norm": 0.35258111357688904,
"learning_rate": 7.990099222205023e-05,
"loss": 2.1052,
"step": 6960
},
{
"epoch": 0.4021463189476114,
"grad_norm": 0.3397604823112488,
"learning_rate": 7.981546652849365e-05,
"loss": 2.0931,
"step": 6970
},
{
"epoch": 0.4027232864066467,
"grad_norm": 0.29654258489608765,
"learning_rate": 7.972980524802509e-05,
"loss": 2.0865,
"step": 6980
},
{
"epoch": 0.40330025386568197,
"grad_norm": 0.30216071009635925,
"learning_rate": 7.964400877019233e-05,
"loss": 2.0926,
"step": 6990
},
{
"epoch": 0.40387722132471726,
"grad_norm": 0.29403412342071533,
"learning_rate": 7.955807748515797e-05,
"loss": 2.1122,
"step": 7000
},
{
"epoch": 0.4044541887837526,
"grad_norm": 0.3340148329734802,
"learning_rate": 7.947201178369761e-05,
"loss": 2.1019,
"step": 7010
},
{
"epoch": 0.4050311562427879,
"grad_norm": 0.28326958417892456,
"learning_rate": 7.938581205719815e-05,
"loss": 2.089,
"step": 7020
},
{
"epoch": 0.4056081237018232,
"grad_norm": 0.2603297829627991,
"learning_rate": 7.929947869765596e-05,
"loss": 2.0979,
"step": 7030
},
{
"epoch": 0.40618509116085855,
"grad_norm": 0.375947505235672,
"learning_rate": 7.92130120976751e-05,
"loss": 2.1027,
"step": 7040
},
{
"epoch": 0.40676205861989384,
"grad_norm": 0.29024848341941833,
"learning_rate": 7.912641265046561e-05,
"loss": 2.0944,
"step": 7050
},
{
"epoch": 0.40733902607892913,
"grad_norm": 0.27378323674201965,
"learning_rate": 7.903968074984155e-05,
"loss": 2.0983,
"step": 7060
},
{
"epoch": 0.4079159935379645,
"grad_norm": 0.38356778025627136,
"learning_rate": 7.89528167902194e-05,
"loss": 2.1056,
"step": 7070
},
{
"epoch": 0.4084929609969998,
"grad_norm": 0.38269197940826416,
"learning_rate": 7.886582116661615e-05,
"loss": 2.098,
"step": 7080
},
{
"epoch": 0.40906992845603507,
"grad_norm": 0.4965763986110687,
"learning_rate": 7.877869427464755e-05,
"loss": 2.1038,
"step": 7090
},
{
"epoch": 0.40964689591507036,
"grad_norm": 0.28945809602737427,
"learning_rate": 7.869143651052629e-05,
"loss": 2.0898,
"step": 7100
},
{
"epoch": 0.4102238633741057,
"grad_norm": 0.30477508902549744,
"learning_rate": 7.86040482710602e-05,
"loss": 2.1105,
"step": 7110
},
{
"epoch": 0.410800830833141,
"grad_norm": 0.2908862829208374,
"learning_rate": 7.851652995365044e-05,
"loss": 2.1111,
"step": 7120
},
{
"epoch": 0.4113777982921763,
"grad_norm": 0.3085802495479584,
"learning_rate": 7.842888195628977e-05,
"loss": 2.0996,
"step": 7130
},
{
"epoch": 0.41195476575121165,
"grad_norm": 0.3265917897224426,
"learning_rate": 7.834110467756059e-05,
"loss": 2.0904,
"step": 7140
},
{
"epoch": 0.41253173321024694,
"grad_norm": 0.37845543026924133,
"learning_rate": 7.825319851663328e-05,
"loss": 2.0892,
"step": 7150
},
{
"epoch": 0.41310870066928224,
"grad_norm": 0.3401089906692505,
"learning_rate": 7.816516387326425e-05,
"loss": 2.1042,
"step": 7160
},
{
"epoch": 0.4136856681283176,
"grad_norm": 0.28092026710510254,
"learning_rate": 7.807700114779424e-05,
"loss": 2.0867,
"step": 7170
},
{
"epoch": 0.4142626355873529,
"grad_norm": 0.3313417434692383,
"learning_rate": 7.798871074114642e-05,
"loss": 2.0914,
"step": 7180
},
{
"epoch": 0.4148396030463882,
"grad_norm": 0.3220118582248688,
"learning_rate": 7.790029305482464e-05,
"loss": 2.1001,
"step": 7190
},
{
"epoch": 0.4154165705054235,
"grad_norm": 0.4763105809688568,
"learning_rate": 7.781174849091147e-05,
"loss": 2.0942,
"step": 7200
},
{
"epoch": 0.4159935379644588,
"grad_norm": 0.29946181178092957,
"learning_rate": 7.772307745206657e-05,
"loss": 2.104,
"step": 7210
},
{
"epoch": 0.4165705054234941,
"grad_norm": 0.3090571165084839,
"learning_rate": 7.763428034152466e-05,
"loss": 2.0946,
"step": 7220
},
{
"epoch": 0.4171474728825294,
"grad_norm": 0.34957677125930786,
"learning_rate": 7.754535756309383e-05,
"loss": 2.1011,
"step": 7230
},
{
"epoch": 0.41772444034156475,
"grad_norm": 0.3043264150619507,
"learning_rate": 7.745630952115363e-05,
"loss": 2.1069,
"step": 7240
},
{
"epoch": 0.41830140780060004,
"grad_norm": 0.30251699686050415,
"learning_rate": 7.736713662065322e-05,
"loss": 2.0902,
"step": 7250
},
{
"epoch": 0.41887837525963534,
"grad_norm": 0.3313656449317932,
"learning_rate": 7.727783926710962e-05,
"loss": 2.1037,
"step": 7260
},
{
"epoch": 0.4194553427186707,
"grad_norm": 0.4356120526790619,
"learning_rate": 7.718841786660577e-05,
"loss": 2.1033,
"step": 7270
},
{
"epoch": 0.420032310177706,
"grad_norm": 0.27830976247787476,
"learning_rate": 7.709887282578871e-05,
"loss": 2.0909,
"step": 7280
},
{
"epoch": 0.4206092776367413,
"grad_norm": 0.32721787691116333,
"learning_rate": 7.700920455186777e-05,
"loss": 2.0885,
"step": 7290
},
{
"epoch": 0.4211862450957766,
"grad_norm": 0.29251882433891296,
"learning_rate": 7.691941345261267e-05,
"loss": 2.1043,
"step": 7300
},
{
"epoch": 0.4217632125548119,
"grad_norm": 0.3718348443508148,
"learning_rate": 7.682949993635164e-05,
"loss": 2.0776,
"step": 7310
},
{
"epoch": 0.4223401800138472,
"grad_norm": 0.3022097945213318,
"learning_rate": 7.673946441196967e-05,
"loss": 2.0869,
"step": 7320
},
{
"epoch": 0.4229171474728825,
"grad_norm": 0.35751357674598694,
"learning_rate": 7.664930728890656e-05,
"loss": 2.096,
"step": 7330
},
{
"epoch": 0.42349411493191785,
"grad_norm": 0.2946200966835022,
