groot_orange_pick_10_10000 / trainer_state.json
yuta0x89's picture
Upload folder using huggingface_hub
90b831b verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 36.101083032490976,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.036101083032490974,
"grad_norm": 14.303378105163574,
"learning_rate": 1.8e-06,
"loss": 1.3221,
"step": 10
},
{
"epoch": 0.07220216606498195,
"grad_norm": 2.9238288402557373,
"learning_rate": 3.8e-06,
"loss": 1.0346,
"step": 20
},
{
"epoch": 0.10830324909747292,
"grad_norm": 3.4732825756073,
"learning_rate": 5.8e-06,
"loss": 0.5915,
"step": 30
},
{
"epoch": 0.1444043321299639,
"grad_norm": 1.5753566026687622,
"learning_rate": 7.8e-06,
"loss": 0.4362,
"step": 40
},
{
"epoch": 0.18050541516245489,
"grad_norm": 1.8541886806488037,
"learning_rate": 9.800000000000001e-06,
"loss": 0.336,
"step": 50
},
{
"epoch": 0.21660649819494585,
"grad_norm": 1.878593921661377,
"learning_rate": 1.18e-05,
"loss": 0.2899,
"step": 60
},
{
"epoch": 0.2527075812274368,
"grad_norm": 1.4575947523117065,
"learning_rate": 1.3800000000000002e-05,
"loss": 0.2418,
"step": 70
},
{
"epoch": 0.2888086642599278,
"grad_norm": 1.1900250911712646,
"learning_rate": 1.58e-05,
"loss": 0.2319,
"step": 80
},
{
"epoch": 0.3249097472924188,
"grad_norm": 1.4874290227890015,
"learning_rate": 1.78e-05,
"loss": 0.2111,
"step": 90
},
{
"epoch": 0.36101083032490977,
"grad_norm": 1.5714935064315796,
"learning_rate": 1.9800000000000004e-05,
"loss": 0.2215,
"step": 100
},
{
"epoch": 0.3971119133574007,
"grad_norm": 1.2372921705245972,
"learning_rate": 2.18e-05,
"loss": 0.1957,
"step": 110
},
{
"epoch": 0.4332129963898917,
"grad_norm": 1.1309036016464233,
"learning_rate": 2.38e-05,
"loss": 0.1844,
"step": 120
},
{
"epoch": 0.4693140794223827,
"grad_norm": 1.808655023574829,
"learning_rate": 2.58e-05,
"loss": 0.1993,
"step": 130
},
{
"epoch": 0.5054151624548736,
"grad_norm": 1.845122218132019,
"learning_rate": 2.7800000000000005e-05,
"loss": 0.1853,
"step": 140
},
{
"epoch": 0.5415162454873647,
"grad_norm": 1.3696422576904297,
"learning_rate": 2.98e-05,
"loss": 0.1659,
"step": 150
},
{
"epoch": 0.5776173285198556,
"grad_norm": 1.2619764804840088,
"learning_rate": 3.18e-05,
"loss": 0.1776,
"step": 160
},
{
"epoch": 0.6137184115523465,
"grad_norm": 1.7037168741226196,
"learning_rate": 3.38e-05,
"loss": 0.1587,
"step": 170
},
{
"epoch": 0.6498194945848376,
"grad_norm": 2.3919386863708496,
"learning_rate": 3.58e-05,
"loss": 0.1796,
"step": 180
},
{
"epoch": 0.6859205776173285,
"grad_norm": 1.7829045057296753,
"learning_rate": 3.7800000000000004e-05,
"loss": 0.1604,
"step": 190
},
{
"epoch": 0.7220216606498195,
"grad_norm": 0.6636906862258911,
"learning_rate": 3.9800000000000005e-05,
"loss": 0.1391,
"step": 200
},
{
"epoch": 0.7581227436823105,
"grad_norm": 1.4346169233322144,
"learning_rate": 4.18e-05,
"loss": 0.1466,
"step": 210
},
{
"epoch": 0.7942238267148014,
"grad_norm": 1.161212682723999,
"learning_rate": 4.38e-05,
"loss": 0.1553,
"step": 220
},
{
"epoch": 0.8303249097472925,
"grad_norm": 1.3307301998138428,
"learning_rate": 4.58e-05,
"loss": 0.1374,
"step": 230
},
{
"epoch": 0.8664259927797834,
"grad_norm": 1.3235535621643066,
"learning_rate": 4.78e-05,
"loss": 0.1443,
"step": 240
},
{
"epoch": 0.9025270758122743,
"grad_norm": 0.7945713996887207,
"learning_rate": 4.9800000000000004e-05,
"loss": 0.1454,
"step": 250
},
{
"epoch": 0.9386281588447654,
"grad_norm": 0.9834956526756287,
"learning_rate": 5.1800000000000005e-05,
"loss": 0.1352,
"step": 260
},
{
"epoch": 0.9747292418772563,
"grad_norm": 1.1631721258163452,
"learning_rate": 5.380000000000001e-05,
"loss": 0.1445,
"step": 270
},
{
"epoch": 1.0108303249097472,
"grad_norm": 1.0873337984085083,
"learning_rate": 5.580000000000001e-05,
"loss": 0.1174,
"step": 280
},
{
"epoch": 1.0469314079422383,
"grad_norm": 0.8637306690216064,
"learning_rate": 5.7799999999999995e-05,
"loss": 0.1338,
"step": 290
},
{
"epoch": 1.0830324909747293,
"grad_norm": 0.9020954370498657,
"learning_rate": 5.9800000000000003e-05,
"loss": 0.1227,
"step": 300
},
{
"epoch": 1.1191335740072201,
"grad_norm": 0.8427388668060303,
"learning_rate": 6.18e-05,
"loss": 0.1293,
"step": 310
},
{
"epoch": 1.1552346570397112,
"grad_norm": 0.8479349613189697,
"learning_rate": 6.38e-05,
"loss": 0.1245,
"step": 320
},
{
"epoch": 1.1913357400722022,
"grad_norm": 1.5972744226455688,
"learning_rate": 6.58e-05,
"loss": 0.1355,
"step": 330
},
{
"epoch": 1.2274368231046933,
"grad_norm": 1.494515061378479,
"learning_rate": 6.780000000000001e-05,
"loss": 0.1435,
"step": 340
},
{
"epoch": 1.263537906137184,
"grad_norm": 0.625743567943573,
"learning_rate": 6.98e-05,
"loss": 0.1228,
"step": 350
},
{
"epoch": 1.2996389891696751,
"grad_norm": 0.8733342289924622,
"learning_rate": 7.18e-05,
"loss": 0.1366,
"step": 360
},
{
"epoch": 1.335740072202166,
"grad_norm": 1.4155577421188354,
"learning_rate": 7.38e-05,
"loss": 0.1364,
"step": 370
},
{
"epoch": 1.371841155234657,
"grad_norm": 1.096773386001587,
"learning_rate": 7.58e-05,
"loss": 0.1194,
"step": 380
},
{
"epoch": 1.407942238267148,
"grad_norm": 0.7929869294166565,
"learning_rate": 7.780000000000001e-05,
"loss": 0.1149,
"step": 390
},
{
"epoch": 1.444043321299639,
"grad_norm": 0.6651306748390198,
"learning_rate": 7.98e-05,
"loss": 0.1162,
"step": 400
},
{
"epoch": 1.48014440433213,
"grad_norm": 0.9790053963661194,
"learning_rate": 8.18e-05,
"loss": 0.1226,
"step": 410
},
{
"epoch": 1.516245487364621,
"grad_norm": 0.9229372143745422,
"learning_rate": 8.38e-05,
"loss": 0.1025,
"step": 420
},
{
"epoch": 1.5523465703971118,
"grad_norm": 1.3593957424163818,
"learning_rate": 8.58e-05,
"loss": 0.1232,
"step": 430
},
{
"epoch": 1.5884476534296028,
"grad_norm": 0.9937139749526978,
"learning_rate": 8.78e-05,
"loss": 0.1098,
"step": 440
},
{
"epoch": 1.6245487364620939,
"grad_norm": 0.8832342624664307,
"learning_rate": 8.98e-05,
"loss": 0.1116,
"step": 450
},
{
"epoch": 1.660649819494585,
"grad_norm": 0.6249854564666748,
"learning_rate": 9.180000000000001e-05,
"loss": 0.1129,
"step": 460
},
{
"epoch": 1.696750902527076,
"grad_norm": 0.9037442803382874,
"learning_rate": 9.38e-05,
"loss": 0.1034,
"step": 470
},
{
"epoch": 1.7328519855595668,
"grad_norm": 0.5689355731010437,
"learning_rate": 9.58e-05,
"loss": 0.1103,
"step": 480
},
{
"epoch": 1.7689530685920578,
"grad_norm": 0.8034074306488037,
"learning_rate": 9.78e-05,
"loss": 0.1144,
"step": 490
},
{
"epoch": 1.8050541516245486,
"grad_norm": 0.6319280862808228,
"learning_rate": 9.98e-05,
"loss": 0.1048,
"step": 500
},
{
"epoch": 1.8411552346570397,
"grad_norm": 0.7947399020195007,
"learning_rate": 9.9999778549206e-05,
"loss": 0.1149,
"step": 510
},
{
"epoch": 1.8772563176895307,
"grad_norm": 0.8703307509422302,
"learning_rate": 9.999901304280685e-05,
"loss": 0.1105,
"step": 520
},
{
"epoch": 1.9133574007220218,
"grad_norm": 0.7093669772148132,
"learning_rate": 9.999770075521164e-05,
"loss": 0.1051,
"step": 530
},
{
"epoch": 1.9494584837545126,
"grad_norm": 0.8823049068450928,
"learning_rate": 9.99958417007713e-05,
"loss": 0.0929,
"step": 540
},
{
"epoch": 1.9855595667870036,
"grad_norm": 0.65891033411026,
"learning_rate": 9.999343589981615e-05,
"loss": 0.1047,
"step": 550
},
{
"epoch": 2.0216606498194944,
"grad_norm": 0.4699995219707489,
"learning_rate": 9.999048337865568e-05,
"loss": 0.0993,
"step": 560
},
{
"epoch": 2.0577617328519855,
"grad_norm": 0.7377736568450928,
"learning_rate": 9.998698416957815e-05,
"loss": 0.098,
"step": 570
},
{
"epoch": 2.0938628158844765,
"grad_norm": 0.7579376697540283,
"learning_rate": 9.998293831085037e-05,
"loss": 0.0952,
"step": 580
},
{
"epoch": 2.1299638989169676,
"grad_norm": 0.6771948337554932,
"learning_rate": 9.997834584671719e-05,
"loss": 0.1038,
"step": 590
},
{
"epoch": 2.1660649819494586,
"grad_norm": 1.0296669006347656,
"learning_rate": 9.997320682740107e-05,
"loss": 0.1115,
"step": 600
},
{
"epoch": 2.2021660649819497,
"grad_norm": 0.7494194507598877,
"learning_rate": 9.996752130910149e-05,
"loss": 0.1123,
"step": 610
},
{
"epoch": 2.2382671480144403,
"grad_norm": 0.635129988193512,
"learning_rate": 9.99612893539944e-05,
"loss": 0.0947,
"step": 620
},
{
"epoch": 2.2743682310469313,
"grad_norm": 0.8458411693572998,
"learning_rate": 9.995451103023144e-05,
"loss": 0.0962,
"step": 630
},
{
"epoch": 2.3104693140794224,
"grad_norm": 0.5293542742729187,
"learning_rate": 9.994718641193928e-05,
"loss": 0.1116,
"step": 640
},
{
"epoch": 2.3465703971119134,
"grad_norm": 0.7339059710502625,
"learning_rate": 9.993931557921874e-05,
"loss": 0.104,
"step": 650
},
{
"epoch": 2.3826714801444044,
"grad_norm": 0.8779047727584839,
"learning_rate": 9.993089861814402e-05,
"loss": 0.0947,
"step": 660
},
{
"epoch": 2.4187725631768955,
"grad_norm": 0.7545035481452942,
"learning_rate": 9.992193562076166e-05,
"loss": 0.1057,
"step": 670
},
{
"epoch": 2.4548736462093865,
"grad_norm": 0.6147914528846741,
"learning_rate": 9.991242668508954e-05,
"loss": 0.104,
"step": 680
},
{
"epoch": 2.490974729241877,
"grad_norm": 0.6101718544960022,
"learning_rate": 9.990237191511587e-05,
"loss": 0.0784,
"step": 690
},
{
"epoch": 2.527075812274368,
"grad_norm": 0.7269706726074219,
"learning_rate": 9.989177142079802e-05,
"loss": 0.0927,
"step": 700
},
{
"epoch": 2.563176895306859,
"grad_norm": 0.44793209433555603,
"learning_rate": 9.988062531806126e-05,
"loss": 0.0988,
"step": 710
},
{
"epoch": 2.5992779783393503,
"grad_norm": 0.8228976726531982,
"learning_rate": 9.986893372879762e-05,
"loss": 0.0946,
"step": 720
},
{
"epoch": 2.6353790613718413,
"grad_norm": 0.9659225344657898,
"learning_rate": 9.985669678086443e-05,
"loss": 0.0967,
"step": 730
},
{
"epoch": 2.671480144404332,
"grad_norm": 0.5384042263031006,
"learning_rate": 9.984391460808298e-05,
"loss": 0.0889,
"step": 740
},
{
"epoch": 2.707581227436823,
"grad_norm": 0.6501038670539856,
"learning_rate": 9.983058735023709e-05,
"loss": 0.0921,
"step": 750
},
{
"epoch": 2.743682310469314,
"grad_norm": 0.7665195465087891,
"learning_rate": 9.98167151530715e-05,
"loss": 0.0787,
"step": 760
},
{
"epoch": 2.779783393501805,
"grad_norm": 0.7574978470802307,
"learning_rate": 9.980229816829034e-05,
"loss": 0.088,
"step": 770
},
{
"epoch": 2.815884476534296,
"grad_norm": 0.4293161928653717,
"learning_rate": 9.978733655355544e-05,
"loss": 0.0898,
"step": 780
},
{
"epoch": 2.851985559566787,
"grad_norm": 0.6935308575630188,
"learning_rate": 9.977183047248464e-05,
"loss": 0.085,
"step": 790
},
{
"epoch": 2.888086642599278,
"grad_norm": 0.8048137426376343,
"learning_rate": 9.975578009464992e-05,
"loss": 0.0816,
"step": 800
},
{
"epoch": 2.9241877256317688,
"grad_norm": 0.7156538963317871,
"learning_rate": 9.97391855955757e-05,
"loss": 0.0889,
"step": 810
},
{
"epoch": 2.96028880866426,
"grad_norm": 0.43860772252082825,
"learning_rate": 9.972204715673669e-05,
"loss": 0.0949,
"step": 820
},
{
"epoch": 2.996389891696751,
"grad_norm": 0.6429822444915771,
"learning_rate": 9.970436496555617e-05,
"loss": 0.0814,
"step": 830
},
{
"epoch": 3.032490974729242,
"grad_norm": 0.8840290307998657,
"learning_rate": 9.968613921540373e-05,
"loss": 0.0828,
"step": 840
},
{
"epoch": 3.068592057761733,
"grad_norm": 1.0691182613372803,
"learning_rate": 9.966737010559326e-05,
"loss": 0.1034,
"step": 850
},
{
"epoch": 3.104693140794224,
"grad_norm": 0.7965003252029419,
"learning_rate": 9.964805784138072e-05,
"loss": 0.0855,
"step": 860
},
{
"epoch": 3.140794223826715,
"grad_norm": 0.5022170543670654,
"learning_rate": 9.962820263396195e-05,
"loss": 0.0844,
"step": 870
},
{
"epoch": 3.1768953068592056,
"grad_norm": 0.6797301769256592,
"learning_rate": 9.960780470047033e-05,
"loss": 0.0816,
"step": 880
},
{
"epoch": 3.2129963898916967,
"grad_norm": 0.5124582052230835,
"learning_rate": 9.958686426397437e-05,
"loss": 0.0786,
"step": 890
},
{
"epoch": 3.2490974729241877,
"grad_norm": 0.554582417011261,
"learning_rate": 9.956538155347534e-05,
"loss": 0.0934,
"step": 900
},
{
"epoch": 3.2851985559566788,
"grad_norm": 0.6361289024353027,
"learning_rate": 9.95433568039047e-05,
"loss": 0.0779,
"step": 910
},
{
"epoch": 3.32129963898917,
"grad_norm": 0.5479450225830078,
"learning_rate": 9.952079025612162e-05,
"loss": 0.077,
"step": 920
},
{
"epoch": 3.357400722021661,
"grad_norm": 0.6115745306015015,
"learning_rate": 9.949768215691022e-05,
"loss": 0.0794,
"step": 930
},
{
"epoch": 3.3935018050541514,
"grad_norm": 0.6622182726860046,
"learning_rate": 9.9474032758977e-05,
"loss": 0.081,
"step": 940
},
{
"epoch": 3.4296028880866425,
"grad_norm": 0.5314200520515442,
"learning_rate": 9.944984232094794e-05,
"loss": 0.0867,
"step": 950
},
{
"epoch": 3.4657039711191335,
"grad_norm": 0.7165815830230713,
"learning_rate": 9.942511110736584e-05,
"loss": 0.086,
"step": 960
},
{
"epoch": 3.5018050541516246,
"grad_norm": 0.9266943335533142,
"learning_rate": 9.939983938868726e-05,
"loss": 0.0811,
"step": 970
},
{
"epoch": 3.5379061371841156,
"grad_norm": 0.669084370136261,
"learning_rate": 9.93740274412797e-05,
"loss": 0.0797,
"step": 980
},
{
"epoch": 3.5740072202166067,
"grad_norm": 0.8197455406188965,
"learning_rate": 9.934767554741846e-05,
"loss": 0.0913,
"step": 990
},
{
"epoch": 3.6101083032490973,
"grad_norm": 0.42351463437080383,
"learning_rate": 9.932078399528361e-05,
"loss": 0.0812,
"step": 1000
},
{
"epoch": 3.6462093862815883,
"grad_norm": 0.5610443353652954,
"learning_rate": 9.929335307895689e-05,
"loss": 0.0775,
"step": 1010
},
{
"epoch": 3.6823104693140793,
"grad_norm": 0.6975977420806885,
"learning_rate": 9.926538309841839e-05,
"loss": 0.0826,
"step": 1020
},
{
"epoch": 3.7184115523465704,
"grad_norm": 0.6497360467910767,
"learning_rate": 9.923687435954334e-05,
"loss": 0.0769,
"step": 1030
},
{
"epoch": 3.7545126353790614,
"grad_norm": 0.5392345786094666,
"learning_rate": 9.920782717409873e-05,
"loss": 0.0814,
"step": 1040
},
{
"epoch": 3.7906137184115525,
"grad_norm": 0.4396078586578369,
"learning_rate": 9.917824185973994e-05,
"loss": 0.0716,
"step": 1050
},
{
"epoch": 3.8267148014440435,
"grad_norm": 0.7224811911582947,
"learning_rate": 9.914811874000723e-05,
"loss": 0.0863,
"step": 1060
},
{
"epoch": 3.862815884476534,
"grad_norm": 0.3890010118484497,
"learning_rate": 9.911745814432218e-05,
"loss": 0.0862,
"step": 1070
},
{
"epoch": 3.898916967509025,
"grad_norm": 0.5354512929916382,
"learning_rate": 9.90862604079842e-05,
"loss": 0.0785,
"step": 1080
},
{
"epoch": 3.935018050541516,
"grad_norm": 0.8041565418243408,
"learning_rate": 9.90545258721667e-05,