"learning_rate": 7.655902897715509e-05,
"loss": 2.0876,
"step": 7340
},
{
"epoch": 0.42407108239095315,
"grad_norm": 0.41840797662734985,
"learning_rate": 7.646862988725916e-05,
"loss": 2.0929,
"step": 7350
},
{
"epoch": 0.42464804984998844,
"grad_norm": 0.3791830539703369,
"learning_rate": 7.637811043031186e-05,
"loss": 2.0857,
"step": 7360
},
{
"epoch": 0.4252250173090238,
"grad_norm": 0.31222841143608093,
"learning_rate": 7.628747101795374e-05,
"loss": 2.0847,
"step": 7370
},
{
"epoch": 0.4258019847680591,
"grad_norm": 0.33404964208602905,
"learning_rate": 7.619671206237078e-05,
"loss": 2.0947,
"step": 7380
},
{
"epoch": 0.4263789522270944,
"grad_norm": 0.40400052070617676,
"learning_rate": 7.610583397629261e-05,
"loss": 2.0963,
"step": 7390
},
{
"epoch": 0.4269559196861297,
"grad_norm": 0.282111257314682,
"learning_rate": 7.601483717299062e-05,
"loss": 2.0824,
"step": 7400
},
{
"epoch": 0.427532887145165,
"grad_norm": 0.38018685579299927,
"learning_rate": 7.592372206627605e-05,
"loss": 2.1036,
"step": 7410
},
{
"epoch": 0.4281098546042003,
"grad_norm": 0.3515368700027466,
"learning_rate": 7.583248907049817e-05,
"loss": 2.0904,
"step": 7420
},
{
"epoch": 0.4286868220632356,
"grad_norm": 0.3954324424266815,
"learning_rate": 7.57411386005423e-05,
"loss": 2.0919,
"step": 7430
},
{
"epoch": 0.42926378952227096,
"grad_norm": 0.44705700874328613,
"learning_rate": 7.564967107182801e-05,
"loss": 2.0923,
"step": 7440
},
{
"epoch": 0.42984075698130625,
"grad_norm": 0.2849537432193756,
"learning_rate": 7.555808690030719e-05,
"loss": 2.0977,
"step": 7450
},
{
"epoch": 0.43041772444034154,
"grad_norm": 0.356541246175766,
"learning_rate": 7.546638650246218e-05,
"loss": 2.1011,
"step": 7460
},
{
"epoch": 0.4309946918993769,
"grad_norm": 0.30027782917022705,
"learning_rate": 7.537457029530386e-05,
"loss": 2.0788,
"step": 7470
},
{
"epoch": 0.4315716593584122,
"grad_norm": 0.39448121190071106,
"learning_rate": 7.528263869636976e-05,
"loss": 2.0994,
"step": 7480
},
{
"epoch": 0.4321486268174475,
"grad_norm": 0.2679554224014282,
"learning_rate": 7.51905921237221e-05,
"loss": 2.0924,
"step": 7490
},
{
"epoch": 0.43272559427648283,
"grad_norm": 0.47223103046417236,
"learning_rate": 7.509843099594606e-05,
"loss": 2.0859,
"step": 7500
},
{
"epoch": 0.4333025617355181,
"grad_norm": 0.4030632972717285,
"learning_rate": 7.500615573214766e-05,
"loss": 2.095,
"step": 7510
},
{
"epoch": 0.4338795291945534,
"grad_norm": 0.2585602402687073,
"learning_rate": 7.491376675195199e-05,
"loss": 2.0963,
"step": 7520
},
{
"epoch": 0.43445649665358876,
"grad_norm": 0.24894532561302185,
"learning_rate": 7.482126447550132e-05,
"loss": 2.0907,
"step": 7530
},
{
"epoch": 0.43503346411262406,
"grad_norm": 0.30701470375061035,
"learning_rate": 7.472864932345307e-05,
"loss": 2.0904,
"step": 7540
},
{
"epoch": 0.43561043157165935,
"grad_norm": 0.303641676902771,
"learning_rate": 7.4635921716978e-05,
"loss": 2.0969,
"step": 7550
},
{
"epoch": 0.43618739903069464,
"grad_norm": 0.2713119685649872,
"learning_rate": 7.454308207775824e-05,
"loss": 2.0691,
"step": 7560
},
{
"epoch": 0.43676436648973,
"grad_norm": 0.3004838526248932,
"learning_rate": 7.445013082798542e-05,
"loss": 2.0791,
"step": 7570
},
{
"epoch": 0.4373413339487653,
"grad_norm": 0.31855785846710205,
"learning_rate": 7.435706839035869e-05,
"loss": 2.0921,
"step": 7580
},
{
"epoch": 0.4379183014078006,
"grad_norm": 0.2251572161912918,
"learning_rate": 7.426389518808287e-05,
"loss": 2.0784,
"step": 7590
},
{
"epoch": 0.43849526886683593,
"grad_norm": 0.33400654792785645,
"learning_rate": 7.417061164486644e-05,
"loss": 2.0949,
"step": 7600
},
{
"epoch": 0.4390722363258712,
"grad_norm": 0.27124184370040894,
"learning_rate": 7.407721818491975e-05,
"loss": 2.0815,
"step": 7610
},
{
"epoch": 0.4396492037849065,
"grad_norm": 0.3008407950401306,
"learning_rate": 7.398371523295286e-05,
"loss": 2.0848,
"step": 7620
},
{
"epoch": 0.44022617124394187,
"grad_norm": 0.2795826196670532,
"learning_rate": 7.389010321417385e-05,
"loss": 2.0879,
"step": 7630
},
{
"epoch": 0.44080313870297716,
"grad_norm": 0.2528461515903473,
"learning_rate": 7.379638255428677e-05,
"loss": 2.0914,
"step": 7640
},
{
"epoch": 0.44138010616201245,
"grad_norm": 0.5015778541564941,
"learning_rate": 7.370255367948971e-05,
"loss": 2.0744,
"step": 7650
},
{
"epoch": 0.44195707362104775,
"grad_norm": 0.267972856760025,
"learning_rate": 7.360861701647287e-05,
"loss": 2.0914,
"step": 7660
},
{
"epoch": 0.4425340410800831,
"grad_norm": 0.31362441182136536,
"learning_rate": 7.351457299241659e-05,
"loss": 2.0884,
"step": 7670
},
{
"epoch": 0.4431110085391184,
"grad_norm": 0.2507345378398895,
"learning_rate": 7.342042203498951e-05,
"loss": 2.0877,
"step": 7680
},
{
"epoch": 0.4436879759981537,
"grad_norm": 0.26686784625053406,
"learning_rate": 7.33261645723465e-05,
"loss": 2.0786,
"step": 7690
},
{
"epoch": 0.44426494345718903,
"grad_norm": 0.2581067681312561,
"learning_rate": 7.323180103312677e-05,
"loss": 2.09,
"step": 7700
},
{
"epoch": 0.4448419109162243,
"grad_norm": 0.27772027254104614,
"learning_rate": 7.313733184645193e-05,
"loss": 2.0982,
"step": 7710
},
{
"epoch": 0.4454188783752596,
"grad_norm": 0.3083910048007965,
"learning_rate": 7.304275744192402e-05,