"loss": 0.0666,
"step": 1090
},
{
"epoch": 3.9711191335740073,
"grad_norm": 0.6462880969047546,
"learning_rate": 9.90222548839135e-05,
"loss": 0.0742,
"step": 1100
},
{
"epoch": 4.007220216606498,
"grad_norm": 0.5570405125617981,
"learning_rate": 9.898944779613495e-05,
"loss": 0.0762,
"step": 1110
},
{
"epoch": 4.043321299638989,
"grad_norm": 0.6631398797035217,
"learning_rate": 9.89561049676041e-05,
"loss": 0.0689,
"step": 1120
},
{
"epoch": 4.07942238267148,
"grad_norm": 0.5336405038833618,
"learning_rate": 9.89222267629528e-05,
"loss": 0.0814,
"step": 1130
},
{
"epoch": 4.115523465703971,
"grad_norm": 0.6243626475334167,
"learning_rate": 9.888781355266763e-05,
"loss": 0.0752,
"step": 1140
},
{
"epoch": 4.1516245487364625,
"grad_norm": 0.36177146434783936,
"learning_rate": 9.885286571308598e-05,
"loss": 0.0714,
"step": 1150
},
{
"epoch": 4.187725631768953,
"grad_norm": 0.703779935836792,
"learning_rate": 9.881738362639182e-05,
"loss": 0.0727,
"step": 1160
},
{
"epoch": 4.223826714801444,
"grad_norm": 0.43657398223876953,
"learning_rate": 9.878136768061154e-05,
"loss": 0.0728,
"step": 1170
},
{
"epoch": 4.259927797833935,
"grad_norm": 0.5111600160598755,
"learning_rate": 9.874481826960979e-05,
"loss": 0.0674,
"step": 1180
},
{
"epoch": 4.296028880866426,
"grad_norm": 0.5064379572868347,
"learning_rate": 9.870773579308503e-05,
"loss": 0.075,
"step": 1190
},
{
"epoch": 4.332129963898917,
"grad_norm": 0.5362913608551025,
"learning_rate": 9.867012065656533e-05,
"loss": 0.071,
"step": 1200
},
{
"epoch": 4.368231046931408,
"grad_norm": 0.7076780200004578,
"learning_rate": 9.863197327140376e-05,
"loss": 0.0705,
"step": 1210
},
{
"epoch": 4.404332129963899,
"grad_norm": 0.7501161694526672,
"learning_rate": 9.859329405477403e-05,
"loss": 0.066,
"step": 1220
},
{
"epoch": 4.44043321299639,
"grad_norm": 0.5108727812767029,
"learning_rate": 9.855408342966585e-05,
"loss": 0.0722,
"step": 1230
},
{
"epoch": 4.4765342960288805,
"grad_norm": 0.5271785259246826,
"learning_rate": 9.851434182488033e-05,
"loss": 0.0653,
"step": 1240
},
{
"epoch": 4.512635379061372,
"grad_norm": 0.4142615795135498,
"learning_rate": 9.84740696750253e-05,
"loss": 0.0753,
"step": 1250
},
{
"epoch": 4.548736462093863,
"grad_norm": 0.7539293169975281,
"learning_rate": 9.843326742051055e-05,
"loss": 0.0725,
"step": 1260
},
{
"epoch": 4.584837545126354,
"grad_norm": 0.3517594039440155,
"learning_rate": 9.839193550754297e-05,
"loss": 0.0701,
"step": 1270
},
{
"epoch": 4.620938628158845,
"grad_norm": 0.5749979615211487,
"learning_rate": 9.835007438812177e-05,
"loss": 0.0796,
"step": 1280
},
{
"epoch": 4.657039711191336,
"grad_norm": 0.5653839111328125,
"learning_rate": 9.830768452003341e-05,
"loss": 0.0827,
"step": 1290
},
{
"epoch": 4.693140794223827,
"grad_norm": 0.7646329402923584,
"learning_rate": 9.826476636684671e-05,
"loss": 0.066,
"step": 1300
},
{
"epoch": 4.729241877256317,
"grad_norm": 0.556329607963562,
"learning_rate": 9.822132039790773e-05,
"loss": 0.0716,
"step": 1310
},
{
"epoch": 4.765342960288809,
"grad_norm": 0.522186815738678,
"learning_rate": 9.817734708833461e-05,
"loss": 0.0792,
"step": 1320
},
{
"epoch": 4.8014440433212995,
"grad_norm": 0.45138314366340637,
"learning_rate": 9.813284691901243e-05,
"loss": 0.0677,
"step": 1330
},
{
"epoch": 4.837545126353791,
"grad_norm": 0.5397866368293762,
"learning_rate": 9.808782037658792e-05,
"loss": 0.0712,
"step": 1340
},
{
"epoch": 4.873646209386282,
"grad_norm": 0.6981330513954163,
"learning_rate": 9.804226795346411e-05,
"loss": 0.0702,
"step": 1350
},
{
"epoch": 4.909747292418773,
"grad_norm": 0.5204023122787476,
"learning_rate": 9.799619014779503e-05,
"loss": 0.0725,
"step": 1360
},
{
"epoch": 4.945848375451264,
"grad_norm": 0.29214534163475037,
"learning_rate": 9.794958746348013e-05,
"loss": 0.0635,
"step": 1370
},
{
"epoch": 4.981949458483754,
"grad_norm": 0.6972929835319519,
"learning_rate": 9.790246041015896e-05,
"loss": 0.0744,
"step": 1380
},
{
"epoch": 5.018050541516246,
"grad_norm": 0.6489711403846741,
"learning_rate": 9.785480950320538e-05,
"loss": 0.056,
"step": 1390
},
{
"epoch": 5.054151624548736,
"grad_norm": 0.5290822386741638,
"learning_rate": 9.78066352637221e-05,
"loss": 0.0696,
"step": 1400
},
{
"epoch": 5.090252707581228,
"grad_norm": 0.36434194445610046,
"learning_rate": 9.775793821853488e-05,
"loss": 0.0659,
"step": 1410
},
{
"epoch": 5.126353790613718,
"grad_norm": 0.4081863462924957,
"learning_rate": 9.77087189001868e-05,
"loss": 0.0745,
"step": 1420
},
{
"epoch": 5.162454873646209,
"grad_norm": 0.41883745789527893,
"learning_rate": 9.765897784693243e-05,
"loss": 0.0647,
"step": 1430
},
{
"epoch": 5.1985559566787005,
"grad_norm": 0.3547934889793396,
"learning_rate": 9.760871560273197e-05,
"loss": 0.0666,
"step": 1440
},
{
"epoch": 5.234657039711191,
"grad_norm": 0.49450764060020447,
"learning_rate": 9.755793271724526e-05,
"loss": 0.0624,
"step": 1450
},
{
"epoch": 5.270758122743683,
"grad_norm": 0.5350094437599182,
"learning_rate": 9.750662974582584e-05,
"loss": 0.0655,
"step": 1460
},
{
"epoch": 5.306859205776173,
"grad_norm": 0.508855938911438,
"learning_rate": 9.745480724951473e-05,
"loss": 0.0669,
"step": 1470
},
{
"epoch": 5.342960288808664,
"grad_norm": 0.5416754484176636,
"learning_rate": 9.740246579503447e-05,
"loss": 0.075,
"step": 1480
},
{
"epoch": 5.379061371841155,
"grad_norm": 0.5284398794174194,
"learning_rate": 9.734960595478284e-05,
"loss": 0.065,
"step": 1490
},
{
"epoch": 5.415162454873646,
"grad_norm": 0.6305639147758484,
"learning_rate": 9.729622830682657e-05,
"loss": 0.0798,
"step": 1500
},
{
"epoch": 5.451263537906137,
"grad_norm": 0.3761369585990906,
"learning_rate": 9.724233343489504e-05,
"loss": 0.0683,
"step": 1510
},
{
"epoch": 5.487364620938628,
"grad_norm": 0.513361930847168,
"learning_rate": 9.718792192837396e-05,
"loss": 0.0679,
"step": 1520
},
{
"epoch": 5.5234657039711195,
"grad_norm": 0.5918077230453491,
"learning_rate": 9.713299438229886e-05,
"loss": 0.0663,
"step": 1530
},
{
"epoch": 5.55956678700361,
"grad_norm": 0.408236026763916,
"learning_rate": 9.707755139734855e-05,
"loss": 0.0693,
"step": 1540
},
{
"epoch": 5.595667870036101,
"grad_norm": 0.4290209412574768,
"learning_rate": 9.702159357983866e-05,
"loss": 0.0704,
"step": 1550
},
{
"epoch": 5.631768953068592,
"grad_norm": 0.5133281350135803,
"learning_rate": 9.696512154171492e-05,
"loss": 0.0645,
"step": 1560
},
{
"epoch": 5.667870036101083,
"grad_norm": 0.5005992650985718,
"learning_rate": 9.690813590054645e-05,
"loss": 0.0687,
"step": 1570
},
{
"epoch": 5.703971119133574,
"grad_norm": 0.5137617588043213,
"learning_rate": 9.685063727951914e-05,
"loss": 0.0595,
"step": 1580
},
{
"epoch": 5.740072202166065,
"grad_norm": 0.34332308173179626,
"learning_rate": 9.679262630742865e-05,
"loss": 0.0663,
"step": 1590
},
{
"epoch": 5.776173285198556,
"grad_norm": 0.5117312669754028,
"learning_rate": 9.673410361867373e-05,
"loss": 0.066,
"step": 1600
},
{
"epoch": 5.812274368231047,
"grad_norm": 0.4463912844657898,
"learning_rate": 9.667506985324909e-05,
"loss": 0.0633,
"step": 1610
},
{
"epoch": 5.8483754512635375,
"grad_norm": 0.32954299449920654,
"learning_rate": 9.661552565673855e-05,
"loss": 0.058,
"step": 1620
},
{
"epoch": 5.884476534296029,
"grad_norm": 0.3394154906272888,
"learning_rate": 9.655547168030789e-05,
"loss": 0.0629,
"step": 1630
},
{
"epoch": 5.92057761732852,
"grad_norm": 0.6172398924827576,
"learning_rate": 9.649490858069777e-05,
"loss": 0.0689,
"step": 1640
},
{
"epoch": 5.956678700361011,
"grad_norm": 0.604339599609375,
"learning_rate": 9.643383702021658e-05,
"loss": 0.0659,
"step": 1650
},
{
"epoch": 5.992779783393502,
"grad_norm": 0.4912291467189789,
"learning_rate": 9.637225766673307e-05,
"loss": 0.0559,
"step": 1660
},
{
"epoch": 6.028880866425993,
"grad_norm": 0.5740599036216736,
"learning_rate": 9.631017119366922e-05,
"loss": 0.0622,
"step": 1670
},
{
"epoch": 6.064981949458484,
"grad_norm": 0.627409040927887,
"learning_rate": 9.624757827999273e-05,
"loss": 0.0639,
"step": 1680
},
{
"epoch": 6.101083032490974,
"grad_norm": 0.349636971950531,
"learning_rate": 9.618447961020971e-05,
"loss": 0.0574,
"step": 1690
},
{
"epoch": 6.137184115523466,
"grad_norm": 0.46503910422325134,
"learning_rate": 9.612087587435707e-05,
"loss": 0.0583,
"step": 1700
},
{
"epoch": 6.1732851985559565,
"grad_norm": 0.5489641427993774,
"learning_rate": 9.605676776799508e-05,
"loss": 0.0666,
"step": 1710
},
{
"epoch": 6.209386281588448,
"grad_norm": 0.38025182485580444,
"learning_rate": 9.599215599219973e-05,
"loss": 0.0671,
"step": 1720
},
{
"epoch": 6.245487364620939,
"grad_norm": 0.37651264667510986,
"learning_rate": 9.592704125355505e-05,
"loss": 0.0617,
"step": 1730
},
{
"epoch": 6.28158844765343,
"grad_norm": 0.5005856156349182,
"learning_rate": 9.586142426414538e-05,
"loss": 0.0543,
"step": 1740
},
{
"epoch": 6.317689530685921,
"grad_norm": 0.5020721554756165,
"learning_rate": 9.57953057415476e-05,
"loss": 0.0586,
"step": 1750
},
{
"epoch": 6.353790613718411,
"grad_norm": 0.4689803719520569,
"learning_rate": 9.572868640882328e-05,
"loss": 0.0696,
"step": 1760
},
{
"epoch": 6.389891696750903,
"grad_norm": 0.5423982739448547,
"learning_rate": 9.56615669945108e-05,
"loss": 0.0611,
"step": 1770
},
{
"epoch": 6.425992779783393,
"grad_norm": 0.5350184440612793,
"learning_rate": 9.55939482326173e-05,
"loss": 0.0651,
"step": 1780
},
{
"epoch": 6.462093862815885,
"grad_norm": 0.41926878690719604,
"learning_rate": 9.552583086261069e-05,
"loss": 0.0682,
"step": 1790
},
{
"epoch": 6.498194945848375,
"grad_norm": 0.6264155507087708,
"learning_rate": 9.545721562941168e-05,
"loss": 0.061,
"step": 1800
},
{
"epoch": 6.534296028880867,
"grad_norm": 0.347290962934494,
"learning_rate": 9.538810328338543e-05,
"loss": 0.0645,
"step": 1810
},
{
"epoch": 6.5703971119133575,
"grad_norm": 0.4246738851070404,
"learning_rate": 9.531849458033349e-05,
"loss": 0.0564,
"step": 1820
},
{
"epoch": 6.606498194945848,
"grad_norm": 0.6301547884941101,
"learning_rate": 9.524839028148547e-05,
"loss": 0.0621,
"step": 1830
},
{
"epoch": 6.64259927797834,
"grad_norm": 0.331828773021698,
"learning_rate": 9.517779115349077e-05,
"loss": 0.0582,
"step": 1840
},
{
"epoch": 6.67870036101083,
"grad_norm": 0.535174548625946,
"learning_rate": 9.510669796841014e-05,
"loss": 0.0577,
"step": 1850
},
{
"epoch": 6.714801444043322,
"grad_norm": 0.3865920603275299,
"learning_rate": 9.503511150370727e-05,
"loss": 0.0663,
"step": 1860
},
{
"epoch": 6.750902527075812,
"grad_norm": 0.4259576201438904,
"learning_rate": 9.496303254224024e-05,
"loss": 0.0622,
"step": 1870
},
{
"epoch": 6.787003610108303,
"grad_norm": 0.4546835422515869,
"learning_rate": 9.489046187225306e-05,
"loss": 0.0562,
"step": 1880
},
{
"epoch": 6.823104693140794,
"grad_norm": 0.5510967373847961,
"learning_rate": 9.481740028736692e-05,
"loss": 0.055,
"step": 1890
},
{
"epoch": 6.859205776173285,
"grad_norm": 0.4583161473274231,
"learning_rate": 9.474384858657164e-05,
"loss": 0.0559,
"step": 1900
},
{
"epoch": 6.8953068592057765,
"grad_norm": 0.4724719226360321,
"learning_rate": 9.466980757421679e-05,
"loss": 0.0663,
"step": 1910
},
{
"epoch": 6.931407942238267,
"grad_norm": 0.4043436348438263,
"learning_rate": 9.459527806000305e-05,
"loss": 0.063,
"step": 1920
},
{
"epoch": 6.967509025270758,
"grad_norm": 0.350629985332489,
"learning_rate": 9.452026085897325e-05,
"loss": 0.0623,
"step": 1930
},
{
"epoch": 7.003610108303249,
"grad_norm": 0.5395665764808655,
"learning_rate": 9.444475679150348e-05,
"loss": 0.061,
"step": 1940
},
{
"epoch": 7.03971119133574,
"grad_norm": 0.6149357557296753,
"learning_rate": 9.436876668329411e-05,
"loss": 0.0625,
"step": 1950
},
{
"epoch": 7.075812274368231,
"grad_norm": 0.5474592447280884,
"learning_rate": 9.429229136536079e-05,
"loss": 0.0541,
"step": 1960
},
{
"epoch": 7.111913357400722,
"grad_norm": 0.5828573107719421,
"learning_rate": 9.421533167402534e-05,
"loss": 0.0631,
"step": 1970
},
{
"epoch": 7.148014440433213,
"grad_norm": 0.3550461232662201,
"learning_rate": 9.413788845090666e-05,
"loss": 0.0602,
"step": 1980
},
{
"epoch": 7.184115523465704,
"grad_norm": 0.3457200825214386,
"learning_rate": 9.405996254291136e-05,
"loss": 0.0585,
"step": 1990
},
{
"epoch": 7.2202166064981945,
"grad_norm": 0.4495382606983185,
"learning_rate": 9.398155480222474e-05,
"loss": 0.0461,
"step": 2000
},
{
"epoch": 7.256317689530686,
"grad_norm": 0.49206307530403137,
"learning_rate": 9.390266608630128e-05,
"loss": 0.0526,
"step": 2010
},
{
"epoch": 7.292418772563177,
"grad_norm": 0.37768569588661194,
"learning_rate": 9.38232972578553e-05,
"loss": 0.0567,
"step": 2020
},
{
"epoch": 7.328519855595668,
"grad_norm": 0.28732579946517944,
"learning_rate": 9.374344918485164e-05,
"loss": 0.0624,
"step": 2030
},
{
"epoch": 7.364620938628159,
"grad_norm": 0.3371933698654175,
"learning_rate": 9.366312274049602e-05,
"loss": 0.0631,
"step": 2040
},
{
"epoch": 7.40072202166065,
"grad_norm": 0.473911315202713,
"learning_rate": 9.358231880322554e-05,
"loss": 0.0646,
"step": 2050
},
{
"epoch": 7.436823104693141,
"grad_norm": 0.3393310010433197,
"learning_rate": 9.350103825669916e-05,
"loss": 0.0589,
"step": 2060
},
{
"epoch": 7.472924187725631,
"grad_norm": 0.37600380182266235,
"learning_rate": 9.341928198978787e-05,
"loss": 0.0513,
"step": 2070
},
{
"epoch": 7.509025270758123,
"grad_norm": 0.3177405595779419,
"learning_rate": 9.333705089656512e-05,
"loss": 0.0512,
"step": 2080
},
{
"epoch": 7.5451263537906135,
"grad_norm": 0.49998438358306885,
"learning_rate": 9.325434587629698e-05,
"loss": 0.06,
"step": 2090
},
{
"epoch": 7.581227436823105,
"grad_norm": 0.45119839906692505,
"learning_rate": 9.31711678334323e-05,
"loss": 0.0575,
"step": 2100
},
{
"epoch": 7.617328519855596,
"grad_norm": 0.29995957016944885,
"learning_rate": 9.308751767759282e-05,
"loss": 0.0585,
"step": 2110
},
{
"epoch": 7.653429602888087,
"grad_norm": 0.5159680247306824,
"learning_rate": 9.300339632356325e-05,
"loss": 0.0541,
"step": 2120
},
{
"epoch": 7.689530685920578,
"grad_norm": 0.5026693344116211,
"learning_rate": 9.291880469128124e-05,
"loss": 0.0634,
"step": 2130
},
{
"epoch": 7.725631768953068,
"grad_norm": 0.38263097405433655,
"learning_rate": 9.283374370582732e-05,
"loss": 0.0577,
"step": 2140
},
{
"epoch": 7.76173285198556,
"grad_norm": 0.4568621814250946,
"learning_rate": 9.274821429741482e-05,
"loss": 0.0625,
"step": 2150
},
{
"epoch": 7.79783393501805,
"grad_norm": 0.4336842894554138,
"learning_rate": 9.266221740137961e-05,