"loss": 2.0919,
"step": 7720
},
{
"epoch": 0.44599584583429497,
"grad_norm": 0.29059112071990967,
"learning_rate": 7.294807824962358e-05,
"loss": 2.0845,
"step": 7730
},
{
"epoch": 0.44657281329333026,
"grad_norm": 0.3928695321083069,
"learning_rate": 7.285329470010765e-05,
"loss": 2.0792,
"step": 7740
},
{
"epoch": 0.44714978075236556,
"grad_norm": 0.379013329744339,
"learning_rate": 7.275840722440784e-05,
"loss": 2.0873,
"step": 7750
},
{
"epoch": 0.4477267482114009,
"grad_norm": 0.2709653079509735,
"learning_rate": 7.266341625402836e-05,
"loss": 2.0761,
"step": 7760
},
{
"epoch": 0.4483037156704362,
"grad_norm": 0.2680635452270508,
"learning_rate": 7.256832222094413e-05,
"loss": 2.0824,
"step": 7770
},
{
"epoch": 0.4488806831294715,
"grad_norm": 0.22318679094314575,
"learning_rate": 7.247312555759867e-05,
"loss": 2.108,
"step": 7780
},
{
"epoch": 0.4494576505885068,
"grad_norm": 0.27637118101119995,
"learning_rate": 7.237782669690226e-05,
"loss": 2.091,
"step": 7790
},
{
"epoch": 0.45003461804754213,
"grad_norm": 0.24532774090766907,
"learning_rate": 7.22824260722299e-05,
"loss": 2.0868,
"step": 7800
},
{
"epoch": 0.45061158550657743,
"grad_norm": 0.3002353310585022,
"learning_rate": 7.21869241174194e-05,
"loss": 2.0886,
"step": 7810
},
{
"epoch": 0.4511885529656127,
"grad_norm": 0.2682434618473053,
"learning_rate": 7.209132126676934e-05,
"loss": 2.081,
"step": 7820
},
{
"epoch": 0.45176552042464807,
"grad_norm": 0.35407841205596924,
"learning_rate": 7.199561795503715e-05,
"loss": 2.0778,
"step": 7830
},
{
"epoch": 0.45234248788368336,
"grad_norm": 0.27851682901382446,
"learning_rate": 7.189981461743706e-05,
"loss": 2.0962,
"step": 7840
},
{
"epoch": 0.45291945534271866,
"grad_norm": 0.30708226561546326,
"learning_rate": 7.180391168963828e-05,
"loss": 2.0771,
"step": 7850
},
{
"epoch": 0.453496422801754,
"grad_norm": 0.2830917537212372,
"learning_rate": 7.170790960776279e-05,
"loss": 2.0888,
"step": 7860
},
{
"epoch": 0.4540733902607893,
"grad_norm": 0.4283241629600525,
"learning_rate": 7.161180880838355e-05,
"loss": 2.0811,
"step": 7870
},
{
"epoch": 0.4546503577198246,
"grad_norm": 0.48016196489334106,
"learning_rate": 7.15156097285224e-05,
"loss": 2.0915,
"step": 7880
},
{
"epoch": 0.4552273251788599,
"grad_norm": 0.2744823396205902,
"learning_rate": 7.141931280564815e-05,
"loss": 2.0774,
"step": 7890
},
{
"epoch": 0.45580429263789524,
"grad_norm": 0.2494378238916397,
"learning_rate": 7.132291847767455e-05,
"loss": 2.0788,
"step": 7900
},
{
"epoch": 0.45638126009693053,
"grad_norm": 0.2590774893760681,
"learning_rate": 7.122642718295826e-05,
"loss": 2.0733,
"step": 7910
},
{
"epoch": 0.4569582275559658,
"grad_norm": 0.37552034854888916,
"learning_rate": 7.112983936029694e-05,
"loss": 2.0732,
"step": 7920
},
{
"epoch": 0.4575351950150012,
"grad_norm": 0.3342891037464142,
"learning_rate": 7.103315544892723e-05,
"loss": 2.0859,
"step": 7930
},
{
"epoch": 0.45811216247403647,
"grad_norm": 0.31473156809806824,
"learning_rate": 7.093637588852267e-05,
"loss": 2.0821,
"step": 7940
},
{
"epoch": 0.45868912993307176,
"grad_norm": 0.3879741132259369,
"learning_rate": 7.083950111919187e-05,
"loss": 2.0668,
"step": 7950
},
{
"epoch": 0.4592660973921071,
"grad_norm": 0.2961817979812622,
"learning_rate": 7.074253158147629e-05,
"loss": 2.0782,
"step": 7960
},
{
"epoch": 0.4598430648511424,
"grad_norm": 0.2857101559638977,
"learning_rate": 7.064546771634844e-05,
"loss": 2.0788,
"step": 7970
},
{
"epoch": 0.4604200323101777,
"grad_norm": 0.3533235788345337,
"learning_rate": 7.054830996520974e-05,
"loss": 2.0796,
"step": 7980
},
{
"epoch": 0.460996999769213,
"grad_norm": 0.3479328453540802,
"learning_rate": 7.045105876988858e-05,
"loss": 2.073,
"step": 7990
},
{
"epoch": 0.46157396722824834,
"grad_norm": 0.3560927212238312,
"learning_rate": 7.035371457263828e-05,
"loss": 2.0701,
"step": 8000
},
{
"epoch": 0.46215093468728363,
"grad_norm": 0.48714378476142883,
"learning_rate": 7.02562778161351e-05,
"loss": 2.078,
"step": 8010
},
{
"epoch": 0.4627279021463189,
"grad_norm": 0.2729594111442566,
"learning_rate": 7.015874894347622e-05,
"loss": 2.0734,
"step": 8020
},
{
"epoch": 0.4633048696053543,
"grad_norm": 0.3087742328643799,
"learning_rate": 7.006112839817772e-05,
"loss": 2.0764,
"step": 8030
},
{
"epoch": 0.46388183706438957,
"grad_norm": 0.34214138984680176,
"learning_rate": 6.996341662417255e-05,
"loss": 2.071,
"step": 8040
},
{
"epoch": 0.46445880452342486,
"grad_norm": 0.3290503919124603,
"learning_rate": 6.986561406580853e-05,
"loss": 2.0635,
"step": 8050
},
{
"epoch": 0.4650357719824602,
"grad_norm": 0.2952798008918762,
"learning_rate": 6.976772116784634e-05,
"loss": 2.0825,
"step": 8060
},
{
"epoch": 0.4656127394414955,
"grad_norm": 0.3497956693172455,
"learning_rate": 6.966973837545747e-05,
"loss": 2.0729,
"step": 8070
},
{
"epoch": 0.4661897069005308,
"grad_norm": 0.3718845546245575,
"learning_rate": 6.957166613422219e-05,
"loss": 2.0671,
"step": 8080
},
{
"epoch": 0.46676667435956615,
"grad_norm": 0.24571186304092407,
"learning_rate": 6.947350489012761e-05,
"loss": 2.076,
"step": 8090
},
{
"epoch": 0.46734364181860144,
"grad_norm": 0.26415371894836426,
"learning_rate": 6.93752550895655e-05,
"loss": 2.0814,
"step": 8100
},
{
"epoch": 0.46792060927763673,
"grad_norm": 0.3642018735408783,
"learning_rate": 6.927691717933039e-05,
"loss": 2.0691,
"step": 8110
},
{
"epoch": 0.46849757673667203,
"grad_norm": 0.42086517810821533,
"learning_rate": 6.917849160661747e-05,
"loss": 2.0752,
"step": 8120
},
{
"epoch": 0.4690745441957074,
"grad_norm": 0.29952847957611084,
"learning_rate": 6.907997881902058e-05,
"loss": 2.0783,
"step": 8130
},
{
"epoch": 0.46965151165474267,
"grad_norm": 0.3071637451648712,
"learning_rate": 6.898137926453023e-05,
"loss": 2.066,
"step": 8140
},
{
"epoch": 0.47022847911377796,
"grad_norm": 0.2822725474834442,
"learning_rate": 6.888269339153144e-05,
"loss": 2.0866,
"step": 8150
},
{
"epoch": 0.4708054465728133,
"grad_norm": 0.32353177666664124,
"learning_rate": 6.878392164880175e-05,
"loss": 2.0795,
"step": 8160
},
{
"epoch": 0.4713824140318486,
"grad_norm": 0.2756877839565277,
"learning_rate": 6.868506448550928e-05,
"loss": 2.0911,
"step": 8170
},
{
"epoch": 0.4719593814908839,
"grad_norm": 0.29774340987205505,
"learning_rate": 6.858612235121054e-05,
"loss": 2.0805,
"step": 8180
},
{
"epoch": 0.47253634894991925,
"grad_norm": 0.2543671727180481,
"learning_rate": 6.848709569584842e-05,
"loss": 2.077,
"step": 8190
},
{
"epoch": 0.47311331640895454,
"grad_norm": 0.6061549782752991,
"learning_rate": 6.838798496975027e-05,
"loss": 2.0623,
"step": 8200
},
{
"epoch": 0.47369028386798984,
"grad_norm": 0.26145073771476746,
"learning_rate": 6.82887906236257e-05,
"loss": 2.0843,
"step": 8210
},
{
"epoch": 0.47426725132702513,
"grad_norm": 0.30414700508117676,
"learning_rate": 6.818951310856456e-05,
"loss": 2.0652,
"step": 8220
},
{
"epoch": 0.4748442187860605,
"grad_norm": 0.2926621735095978,
"learning_rate": 6.809015287603493e-05,
"loss": 2.0845,
"step": 8230
},
{
"epoch": 0.4754211862450958,
"grad_norm": 0.27972277998924255,
"learning_rate": 6.799071037788108e-05,
"loss": 2.0572,
"step": 8240
},
{
"epoch": 0.47599815370413107,
"grad_norm": 0.3105740249156952,
"learning_rate": 6.789118606632134e-05,
"loss": 2.0678,
"step": 8250
},
{
"epoch": 0.4765751211631664,
"grad_norm": 0.32389092445373535,
"learning_rate": 6.779158039394616e-05,
"loss": 2.0793,
"step": 8260
},
{
"epoch": 0.4771520886222017,
"grad_norm": 0.3562007248401642,
"learning_rate": 6.769189381371591e-05,
"loss": 2.0802,
"step": 8270
},
{
"epoch": 0.477729056081237,
"grad_norm": 0.2581557035446167,
"learning_rate": 6.759212677895893e-05,
"loss": 2.0635,
"step": 8280
},
{
"epoch": 0.47830602354027235,
"grad_norm": 0.24171455204486847,
"learning_rate": 6.749227974336938e-05,
"loss": 2.07,
"step": 8290
},
{
"epoch": 0.47888299099930765,
"grad_norm": 0.2817774713039398,
"learning_rate": 6.739235316100532e-05,
"loss": 2.0836,
"step": 8300
},
{
"epoch": 0.47945995845834294,
"grad_norm": 0.2664335370063782,
"learning_rate": 6.729234748628643e-05,
"loss": 2.0824,
"step": 8310
},
{
"epoch": 0.48003692591737823,
"grad_norm": 0.3076491057872772,
"learning_rate": 6.719226317399219e-05,
"loss": 2.0858,
"step": 8320
},
{
"epoch": 0.4806138933764136,
"grad_norm": 0.37371984124183655,
"learning_rate": 6.709210067925961e-05,
"loss": 2.075,
"step": 8330
},
{
"epoch": 0.4811908608354489,
"grad_norm": 0.29041746258735657,
"learning_rate": 6.699186045758124e-05,
"loss": 2.0636,
"step": 8340
},
{
"epoch": 0.48176782829448417,
"grad_norm": 0.3501351773738861,
"learning_rate": 6.689154296480313e-05,
"loss": 2.0737,
"step": 8350
},
{
"epoch": 0.4823447957535195,
"grad_norm": 0.3749564588069916,
"learning_rate": 6.679114865712271e-05,
"loss": 2.075,
"step": 8360
},
{
"epoch": 0.4829217632125548,
"grad_norm": 0.364692747592926,
"learning_rate": 6.669067799108671e-05,
"loss": 2.0708,
"step": 8370
},
{
"epoch": 0.4834987306715901,
"grad_norm": 0.34302058815956116,
"learning_rate": 6.659013142358912e-05,
"loss": 2.0702,
"step": 8380
},
{
"epoch": 0.48407569813062545,
"grad_norm": 0.33347088098526,
"learning_rate": 6.648950941186909e-05,
"loss": 2.0759,
"step": 8390
},
{
"epoch": 0.48465266558966075,
"grad_norm": 0.2980937063694,
"learning_rate": 6.638881241350884e-05,
"loss": 2.0607,
"step": 8400
},
{
"epoch": 0.48522963304869604,
"grad_norm": 0.2834320664405823,
"learning_rate": 6.628804088643165e-05,
"loss": 2.0604,
"step": 8410
},
{
"epoch": 0.4858066005077314,
"grad_norm": 0.2671113908290863,
"learning_rate": 6.618719528889966e-05,
"loss": 2.0687,
"step": 8420
},
{
"epoch": 0.4863835679667667,
"grad_norm": 0.4396528899669647,
"learning_rate": 6.608627607951185e-05,
"loss": 2.0689,
"step": 8430
},
{
"epoch": 0.486960535425802,
"grad_norm": 0.31399500370025635,
"learning_rate": 6.598528371720201e-05,
"loss": 2.0692,
"step": 8440
},
{
"epoch": 0.48753750288483727,
"grad_norm": 0.23927593231201172,
"learning_rate": 6.588421866123653e-05,
"loss": 2.0836,
"step": 8450
},
{
"epoch": 0.4881144703438726,
"grad_norm": 0.32067403197288513,
"learning_rate": 6.578308137121242e-05,
"loss": 2.0747,
"step": 8460
},
{
"epoch": 0.4886914378029079,
"grad_norm": 0.25653791427612305,
"learning_rate": 6.56818723070552e-05,
"loss": 2.0697,
"step": 8470
},
{
"epoch": 0.4892684052619432,
"grad_norm": 0.35823753476142883,
"learning_rate": 6.558059192901667e-05,
"loss": 2.0654,
"step": 8480
},
{
"epoch": 0.48984537272097856,