"loss": 0.0599,
"step": 2160
},
{
"epoch": 7.833935018050542,
"grad_norm": 0.31705576181411743,
"learning_rate": 9.257575395817001e-05,
"loss": 0.0509,
"step": 2170
},
{
"epoch": 7.870036101083032,
"grad_norm": 0.5223832726478577,
"learning_rate": 9.248882491333637e-05,
"loss": 0.052,
"step": 2180
},
{
"epoch": 7.906137184115524,
"grad_norm": 0.5156125426292419,
"learning_rate": 9.240143121752076e-05,
"loss": 0.0494,
"step": 2190
},
{
"epoch": 7.9422382671480145,
"grad_norm": 0.49427103996276855,
"learning_rate": 9.23135738264467e-05,
"loss": 0.0535,
"step": 2200
},
{
"epoch": 7.978339350180505,
"grad_norm": 0.37729743123054504,
"learning_rate": 9.222525370090849e-05,
"loss": 0.0531,
"step": 2210
},
{
"epoch": 8.014440433212997,
"grad_norm": 0.4160604178905487,
"learning_rate": 9.213647180676088e-05,
"loss": 0.0458,
"step": 2220
},
{
"epoch": 8.050541516245488,
"grad_norm": 0.5239390134811401,
"learning_rate": 9.204722911490846e-05,
"loss": 0.0504,
"step": 2230
},
{
"epoch": 8.086642599277978,
"grad_norm": 0.33880814909935,
"learning_rate": 9.1957526601295e-05,
"loss": 0.0506,
"step": 2240
},
{
"epoch": 8.12274368231047,
"grad_norm": 0.4841059446334839,
"learning_rate": 9.186736524689281e-05,
"loss": 0.0597,
"step": 2250
},
{
"epoch": 8.15884476534296,
"grad_norm": 0.48662877082824707,
"learning_rate": 9.177674603769204e-05,
"loss": 0.05,
"step": 2260
},
{
"epoch": 8.19494584837545,
"grad_norm": 0.3764309585094452,
"learning_rate": 9.168566996468983e-05,
"loss": 0.0505,
"step": 2270
},
{
"epoch": 8.231046931407942,
"grad_norm": 0.30875930190086365,
"learning_rate": 9.159413802387951e-05,
"loss": 0.0515,
"step": 2280
},
{
"epoch": 8.267148014440433,
"grad_norm": 0.5545517802238464,
"learning_rate": 9.150215121623974e-05,
"loss": 0.0596,
"step": 2290
},
{
"epoch": 8.303249097472925,
"grad_norm": 0.47426745295524597,
"learning_rate": 9.140971054772349e-05,
"loss": 0.0488,
"step": 2300
},
{
"epoch": 8.339350180505415,
"grad_norm": 0.37424319982528687,
"learning_rate": 9.131681702924713e-05,
"loss": 0.0629,
"step": 2310
},
{
"epoch": 8.375451263537906,
"grad_norm": 0.3619486391544342,
"learning_rate": 9.122347167667926e-05,
"loss": 0.0504,
"step": 2320
},
{
"epoch": 8.411552346570398,
"grad_norm": 0.41219672560691833,
"learning_rate": 9.112967551082973e-05,
"loss": 0.0611,
"step": 2330
},
{
"epoch": 8.447653429602887,
"grad_norm": 0.5103853344917297,
"learning_rate": 9.103542955743835e-05,
"loss": 0.0548,
"step": 2340
},
{
"epoch": 8.483754512635379,
"grad_norm": 0.5853766798973083,
"learning_rate": 9.094073484716381e-05,
"loss": 0.0575,
"step": 2350
},
{
"epoch": 8.51985559566787,
"grad_norm": 0.3751297891139984,
"learning_rate": 9.084559241557226e-05,
"loss": 0.0539,
"step": 2360
},
{
"epoch": 8.555956678700362,
"grad_norm": 0.6681078672409058,
"learning_rate": 9.075000330312608e-05,
"loss": 0.0579,
"step": 2370
},
{
"epoch": 8.592057761732852,
"grad_norm": 0.44821444153785706,
"learning_rate": 9.065396855517253e-05,
"loss": 0.0475,
"step": 2380
},
{
"epoch": 8.628158844765343,
"grad_norm": 0.42596372961997986,
"learning_rate": 9.055748922193219e-05,
"loss": 0.0551,
"step": 2390
},
{
"epoch": 8.664259927797834,
"grad_norm": 0.5802203416824341,
"learning_rate": 9.046056635848761e-05,
"loss": 0.0591,
"step": 2400
},
{
"epoch": 8.700361010830324,
"grad_norm": 0.27570971846580505,
"learning_rate": 9.036320102477169e-05,
"loss": 0.05,
"step": 2410
},
{
"epoch": 8.736462093862816,
"grad_norm": 0.4175958037376404,
"learning_rate": 9.02653942855561e-05,
"loss": 0.0569,
"step": 2420
},
{
"epoch": 8.772563176895307,
"grad_norm": 0.2719372510910034,
"learning_rate": 9.016714721043971e-05,
"loss": 0.0473,
"step": 2430
},
{
"epoch": 8.808664259927799,
"grad_norm": 0.3605029284954071,
"learning_rate": 9.006846087383675e-05,
"loss": 0.0531,
"step": 2440
},
{
"epoch": 8.844765342960288,
"grad_norm": 0.41065704822540283,
"learning_rate": 8.996933635496523e-05,
"loss": 0.061,
"step": 2450
},
{
"epoch": 8.88086642599278,
"grad_norm": 0.4829605519771576,
"learning_rate": 8.986977473783498e-05,
"loss": 0.0625,
"step": 2460
},
{
"epoch": 8.916967509025271,
"grad_norm": 0.3795032799243927,
"learning_rate": 8.97697771112359e-05,
"loss": 0.0521,
"step": 2470
},
{
"epoch": 8.953068592057761,
"grad_norm": 0.45088741183280945,
"learning_rate": 8.966934456872602e-05,
"loss": 0.0563,
"step": 2480
},
{
"epoch": 8.989169675090253,
"grad_norm": 0.3140650689601898,
"learning_rate": 8.95684782086195e-05,
"loss": 0.0533,
"step": 2490
},
{
"epoch": 9.025270758122744,
"grad_norm": 0.33418217301368713,
"learning_rate": 8.946717913397476e-05,
"loss": 0.0496,
"step": 2500
},
{
"epoch": 9.061371841155236,
"grad_norm": 0.4059543013572693,
"learning_rate": 8.93654484525822e-05,
"loss": 0.0529,
"step": 2510
},
{
"epoch": 9.097472924187725,
"grad_norm": 0.41044583916664124,
"learning_rate": 8.926328727695226e-05,
"loss": 0.0453,
"step": 2520
},
{
"epoch": 9.133574007220217,
"grad_norm": 0.40976062417030334,
"learning_rate": 8.916069672430319e-05,
"loss": 0.0401,
"step": 2530
},
{
"epoch": 9.169675090252708,
"grad_norm": 0.34137043356895447,
"learning_rate": 8.905767791654884e-05,
"loss": 0.0498,
"step": 2540
},
{
"epoch": 9.205776173285198,
"grad_norm": 0.49096620082855225,
"learning_rate": 8.895423198028638e-05,
"loss": 0.0556,
"step": 2550
},
{
"epoch": 9.24187725631769,
"grad_norm": 0.43471184372901917,
"learning_rate": 8.885036004678402e-05,
"loss": 0.0492,
"step": 2560
},
{
"epoch": 9.277978339350181,
"grad_norm": 0.3824317455291748,
"learning_rate": 8.874606325196857e-05,
"loss": 0.052,
"step": 2570
},
{
"epoch": 9.314079422382672,
"grad_norm": 0.6544570326805115,
"learning_rate": 8.864134273641304e-05,
"loss": 0.0515,
"step": 2580
},
{
"epoch": 9.350180505415162,
"grad_norm": 0.481044203042984,
"learning_rate": 8.853619964532427e-05,
"loss": 0.0538,
"step": 2590
},
{
"epoch": 9.386281588447654,
"grad_norm": 0.3220379948616028,
"learning_rate": 8.843063512853019e-05,
"loss": 0.042,
"step": 2600
},
{
"epoch": 9.422382671480145,
"grad_norm": 0.3647150695323944,
"learning_rate": 8.832465034046749e-05,
"loss": 0.0443,
"step": 2610
},
{
"epoch": 9.458483754512635,
"grad_norm": 0.37365350127220154,
"learning_rate": 8.821824644016882e-05,
"loss": 0.0549,
"step": 2620
},
{
"epoch": 9.494584837545126,
"grad_norm": 0.33478137850761414,
"learning_rate": 8.811142459125019e-05,
"loss": 0.0515,
"step": 2630
},
{
"epoch": 9.530685920577618,
"grad_norm": 0.3163363039493561,
"learning_rate": 8.800418596189822e-05,
"loss": 0.0438,
"step": 2640
},
{
"epoch": 9.566787003610107,
"grad_norm": 0.46505793929100037,
"learning_rate": 8.789653172485737e-05,
"loss": 0.0502,
"step": 2650
},
{
"epoch": 9.602888086642599,
"grad_norm": 0.3785809874534607,
"learning_rate": 8.778846305741715e-05,
"loss": 0.0597,
"step": 2660
},
{
"epoch": 9.63898916967509,
"grad_norm": 0.3519617021083832,
"learning_rate": 8.767998114139918e-05,
"loss": 0.0528,
"step": 2670
},
{
"epoch": 9.675090252707582,
"grad_norm": 0.4126377999782562,
"learning_rate": 8.757108716314429e-05,
"loss": 0.0529,
"step": 2680
},
{
"epoch": 9.711191335740072,
"grad_norm": 0.34473440051078796,
"learning_rate": 8.746178231349962e-05,
"loss": 0.0496,
"step": 2690
},
{
"epoch": 9.747292418772563,
"grad_norm": 0.2740600109100342,
"learning_rate": 8.735206778780549e-05,
"loss": 0.0529,
"step": 2700
},
{
"epoch": 9.783393501805055,
"grad_norm": 0.42622506618499756,
"learning_rate": 8.724194478588234e-05,
"loss": 0.0507,
"step": 2710
},
{
"epoch": 9.819494584837544,
"grad_norm": 0.32149437069892883,
"learning_rate": 8.713141451201772e-05,
"loss": 0.0488,
"step": 2720
},
{
"epoch": 9.855595667870036,
"grad_norm": 0.37112313508987427,
"learning_rate": 8.702047817495295e-05,
"loss": 0.0503,
"step": 2730
},
{
"epoch": 9.891696750902527,
"grad_norm": 0.41299912333488464,
"learning_rate": 8.69091369878701e-05,
"loss": 0.0499,
"step": 2740
},
{
"epoch": 9.927797833935019,
"grad_norm": 0.4940550923347473,
"learning_rate": 8.679739216837849e-05,
"loss": 0.0531,
"step": 2750
},
{
"epoch": 9.963898916967509,
"grad_norm": 0.44831499457359314,
"learning_rate": 8.66852449385016e-05,
"loss": 0.0483,
"step": 2760
},
{
"epoch": 10.0,
"grad_norm": 0.365975022315979,
"learning_rate": 8.657269652466356e-05,
"loss": 0.047,
"step": 2770
},
{
"epoch": 10.036101083032491,
"grad_norm": 0.285571813583374,
"learning_rate": 8.645974815767577e-05,
"loss": 0.0485,
"step": 2780
},
{
"epoch": 10.072202166064981,
"grad_norm": 0.4423803985118866,
"learning_rate": 8.634640107272351e-05,
"loss": 0.0443,
"step": 2790
},
{
"epoch": 10.108303249097473,
"grad_norm": 0.4387251138687134,
"learning_rate": 8.623265650935234e-05,
"loss": 0.0466,
"step": 2800
},
{
"epoch": 10.144404332129964,
"grad_norm": 0.30347326397895813,
"learning_rate": 8.611851571145456e-05,
"loss": 0.0504,
"step": 2810
},
{
"epoch": 10.180505415162456,
"grad_norm": 0.30560502409935,
"learning_rate": 8.600397992725566e-05,
"loss": 0.0485,
"step": 2820
},
{
"epoch": 10.216606498194945,
"grad_norm": 0.5056556463241577,
"learning_rate": 8.588905040930061e-05,
"loss": 0.0523,
"step": 2830
},
{
"epoch": 10.252707581227437,
"grad_norm": 0.3621695339679718,
"learning_rate": 8.577372841444022e-05,
"loss": 0.0527,
"step": 2840
},
{
"epoch": 10.288808664259928,
"grad_norm": 0.3315909504890442,
"learning_rate": 8.565801520381736e-05,
"loss": 0.0459,
"step": 2850
},
{
"epoch": 10.324909747292418,
"grad_norm": 0.2972622215747833,
"learning_rate": 8.554191204285313e-05,
"loss": 0.0499,
"step": 2860
},
{
"epoch": 10.36101083032491,
"grad_norm": 0.2993919551372528,
"learning_rate": 8.542542020123315e-05,
"loss": 0.0472,
"step": 2870
},
{
"epoch": 10.397111913357401,
"grad_norm": 0.447722852230072,
"learning_rate": 8.530854095289347e-05,
"loss": 0.0448,
"step": 2880
},
{
"epoch": 10.433212996389893,
"grad_norm": 0.5251609086990356,
"learning_rate": 8.519127557600688e-05,
"loss": 0.0434,
"step": 2890
},
{
"epoch": 10.469314079422382,
"grad_norm": 0.3675687909126282,
"learning_rate": 8.507362535296871e-05,
"loss": 0.0464,
"step": 2900
},
{
"epoch": 10.505415162454874,
"grad_norm": 0.3707270324230194,
"learning_rate": 8.495559157038299e-05,
"loss": 0.0613,
"step": 2910
},
{
"epoch": 10.541516245487365,
"grad_norm": 0.5130832195281982,
"learning_rate": 8.483717551904823e-05,
"loss": 0.0503,
"step": 2920
},
{
"epoch": 10.577617328519855,
"grad_norm": 0.35453906655311584,
"learning_rate": 8.47183784939434e-05,
"loss": 0.0469,
"step": 2930
},
{
"epoch": 10.613718411552346,
"grad_norm": 0.3935864269733429,
"learning_rate": 8.459920179421374e-05,
"loss": 0.0532,
"step": 2940
},
{
"epoch": 10.649819494584838,
"grad_norm": 0.3294016718864441,
"learning_rate": 8.447964672315656e-05,
"loss": 0.0484,
"step": 2950
},
{
"epoch": 10.685920577617328,
"grad_norm": 0.5071790814399719,
"learning_rate": 8.435971458820692e-05,
"loss": 0.0555,
"step": 2960
},
{
"epoch": 10.722021660649819,
"grad_norm": 0.30743515491485596,
"learning_rate": 8.423940670092345e-05,
"loss": 0.0462,
"step": 2970
},
{
"epoch": 10.75812274368231,
"grad_norm": 0.3643742799758911,
"learning_rate": 8.411872437697394e-05,
"loss": 0.0473,
"step": 2980
},
{
"epoch": 10.794223826714802,
"grad_norm": 0.47820091247558594,
"learning_rate": 8.399766893612096e-05,
"loss": 0.0451,
"step": 2990
},
{
"epoch": 10.830324909747292,
"grad_norm": 0.492313951253891,
"learning_rate": 8.38762417022074e-05,
"loss": 0.055,
"step": 3000
},
{
"epoch": 10.866425992779783,
"grad_norm": 0.2783392667770386,
"learning_rate": 8.375444400314204e-05,
"loss": 0.0482,
"step": 3010
},
{
"epoch": 10.902527075812275,
"grad_norm": 0.20478548109531403,
"learning_rate": 8.3632277170885e-05,
"loss": 0.0559,
"step": 3020
},
{
"epoch": 10.938628158844764,
"grad_norm": 0.470162957906723,
"learning_rate": 8.350974254143318e-05,
"loss": 0.0484,
"step": 3030
},
{
"epoch": 10.974729241877256,
"grad_norm": 0.3410859704017639,
"learning_rate": 8.338684145480566e-05,
"loss": 0.0474,
"step": 3040
},
{
"epoch": 11.010830324909747,
"grad_norm": 0.32766473293304443,
"learning_rate": 8.326357525502904e-05,
"loss": 0.052,
"step": 3050
},
{
"epoch": 11.046931407942239,
"grad_norm": 0.21446117758750916,
"learning_rate": 8.313994529012273e-05,
"loss": 0.0413,
"step": 3060
},
{
"epoch": 11.083032490974729,
"grad_norm": 0.6218424439430237,
"learning_rate": 8.301595291208422e-05,
"loss": 0.0526,
"step": 3070
},
{
"epoch": 11.11913357400722,
"grad_norm": 0.38401177525520325,
"learning_rate": 8.289159947687427e-05,
"loss": 0.0486,
"step": 3080
},
{
"epoch": 11.155234657039712,
"grad_norm": 0.41590479016304016,
"learning_rate": 8.276688634440216e-05,
"loss": 0.0497,
"step": 3090
},
{
"epoch": 11.191335740072201,
"grad_norm": 0.5154473185539246,
"learning_rate": 8.26418148785107e-05,
"loss": 0.0468,
"step": 3100
},
{
"epoch": 11.227436823104693,
"grad_norm": 0.3415977358818054,
"learning_rate": 8.251638644696141e-05,
"loss": 0.0508,
"step": 3110
},
{
"epoch": 11.263537906137184,
"grad_norm": 0.49194249510765076,
"learning_rate": 8.23906024214195e-05,
"loss": 0.0467,
"step": 3120
},
{
"epoch": 11.299638989169676,
"grad_norm": 0.6073770523071289,
"learning_rate": 8.226446417743897e-05,
"loss": 0.048,
"step": 3130
},
{
"epoch": 11.335740072202166,
"grad_norm": 0.37768247723579407,
"learning_rate": 8.213797309444742e-05,
"loss": 0.0509,
"step": 3140
},
{
"epoch": 11.371841155234657,
"grad_norm": 0.3825902044773102,
"learning_rate": 8.201113055573105e-05,
"loss": 0.0481,
"step": 3150
},
{
"epoch": 11.407942238267148,
"grad_norm": 0.45687535405158997,
"learning_rate": 8.188393794841958e-05,
"loss": 0.0407,
"step": 3160
},
{
"epoch": 11.444043321299638,
"grad_norm": 0.32624551653862,
"learning_rate": 8.175639666347094e-05,
"loss": 0.0472,
"step": 3170
},
{
"epoch": 11.48014440433213,
"grad_norm": 0.32584348320961,
"learning_rate": 8.162850809565623e-05,
"loss": 0.043,
"step": 3180
},
{
"epoch": 11.516245487364621,
"grad_norm": 0.4904119670391083,
"learning_rate": 8.150027364354431e-05,
"loss": 0.0477,
"step": 3190
},
{
"epoch": 11.552346570397113,
"grad_norm": 0.40484169125556946,
"learning_rate": 8.137169470948662e-05,
"loss": 0.0491,
"step": 3200
},
{
"epoch": 11.588447653429602,
"grad_norm": 0.296037882566452,
"learning_rate": 8.124277269960179e-05,
"loss": 0.0487,
"step": 3210
},
{
"epoch": 11.624548736462094,
"grad_norm": 0.42236408591270447,
"learning_rate": 8.111350902376023e-05,
"loss": 0.0452,
"step": 3220
},
{
"epoch": 11.660649819494585,
"grad_norm": 0.5262724757194519,
"learning_rate": 8.098390509556883e-05,
"loss": 0.0481,
"step": 3230
},
{
"epoch": 11.696750902527075,
"grad_norm": 0.32716065645217896,
"learning_rate": 8.085396233235536e-05,
"loss": 0.0407,
"step": 3240
},
{
"epoch": 11.732851985559567,
"grad_norm": 0.4551047384738922,
"learning_rate": 8.072368215515306e-05,
"loss": 0.0401,
"step": 3250
},
{
"epoch": 11.768953068592058,
"grad_norm": 0.2930159866809845,
"learning_rate": 8.059306598868506e-05,
"loss": 0.0549,
"step": 3260
},
{
"epoch": 11.80505415162455,
"grad_norm": 0.454457551240921,
"learning_rate": 8.046211526134888e-05,
"loss": 0.0481,
"step": 3270
},
{
"epoch": 11.84115523465704,
"grad_norm": 0.49725404381752014,
"learning_rate": 8.033083140520065e-05,
"loss": 0.0457,
"step": 3280
},
{
"epoch": 11.87725631768953,
"grad_norm": 0.35449522733688354,
"learning_rate": 8.019921585593962e-05,
"loss": 0.0519,
"step": 3290
},
{
"epoch": 11.913357400722022,
"grad_norm": 0.42122408747673035,
"learning_rate": 8.006727005289232e-05,
"loss": 0.042,
"step": 3300
},
{
"epoch": 11.949458483754512,
"grad_norm": 0.4187127649784088,
"learning_rate": 7.993499543899692e-05,
"loss": 0.0496,
"step": 3310
},
{
"epoch": 11.985559566787003,
"grad_norm": 0.3574303090572357,
"learning_rate": 7.980239346078742e-05,
"loss": 0.0407,
"step": 3320
},
{
"epoch": 12.021660649819495,
"grad_norm": 0.34231457114219666,
"learning_rate": 7.966946556837778e-05,
"loss": 0.0417,
"step": 3330
},
{
"epoch": 12.057761732851986,
"grad_norm": 0.5687417984008789,
"learning_rate": 7.953621321544616e-05,
"loss": 0.0485,
"step": 3340
},
{
"epoch": 12.093862815884476,
"grad_norm": 0.5793416500091553,
"learning_rate": 7.940263785921896e-05,
"loss": 0.0513,
"step": 3350
},
{
"epoch": 12.129963898916968,
"grad_norm": 0.3382067382335663,
"learning_rate": 7.926874096045482e-05,
"loss": 0.0468,
"step": 3360
},
{
"epoch": 12.166064981949459,
"grad_norm": 0.3719868063926697,
"learning_rate": 7.913452398342881e-05,
"loss": 0.0554,
"step": 3370
},
{
"epoch": 12.202166064981949,
"grad_norm": 0.3792908489704132,
"learning_rate": 7.89999883959163e-05,
"loss": 0.0548,
"step": 3380
},
{
"epoch": 12.23826714801444,
"grad_norm": 0.2616323232650757,
"learning_rate": 7.886513566917687e-05,
"loss": 0.0451,
"step": 3390
},
{
"epoch": 12.274368231046932,
"grad_norm": 0.3615397810935974,
"learning_rate": 7.872996727793838e-05,
"loss": 0.0494,
"step": 3400
},
{
"epoch": 12.310469314079423,
"grad_norm": 0.36225149035453796,
"learning_rate": 7.859448470038069e-05,
"loss": 0.0466,
"step": 3410
},
{
"epoch": 12.346570397111913,
"grad_norm": 0.2508832514286041,
"learning_rate": 7.845868941811956e-05,
"loss": 0.0481,
"step": 3420
},
{
"epoch": 12.382671480144404,
"grad_norm": 0.34768667817115784,
"learning_rate": 7.832258291619043e-05,
"loss": 0.0404,
"step": 3430
},
{
"epoch": 12.418772563176896,
"grad_norm": 0.27604231238365173,
"learning_rate": 7.81861666830322e-05,
"loss": 0.0385,
"step": 3440
},
{
"epoch": 12.454873646209386,
"grad_norm": 0.2694339454174042,
"learning_rate": 7.804944221047097e-05,
"loss": 0.0444,
"step": 3450
},
{
"epoch": 12.490974729241877,
"grad_norm": 0.3244231343269348,
"learning_rate": 7.791241099370364e-05,
"loss": 0.0498,
"step": 3460
},
{
"epoch": 12.527075812274369,
"grad_norm": 0.32860997319221497,
"learning_rate": 7.777507453128163e-05,
"loss": 0.0453,
"step": 3470
},
{
"epoch": 12.56317689530686,
"grad_norm": 0.37944501638412476,
"learning_rate": 7.763743432509451e-05,
"loss": 0.0386,
"step": 3480
},
{
"epoch": 12.59927797833935,
"grad_norm": 0.2797277271747589,
"learning_rate": 7.749949188035353e-05,
"loss": 0.0391,
"step": 3490
},
{
"epoch": 12.635379061371841,
"grad_norm": 0.3597061336040497,
"learning_rate": 7.736124870557516e-05,
"loss": 0.0494,
"step": 3500
},
{
"epoch": 12.671480144404333,
"grad_norm": 0.3266807794570923,
"learning_rate": 7.722270631256459e-05,
"loss": 0.0416,
"step": 3510
},
{
"epoch": 12.707581227436823,
"grad_norm": 0.36057159304618835,
"learning_rate": 7.708386621639925e-05,
"loss": 0.0434,
"step": 3520
},
{
"epoch": 12.743682310469314,
"grad_norm": 0.4390406012535095,
"learning_rate": 7.694472993541219e-05,
"loss": 0.045,
"step": 3530
},
{
"epoch": 12.779783393501805,
"grad_norm": 0.32979968190193176,
"learning_rate": 7.680529899117547e-05,
"loss": 0.0504,
"step": 3540
},
{
"epoch": 12.815884476534297,
"grad_norm": 0.4777357280254364,
"learning_rate": 7.666557490848358e-05,
"loss": 0.0474,
"step": 3550
},
{
"epoch": 12.851985559566787,
"grad_norm": 0.27603527903556824,
"learning_rate": 7.65255592153367e-05,
"loss": 0.0463,
"step": 3560
},
{
"epoch": 12.888086642599278,
"grad_norm": 0.4240185618400574,
"learning_rate": 7.638525344292402e-05,
"loss": 0.0479,
"step": 3570
},
{
"epoch": 12.92418772563177,
"grad_norm": 0.35840970277786255,
"learning_rate": 7.624465912560697e-05,
"loss": 0.0452,
"step": 3580
},
{
"epoch": 12.96028880866426,
"grad_norm": 0.25613051652908325,
"learning_rate": 7.610377780090249e-05,
"loss": 0.0413,
"step": 3590
},
{
"epoch": 12.99638989169675,
"grad_norm": 0.24853578209877014,
"learning_rate": 7.596261100946618e-05,
"loss": 0.0495,
"step": 3600
},
{
"epoch": 13.032490974729242,
"grad_norm": 0.4584263861179352,
"learning_rate": 7.582116029507542e-05,
"loss": 0.0397,
"step": 3610
},
{
"epoch": 13.068592057761732,
"grad_norm": 0.34161871671676636,
"learning_rate": 7.56794272046126e-05,
"loss": 0.0469,
"step": 3620
},
{
"epoch": 13.104693140794224,
"grad_norm": 0.39896032214164734,
"learning_rate": 7.55374132880481e-05,
"loss": 0.0387,
"step": 3630
},
{
"epoch": 13.140794223826715,
"grad_norm": 0.3332551121711731,
"learning_rate": 7.539512009842333e-05,
"loss": 0.044,
"step": 3640
},
{
"epoch": 13.176895306859207,
"grad_norm": 0.3252636194229126,
"learning_rate": 7.525254919183382e-05,
"loss": 0.0451,
"step": 3650
},
{
"epoch": 13.212996389891696,
"grad_norm": 0.2662648856639862,
"learning_rate": 7.510970212741215e-05,
"loss": 0.0408,
"step": 3660
},
{
"epoch": 13.249097472924188,
"grad_norm": 0.37176191806793213,
"learning_rate": 7.496658046731096e-05,
"loss": 0.037,
"step": 3670
},
{
"epoch": 13.28519855595668,
"grad_norm": 0.3949223756790161,
"learning_rate": 7.482318577668578e-05,
"loss": 0.0465,
"step": 3680
},
{
"epoch": 13.321299638989169,
"grad_norm": 0.2646092176437378,
"learning_rate": 7.467951962367796e-05,
"loss": 0.0434,
"step": 3690
},
{
"epoch": 13.35740072202166,
"grad_norm": 0.3295803368091583,
"learning_rate": 7.453558357939755e-05,
"loss": 0.0393,
"step": 3700
},
{
"epoch": 13.393501805054152,
"grad_norm": 0.39427492022514343,
"learning_rate": 7.439137921790606e-05,
"loss": 0.0497,
"step": 3710
},
{
"epoch": 13.429602888086643,
"grad_norm": 0.35017260909080505,
"learning_rate": 7.42469081161993e-05,
"loss": 0.0395,
"step": 3720
},
{
"epoch": 13.465703971119133,
"grad_norm": 0.30179521441459656,
"learning_rate": 7.410217185419006e-05,
"loss": 0.0398,
"step": 3730
},
{
"epoch": 13.501805054151625,
"grad_norm": 0.41527479887008667,
"learning_rate": 7.395717201469095e-05,
"loss": 0.0423,
"step": 3740
},
{
"epoch": 13.537906137184116,
"grad_norm": 0.3234764039516449,
"learning_rate": 7.381191018339696e-05,
"loss": 0.0377,
"step": 3750
},
{
"epoch": 13.574007220216606,
"grad_norm": 0.3420933783054352,
"learning_rate": 7.36663879488682e-05,
"loss": 0.0424,
"step": 3760
},
{
"epoch": 13.610108303249097,
"grad_norm": 0.3988379240036011,
"learning_rate": 7.352060690251254e-05,
"loss": 0.0483,
"step": 3770
},
{
"epoch": 13.646209386281589,
"grad_norm": 0.28290730714797974,
"learning_rate": 7.337456863856811e-05,
"loss": 0.0447,
"step": 3780
},
{
"epoch": 13.68231046931408,
"grad_norm": 0.37324759364128113,
"learning_rate": 7.3228274754086e-05,
"loss": 0.0469,
"step": 3790
},
{
"epoch": 13.71841155234657,
"grad_norm": 0.2852276563644409,
"learning_rate": 7.308172684891267e-05,
"loss": 0.0441,
"step": 3800
},
{
"epoch": 13.754512635379061,
"grad_norm": 0.18142566084861755,
"learning_rate": 7.293492652567255e-05,
"loss": 0.0388,
"step": 3810
},
{
"epoch": 13.790613718411553,
"grad_norm": 0.3799176514148712,
"learning_rate": 7.278787538975043e-05,
"loss": 0.0499,
"step": 3820
},
{
"epoch": 13.826714801444043,
"grad_norm": 0.3972136974334717,
"learning_rate": 7.2640575049274e-05,
"loss": 0.0454,
"step": 3830
},
{
"epoch": 13.862815884476534,
"grad_norm": 0.2528814673423767,
"learning_rate": 7.249302711509616e-05,
"loss": 0.0391,
"step": 3840
},
{
"epoch": 13.898916967509026,
"grad_norm": 0.3603387475013733,
"learning_rate": 7.23452332007775e-05,
"loss": 0.0408,
"step": 3850
},
{
"epoch": 13.935018050541515,
"grad_norm": 0.29239344596862793,
"learning_rate": 7.219719492256858e-05,
"loss": 0.0452,
"step": 3860
},
{
"epoch": 13.971119133574007,
"grad_norm": 0.3348754942417145,
"learning_rate": 7.20489138993923e-05,
"loss": 0.038,
"step": 3870
},
{
"epoch": 14.007220216606498,
"grad_norm": 0.2634284496307373,
"learning_rate": 7.190039175282614e-05,
"loss": 0.048,
"step": 3880
},
{
"epoch": 14.04332129963899,
"grad_norm": 0.4274432063102722,
"learning_rate": 7.175163010708455e-05,
"loss": 0.0337,
"step": 3890
},
{
"epoch": 14.07942238267148,
"grad_norm": 0.33779147267341614,
"learning_rate": 7.1602630589001e-05,
"loss": 0.0419,
"step": 3900
},
{
"epoch": 14.115523465703971,
"grad_norm": 0.302521288394928,
"learning_rate": 7.14533948280104e-05,
"loss": 0.0381,
"step": 3910
},
{
"epoch": 14.151624548736462,
"grad_norm": 0.33974364399909973,
"learning_rate": 7.130392445613109e-05,
"loss": 0.0391,
"step": 3920
},
{
"epoch": 14.187725631768952,
"grad_norm": 0.41150495409965515,
"learning_rate": 7.115422110794711e-05,
"loss": 0.0462,
"step": 3930
},
{
"epoch": 14.223826714801444,
"grad_norm": 0.2341962456703186,
"learning_rate": 7.100428642059033e-05,
"loss": 0.0442,
"step": 3940
},
{
"epoch": 14.259927797833935,
"grad_norm": 0.3411276042461395,
"learning_rate": 7.08541220337224e-05,
"loss": 0.0475,
"step": 3950
},
{
"epoch": 14.296028880866427,
"grad_norm": 0.3072654902935028,
"learning_rate": 7.070372958951706e-05,
"loss": 0.035,
"step": 3960
},
{
"epoch": 14.332129963898916,
"grad_norm": 0.4435533881187439,
"learning_rate": 7.055311073264194e-05,
"loss": 0.0392,
"step": 3970
},
{
"epoch": 14.368231046931408,
"grad_norm": 0.360043466091156,
"learning_rate": 7.040226711024077e-05,
"loss": 0.0364,
"step": 3980
},
{
"epoch": 14.4043321299639,
"grad_norm": 0.2570934295654297,
"learning_rate": 7.02512003719152e-05,
"loss": 0.0466,
"step": 3990
},
{
"epoch": 14.440433212996389,
"grad_norm": 0.47872069478034973,
"learning_rate": 7.00999121697069e-05,
"loss": 0.0449,
"step": 4000
},
{
"epoch": 14.47653429602888,
"grad_norm": 0.37337303161621094,
"learning_rate": 6.99484041580794e-05,
"loss": 0.0402,
"step": 4010
},
{
"epoch": 14.512635379061372,
"grad_norm": 0.4141935706138611,
"learning_rate": 6.979667799390004e-05,
"loss": 0.0413,
"step": 4020
},
{
"epoch": 14.548736462093864,
"grad_norm": 0.40998849272727966,
"learning_rate": 6.964473533642185e-05,
"loss": 0.0419,
"step": 4030
},
{
"epoch": 14.584837545126353,
"grad_norm": 0.43466073274612427,
"learning_rate": 6.949257784726539e-05,
"loss": 0.0441,
"step": 4040
},
{
"epoch": 14.620938628158845,
"grad_norm": 0.3791906237602234,
"learning_rate": 6.934020719040056e-05,
"loss": 0.0409,
"step": 4050
},
{
"epoch": 14.657039711191336,
"grad_norm": 0.47524070739746094,
"learning_rate": 6.918762503212848e-05,
"loss": 0.0459,
"step": 4060
},
{
"epoch": 14.693140794223826,
"grad_norm": 0.3602631688117981,
"learning_rate": 6.903483304106319e-05,
"loss": 0.0471,
"step": 4070
},
{
"epoch": 14.729241877256317,
"grad_norm": 0.27463483810424805,
"learning_rate": 6.888183288811341e-05,
"loss": 0.0368,
"step": 4080
},
{
"epoch": 14.765342960288809,
"grad_norm": 0.30613768100738525,
"learning_rate": 6.87286262464643e-05,
"loss": 0.044,
"step": 4090
},
{
"epoch": 14.8014440433213,
"grad_norm": 0.3187946379184723,
"learning_rate": 6.857521479155915e-05,
"loss": 0.0453,
"step": 4100
},
{
"epoch": 14.83754512635379,
"grad_norm": 0.31422433257102966,
"learning_rate": 6.842160020108104e-05,
"loss": 0.0366,
"step": 4110
},
{
"epoch": 14.873646209386282,
"grad_norm": 0.24223259091377258,
"learning_rate": 6.826778415493455e-05,
"loss": 0.0372,
"step": 4120
},
{
"epoch": 14.909747292418773,
"grad_norm": 0.4518241286277771,
"learning_rate": 6.811376833522729e-05,
"loss": 0.0475,
"step": 4130
},
{
"epoch": 14.945848375451263,
"grad_norm": 0.31902456283569336,
"learning_rate": 6.795955442625159e-05,
"loss": 0.0446,
"step": 4140
},
{
"epoch": 14.981949458483754,
"grad_norm": 0.36801326274871826,
"learning_rate": 6.780514411446608e-05,
"loss": 0.0394,
"step": 4150
},
{
"epoch": 15.018050541516246,
"grad_norm": 0.44974032044410706,
"learning_rate": 6.765053908847716e-05,
"loss": 0.0354,
"step": 4160
},
{
"epoch": 15.054151624548737,
"grad_norm": 0.16332118213176727,
"learning_rate": 6.749574103902064e-05,
"loss": 0.0408,
"step": 4170
},
{
"epoch": 15.090252707581227,
"grad_norm": 0.3500916659832001,
"learning_rate": 6.734075165894317e-05,
"loss": 0.0381,
"step": 4180
},
{
"epoch": 15.126353790613718,
"grad_norm": 0.3964041769504547,
"learning_rate": 6.71855726431838e-05,
"loss": 0.0475,
"step": 4190
},
{
"epoch": 15.16245487364621,
"grad_norm": 0.35957953333854675,
"learning_rate": 6.703020568875538e-05,
"loss": 0.0362,
"step": 4200
},
{
"epoch": 15.1985559566787,
"grad_norm": 0.2818802297115326,
"learning_rate": 6.687465249472603e-05,
"loss": 0.0376,
"step": 4210
},
{
"epoch": 15.234657039711191,
"grad_norm": 0.2862112820148468,
"learning_rate": 6.671891476220055e-05,
"loss": 0.0347,
"step": 4220
},
{
"epoch": 15.270758122743683,
"grad_norm": 0.21799275279045105,
"learning_rate": 6.656299419430183e-05,
"loss": 0.034,
"step": 4230
},
{
"epoch": 15.306859205776174,
"grad_norm": 0.34679868817329407,
"learning_rate": 6.640689249615223e-05,
"loss": 0.0396,
"step": 4240
},
{
"epoch": 15.342960288808664,
"grad_norm": 0.3367157280445099,
"learning_rate": 6.625061137485491e-05,
"loss": 0.0392,
"step": 4250
},
{
"epoch": 15.379061371841155,
"grad_norm": 0.31001347303390503,
"learning_rate": 6.609415253947517e-05,
"loss": 0.0446,
"step": 4260
},
{
"epoch": 15.415162454873647,
"grad_norm": 0.24991685152053833,
"learning_rate": 6.593751770102178e-05,
"loss": 0.0356,
"step": 4270
},
{
"epoch": 15.451263537906136,
"grad_norm": 0.46583548188209534,
"learning_rate": 6.578070857242823e-05,
"loss": 0.0399,
"step": 4280
},
{
"epoch": 15.487364620938628,