"grad_norm": 0.3746614158153534,
"learning_rate": 6.547924069767304e-05,
"loss": 2.0759,
"step": 8490
},
{
"epoch": 0.49042234018001385,
"grad_norm": 0.284355491399765,
"learning_rate": 6.537781907392274e-05,
"loss": 2.0752,
"step": 8500
},
{
"epoch": 0.49099930763904914,
"grad_norm": 0.5059247016906738,
"learning_rate": 6.527632751898423e-05,
"loss": 2.0586,
"step": 8510
},
{
"epoch": 0.4915762750980845,
"grad_norm": 0.3632018566131592,
"learning_rate": 6.517476649439403e-05,
"loss": 2.0768,
"step": 8520
},
{
"epoch": 0.4921532425571198,
"grad_norm": 0.309618204832077,
"learning_rate": 6.50731364620046e-05,
"loss": 2.0766,
"step": 8530
},
{
"epoch": 0.4927302100161551,
"grad_norm": 0.48799097537994385,
"learning_rate": 6.497143788398215e-05,
"loss": 2.0806,
"step": 8540
},
{
"epoch": 0.4933071774751904,
"grad_norm": 0.25191518664360046,
"learning_rate": 6.486967122280462e-05,
"loss": 2.0661,
"step": 8550
},
{
"epoch": 0.4938841449342257,
"grad_norm": 0.2887539565563202,
"learning_rate": 6.476783694125964e-05,
"loss": 2.0657,
"step": 8560
},
{
"epoch": 0.494461112393261,
"grad_norm": 0.2615150213241577,
"learning_rate": 6.466593550244227e-05,
"loss": 2.0777,
"step": 8570
},
{
"epoch": 0.4950380798522963,
"grad_norm": 0.3972332179546356,
"learning_rate": 6.456396736975297e-05,
"loss": 2.0726,
"step": 8580
},
{
"epoch": 0.49561504731133166,
"grad_norm": 0.2527362108230591,
"learning_rate": 6.446193300689553e-05,
"loss": 2.0695,
"step": 8590
},
{
"epoch": 0.49619201477036695,
"grad_norm": 0.27918368577957153,
"learning_rate": 6.435983287787489e-05,
"loss": 2.0726,
"step": 8600
},
{
"epoch": 0.49676898222940225,
"grad_norm": 0.34241583943367004,
"learning_rate": 6.425766744699507e-05,
"loss": 2.0653,
"step": 8610
},
{
"epoch": 0.4973459496884376,
"grad_norm": 0.33931684494018555,
"learning_rate": 6.415543717885707e-05,
"loss": 2.0674,
"step": 8620
},
{
"epoch": 0.4979229171474729,
"grad_norm": 0.2961905598640442,
"learning_rate": 6.405314253835675e-05,
"loss": 2.057,
"step": 8630
},
{
"epoch": 0.4984998846065082,
"grad_norm": 0.405836284160614,
"learning_rate": 6.395078399068264e-05,
"loss": 2.0448,
"step": 8640
},
{
"epoch": 0.49907685206554353,
"grad_norm": 0.3253948986530304,
"learning_rate": 6.384836200131398e-05,
"loss": 2.0713,
"step": 8650
},
{
"epoch": 0.4996538195245788,
"grad_norm": 0.3465583920478821,
"learning_rate": 6.374587703601846e-05,
"loss": 2.076,
"step": 8660
},
{
"epoch": 0.5002307869836141,
"grad_norm": 0.25997796654701233,
"learning_rate": 6.364332956085014e-05,
"loss": 2.0584,
"step": 8670
},
{
"epoch": 0.5008077544426495,
"grad_norm": 0.31124719977378845,
"learning_rate": 6.35407200421474e-05,
"loss": 2.0722,
"step": 8680
},
{
"epoch": 0.5013847219016847,
"grad_norm": 0.41811326146125793,
"learning_rate": 6.343804894653072e-05,
"loss": 2.0436,
"step": 8690
},
{
"epoch": 0.50196168936072,
"grad_norm": 0.3640015423297882,
"learning_rate": 6.333531674090061e-05,
"loss": 2.0728,
"step": 8700
},
{
"epoch": 0.5025386568197554,
"grad_norm": 0.22485965490341187,
"learning_rate": 6.323252389243553e-05,
"loss": 2.0681,
"step": 8710
},
{
"epoch": 0.5031156242787906,
"grad_norm": 0.2506142854690552,
"learning_rate": 6.312967086858963e-05,
"loss": 2.0669,
"step": 8720
},
{
"epoch": 0.503692591737826,
"grad_norm": 0.3741120398044586,
"learning_rate": 6.302675813709077e-05,
"loss": 2.0548,
"step": 8730
},
{
"epoch": 0.5042695591968613,
"grad_norm": 0.2388111650943756,
"learning_rate": 6.292378616593833e-05,
"loss": 2.075,
"step": 8740
},
{
"epoch": 0.5048465266558966,
"grad_norm": 0.2684873044490814,
"learning_rate": 6.282075542340105e-05,
"loss": 2.0677,
"step": 8750
},
{
"epoch": 0.5054234941149319,
"grad_norm": 0.2846076488494873,
"learning_rate": 6.271766637801499e-05,
"loss": 2.0608,
"step": 8760
},
{
"epoch": 0.5060004615739673,
"grad_norm": 0.24689878523349762,
"learning_rate": 6.261451949858127e-05,
"loss": 2.0522,
"step": 8770
},
{
"epoch": 0.5065774290330025,
"grad_norm": 0.4451148808002472,
"learning_rate": 6.251131525416408e-05,
"loss": 2.0595,
"step": 8780
},
{
"epoch": 0.5071543964920379,
"grad_norm": 0.4091099500656128,
"learning_rate": 6.240805411408844e-05,
"loss": 2.0484,
"step": 8790
},
{
"epoch": 0.5077313639510732,
"grad_norm": 0.24189534783363342,
"learning_rate": 6.230473654793812e-05,
"loss": 2.0602,
"step": 8800
},
{
"epoch": 0.5083083314101084,
"grad_norm": 0.2729124426841736,
"learning_rate": 6.220136302555349e-05,
"loss": 2.0724,
"step": 8810
},
{
"epoch": 0.5088852988691438,
"grad_norm": 0.29464176297187805,
"learning_rate": 6.209793401702938e-05,
"loss": 2.0466,
"step": 8820
},
{
"epoch": 0.509462266328179,
"grad_norm": 0.36005499958992004,
"learning_rate": 6.199444999271295e-05,
"loss": 2.0682,
"step": 8830
},
{
"epoch": 0.5100392337872144,
"grad_norm": 0.33139434456825256,
"learning_rate": 6.189091142320154e-05,
"loss": 2.0666,
"step": 8840
},
{
"epoch": 0.5106162012462497,
"grad_norm": 0.23931476473808289,
"learning_rate": 6.178731877934053e-05,
"loss": 2.065,
"step": 8850
},
{
"epoch": 0.511193168705285,
"grad_norm": 0.4048471748828888,
"learning_rate": 6.168367253222121e-05,
"loss": 2.0681,
"step": 8860
},
{
"epoch": 0.5117701361643203,
"grad_norm": 0.32574188709259033,