"grad_norm": 0.2639727294445038,
"learning_rate": 6.562372686853402e-05,
"loss": 0.0381,
"step": 4290
},
{
"epoch": 15.52346570397112,
"grad_norm": 0.24404628574848175,
"learning_rate": 6.546657430606593e-05,
"loss": 0.0404,
"step": 4300
},
{
"epoch": 15.559566787003611,
"grad_norm": 0.3236408233642578,
"learning_rate": 6.530925260361918e-05,
"loss": 0.0358,
"step": 4310
},
{
"epoch": 15.5956678700361,
"grad_norm": 0.2713198661804199,
"learning_rate": 6.515176348163871e-05,
"loss": 0.0397,
"step": 4320
},
{
"epoch": 15.631768953068592,
"grad_norm": 0.22722287476062775,
"learning_rate": 6.499410866240032e-05,
"loss": 0.0388,
"step": 4330
},
{
"epoch": 15.667870036101084,
"grad_norm": 0.4177260994911194,
"learning_rate": 6.48362898699919e-05,
"loss": 0.0396,
"step": 4340
},
{
"epoch": 15.703971119133573,
"grad_norm": 0.30247387290000916,
"learning_rate": 6.467830883029443e-05,
"loss": 0.0352,
"step": 4350
},
{
"epoch": 15.740072202166065,
"grad_norm": 0.2654128968715668,
"learning_rate": 6.452016727096326e-05,
"loss": 0.034,
"step": 4360
},
{
"epoch": 15.776173285198556,
"grad_norm": 0.35598665475845337,
"learning_rate": 6.436186692140916e-05,
"loss": 0.04,
"step": 4370
},
{
"epoch": 15.812274368231048,
"grad_norm": 0.34090209007263184,
"learning_rate": 6.420340951277938e-05,
"loss": 0.0378,
"step": 4380
},
{
"epoch": 15.848375451263538,
"grad_norm": 0.40665608644485474,
"learning_rate": 6.404479677793874e-05,
"loss": 0.0414,
"step": 4390
},
{
"epoch": 15.884476534296029,
"grad_norm": 0.2219485342502594,
"learning_rate": 6.388603045145075e-05,
"loss": 0.0441,
"step": 4400
},
{
"epoch": 15.92057761732852,
"grad_norm": 0.3980633020401001,
"learning_rate": 6.372711226955843e-05,
"loss": 0.0415,
"step": 4410
},
{
"epoch": 15.95667870036101,
"grad_norm": 0.2842003405094147,
"learning_rate": 6.356804397016564e-05,
"loss": 0.0398,
"step": 4420
},
{
"epoch": 15.992779783393502,
"grad_norm": 0.22785481810569763,
"learning_rate": 6.340882729281779e-05,
"loss": 0.0355,
"step": 4430
},
{
"epoch": 16.028880866425993,
"grad_norm": 0.2921331524848938,
"learning_rate": 6.324946397868294e-05,
"loss": 0.0398,
"step": 4440
},
{
"epoch": 16.064981949458485,
"grad_norm": 0.3846503794193268,
"learning_rate": 6.308995577053276e-05,
"loss": 0.0377,
"step": 4450
},
{
"epoch": 16.101083032490976,
"grad_norm": 0.3269325792789459,
"learning_rate": 6.293030441272347e-05,
"loss": 0.036,
"step": 4460
},
{
"epoch": 16.137184115523464,
"grad_norm": 0.2513223886489868,
"learning_rate": 6.277051165117677e-05,
"loss": 0.0365,
"step": 4470
},
{
"epoch": 16.173285198555956,
"grad_norm": 0.32939809560775757,
"learning_rate": 6.261057923336064e-05,
"loss": 0.0394,
"step": 4480
},
{
"epoch": 16.209386281588447,
"grad_norm": 0.3236745297908783,
"learning_rate": 6.245050890827042e-05,
"loss": 0.0461,
"step": 4490
},
{
"epoch": 16.24548736462094,
"grad_norm": 0.29283466935157776,
"learning_rate": 6.229030242640952e-05,
"loss": 0.0431,
"step": 4500
},
{
"epoch": 16.28158844765343,
"grad_norm": 0.27510249614715576,
"learning_rate": 6.212996153977037e-05,
"loss": 0.0409,
"step": 4510
},
{
"epoch": 16.31768953068592,
"grad_norm": 0.2821718752384186,
"learning_rate": 6.196948800181523e-05,
"loss": 0.039,
"step": 4520
},
{
"epoch": 16.353790613718413,
"grad_norm": 0.34371283650398254,
"learning_rate": 6.180888356745695e-05,
"loss": 0.04,
"step": 4530
},
{
"epoch": 16.3898916967509,
"grad_norm": 0.4911617338657379,
"learning_rate": 6.164814999303995e-05,
"loss": 0.0414,
"step": 4540
},
{
"epoch": 16.425992779783392,
"grad_norm": 0.4056515395641327,
"learning_rate": 6.148728903632081e-05,
"loss": 0.0422,
"step": 4550
},
{
"epoch": 16.462093862815884,
"grad_norm": 0.2657906115055084,
"learning_rate": 6.132630245644921e-05,
"loss": 0.0387,
"step": 4560
},
{
"epoch": 16.498194945848375,
"grad_norm": 0.3080565929412842,
"learning_rate": 6.116519201394857e-05,
"loss": 0.0399,
"step": 4570
},
{
"epoch": 16.534296028880867,
"grad_norm": 0.30755096673965454,
"learning_rate": 6.10039594706969e-05,
"loss": 0.0333,
"step": 4580
},
{
"epoch": 16.57039711191336,
"grad_norm": 0.3727869689464569,
"learning_rate": 6.084260658990744e-05,
"loss": 0.0376,
"step": 4590
},
{
"epoch": 16.60649819494585,
"grad_norm": 0.21919415891170502,
"learning_rate": 6.068113513610943e-05,
"loss": 0.0336,
"step": 4600
},
{
"epoch": 16.642599277978338,
"grad_norm": 0.4546530544757843,
"learning_rate": 6.0519546875128876e-05,
"loss": 0.0431,
"step": 4610
},
{
"epoch": 16.67870036101083,
"grad_norm": 0.523841917514801,
"learning_rate": 6.035784357406906e-05,
"loss": 0.0418,
"step": 4620
},
{
"epoch": 16.71480144404332,
"grad_norm": 0.31432080268859863,
"learning_rate": 6.01960270012914e-05,
"loss": 0.039,
"step": 4630
},
{
"epoch": 16.750902527075812,
"grad_norm": 0.30887314677238464,
"learning_rate": 6.003409892639599e-05,
"loss": 0.0374,
"step": 4640
},
{
"epoch": 16.787003610108304,
"grad_norm": 0.3264867067337036,
"learning_rate": 5.9872061120202336e-05,
"loss": 0.041,
"step": 4650
},
{
"epoch": 16.823104693140795,
"grad_norm": 0.2943725287914276,
"learning_rate": 5.9709915354729914e-05,
"loss": 0.0431,
"step": 4660
},
{
"epoch": 16.859205776173287,
"grad_norm": 0.19398893415927887,
"learning_rate": 5.9547663403178824e-05,
"loss": 0.0368,
"step": 4670
},
{
"epoch": 16.895306859205775,
"grad_norm": 0.3400932252407074,
"learning_rate": 5.9385307039910445e-05,
"loss": 0.0303,
"step": 4680
},
{
"epoch": 16.931407942238266,
"grad_norm": 0.2681838870048523,
"learning_rate": 5.922284804042792e-05,
"loss": 0.0347,
"step": 4690
},
{
"epoch": 16.967509025270758,
"grad_norm": 0.19491979479789734,
"learning_rate": 5.906028818135687e-05,
"loss": 0.0412,
"step": 4700
},
{
"epoch": 17.00361010830325,
"grad_norm": 0.3350265920162201,
"learning_rate": 5.889762924042585e-05,
"loss": 0.0365,
"step": 4710
},
{
"epoch": 17.03971119133574,
"grad_norm": 0.36835935711860657,
"learning_rate": 5.873487299644699e-05,
"loss": 0.0356,
"step": 4720
},
{
"epoch": 17.075812274368232,
"grad_norm": 0.3096553683280945,
"learning_rate": 5.857202122929649e-05,
"loss": 0.0397,
"step": 4730
},
{
"epoch": 17.111913357400724,
"grad_norm": 0.3162147104740143,
"learning_rate": 5.840907571989518e-05,
"loss": 0.0309,
"step": 4740
},
{
"epoch": 17.14801444043321,
"grad_norm": 0.22204160690307617,
"learning_rate": 5.824603825018904e-05,
"loss": 0.0313,
"step": 4750
},
{
"epoch": 17.184115523465703,
"grad_norm": 0.266611248254776,
"learning_rate": 5.808291060312975e-05,
"loss": 0.0354,
"step": 4760
},
{
"epoch": 17.220216606498195,
"grad_norm": 0.3281164765357971,
"learning_rate": 5.7919694562655083e-05,
"loss": 0.0356,
"step": 4770
},
{
"epoch": 17.256317689530686,
"grad_norm": 0.30538710951805115,
"learning_rate": 5.775639191366954e-05,
"loss": 0.0349,
"step": 4780
},
{
"epoch": 17.292418772563177,
"grad_norm": 0.36445218324661255,
"learning_rate": 5.75930044420247e-05,
"loss": 0.0323,
"step": 4790
},
{
"epoch": 17.32851985559567,
"grad_norm": 0.335029274225235,
"learning_rate": 5.74295339344998e-05,
"loss": 0.0375,
"step": 4800
},
{
"epoch": 17.36462093862816,
"grad_norm": 0.255593478679657,
"learning_rate": 5.726598217878211e-05,
"loss": 0.0327,
"step": 4810
},
{
"epoch": 17.40072202166065,
"grad_norm": 0.2580687403678894,
"learning_rate": 5.71023509634474e-05,
"loss": 0.0384,
"step": 4820
},
{
"epoch": 17.43682310469314,
"grad_norm": 0.29046639800071716,
"learning_rate": 5.693864207794049e-05,
"loss": 0.0328,
"step": 4830
},
{
"epoch": 17.47292418772563,
"grad_norm": 0.2708757221698761,
"learning_rate": 5.677485731255545e-05,
"loss": 0.0435,
"step": 4840
},
{
"epoch": 17.509025270758123,
"grad_norm": 0.24068716168403625,
"learning_rate": 5.6610998458416296e-05,
"loss": 0.0357,
"step": 4850
},
{
"epoch": 17.545126353790614,
"grad_norm": 0.33343881368637085,
"learning_rate": 5.644706730745716e-05,
"loss": 0.0398,
"step": 4860
},
{
"epoch": 17.581227436823106,
"grad_norm": 0.30723315477371216,
"learning_rate": 5.628306565240287e-05,
"loss": 0.0365,
"step": 4870
},
{
"epoch": 17.617328519855597,
"grad_norm": 0.2691371738910675,
"learning_rate": 5.611899528674923e-05,
"loss": 0.0389,
"step": 4880
},
{
"epoch": 17.653429602888085,
"grad_norm": 0.4236083924770355,
"learning_rate": 5.595485800474349e-05,
"loss": 0.0315,
"step": 4890
},
{
"epoch": 17.689530685920577,
"grad_norm": 0.2767259478569031,
"learning_rate": 5.579065560136467e-05,
"loss": 0.0346,
"step": 4900
},
{
"epoch": 17.72563176895307,
"grad_norm": 0.2894260585308075,
"learning_rate": 5.562638987230392e-05,
"loss": 0.0333,
"step": 4910
},
{
"epoch": 17.76173285198556,
"grad_norm": 0.22172419726848602,
"learning_rate": 5.546206261394498e-05,
"loss": 0.0308,
"step": 4920
},
{
"epoch": 17.79783393501805,
"grad_norm": 0.32570475339889526,
"learning_rate": 5.529767562334437e-05,
"loss": 0.0339,
"step": 4930
},
{
"epoch": 17.833935018050543,
"grad_norm": 0.4423183798789978,
"learning_rate": 5.5133230698211926e-05,
"loss": 0.0315,
"step": 4940
},
{
"epoch": 17.870036101083034,
"grad_norm": 0.2762327492237091,
"learning_rate": 5.496872963689096e-05,
"loss": 0.0389,
"step": 4950
},
{
"epoch": 17.906137184115522,
"grad_norm": 0.2741451859474182,
"learning_rate": 5.4804174238338756e-05,
"loss": 0.0329,
"step": 4960
},
{
"epoch": 17.942238267148014,
"grad_norm": 0.30256128311157227,
"learning_rate": 5.463956630210678e-05,
"loss": 0.0383,
"step": 4970
},
{
"epoch": 17.978339350180505,
"grad_norm": 0.3132164180278778,
"learning_rate": 5.4474907628321046e-05,
"loss": 0.0411,
"step": 4980
},
{
"epoch": 18.014440433212997,
"grad_norm": 0.3592446446418762,
"learning_rate": 5.431020001766244e-05,
"loss": 0.0316,
"step": 4990
},
{
"epoch": 18.050541516245488,
"grad_norm": 0.45032453536987305,
"learning_rate": 5.4145445271346986e-05,
"loss": 0.0334,
"step": 5000
},
{
"epoch": 18.08664259927798,
"grad_norm": 0.41425392031669617,
"learning_rate": 5.398064519110622e-05,
"loss": 0.0376,
"step": 5010
},
{
"epoch": 18.12274368231047,
"grad_norm": 0.42565903067588806,
"learning_rate": 5.3815801579167394e-05,
"loss": 0.0339,
"step": 5020
},
{
"epoch": 18.15884476534296,
"grad_norm": 0.2825949788093567,
"learning_rate": 5.365091623823382e-05,
"loss": 0.0317,
"step": 5030
},
{
"epoch": 18.19494584837545,
"grad_norm": 0.30478280782699585,
"learning_rate": 5.348599097146521e-05,
"loss": 0.0309,
"step": 5040
},
{
"epoch": 18.231046931407942,
"grad_norm": 0.20079968869686127,
"learning_rate": 5.3321027582457836e-05,
"loss": 0.0358,
"step": 5050
},
{
"epoch": 18.267148014440433,
"grad_norm": 0.3120266795158386,
"learning_rate": 5.315602787522491e-05,
"loss": 0.0343,
"step": 5060
},
{
"epoch": 18.303249097472925,
"grad_norm": 0.3114013075828552,
"learning_rate": 5.299099365417678e-05,
"loss": 0.0356,
"step": 5070
},
{
"epoch": 18.339350180505416,
"grad_norm": 0.2290160357952118,
"learning_rate": 5.2825926724101236e-05,
"loss": 0.0311,
"step": 5080
},
{
"epoch": 18.375451263537904,
"grad_norm": 0.28134140372276306,
"learning_rate": 5.26608288901438e-05,
"loss": 0.0371,
"step": 5090
},
{
"epoch": 18.411552346570396,
"grad_norm": 0.24850766360759735,
"learning_rate": 5.24957019577879e-05,
"loss": 0.0345,
"step": 5100
},
{
"epoch": 18.447653429602887,
"grad_norm": 0.22971196472644806,
"learning_rate": 5.2330547732835266e-05,
"loss": 0.0298,
"step": 5110
},
{
"epoch": 18.48375451263538,
"grad_norm": 0.311047226190567,
"learning_rate": 5.2165368021385996e-05,
"loss": 0.0433,
"step": 5120
},
{
"epoch": 18.51985559566787,
"grad_norm": 0.35204485058784485,
"learning_rate": 5.200016462981897e-05,
"loss": 0.0348,
"step": 5130
},
{
"epoch": 18.555956678700362,
"grad_norm": 0.30155161023139954,
"learning_rate": 5.1834939364772015e-05,
"loss": 0.0309,
"step": 5140
},
{
"epoch": 18.592057761732853,
"grad_norm": 0.2238924354314804,
"learning_rate": 5.166969403312214e-05,
"loss": 0.0324,
"step": 5150
},
{
"epoch": 18.628158844765345,
"grad_norm": 0.3110749125480652,
"learning_rate": 5.1504430441965844e-05,
"loss": 0.0399,
"step": 5160
},
{
"epoch": 18.664259927797833,
"grad_norm": 0.2391263246536255,
"learning_rate": 5.133915039859923e-05,
"loss": 0.0272,
"step": 5170
},
{
"epoch": 18.700361010830324,
"grad_norm": 0.3187720477581024,
"learning_rate": 5.1173855710498444e-05,
"loss": 0.0328,
"step": 5180
},
{
"epoch": 18.736462093862816,
"grad_norm": 0.2643289864063263,
"learning_rate": 5.100854818529967e-05,
"loss": 0.0326,
"step": 5190
},
{
"epoch": 18.772563176895307,
"grad_norm": 0.2958979606628418,
"learning_rate": 5.084322963077951e-05,
"loss": 0.0361,
"step": 5200
},
{
"epoch": 18.8086642599278,
"grad_norm": 0.24935655295848846,
"learning_rate": 5.067790185483522e-05,
"loss": 0.032,
"step": 5210
},
{
"epoch": 18.84476534296029,
"grad_norm": 0.33372852206230164,
"learning_rate": 5.0512566665464844e-05,
"loss": 0.0377,
"step": 5220
},
{
"epoch": 18.880866425992778,
"grad_norm": 0.2626919448375702,
"learning_rate": 5.034722587074755e-05,
"loss": 0.0338,
"step": 5230
},
{
"epoch": 18.91696750902527,
"grad_norm": 0.38296061754226685,
"learning_rate": 5.018188127882375e-05,
"loss": 0.0321,
"step": 5240
},
{
"epoch": 18.95306859205776,
"grad_norm": 0.3037661015987396,
"learning_rate": 5.0016534697875417e-05,
"loss": 0.0376,
"step": 5250
},
{
"epoch": 18.989169675090253,
"grad_norm": 0.25098228454589844,
"learning_rate": 4.9851187936106294e-05,
"loss": 0.036,
"step": 5260
},
{
"epoch": 19.025270758122744,
"grad_norm": 0.28482964634895325,
"learning_rate": 4.968584280172206e-05,
"loss": 0.037,
"step": 5270
},
{
"epoch": 19.061371841155236,
"grad_norm": 0.41901057958602905,
"learning_rate": 4.95205011029106e-05,
"loss": 0.0366,
"step": 5280
},
{
"epoch": 19.097472924187727,
"grad_norm": 0.3747750520706177,
"learning_rate": 4.935516464782227e-05,
"loss": 0.0346,
"step": 5290
},
{
"epoch": 19.133574007220215,
"grad_norm": 0.25415968894958496,
"learning_rate": 4.918983524455003e-05,
"loss": 0.0322,
"step": 5300
},
{
"epoch": 19.169675090252706,
"grad_norm": 0.24878446757793427,
"learning_rate": 4.9024514701109766e-05,
"loss": 0.0333,
"step": 5310
},
{
"epoch": 19.205776173285198,
"grad_norm": 0.2366953045129776,
"learning_rate": 4.885920482542043e-05,
"loss": 0.0333,
"step": 5320
},
{
"epoch": 19.24187725631769,
"grad_norm": 0.3088334798812866,
"learning_rate": 4.869390742528438e-05,
"loss": 0.0364,
"step": 5330
},
{
"epoch": 19.27797833935018,
"grad_norm": 0.29115018248558044,
"learning_rate": 4.852862430836744e-05,
"loss": 0.0379,
"step": 5340
},
{
"epoch": 19.314079422382672,
"grad_norm": 0.30768904089927673,
"learning_rate": 4.836335728217933e-05,
"loss": 0.0321,
"step": 5350
},
{
"epoch": 19.350180505415164,
"grad_norm": 0.27671119570732117,
"learning_rate": 4.819810815405379e-05,
"loss": 0.0339,
"step": 5360
},
{
"epoch": 19.386281588447652,
"grad_norm": 0.26744624972343445,
"learning_rate": 4.803287873112877e-05,
"loss": 0.0369,
"step": 5370
},
{
"epoch": 19.422382671480143,
"grad_norm": 0.2991783916950226,
"learning_rate": 4.786767082032681e-05,
"loss": 0.0297,
"step": 5380
},
{
"epoch": 19.458483754512635,
"grad_norm": 0.328084260225296,
"learning_rate": 4.77024862283351e-05,
"loss": 0.0304,
"step": 5390
},
{
"epoch": 19.494584837545126,
"grad_norm": 0.2546713948249817,
"learning_rate": 4.753732676158593e-05,
"loss": 0.0356,
"step": 5400
},
{
"epoch": 19.530685920577618,
"grad_norm": 0.34368693828582764,
"learning_rate": 4.737219422623672e-05,
"loss": 0.0346,
"step": 5410
},
{
"epoch": 19.56678700361011,
"grad_norm": 0.30635061860084534,
"learning_rate": 4.720709042815044e-05,
"loss": 0.0315,
"step": 5420
},
{
"epoch": 19.6028880866426,
"grad_norm": 0.29206594824790955,
"learning_rate": 4.704201717287578e-05,
"loss": 0.0295,
"step": 5430
},
{
"epoch": 19.63898916967509,
"grad_norm": 0.2934032380580902,
"learning_rate": 4.6876976265627404e-05,
"loss": 0.031,
"step": 5440
},
{
"epoch": 19.67509025270758,
"grad_norm": 0.22749246656894684,
"learning_rate": 4.671196951126626e-05,
"loss": 0.036,
"step": 5450
},
{
"epoch": 19.71119133574007,
"grad_norm": 0.26649293303489685,
"learning_rate": 4.654699871427971e-05,
"loss": 0.0348,
"step": 5460
},
{
"epoch": 19.747292418772563,
"grad_norm": 0.20854540169239044,
"learning_rate": 4.6382065678762034e-05,
"loss": 0.0295,
"step": 5470
},
{
"epoch": 19.783393501805055,
"grad_norm": 0.2517044246196747,
"learning_rate": 4.6217172208394424e-05,
"loss": 0.0307,
"step": 5480
},
{
"epoch": 19.819494584837546,
"grad_norm": 0.4335871934890747,
"learning_rate": 4.605232010642549e-05,
"loss": 0.0315,
"step": 5490
},
{
"epoch": 19.855595667870038,
"grad_norm": 0.2591581344604492,
"learning_rate": 4.588751117565142e-05,
"loss": 0.0327,
"step": 5500
},
{
"epoch": 19.891696750902526,
"grad_norm": 0.27881044149398804,
"learning_rate": 4.5722747218396214e-05,
"loss": 0.0369,
"step": 5510
},
{
"epoch": 19.927797833935017,
"grad_norm": 0.15780337154865265,
"learning_rate": 4.5558030036492194e-05,
"loss": 0.0259,
"step": 5520
},
{
"epoch": 19.96389891696751,
"grad_norm": 0.29680997133255005,
"learning_rate": 4.539336143125999e-05,
"loss": 0.0396,
"step": 5530
},
{
"epoch": 20.0,
"grad_norm": 0.25618883967399597,
"learning_rate": 4.522874320348916e-05,
"loss": 0.0291,
"step": 5540
},
{
"epoch": 20.03610108303249,
"grad_norm": 0.2965971529483795,
"learning_rate": 4.506417715341821e-05,
"loss": 0.0254,
"step": 5550
},
{
"epoch": 20.072202166064983,
"grad_norm": 0.2831667363643646,
"learning_rate": 4.489966508071511e-05,
"loss": 0.0327,
"step": 5560
},
{
"epoch": 20.108303249097474,
"grad_norm": 0.26466771960258484,
"learning_rate": 4.4735208784457575e-05,
"loss": 0.027,
"step": 5570
},
{
"epoch": 20.144404332129962,
"grad_norm": 0.25879523158073425,
"learning_rate": 4.457081006311325e-05,
"loss": 0.0315,
"step": 5580
},
{
"epoch": 20.180505415162454,
"grad_norm": 0.26909783482551575,
"learning_rate": 4.440647071452027e-05,
"loss": 0.0349,
"step": 5590
},
{
"epoch": 20.216606498194945,
"grad_norm": 0.1857617348432541,
"learning_rate": 4.424219253586737e-05,
"loss": 0.0316,
"step": 5600
},
{
"epoch": 20.252707581227437,
"grad_norm": 0.23268231749534607,
"learning_rate": 4.407797732367443e-05,
"loss": 0.0375,
"step": 5610
},
{
"epoch": 20.28880866425993,
"grad_norm": 0.3097105622291565,
"learning_rate": 4.391382687377268e-05,
"loss": 0.0344,
"step": 5620
},
{
"epoch": 20.32490974729242,
"grad_norm": 0.21181944012641907,
"learning_rate": 4.374974298128512e-05,
"loss": 0.0308,
"step": 5630
},
{
"epoch": 20.36101083032491,
"grad_norm": 0.17041045427322388,
"learning_rate": 4.358572744060699e-05,
"loss": 0.0311,
"step": 5640
},
{
"epoch": 20.3971119133574,
"grad_norm": 0.20999184250831604,
"learning_rate": 4.342178204538588e-05,
"loss": 0.0288,
"step": 5650
},
{
"epoch": 20.43321299638989,
"grad_norm": 0.301200807094574,
"learning_rate": 4.325790858850241e-05,
"loss": 0.0314,
"step": 5660
},
{
"epoch": 20.469314079422382,
"grad_norm": 0.23410813510417938,
"learning_rate": 4.309410886205043e-05,
"loss": 0.029,
"step": 5670
},
{
"epoch": 20.505415162454874,
"grad_norm": 0.2804301977157593,
"learning_rate": 4.293038465731752e-05,
"loss": 0.0315,
"step": 5680
},
{
"epoch": 20.541516245487365,
"grad_norm": 0.2343648225069046,
"learning_rate": 4.276673776476533e-05,
"loss": 0.031,
"step": 5690
},
{
"epoch": 20.577617328519857,
"grad_norm": 0.20691975951194763,
"learning_rate": 4.260316997401007e-05,
"loss": 0.0259,
"step": 5700
},
{
"epoch": 20.613718411552348,
"grad_norm": 0.24357058107852936,
"learning_rate": 4.243968307380293e-05,
"loss": 0.0292,
"step": 5710
},
{
"epoch": 20.649819494584836,
"grad_norm": 0.2861655652523041,
"learning_rate": 4.22762788520104e-05,
"loss": 0.0325,
"step": 5720
},
{
"epoch": 20.685920577617328,
"grad_norm": 0.24166567623615265,
"learning_rate": 4.211295909559491e-05,
"loss": 0.0371,
"step": 5730
},
{
"epoch": 20.72202166064982,
"grad_norm": 0.19814735651016235,
"learning_rate": 4.194972559059511e-05,
"loss": 0.0317,
"step": 5740
},
{
"epoch": 20.75812274368231,
"grad_norm": 0.20906837284564972,
"learning_rate": 4.178658012210651e-05,
"loss": 0.0292,
"step": 5750
},
{
"epoch": 20.794223826714802,
"grad_norm": 0.4491354525089264,
"learning_rate": 4.162352447426177e-05,
"loss": 0.031,
"step": 5760
},
{
"epoch": 20.830324909747294,
"grad_norm": 0.28082311153411865,
"learning_rate": 4.146056043021135e-05,
"loss": 0.0295,
"step": 5770
},
{
"epoch": 20.866425992779785,
"grad_norm": 0.36126407980918884,
"learning_rate": 4.1297689772103944e-05,
"loss": 0.035,
"step": 5780
},
{
"epoch": 20.902527075812273,
"grad_norm": 0.2725689709186554,
"learning_rate": 4.113491428106694e-05,
"loss": 0.0343,
"step": 5790
},
{
"epoch": 20.938628158844764,
"grad_norm": 0.31661349534988403,
"learning_rate": 4.0972235737187055e-05,
"loss": 0.0316,
"step": 5800
},
{
"epoch": 20.974729241877256,
"grad_norm": 0.18992729485034943,
"learning_rate": 4.080965591949076e-05,
"loss": 0.0292,
"step": 5810
},
{
"epoch": 21.010830324909747,
"grad_norm": 0.38685137033462524,
"learning_rate": 4.0647176605924924e-05,
"loss": 0.0287,
"step": 5820
},
{
"epoch": 21.04693140794224,
"grad_norm": 0.3436387777328491,
"learning_rate": 4.0484799573337255e-05,
"loss": 0.0336,
"step": 5830
},
{
"epoch": 21.08303249097473,
"grad_norm": 0.3449047803878784,
"learning_rate": 4.032252659745699e-05,
"loss": 0.0319,
"step": 5840
},
{
"epoch": 21.119133574007222,
"grad_norm": 0.27810367941856384,
"learning_rate": 4.016035945287539e-05,
"loss": 0.0345,
"step": 5850
},
{
"epoch": 21.15523465703971,
"grad_norm": 0.26223263144493103,
"learning_rate": 3.999829991302635e-05,
"loss": 0.03,
"step": 5860
},
{
"epoch": 21.1913357400722,
"grad_norm": 0.1911151111125946,
"learning_rate": 3.983634975016707e-05,
"loss": 0.0319,
"step": 5870
},
{
"epoch": 21.227436823104693,
"grad_norm": 0.3983359932899475,
"learning_rate": 3.967451073535854e-05,
"loss": 0.0297,
"step": 5880
},
{
"epoch": 21.263537906137184,
"grad_norm": 0.24901708960533142,
"learning_rate": 3.951278463844633e-05,
"loss": 0.0375,
"step": 5890
},
{
"epoch": 21.299638989169676,
"grad_norm": 0.3323638141155243,
"learning_rate": 3.935117322804111e-05,
"loss": 0.0278,
"step": 5900
},
{
"epoch": 21.335740072202167,
"grad_norm": 0.28690487146377563,
"learning_rate": 3.918967827149938e-05,
"loss": 0.029,
"step": 5910
},
{
"epoch": 21.37184115523466,
"grad_norm": 0.2840134799480438,
"learning_rate": 3.9028301534904094e-05,
"loss": 0.032,
"step": 5920
},
{
"epoch": 21.407942238267147,
"grad_norm": 0.2103012502193451,
"learning_rate": 3.88670447830454e-05,
"loss": 0.0291,
"step": 5930
},
{
"epoch": 21.444043321299638,
"grad_norm": 0.2093581110239029,
"learning_rate": 3.870590977940132e-05,
"loss": 0.0282,
"step": 5940
},
{
"epoch": 21.48014440433213,
"grad_norm": 0.24812541902065277,
"learning_rate": 3.8544898286118404e-05,
"loss": 0.0271,
"step": 5950
},
{
"epoch": 21.51624548736462,
"grad_norm": 0.20950178802013397,
"learning_rate": 3.838401206399257e-05,
"loss": 0.0234,
"step": 5960
},
{
"epoch": 21.552346570397113,
"grad_norm": 0.36595800518989563,
"learning_rate": 3.822325287244975e-05,
"loss": 0.0286,
"step": 5970
},
{
"epoch": 21.588447653429604,
"grad_norm": 0.2616178095340729,
"learning_rate": 3.8062622469526725e-05,
"loss": 0.0222,
"step": 5980
},
{
"epoch": 21.624548736462096,
"grad_norm": 0.20015136897563934,
"learning_rate": 3.790212261185183e-05,
"loss": 0.023,
"step": 5990
},
{
"epoch": 21.660649819494584,
"grad_norm": 0.24901485443115234,
"learning_rate": 3.7741755054625794e-05,
"loss": 0.0342,
"step": 6000
},
{
"epoch": 21.696750902527075,
"grad_norm": 0.19870421290397644,
"learning_rate": 3.758152155160255e-05,
"loss": 0.024,
"step": 6010
},
{
"epoch": 21.732851985559567,
"grad_norm": 0.15553215146064758,
"learning_rate": 3.742142385506999e-05,
"loss": 0.0302,
"step": 6020
},
{
"epoch": 21.768953068592058,
"grad_norm": 0.21830004453659058,
"learning_rate": 3.72614637158309e-05,
"loss": 0.0285,
"step": 6030
},
{
"epoch": 21.80505415162455,
"grad_norm": 0.31320229172706604,
"learning_rate": 3.710164288318371e-05,
"loss": 0.0334,
"step": 6040
},
{
"epoch": 21.84115523465704,
"grad_norm": 0.1947004497051239,
"learning_rate": 3.694196310490345e-05,
"loss": 0.0268,
"step": 6050
},
{
"epoch": 21.87725631768953,
"grad_norm": 0.4148366153240204,
"learning_rate": 3.678242612722259e-05,
"loss": 0.0319,
"step": 6060
},
{
"epoch": 21.91335740072202,
"grad_norm": 0.2147468626499176,
"learning_rate": 3.6623033694811953e-05,
"loss": 0.0286,
"step": 6070
},
{
"epoch": 21.949458483754512,
"grad_norm": 0.26231124997138977,
"learning_rate": 3.6463787550761665e-05,
"loss": 0.0322,
"step": 6080
},
{
"epoch": 21.985559566787003,
"grad_norm": 0.2691902816295624,
"learning_rate": 3.630468943656202e-05,
"loss": 0.0249,
"step": 6090
},
{
"epoch": 22.021660649819495,
"grad_norm": 0.24936966598033905,
"learning_rate": 3.6145741092084523e-05,
"loss": 0.0229,
"step": 6100
},
{
"epoch": 22.057761732851986,
"grad_norm": 0.3215731680393219,
"learning_rate": 3.598694425556278e-05,
"loss": 0.0279,
"step": 6110
},
{
"epoch": 22.093862815884478,
"grad_norm": 0.20099368691444397,
"learning_rate": 3.58283006635736e-05,
"loss": 0.0238,
"step": 6120
},
{
"epoch": 22.129963898916966,
"grad_norm": 0.27343305945396423,
"learning_rate": 3.566981205101781e-05,
"loss": 0.0265,
"step": 6130
},
{
"epoch": 22.166064981949457,
"grad_norm": 0.3141259551048279,
"learning_rate": 3.5511480151101556e-05,
"loss": 0.0242,
"step": 6140
},
{
"epoch": 22.20216606498195,
"grad_norm": 0.28763604164123535,
"learning_rate": 3.5353306695317104e-05,
"loss": 0.0228,
"step": 6150
},
{
"epoch": 22.23826714801444,
"grad_norm": 0.3379686176776886,
"learning_rate": 3.519529341342402e-05,
"loss": 0.03,
"step": 6160
},
{
"epoch": 22.27436823104693,
"grad_norm": 0.3278880715370178,
"learning_rate": 3.503744203343026e-05,
"loss": 0.0312,
"step": 6170
},
{
"epoch": 22.310469314079423,
"grad_norm": 0.2681352198123932,
"learning_rate": 3.487975428157318e-05,
"loss": 0.0286,
"step": 6180
},
{
"epoch": 22.346570397111915,
"grad_norm": 0.41349533200263977,
"learning_rate": 3.472223188230083e-05,
"loss": 0.0281,
"step": 6190
},
{
"epoch": 22.382671480144403,
"grad_norm": 0.3628394305706024,
"learning_rate": 3.4564876558252866e-05,
"loss": 0.0325,
"step": 6200
},
{
"epoch": 22.418772563176894,
"grad_norm": 0.25120222568511963,
"learning_rate": 3.440769003024195e-05,
"loss": 0.027,
"step": 6210
},
{
"epoch": 22.454873646209386,
"grad_norm": 0.31477054953575134,
"learning_rate": 3.425067401723477e-05,
"loss": 0.0251,
"step": 6220
},
{
"epoch": 22.490974729241877,
"grad_norm": 0.37628892064094543,
"learning_rate": 3.409383023633325e-05,
"loss": 0.0288,
"step": 6230
},
{
"epoch": 22.52707581227437,
"grad_norm": 0.30439165234565735,
"learning_rate": 3.3937160402755894e-05,
"loss": 0.0294,
"step": 6240
},
{
"epoch": 22.56317689530686,
"grad_norm": 0.16752374172210693,
"learning_rate": 3.378066622981885e-05,
"loss": 0.0279,
"step": 6250
},
{
"epoch": 22.59927797833935,
"grad_norm": 0.2891589403152466,
"learning_rate": 3.362434942891738e-05,
"loss": 0.032,
"step": 6260
},
{
"epoch": 22.63537906137184,
"grad_norm": 0.211389422416687,
"learning_rate": 3.346821170950693e-05,
"loss": 0.0294,
"step": 6270
},
{
"epoch": 22.67148014440433,
"grad_norm": 0.26439085602760315,
"learning_rate": 3.3312254779084585e-05,
"loss": 0.0286,
"step": 6280
},
{
"epoch": 22.707581227436823,
"grad_norm": 0.23191316425800323,
"learning_rate": 3.315648034317039e-05,
"loss": 0.029,
"step": 6290
},
{
"epoch": 22.743682310469314,
"grad_norm": 0.1382790058851242,
"learning_rate": 3.3000890105288564e-05,
"loss": 0.0255,
"step": 6300
},
{
"epoch": 22.779783393501805,
"grad_norm": 0.3309822380542755,
"learning_rate": 3.284548576694908e-05,
"loss": 0.0279,
"step": 6310
},
{
"epoch": 22.815884476534297,
"grad_norm": 0.2321564257144928,
"learning_rate": 3.2690269027628815e-05,
"loss": 0.0248,
"step": 6320
},
{
"epoch": 22.85198555956679,
"grad_norm": 0.24728421866893768,
"learning_rate": 3.253524158475324e-05,
"loss": 0.0252,
"step": 6330
},
{
"epoch": 22.888086642599276,
"grad_norm": 0.2944360375404358,
"learning_rate": 3.238040513367757e-05,
"loss": 0.0248,
"step": 6340
},
{
"epoch": 22.924187725631768,
"grad_norm": 0.22109302878379822,
"learning_rate": 3.222576136766843e-05,
"loss": 0.0248,
"step": 6350
},
{
"epoch": 22.96028880866426,
"grad_norm": 0.20417636632919312,
"learning_rate": 3.2071311977885324e-05,
"loss": 0.0299,
"step": 6360
},
{
"epoch": 22.99638989169675,
"grad_norm": 0.32738369703292847,
"learning_rate": 3.191705865336197e-05,
"loss": 0.0277,
"step": 6370
},
{
"epoch": 23.032490974729242,
"grad_norm": 0.28265514969825745,
"learning_rate": 3.1763003080988075e-05,
"loss": 0.026,
"step": 6380
},
{
"epoch": 23.068592057761734,
"grad_norm": 0.32340627908706665,
"learning_rate": 3.160914694549063e-05,
"loss": 0.0316,
"step": 6390
},
{
"epoch": 23.104693140794225,
"grad_norm": 0.29524344205856323,
"learning_rate": 3.145549192941573e-05,
"loss": 0.0304,
"step": 6400