"learning_rate": 6.157997315317866e-05,
"loss": 2.0524,
"step": 8870
},
{
"epoch": 0.5123471036233557,
"grad_norm": 0.2571089267730713,
"learning_rate": 6.147622111378953e-05,
"loss": 2.0675,
"step": 8880
},
{
"epoch": 0.5129240710823909,
"grad_norm": 0.30575376749038696,
"learning_rate": 6.137241688586999e-05,
"loss": 2.0588,
"step": 8890
},
{
"epoch": 0.5135010385414263,
"grad_norm": 0.29877418279647827,
"learning_rate": 6.12685609414735e-05,
"loss": 2.0551,
"step": 8900
},
{
"epoch": 0.5140780060004616,
"grad_norm": 0.2831680476665497,
"learning_rate": 6.116465375288874e-05,
"loss": 2.0544,
"step": 8910
},
{
"epoch": 0.5146549734594968,
"grad_norm": 0.24778659641742706,
"learning_rate": 6.106069579263739e-05,
"loss": 2.0718,
"step": 8920
},
{
"epoch": 0.5152319409185322,
"grad_norm": 0.2545316517353058,
"learning_rate": 6.095668753347203e-05,
"loss": 2.0755,
"step": 8930
},
{
"epoch": 0.5158089083775675,
"grad_norm": 0.2952696979045868,
"learning_rate": 6.085262944837401e-05,
"loss": 2.0659,
"step": 8940
},
{
"epoch": 0.5163858758366028,
"grad_norm": 0.3033768832683563,
"learning_rate": 6.0748522010551215e-05,
"loss": 2.0602,
"step": 8950
},
{
"epoch": 0.5169628432956381,
"grad_norm": 0.28780362010002136,
"learning_rate": 6.0644365693435966e-05,
"loss": 2.0636,
"step": 8960
},
{
"epoch": 0.5175398107546735,
"grad_norm": 0.3159029483795166,
"learning_rate": 6.05401609706829e-05,
"loss": 2.0676,
"step": 8970
},
{
"epoch": 0.5181167782137087,
"grad_norm": 0.27421289682388306,
"learning_rate": 6.043590831616677e-05,
"loss": 2.0687,
"step": 8980
},
{
"epoch": 0.5186937456727441,
"grad_norm": 0.27115151286125183,
"learning_rate": 6.033160820398025e-05,
"loss": 2.0566,
"step": 8990
},
{
"epoch": 0.5192707131317794,
"grad_norm": 0.2915196120738983,
"learning_rate": 6.022726110843192e-05,
"loss": 2.0522,
"step": 9000
},
{
"epoch": 0.5198476805908147,
"grad_norm": 0.2636251151561737,
"learning_rate": 6.0122867504043946e-05,
"loss": 2.0512,
"step": 9010
},
{
"epoch": 0.52042464804985,
"grad_norm": 0.23821695148944855,
"learning_rate": 6.001842786555002e-05,
"loss": 2.0418,
"step": 9020
},
{
"epoch": 0.5210016155088854,
"grad_norm": 0.3464409410953522,
"learning_rate": 5.991394266789319e-05,
"loss": 2.0717,
"step": 9030
},
{
"epoch": 0.5215785829679206,
"grad_norm": 0.315859854221344,
"learning_rate": 5.9809412386223615e-05,
"loss": 2.0537,
"step": 9040
},
{
"epoch": 0.5221555504269559,
"grad_norm": 0.32564985752105713,
"learning_rate": 5.9704837495896604e-05,
"loss": 2.0756,
"step": 9050
},
{
"epoch": 0.5227325178859912,
"grad_norm": 0.32612520456314087,
"learning_rate": 5.9600218472470196e-05,
"loss": 2.0571,
"step": 9060
},
{
"epoch": 0.5233094853450265,
"grad_norm": 0.25429829955101013,
"learning_rate": 5.9495555791703214e-05,
"loss": 2.0685,
"step": 9070
},
{
"epoch": 0.5238864528040619,
"grad_norm": 0.2449975311756134,
"learning_rate": 5.939084992955297e-05,
"loss": 2.0534,
"step": 9080
},
{
"epoch": 0.5244634202630971,
"grad_norm": 0.4133405089378357,
"learning_rate": 5.928610136217317e-05,
"loss": 2.0582,
"step": 9090
},
{
"epoch": 0.5250403877221325,
"grad_norm": 0.3849872350692749,
"learning_rate": 5.918131056591169e-05,
"loss": 2.0663,
"step": 9100
},
{
"epoch": 0.5256173551811678,
"grad_norm": 0.41498827934265137,
"learning_rate": 5.907647801730848e-05,
"loss": 2.0535,
"step": 9110
},
{
"epoch": 0.526194322640203,
"grad_norm": 0.34519997239112854,
"learning_rate": 5.8971604193093324e-05,
"loss": 2.0579,
"step": 9120
},
{
"epoch": 0.5267712900992384,
"grad_norm": 0.2998815178871155,
"learning_rate": 5.8866689570183766e-05,
"loss": 2.068,
"step": 9130
},
{
"epoch": 0.5273482575582737,
"grad_norm": 0.24292801320552826,
"learning_rate": 5.876173462568282e-05,
"loss": 2.0493,
"step": 9140
},
{
"epoch": 0.527925225017309,
"grad_norm": 0.29909130930900574,
"learning_rate": 5.86567398368769e-05,
"loss": 2.0573,
"step": 9150
},
{
"epoch": 0.5285021924763443,
"grad_norm": 0.256276935338974,
"learning_rate": 5.8551705681233604e-05,
"loss": 2.0545,
"step": 9160
},
{
"epoch": 0.5290791599353797,
"grad_norm": 0.26847782731056213,
"learning_rate": 5.8446632636399534e-05,
"loss": 2.0723,
"step": 9170
},
{
"epoch": 0.5296561273944149,
"grad_norm": 0.2674199342727661,
"learning_rate": 5.8341521180198177e-05,
"loss": 2.067,
"step": 9180
},
{
"epoch": 0.5302330948534503,
"grad_norm": 0.26743456721305847,
"learning_rate": 5.82363717906277e-05,
"loss": 2.0697,
"step": 9190
},
{
"epoch": 0.5308100623124856,
"grad_norm": 0.302480012178421,
"learning_rate": 5.81311849458587e-05,
"loss": 2.0561,
"step": 9200
},
{
"epoch": 0.5313870297715209,
"grad_norm": 0.30073806643486023,
"learning_rate": 5.802596112423221e-05,
"loss": 2.0624,
"step": 9210
},
{
"epoch": 0.5319639972305562,
"grad_norm": 0.38998448848724365,
"learning_rate": 5.7920700804257346e-05,
"loss": 2.0574,
"step": 9220
},
{
"epoch": 0.5325409646895916,
"grad_norm": 0.4289480745792389,
"learning_rate": 5.78154044646092e-05,
"loss": 2.0635,
"step": 9230
},
{
"epoch": 0.5331179321486268,
"grad_norm": 0.2888721823692322,
"learning_rate": 5.771007258412669e-05,
"loss": 2.0494,
"step": 9240
},
{
"epoch": 0.5336948996076621,
"grad_norm": 0.30417218804359436,
"learning_rate": 5.760470564181038e-05,