},
{
"epoch": 23.140794223826713,
"grad_norm": 0.3630560040473938,
"learning_rate": 3.130203971310999e-05,
"loss": 0.0273,
"step": 6410
},
{
"epoch": 23.176895306859205,
"grad_norm": 0.3041969835758209,
"learning_rate": 3.114879197470225e-05,
"loss": 0.0286,
"step": 6420
},
{
"epoch": 23.212996389891696,
"grad_norm": 0.35867491364479065,
"learning_rate": 3.0995750390085285e-05,
"loss": 0.0215,
"step": 6430
},
{
"epoch": 23.249097472924188,
"grad_norm": 0.2925831079483032,
"learning_rate": 3.084291663289728e-05,
"loss": 0.0221,
"step": 6440
},
{
"epoch": 23.28519855595668,
"grad_norm": 0.20823247730731964,
"learning_rate": 3.069029237450375e-05,
"loss": 0.0245,
"step": 6450
},
{
"epoch": 23.32129963898917,
"grad_norm": 0.1792321801185608,
"learning_rate": 3.053787928397911e-05,
"loss": 0.0187,
"step": 6460
},
{
"epoch": 23.357400722021662,
"grad_norm": 0.20706138014793396,
"learning_rate": 3.0385679028088526e-05,
"loss": 0.0264,
"step": 6470
},
{
"epoch": 23.39350180505415,
"grad_norm": 0.1782984733581543,
"learning_rate": 3.023369327126959e-05,
"loss": 0.0284,
"step": 6480
},
{
"epoch": 23.42960288808664,
"grad_norm": 0.25688961148262024,
"learning_rate": 3.0081923675614198e-05,
"loss": 0.0233,
"step": 6490
},
{
"epoch": 23.465703971119133,
"grad_norm": 0.2861765921115875,
"learning_rate": 2.993037190085034e-05,
"loss": 0.0246,
"step": 6500
},
{
"epoch": 23.501805054151625,
"grad_norm": 0.21940988302230835,
"learning_rate": 2.977903960432392e-05,
"loss": 0.0253,
"step": 6510
},
{
"epoch": 23.537906137184116,
"grad_norm": 0.24884116649627686,
"learning_rate": 2.9627928440980722e-05,
"loss": 0.0259,
"step": 6520
},
{
"epoch": 23.574007220216608,
"grad_norm": 0.33147934079170227,
"learning_rate": 2.9477040063348183e-05,
"loss": 0.0282,
"step": 6530
},
{
"epoch": 23.6101083032491,
"grad_norm": 0.2484951764345169,
"learning_rate": 2.9326376121517456e-05,
"loss": 0.0279,
"step": 6540
},
{
"epoch": 23.646209386281587,
"grad_norm": 0.24542994797229767,
"learning_rate": 2.9175938263125236e-05,
"loss": 0.0238,
"step": 6550
},
{
"epoch": 23.68231046931408,
"grad_norm": 0.2500183880329132,
"learning_rate": 2.9025728133335873e-05,
"loss": 0.0301,
"step": 6560
},
{
"epoch": 23.71841155234657,
"grad_norm": 0.2633858323097229,
"learning_rate": 2.8875747374823288e-05,
"loss": 0.0286,
"step": 6570
},
{
"epoch": 23.75451263537906,
"grad_norm": 0.26278653740882874,
"learning_rate": 2.872599762775298e-05,
"loss": 0.0306,
"step": 6580
},
{
"epoch": 23.790613718411553,
"grad_norm": 0.1902109533548355,
"learning_rate": 2.857648052976425e-05,
"loss": 0.0253,
"step": 6590
},
{
"epoch": 23.826714801444044,
"grad_norm": 0.19105219841003418,
"learning_rate": 2.8427197715952047e-05,
"loss": 0.0246,
"step": 6600
},
{
"epoch": 23.862815884476536,
"grad_norm": 0.46645423769950867,
"learning_rate": 2.8278150818849393e-05,
"loss": 0.0248,
"step": 6610
},
{
"epoch": 23.898916967509024,
"grad_norm": 0.2106551229953766,
"learning_rate": 2.812934146840922e-05,
"loss": 0.0254,
"step": 6620
},
{
"epoch": 23.935018050541515,
"grad_norm": 0.2511257231235504,
"learning_rate": 2.7980771291986764e-05,
"loss": 0.0285,
"step": 6630
},
{
"epoch": 23.971119133574007,
"grad_norm": 0.25811731815338135,
"learning_rate": 2.783244191432167e-05,
"loss": 0.0253,
"step": 6640
},
{
"epoch": 24.0072202166065,
"grad_norm": 0.18597958981990814,
"learning_rate": 2.768435495752022e-05,
"loss": 0.0231,
"step": 6650
},
{
"epoch": 24.04332129963899,
"grad_norm": 0.25478193163871765,
"learning_rate": 2.753651204103771e-05,
"loss": 0.0277,
"step": 6660
},
{
"epoch": 24.07942238267148,
"grad_norm": 0.26418277621269226,
"learning_rate": 2.7388914781660523e-05,
"loss": 0.022,
"step": 6670
},
{
"epoch": 24.115523465703973,
"grad_norm": 0.24497416615486145,
"learning_rate": 2.7241564793488693e-05,
"loss": 0.0249,
"step": 6680
},
{
"epoch": 24.15162454873646,
"grad_norm": 0.3049469292163849,
"learning_rate": 2.7094463687918037e-05,
"loss": 0.0221,
"step": 6690
},
{
"epoch": 24.187725631768952,
"grad_norm": 0.22437645494937897,
"learning_rate": 2.694761307362268e-05,
"loss": 0.0227,
"step": 6700
},
{
"epoch": 24.223826714801444,
"grad_norm": 0.227764293551445,
"learning_rate": 2.6801014556537467e-05,
"loss": 0.0218,
"step": 6710
},
{
"epoch": 24.259927797833935,
"grad_norm": 0.1991446614265442,
"learning_rate": 2.6654669739840243e-05,
"loss": 0.0256,
"step": 6720
},
{
"epoch": 24.296028880866427,
"grad_norm": 0.27192482352256775,
"learning_rate": 2.650858022393451e-05,
"loss": 0.0249,
"step": 6730
},
{
"epoch": 24.332129963898918,
"grad_norm": 0.28191351890563965,
"learning_rate": 2.6362747606431747e-05,
"loss": 0.0316,
"step": 6740
},
{
"epoch": 24.36823104693141,
"grad_norm": 0.19530995190143585,
"learning_rate": 2.6217173482134172e-05,
"loss": 0.0236,
"step": 6750
},
{
"epoch": 24.404332129963898,
"grad_norm": 0.29901766777038574,
"learning_rate": 2.6071859443017044e-05,
"loss": 0.0184,
"step": 6760
},
{
"epoch": 24.44043321299639,
"grad_norm": 0.1920684278011322,
"learning_rate": 2.5926807078211414e-05,
"loss": 0.0247,
"step": 6770
},
{
"epoch": 24.47653429602888,
"grad_norm": 0.2236887365579605,
"learning_rate": 2.5782017973986728e-05,
"loss": 0.0234,
"step": 6780
},
{
"epoch": 24.512635379061372,
"grad_norm": 0.22846749424934387,
"learning_rate": 2.5637493713733374e-05,
"loss": 0.0234,
"step": 6790
},
{
"epoch": 24.548736462093864,
"grad_norm": 0.29160743951797485,
"learning_rate": 2.549323587794559e-05,
"loss": 0.0282,
"step": 6800
},
{
"epoch": 24.584837545126355,
"grad_norm": 0.3081692159175873,
"learning_rate": 2.5349246044203895e-05,
"loss": 0.0215,
"step": 6810
},
{
"epoch": 24.620938628158846,
"grad_norm": 0.16789400577545166,
"learning_rate": 2.520552578715808e-05,
"loss": 0.0203,
"step": 6820
},
{
"epoch": 24.657039711191334,
"grad_norm": 0.2446165233850479,
"learning_rate": 2.506207667850981e-05,
"loss": 0.0234,
"step": 6830
},
{
"epoch": 24.693140794223826,
"grad_norm": 0.28270697593688965,
"learning_rate": 2.4918900286995555e-05,
"loss": 0.0255,
"step": 6840
},
{
"epoch": 24.729241877256317,
"grad_norm": 0.19977334141731262,
"learning_rate": 2.4775998178369458e-05,
"loss": 0.0211,
"step": 6850
},
{
"epoch": 24.76534296028881,
"grad_norm": 0.3588809370994568,
"learning_rate": 2.4633371915386017e-05,
"loss": 0.0256,
"step": 6860
},
{
"epoch": 24.8014440433213,
"grad_norm": 0.2613070011138916,
"learning_rate": 2.4491023057783235e-05,
"loss": 0.03,
"step": 6870
},
{
"epoch": 24.837545126353792,
"grad_norm": 0.23739217221736908,
"learning_rate": 2.4348953162265375e-05,
"loss": 0.0265,
"step": 6880
},
{
"epoch": 24.87364620938628,
"grad_norm": 0.2952808439731598,
"learning_rate": 2.420716378248607e-05,
"loss": 0.03,
"step": 6890
},
{
"epoch": 24.90974729241877,
"grad_norm": 0.24205149710178375,
"learning_rate": 2.4065656469031266e-05,
"loss": 0.0187,
"step": 6900
},
{
"epoch": 24.945848375451263,
"grad_norm": 0.25526392459869385,
"learning_rate": 2.3924432769402268e-05,
"loss": 0.0281,
"step": 6910
},
{
"epoch": 24.981949458483754,
"grad_norm": 0.2091396152973175,
"learning_rate": 2.3783494227998844e-05,
"loss": 0.0291,
"step": 6920
},
{
"epoch": 25.018050541516246,
"grad_norm": 0.28312212228775024,
"learning_rate": 2.3642842386102264e-05,
"loss": 0.0264,
"step": 6930
},
{
"epoch": 25.054151624548737,
"grad_norm": 0.19178296625614166,
"learning_rate": 2.3502478781858567e-05,
"loss": 0.0215,
"step": 6940
},
{
"epoch": 25.09025270758123,
"grad_norm": 0.24188187718391418,
"learning_rate": 2.3362404950261628e-05,
"loss": 0.0243,
"step": 6950
},
{
"epoch": 25.126353790613717,
"grad_norm": 0.1949673891067505,
"learning_rate": 2.3222622423136458e-05,
"loss": 0.0244,
"step": 6960
},
{
"epoch": 25.162454873646208,
"grad_norm": 0.19260956346988678,
"learning_rate": 2.3083132729122332e-05,
"loss": 0.0176,
"step": 6970
},
{
"epoch": 25.1985559566787,
"grad_norm": 0.2616089880466461,
"learning_rate": 2.294393739365621e-05,
"loss": 0.0266,
"step": 6980
},
{
"epoch": 25.23465703971119,
"grad_norm": 0.20548264682292938,
"learning_rate": 2.2805037938956e-05,
"loss": 0.0246,
"step": 6990
},
{
"epoch": 25.270758122743683,
"grad_norm": 0.23555199801921844,
"learning_rate": 2.266643588400386e-05,
"loss": 0.022,
"step": 7000
},
{
"epoch": 25.306859205776174,
"grad_norm": 0.24743910133838654,
"learning_rate": 2.252813274452969e-05,
"loss": 0.0277,
"step": 7010
},
{
"epoch": 25.342960288808666,
"grad_norm": 0.21370252966880798,
"learning_rate": 2.2390130032994427e-05,
"loss": 0.0181,
"step": 7020
},
{
"epoch": 25.379061371841154,
"grad_norm": 0.17858874797821045,
"learning_rate": 2.2252429258573633e-05,
"loss": 0.023,
"step": 7030
},
{
"epoch": 25.415162454873645,
"grad_norm": 0.2576480805873871,
"learning_rate": 2.2115031927140904e-05,
"loss": 0.0244,
"step": 7040
},
{
"epoch": 25.451263537906136,
"grad_norm": 0.3007279932498932,
"learning_rate": 2.1977939541251463e-05,
"loss": 0.0285,
"step": 7050
},
{
"epoch": 25.487364620938628,
"grad_norm": 0.3047775626182556,
"learning_rate": 2.1841153600125684e-05,
"loss": 0.0265,
"step": 7060
},
{
"epoch": 25.52346570397112,
"grad_norm": 0.294158935546875,
"learning_rate": 2.170467559963267e-05,
"loss": 0.023,
"step": 7070
},
{
"epoch": 25.55956678700361,
"grad_norm": 0.16885989904403687,
"learning_rate": 2.1568507032273982e-05,
"loss": 0.0196,
"step": 7080
},
{
"epoch": 25.595667870036102,
"grad_norm": 0.19250856339931488,
"learning_rate": 2.1432649387167264e-05,
"loss": 0.0229,
"step": 7090
},
{
"epoch": 25.63176895306859,
"grad_norm": 0.20653019845485687,
"learning_rate": 2.1297104150029973e-05,
"loss": 0.0206,
"step": 7100
},
{
"epoch": 25.667870036101082,
"grad_norm": 0.24652233719825745,
"learning_rate": 2.116187280316307e-05,
"loss": 0.0237,
"step": 7110
},
{
"epoch": 25.703971119133573,
"grad_norm": 0.19491852819919586,
"learning_rate": 2.1026956825434908e-05,
"loss": 0.0216,
"step": 7120
},
{
"epoch": 25.740072202166065,
"grad_norm": 0.2192380279302597,
"learning_rate": 2.0892357692265017e-05,
"loss": 0.0255,
"step": 7130
},
{
"epoch": 25.776173285198556,
"grad_norm": 0.3004741072654724,
"learning_rate": 2.0758076875607947e-05,
"loss": 0.0227,
"step": 7140
},
{
"epoch": 25.812274368231048,
"grad_norm": 0.1906908005475998,
"learning_rate": 2.0624115843937207e-05,
"loss": 0.0233,
"step": 7150
},
{
"epoch": 25.84837545126354,
"grad_norm": 0.23217976093292236,
"learning_rate": 2.0490476062229157e-05,
"loss": 0.0248,
"step": 7160
},
{
"epoch": 25.884476534296027,
"grad_norm": 0.2310824692249298,
"learning_rate": 2.035715899194704e-05,
"loss": 0.0233,
"step": 7170
},
{
"epoch": 25.92057761732852,
"grad_norm": 0.15220053493976593,
"learning_rate": 2.022416609102499e-05,
"loss": 0.0207,
"step": 7180
},
{
"epoch": 25.95667870036101,
"grad_norm": 0.3205597996711731,
"learning_rate": 2.009149881385205e-05,
"loss": 0.0269,
"step": 7190
},
{
"epoch": 25.9927797833935,
"grad_norm": 0.18352051079273224,
"learning_rate": 1.995915861125634e-05,
"loss": 0.0227,
"step": 7200
},
{
"epoch": 26.028880866425993,
"grad_norm": 0.22951056063175201,
"learning_rate": 1.9827146930489065e-05,
"loss": 0.0196,
"step": 7210
},
{
"epoch": 26.064981949458485,
"grad_norm": 0.21394377946853638,
"learning_rate": 1.9695465215208848e-05,
"loss": 0.0216,
"step": 7220
},
{
"epoch": 26.101083032490976,
"grad_norm": 0.21465672552585602,
"learning_rate": 1.9564114905465813e-05,
"loss": 0.0216,
"step": 7230
},
{
"epoch": 26.137184115523464,
"grad_norm": 0.16474515199661255,
"learning_rate": 1.9433097437685936e-05,
"loss": 0.0199,
"step": 7240
},
{
"epoch": 26.173285198555956,
"grad_norm": 0.23230904340744019,
"learning_rate": 1.930241424465521e-05,
"loss": 0.0196,
"step": 7250
},
{
"epoch": 26.209386281588447,
"grad_norm": 0.24260899424552917,
"learning_rate": 1.9172066755504115e-05,
"loss": 0.0241,
"step": 7260
},
{
"epoch": 26.24548736462094,
"grad_norm": 0.25509822368621826,
"learning_rate": 1.9042056395691914e-05,
"loss": 0.0226,
"step": 7270
},
{
"epoch": 26.28158844765343,
"grad_norm": 0.16365303099155426,
"learning_rate": 1.8912384586991066e-05,
"loss": 0.0218,
"step": 7280
},
{
"epoch": 26.31768953068592,
"grad_norm": 0.21962285041809082,
"learning_rate": 1.8783052747471717e-05,
"loss": 0.0192,
"step": 7290
},
{
"epoch": 26.353790613718413,
"grad_norm": 0.2750597298145294,
"learning_rate": 1.865406229148611e-05,
"loss": 0.0259,
"step": 7300
},
{
"epoch": 26.3898916967509,
"grad_norm": 0.13160721957683563,
"learning_rate": 1.8525414629653233e-05,
"loss": 0.0224,
"step": 7310
},
{
"epoch": 26.425992779783392,
"grad_norm": 0.15132488310337067,
"learning_rate": 1.8397111168843255e-05,
"loss": 0.0225,
"step": 7320
},
{
"epoch": 26.462093862815884,
"grad_norm": 0.27114179730415344,
"learning_rate": 1.8269153312162323e-05,
"loss": 0.0233,
"step": 7330
},
{
"epoch": 26.498194945848375,
"grad_norm": 0.20210124552249908,
"learning_rate": 1.8141542458937054e-05,
"loss": 0.0249,
"step": 7340
},
{
"epoch": 26.534296028880867,
"grad_norm": 0.22995750606060028,
"learning_rate": 1.8014280004699268e-05,
"loss": 0.0208,
"step": 7350
},
{
"epoch": 26.57039711191336,
"grad_norm": 0.21464337408542633,
"learning_rate": 1.788736734117078e-05,
"loss": 0.0176,
"step": 7360
},
{
"epoch": 26.60649819494585,
"grad_norm": 0.23118972778320312,
"learning_rate": 1.7760805856248152e-05,
"loss": 0.0212,
"step": 7370
},
{
"epoch": 26.642599277978338,
"grad_norm": 0.19908009469509125,
"learning_rate": 1.7634596933987518e-05,
"loss": 0.0247,
"step": 7380
},
{
"epoch": 26.67870036101083,
"grad_norm": 0.21587587893009186,
"learning_rate": 1.7508741954589404e-05,
"loss": 0.0286,
"step": 7390
},
{
"epoch": 26.71480144404332,
"grad_norm": 0.1667470782995224,
"learning_rate": 1.7383242294383717e-05,
"loss": 0.0186,
"step": 7400
},
{
"epoch": 26.750902527075812,
"grad_norm": 0.21209971606731415,
"learning_rate": 1.7258099325814632e-05,
"loss": 0.0199,
"step": 7410
},
{
"epoch": 26.787003610108304,
"grad_norm": 0.19643503427505493,
"learning_rate": 1.7133314417425594e-05,
"loss": 0.0215,
"step": 7420
},
{
"epoch": 26.823104693140795,
"grad_norm": 0.15766191482543945,
"learning_rate": 1.7008888933844408e-05,
"loss": 0.021,
"step": 7430
},
{
"epoch": 26.859205776173287,
"grad_norm": 0.26617512106895447,
"learning_rate": 1.6884824235768172e-05,
"loss": 0.0254,
"step": 7440
},
{
"epoch": 26.895306859205775,
"grad_norm": 0.19353647530078888,
"learning_rate": 1.6761121679948592e-05,
"loss": 0.0197,
"step": 7450
},
{
"epoch": 26.931407942238266,