"loss": 2.0512,
"step": 9250
},
{
"epoch": 0.5342718670666975,
"grad_norm": 0.2877718508243561,
"learning_rate": 5.749930411682022e-05,
"loss": 2.0628,
"step": 9260
},
{
"epoch": 0.5348488345257327,
"grad_norm": 0.24942269921302795,
"learning_rate": 5.739386848847346e-05,
"loss": 2.0603,
"step": 9270
},
{
"epoch": 0.5354258019847681,
"grad_norm": 0.2586665749549866,
"learning_rate": 5.728839923624243e-05,
"loss": 2.0637,
"step": 9280
},
{
"epoch": 0.5360027694438033,
"grad_norm": 0.2691705822944641,
"learning_rate": 5.7182896839752374e-05,
"loss": 2.0582,
"step": 9290
},
{
"epoch": 0.5365797369028387,
"grad_norm": 0.24790114164352417,
"learning_rate": 5.7077361778779246e-05,
"loss": 2.0675,
"step": 9300
},
{
"epoch": 0.537156704361874,
"grad_norm": 0.2401387095451355,
"learning_rate": 5.697179453324758e-05,
"loss": 2.0668,
"step": 9310
},
{
"epoch": 0.5377336718209093,
"grad_norm": 0.24574357271194458,
"learning_rate": 5.686619558322821e-05,
"loss": 2.0511,
"step": 9320
},
{
"epoch": 0.5383106392799446,
"grad_norm": 0.3961896002292633,
"learning_rate": 5.676056540893618e-05,
"loss": 2.0449,
"step": 9330
},
{
"epoch": 0.53888760673898,
"grad_norm": 0.35582342743873596,
"learning_rate": 5.665490449072854e-05,
"loss": 2.0737,
"step": 9340
},
{
"epoch": 0.5394645741980152,
"grad_norm": 0.2613429129123688,
"learning_rate": 5.6549213309102136e-05,
"loss": 2.0627,
"step": 9350
},
{
"epoch": 0.5400415416570505,
"grad_norm": 0.24746835231781006,
"learning_rate": 5.6443492344691426e-05,
"loss": 2.0437,
"step": 9360
},
{
"epoch": 0.5406185091160859,
"grad_norm": 0.2716011106967926,
"learning_rate": 5.633774207826632e-05,
"loss": 2.0565,
"step": 9370
},
{
"epoch": 0.5411954765751211,
"grad_norm": 0.2716337740421295,
"learning_rate": 5.623196299072999e-05,
"loss": 2.0446,
"step": 9380
},
{
"epoch": 0.5417724440341565,
"grad_norm": 0.26952430605888367,
"learning_rate": 5.612615556311665e-05,
"loss": 2.0628,
"step": 9390
},
{
"epoch": 0.5423494114931918,
"grad_norm": 0.23789243400096893,
"learning_rate": 5.602032027658941e-05,
"loss": 2.0583,
"step": 9400
},
{
"epoch": 0.5429263789522271,
"grad_norm": 0.3613603413105011,
"learning_rate": 5.591445761243806e-05,
"loss": 2.0637,
"step": 9410
},
{
"epoch": 0.5435033464112624,
"grad_norm": 0.5396514534950256,
"learning_rate": 5.580856805207687e-05,
"loss": 2.0623,
"step": 9420
},
{
"epoch": 0.5440803138702978,
"grad_norm": 0.2915268838405609,
"learning_rate": 5.5702652077042485e-05,
"loss": 2.0632,
"step": 9430
},
{
"epoch": 0.544657281329333,
"grad_norm": 0.22387048602104187,
"learning_rate": 5.5596710168991594e-05,
"loss": 2.0496,
"step": 9440
},
{
"epoch": 0.5452342487883683,
"grad_norm": 0.27139759063720703,
"learning_rate": 5.549074280969887e-05,
"loss": 2.0523,
"step": 9450
},
{
"epoch": 0.5458112162474037,
"grad_norm": 0.2295805811882019,
"learning_rate": 5.538475048105471e-05,
"loss": 2.0526,
"step": 9460
},
{
"epoch": 0.5463881837064389,
"grad_norm": 0.2620329260826111,
"learning_rate": 5.527873366506308e-05,
"loss": 2.039,
"step": 9470
},
{
"epoch": 0.5469651511654743,
"grad_norm": 0.25486627221107483,
"learning_rate": 5.517269284383923e-05,
"loss": 2.0497,
"step": 9480
},
{
"epoch": 0.5475421186245096,
"grad_norm": 0.30340516567230225,
"learning_rate": 5.506662849960769e-05,
"loss": 2.0385,
"step": 9490
},
{
"epoch": 0.5481190860835449,
"grad_norm": 0.2664933204650879,
"learning_rate": 5.496054111469987e-05,
"loss": 2.046,
"step": 9500
},
{
"epoch": 0.5486960535425802,
"grad_norm": 0.33765774965286255,
"learning_rate": 5.485443117155198e-05,
"loss": 2.0455,
"step": 9510
},
{
"epoch": 0.5492730210016155,
"grad_norm": 0.26019567251205444,
"learning_rate": 5.474829915270283e-05,
"loss": 2.0551,
"step": 9520
},
{
"epoch": 0.5498499884606508,
"grad_norm": 0.3324459195137024,
"learning_rate": 5.464214554079162e-05,
"loss": 2.0465,
"step": 9530
},
{
"epoch": 0.5504269559196862,
"grad_norm": 0.23958685994148254,
"learning_rate": 5.453597081855574e-05,
"loss": 2.0526,
"step": 9540
},
{
"epoch": 0.5510039233787214,
"grad_norm": 0.24503254890441895,
"learning_rate": 5.442977546882856e-05,
"loss": 2.054,
"step": 9550
},
{
"epoch": 0.5515808908377567,
"grad_norm": 0.29835647344589233,
"learning_rate": 5.432355997453729e-05,
"loss": 2.0557,
"step": 9560
},
{
"epoch": 0.5521578582967921,
"grad_norm": 0.2821817398071289,
"learning_rate": 5.421732481870073e-05,
"loss": 2.0621,
"step": 9570
},
{
"epoch": 0.5527348257558273,
"grad_norm": 0.2836339473724365,
"learning_rate": 5.411107048442708e-05,
"loss": 2.0618,
"step": 9580
},
{
"epoch": 0.5533117932148627,
"grad_norm": 0.23530443012714386,
"learning_rate": 5.4004797454911784e-05,
"loss": 2.0579,
"step": 9590
},
{
"epoch": 0.553888760673898,
"grad_norm": 0.27028971910476685,
"learning_rate": 5.389850621343525e-05,
"loss": 2.0593,
"step": 9600
},
{
"epoch": 0.5544657281329333,
"grad_norm": 0.26975545287132263,
"learning_rate": 5.379219724336077e-05,
"loss": 2.049,
"step": 9610
},
{
"epoch": 0.5550426955919686,
"grad_norm": 0.2712181806564331,
"learning_rate": 5.368587102813224e-05,
"loss": 2.0533,
"step": 9620
},
{
"epoch": 0.555619663051004,
"grad_norm": 0.2469034343957901,
"learning_rate": 5.357952805127194e-05,
"loss": 2.0419,
"step": 9630
},
{
"epoch": 0.5561966305100392,
"grad_norm": 0.3148432970046997,
"learning_rate": 5.347316879637844e-05,
"loss": 2.0518,
"step": 9640
},
{
"epoch": 0.5567735979690746,
"grad_norm": 0.2899792492389679,
"learning_rate": 5.3366793747124245e-05,
"loss": 2.0508,
"step": 9650
},
{
"epoch": 0.5573505654281099,
"grad_norm": 0.2373170405626297,
"learning_rate": 5.3260403387253764e-05,
"loss": 2.0638,
"step": 9660
},
{
"epoch": 0.5579275328871451,
"grad_norm": 0.29644837975502014,
"learning_rate": 5.315399820058102e-05,
"loss": 2.0491,
"step": 9670
},
{
"epoch": 0.5585045003461805,
"grad_norm": 0.3238583505153656,
"learning_rate": 5.304757867098742e-05,
"loss": 2.0491,
"step": 9680
},
{
"epoch": 0.5590814678052158,
"grad_norm": 0.328906774520874,
"learning_rate": 5.294114528241963e-05,
"loss": 2.0633,
"step": 9690
},
{
"epoch": 0.5596584352642511,
"grad_norm": 0.3107961118221283,
"learning_rate": 5.283469851888735e-05,
"loss": 2.0385,
"step": 9700
},
{
"epoch": 0.5602354027232864,
"grad_norm": 0.3088037073612213,
"learning_rate": 5.272823886446107e-05,
"loss": 2.063,
"step": 9710
},
{
"epoch": 0.5608123701823217,
"grad_norm": 0.2720682919025421,
"learning_rate": 5.262176680326991e-05,
"loss": 2.0408,
"step": 9720
},
{
"epoch": 0.561389337641357,
"grad_norm": 0.2519686818122864,
"learning_rate": 5.251528281949942e-05,
"loss": 2.0475,
"step": 9730
},
{
"epoch": 0.5619663051003924,
"grad_norm": 0.2880825102329254,
"learning_rate": 5.240878739738938e-05,
"loss": 2.0548,
"step": 9740
},
{
"epoch": 0.5625432725594276,
"grad_norm": 0.3425326645374298,
"learning_rate": 5.230228102123157e-05,
"loss": 2.0552,
"step": 9750
},
{
"epoch": 0.563120240018463,
"grad_norm": 0.23194807767868042,
"learning_rate": 5.219576417536757e-05,
"loss": 2.0372,
"step": 9760
},
{
"epoch": 0.5636972074774983,
"grad_norm": 0.23607051372528076,
"learning_rate": 5.2089237344186604e-05,
"loss": 2.0448,
"step": 9770
},
{
"epoch": 0.5642741749365335,
"grad_norm": 0.31973376870155334,
"learning_rate": 5.1982701012123293e-05,
"loss": 2.043,
"step": 9780
},
{
"epoch": 0.5648511423955689,
"grad_norm": 0.3390558063983917,
"learning_rate": 5.1876155663655423e-05,
"loss": 2.05,
"step": 9790
},
{
"epoch": 0.5654281098546042,
"grad_norm": 0.24147625267505646,
"learning_rate": 5.176960178330186e-05,
"loss": 2.0488,
"step": 9800
},
{
"epoch": 0.5660050773136395,
"grad_norm": 0.226022869348526,
"learning_rate": 5.166303985562021e-05,
"loss": 2.0434,
"step": 9810
},
{
"epoch": 0.5665820447726748,
"grad_norm": 0.2638387680053711,
"learning_rate": 5.155647036520469e-05,
"loss": 2.0366,
"step": 9820
},
{
"epoch": 0.5671590122317102,
"grad_norm": 0.31156018376350403,
"learning_rate": 5.144989379668391e-05,
"loss": 2.0387,
"step": 9830
},
{
"epoch": 0.5677359796907454,
"grad_norm": 0.23942683637142181,
"learning_rate": 5.1343310634718665e-05,
"loss": 2.0422,
"step": 9840
},
{
"epoch": 0.5683129471497808,
"grad_norm": 0.29752033948898315,
"learning_rate": 5.123672136399975e-05,
"loss": 2.0551,
"step": 9850
},
{
"epoch": 0.5688899146088161,
"grad_norm": 0.26006224751472473,
"learning_rate": 5.113012646924572e-05,
"loss": 2.0375,
"step": 9860
},
{
"epoch": 0.5694668820678513,
"grad_norm": 0.31031838059425354,
"learning_rate": 5.102352643520072e-05,
"loss": 2.0491,
"step": 9870
},
{
"epoch": 0.5700438495268867,
"grad_norm": 0.2978913187980652,
"learning_rate": 5.091692174663224e-05,
"loss": 2.0709,
"step": 9880
},
{
"epoch": 0.570620816985922,
"grad_norm": 0.29858845472335815,
"learning_rate": 5.0810312888328985e-05,
"loss": 2.0553,
"step": 9890
},
{
"epoch": 0.5711977844449573,
"grad_norm": 0.2596759498119354,
"learning_rate": 5.070370034509856e-05,
"loss": 2.0661,
"step": 9900
},
{
"epoch": 0.5717747519039926,
"grad_norm": 0.2523353695869446,
"learning_rate": 5.0597084601765376e-05,
"loss": 2.0336,
"step": 9910
},
{
"epoch": 0.572351719363028,
"grad_norm": 0.2640801668167114,
"learning_rate": 5.049046614316837e-05,
"loss": 2.0434,
"step": 9920
},
{
"epoch": 0.5729286868220632,
"grad_norm": 0.3435620069503784,
"learning_rate": 5.038384545415885e-05,
"loss": 2.0458,
"step": 9930
},
{
"epoch": 0.5735056542810986,
"grad_norm": 0.2811577022075653,
"learning_rate": 5.0277223019598265e-05,
"loss": 2.0498,
"step": 9940
},
{
"epoch": 0.5740826217401338,
"grad_norm": 0.33941999077796936,
"learning_rate": 5.0170599324355974e-05,
"loss": 2.0369,
"step": 9950
},
{
"epoch": 0.5746595891991692,
"grad_norm": 0.2926217019557953,
"learning_rate": 5.0063974853307086e-05,
"loss": 2.0558,
"step": 9960
},
{
"epoch": 0.5752365566582045,
"grad_norm": 0.2736097276210785,
"learning_rate": 4.9957350091330217e-05,
"loss": 2.0393,
"step": 9970
},
{
"epoch": 0.5758135241172397,
"grad_norm": 0.23603315651416779,
"learning_rate": 4.985072552330537e-05,
"loss": 2.0421,
"step": 9980
},
{
"epoch": 0.5763904915762751,
"grad_norm": 0.2496812641620636,
"learning_rate": 4.97441016341116e-05,
"loss": 2.0536,
"step": 9990
},
{
"epoch": 0.5769674590353104,
"grad_norm": 0.2511531114578247,
"learning_rate": 4.963747890862491e-05,
"loss": 2.0478,
"step": 10000
}
],
"logging_steps": 10,
"max_steps": 17332,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 6.182962763871027e+19,
"train_batch_size": 96,
"trial_name": null,
"trial_params": null
}