"grad_norm": 0.19385726749897003,
"learning_rate": 1.663778261917695e-05,
"loss": 0.0199,
"step": 7460
},
{
"epoch": 26.967509025270758,
"grad_norm": 0.17210079729557037,
"learning_rate": 1.651480840226952e-05,
"loss": 0.0247,
"step": 7470
},
{
"epoch": 27.00361010830325,
"grad_norm": 0.2663367986679077,
"learning_rate": 1.639220037405258e-05,
"loss": 0.0206,
"step": 7480
},
{
"epoch": 27.03971119133574,
"grad_norm": 0.2045394480228424,
"learning_rate": 1.6269959875347906e-05,
"loss": 0.0205,
"step": 7490
},
{
"epoch": 27.075812274368232,
"grad_norm": 0.1802651584148407,
"learning_rate": 1.614808824295802e-05,
"loss": 0.021,
"step": 7500
},
{
"epoch": 27.111913357400724,
"grad_norm": 0.1829817295074463,
"learning_rate": 1.602658680965152e-05,
"loss": 0.0159,
"step": 7510
},
{
"epoch": 27.14801444043321,
"grad_norm": 0.2554306089878082,
"learning_rate": 1.5905456904148686e-05,
"loss": 0.0225,
"step": 7520
},
{
"epoch": 27.184115523465703,
"grad_norm": 0.24635805189609528,
"learning_rate": 1.57846998511067e-05,
"loss": 0.0213,
"step": 7530
},
{
"epoch": 27.220216606498195,
"grad_norm": 0.3170551061630249,
"learning_rate": 1.566431697110538e-05,
"loss": 0.0251,
"step": 7540
},
{
"epoch": 27.256317689530686,
"grad_norm": 0.22387059032917023,
"learning_rate": 1.554430958063259e-05,
"loss": 0.018,
"step": 7550
},
{
"epoch": 27.292418772563177,
"grad_norm": 0.26271986961364746,
"learning_rate": 1.5424678992069912e-05,
"loss": 0.0211,
"step": 7560
},
{
"epoch": 27.32851985559567,
"grad_norm": 0.23111337423324585,
"learning_rate": 1.5305426513678362e-05,
"loss": 0.0174,
"step": 7570
},
{
"epoch": 27.36462093862816,
"grad_norm": 0.17635098099708557,
"learning_rate": 1.518655344958388e-05,
"loss": 0.0212,
"step": 7580
},
{
"epoch": 27.40072202166065,
"grad_norm": 0.226985365152359,
"learning_rate": 1.5068061099763275e-05,
"loss": 0.0191,
"step": 7590
},
{
"epoch": 27.43682310469314,
"grad_norm": 0.2475198656320572,
"learning_rate": 1.494995076002988e-05,
"loss": 0.0211,
"step": 7600
},
{
"epoch": 27.47292418772563,
"grad_norm": 0.24423539638519287,
"learning_rate": 1.4832223722019456e-05,
"loss": 0.0242,
"step": 7610
},
{
"epoch": 27.509025270758123,
"grad_norm": 0.20034530758857727,
"learning_rate": 1.4714881273176035e-05,
"loss": 0.0202,
"step": 7620
},
{
"epoch": 27.545126353790614,
"grad_norm": 0.233820378780365,
"learning_rate": 1.4597924696737835e-05,
"loss": 0.0188,
"step": 7630
},
{
"epoch": 27.581227436823106,
"grad_norm": 0.1908191740512848,
"learning_rate": 1.4481355271723252e-05,
"loss": 0.0215,
"step": 7640
},
{
"epoch": 27.617328519855597,
"grad_norm": 0.2116384655237198,
"learning_rate": 1.4365174272916809e-05,
"loss": 0.0244,
"step": 7650
},
{
"epoch": 27.653429602888085,
"grad_norm": 0.2143145352602005,
"learning_rate": 1.4249382970855319e-05,
"loss": 0.0186,
"step": 7660
},
{
"epoch": 27.689530685920577,
"grad_norm": 0.1564183533191681,
"learning_rate": 1.4133982631813903e-05,
"loss": 0.0196,
"step": 7670
},
{
"epoch": 27.72563176895307,
"grad_norm": 0.23016080260276794,
"learning_rate": 1.4018974517792194e-05,
"loss": 0.0162,
"step": 7680
},
{
"epoch": 27.76173285198556,
"grad_norm": 0.23236685991287231,
"learning_rate": 1.390435988650048e-05,
"loss": 0.0173,
"step": 7690
},
{
"epoch": 27.79783393501805,
"grad_norm": 0.1841185986995697,
"learning_rate": 1.3790139991346006e-05,
"loss": 0.0174,
"step": 7700
},
{
"epoch": 27.833935018050543,
"grad_norm": 0.20582100749015808,
"learning_rate": 1.367631608141926e-05,
"loss": 0.0213,
"step": 7710
},
{
"epoch": 27.870036101083034,
"grad_norm": 0.1874004751443863,
"learning_rate": 1.3562889401480278e-05,
"loss": 0.0207,
"step": 7720
},
{
"epoch": 27.906137184115522,
"grad_norm": 0.20177176594734192,
"learning_rate": 1.3449861191945074e-05,
"loss": 0.022,
"step": 7730
},
{
"epoch": 27.942238267148014,
"grad_norm": 0.18946006894111633,
"learning_rate": 1.3337232688872009e-05,
"loss": 0.0176,
"step": 7740
},
{
"epoch": 27.978339350180505,
"grad_norm": 0.2934093773365021,
"learning_rate": 1.3225005123948364e-05,
"loss": 0.023,
"step": 7750
},
{
"epoch": 28.014440433212997,
"grad_norm": 0.19924205541610718,
"learning_rate": 1.311317972447681e-05,
"loss": 0.0189,
"step": 7760
},
{
"epoch": 28.050541516245488,
"grad_norm": 0.22900228202342987,
"learning_rate": 1.3001757713361996e-05,
"loss": 0.0174,
"step": 7770
},
{
"epoch": 28.08664259927798,
"grad_norm": 0.23986896872520447,
"learning_rate": 1.2890740309097204e-05,
"loss": 0.0193,
"step": 7780
},
{
"epoch": 28.12274368231047,
"grad_norm": 0.17657983303070068,
"learning_rate": 1.2780128725750944e-05,
"loss": 0.0225,
"step": 7790
},
{
"epoch": 28.15884476534296,
"grad_norm": 0.23977473378181458,
"learning_rate": 1.266992417295379e-05,
"loss": 0.0206,
"step": 7800
},
{
"epoch": 28.19494584837545,
"grad_norm": 0.1975516378879547,
"learning_rate": 1.2560127855885073e-05,
"loss": 0.0192,
"step": 7810
},
{
"epoch": 28.231046931407942,
"grad_norm": 0.22167274355888367,
"learning_rate": 1.2450740975259745e-05,
"loss": 0.0244,
"step": 7820
},
{
"epoch": 28.267148014440433,
"grad_norm": 0.24081934988498688,
"learning_rate": 1.234176472731517e-05,
"loss": 0.0266,
"step": 7830
},
{
"epoch": 28.303249097472925,
"grad_norm": 0.32553035020828247,
"learning_rate": 1.2233200303798158e-05,
"loss": 0.0192,
"step": 7840
},
{
"epoch": 28.339350180505416,
"grad_norm": 0.24330414831638336,
"learning_rate": 1.2125048891951846e-05,
"loss": 0.0203,
"step": 7850
},
{
"epoch": 28.375451263537904,
"grad_norm": 0.15358476340770721,
"learning_rate": 1.2017311674502745e-05,
"loss": 0.0165,
"step": 7860
},
{
"epoch": 28.411552346570396,
"grad_norm": 0.20697100460529327,
"learning_rate": 1.1909989829647822e-05,
"loss": 0.0182,
"step": 7870
},
{
"epoch": 28.447653429602887,
"grad_norm": 0.21491262316703796,
"learning_rate": 1.1803084531041553e-05,
"loss": 0.0232,
"step": 7880
},
{
"epoch": 28.48375451263538,
"grad_norm": 0.2125055491924286,
"learning_rate": 1.1696596947783162e-05,
"loss": 0.0216,
"step": 7890
},
{
"epoch": 28.51985559566787,
"grad_norm": 0.2420768290758133,
"learning_rate": 1.1590528244403803e-05,
"loss": 0.0206,
"step": 7900
},
{
"epoch": 28.555956678700362,
"grad_norm": 0.18537470698356628,
"learning_rate": 1.148487958085382e-05,
"loss": 0.022,
"step": 7910
},
{
"epoch": 28.592057761732853,
"grad_norm": 0.17309150099754333,
"learning_rate": 1.1379652112490086e-05,
"loss": 0.0219,
"step": 7920
},
{
"epoch": 28.628158844765345,
"grad_norm": 0.14592359960079193,
"learning_rate": 1.1274846990063315e-05,
"loss": 0.0187,
"step": 7930
},
{
"epoch": 28.664259927797833,
"grad_norm": 0.2157330960035324,
"learning_rate": 1.117046535970554e-05,
"loss": 0.0221,
"step": 7940
},
{
"epoch": 28.700361010830324,
"grad_norm": 0.19493581354618073,
"learning_rate": 1.106650836291755e-05,
"loss": 0.0273,
"step": 7950
},
{
"epoch": 28.736462093862816,
"grad_norm": 0.18499520421028137,
"learning_rate": 1.0962977136556418e-05,
"loss": 0.0278,
"step": 7960
},
{
"epoch": 28.772563176895307,
"grad_norm": 0.18380481004714966,
"learning_rate": 1.0859872812823024e-05,
"loss": 0.022,
"step": 7970
},
{
"epoch": 28.8086642599278,
"grad_norm": 0.2502525746822357,
"learning_rate": 1.0757196519249747e-05,
"loss": 0.022,
"step": 7980
},
{
"epoch": 28.84476534296029,
"grad_norm": 0.19856320321559906,
"learning_rate": 1.0654949378688077e-05,
"loss": 0.0186,
"step": 7990
},
{
"epoch": 28.880866425992778,
"grad_norm": 0.31073760986328125,
"learning_rate": 1.0553132509296376e-05,
"loss": 0.0218,
"step": 8000
},
{
"epoch": 28.91696750902527,
"grad_norm": 0.1758049577474594,
"learning_rate": 1.0451747024527613e-05,
"loss": 0.0189,
"step": 8010
},
{
"epoch": 28.95306859205776,
"grad_norm": 0.16876867413520813,
"learning_rate": 1.0350794033117189e-05,
"loss": 0.0238,
"step": 8020
},
{
"epoch": 28.989169675090253,
"grad_norm": 0.30022498965263367,
"learning_rate": 1.0250274639070856e-05,
"loss": 0.0206,
"step": 8030
},
{
"epoch": 29.025270758122744,
"grad_norm": 0.12251915037631989,
"learning_rate": 1.0150189941652599e-05,
"loss": 0.0173,
"step": 8040
},
{
"epoch": 29.061371841155236,
"grad_norm": 0.21832774579524994,
"learning_rate": 1.0050541035372635e-05,
"loss": 0.0157,
"step": 8050
},
{
"epoch": 29.097472924187727,
"grad_norm": 0.22191844880580902,
"learning_rate": 9.951329009975458e-06,
"loss": 0.0191,
"step": 8060
},
{
"epoch": 29.133574007220215,
"grad_norm": 0.1808042973279953,
"learning_rate": 9.852554950427845e-06,
"loss": 0.0228,
"step": 8070
},
{
"epoch": 29.169675090252706,
"grad_norm": 0.1901533305644989,
"learning_rate": 9.754219936907105e-06,
"loss": 0.0196,
"step": 8080
},
{
"epoch": 29.205776173285198,
"grad_norm": 0.29592928290367126,
"learning_rate": 9.656325044789194e-06,
"loss": 0.0154,
"step": 8090
},
{
"epoch": 29.24187725631769,
"grad_norm": 0.21368078887462616,
"learning_rate": 9.55887134463697e-06,
"loss": 0.0159,
"step": 8100
},
{
"epoch": 29.27797833935018,
"grad_norm": 0.2306506484746933,
"learning_rate": 9.461859902188475e-06,
"loss": 0.0214,
"step": 8110
},
{
"epoch": 29.314079422382672,
"grad_norm": 0.1923961043357849,
"learning_rate": 9.365291778345303e-06,
"loss": 0.0186,
"step": 8120
},
{
"epoch": 29.350180505415164,
"grad_norm": 0.184268981218338,
"learning_rate": 9.269168029160991e-06,
"loss": 0.0175,
"step": 8130
},
{
"epoch": 29.386281588447652,
"grad_norm": 0.28840163350105286,
"learning_rate": 9.173489705829447e-06,
"loss": 0.019,
"step": 8140
},
{
"epoch": 29.422382671480143,
"grad_norm": 0.1663977950811386,
"learning_rate": 9.078257854673516e-06,
"loss": 0.022,
"step": 8150
},
{
"epoch": 29.458483754512635,
"grad_norm": 0.19515427947044373,
"learning_rate": 8.983473517133429e-06,
"loss": 0.0221,
"step": 8160
},
{
"epoch": 29.494584837545126,
"grad_norm": 0.1358731985092163,
"learning_rate": 8.889137729755537e-06,
"loss": 0.0174,
"step": 8170
},
{
"epoch": 29.530685920577618,
"grad_norm": 0.1901516616344452,
"learning_rate": 8.79525152418087e-06,
"loss": 0.0186,
"step": 8180
},
{
"epoch": 29.56678700361011,
"grad_norm": 0.18884000182151794,
"learning_rate": 8.701815927133961e-06,
"loss": 0.0193,
"step": 8190
},
{
"epoch": 29.6028880866426,
"grad_norm": 0.19345715641975403,
"learning_rate": 8.608831960411534e-06,
"loss": 0.0199,
"step": 8200
},
{
"epoch": 29.63898916967509,
"grad_norm": 0.19965921342372894,
"learning_rate": 8.516300640871321e-06,
"loss": 0.0192,
"step": 8210
},
{
"epoch": 29.67509025270758,
"grad_norm": 0.16329129040241241,
"learning_rate": 8.424222980421038e-06,
"loss": 0.0196,
"step": 8220
},
{
"epoch": 29.71119133574007,
"grad_norm": 0.23898659646511078,
"learning_rate": 8.332599986007184e-06,
"loss": 0.0199,
"step": 8230
},
{
"epoch": 29.747292418772563,
"grad_norm": 0.15860313177108765,
"learning_rate": 8.241432659604203e-06,
"loss": 0.0157,
"step": 8240
},
{
"epoch": 29.783393501805055,
"grad_norm": 0.14782391488552094,
"learning_rate": 8.150721998203331e-06,
"loss": 0.0164,
"step": 8250
},
{
"epoch": 29.819494584837546,
"grad_norm": 0.26512598991394043,
"learning_rate": 8.06046899380184e-06,
"loss": 0.0185,
"step": 8260
},
{
"epoch": 29.855595667870038,
"grad_norm": 0.2540733814239502,
"learning_rate": 7.970674633392133e-06,
"loss": 0.0233,
"step": 8270
},
{
"epoch": 29.891696750902526,
"grad_norm": 0.20593500137329102,
"learning_rate": 7.881339898950924e-06,
"loss": 0.0205,
"step": 8280
},
{
"epoch": 29.927797833935017,
"grad_norm": 0.13345380127429962,
"learning_rate": 7.792465767428597e-06,
"loss": 0.0204,
"step": 8290
},
{
"epoch": 29.96389891696751,
"grad_norm": 0.1564672291278839,
"learning_rate": 7.704053210738376e-06,
"loss": 0.0163,
"step": 8300
},
{
"epoch": 30.0,
"grad_norm": 0.2338448166847229,
"learning_rate": 7.6161031957458494e-06,
"loss": 0.0209,
"step": 8310
},
{
"epoch": 30.03610108303249,
"grad_norm": 0.27858781814575195,
"learning_rate": 7.5286166842582605e-06,
"loss": 0.0189,
"step": 8320
},
{
"epoch": 30.072202166064983,
"grad_norm": 0.2322964072227478,
"learning_rate": 7.4415946330140814e-06,
"loss": 0.0202,
"step": 8330
},
{
"epoch": 30.108303249097474,
"grad_norm": 0.2365698665380478,
"learning_rate": 7.3550379936725644e-06,
"loss": 0.0183,
"step": 8340
},
{
"epoch": 30.144404332129962,
"grad_norm": 0.2730034589767456,
"learning_rate": 7.2689477128032035e-06,
"loss": 0.0168,
"step": 8350
},
{
"epoch": 30.180505415162454,
"grad_norm": 0.2949770390987396,
"learning_rate": 7.183324731875551e-06,
"loss": 0.0189,
"step": 8360
},
{
"epoch": 30.216606498194945,
"grad_norm": 0.13929913938045502,
"learning_rate": 7.098169987248782e-06,
"loss": 0.0127,
"step": 8370
},
{
"epoch": 30.252707581227437,
"grad_norm": 0.2416556179523468,
"learning_rate": 7.013484410161553e-06,
"loss": 0.0174,
"step": 8380
},
{
"epoch": 30.28880866425993,
"grad_norm": 0.20826245844364166,
"learning_rate": 6.92926892672176e-06,
"loss": 0.0179,
"step": 8390
},
{
"epoch": 30.32490974729242,
"grad_norm": 0.2940756380558014,
"learning_rate": 6.845524457896446e-06,
"loss": 0.0199,
"step": 8400
},
{
"epoch": 30.36101083032491,
"grad_norm": 0.1727730631828308,
"learning_rate": 6.7622519195017165e-06,
"loss": 0.0147,
"step": 8410
},
{
"epoch": 30.3971119133574,
"grad_norm": 0.17099951207637787,
"learning_rate": 6.679452222192684e-06,
"loss": 0.0208,
"step": 8420
},
{
"epoch": 30.43321299638989,
"grad_norm": 0.17958728969097137,
"learning_rate": 6.597126271453579e-06,
"loss": 0.0173,
"step": 8430
},
{
"epoch": 30.469314079422382,
"grad_norm": 0.17829561233520508,
"learning_rate": 6.51527496758782e-06,
"loss": 0.0168,
"step": 8440
},
{
"epoch": 30.505415162454874,
"grad_norm": 0.1982552409172058,
"learning_rate": 6.433899205708155e-06,
"loss": 0.0154,
"step": 8450
},
{
"epoch": 30.541516245487365,
"grad_norm": 0.26757529377937317,
"learning_rate": 6.352999875726856e-06,
"loss": 0.0169,
"step": 8460
},
{
"epoch": 30.577617328519857,
"grad_norm": 0.15877163410186768,
"learning_rate": 6.272577862346052e-06,
"loss": 0.0144,
"step": 8470
},
{
"epoch": 30.613718411552348,
"grad_norm": 0.26375260949134827,
"learning_rate": 6.192634045047996e-06,
"loss": 0.0179,
"step": 8480
},
{
"epoch": 30.649819494584836,
"grad_norm": 0.2675758898258209,
"learning_rate": 6.113169298085458e-06,
"loss": 0.0249,
"step": 8490
},
{
"epoch": 30.685920577617328,
"grad_norm": 0.17940880358219147,
"learning_rate": 6.034184490472195e-06,
"loss": 0.0157,
"step": 8500
},
{
"epoch": 30.72202166064982,
"grad_norm": 0.22204822301864624,
"learning_rate": 5.955680485973386e-06,
"loss": 0.0198,
"step": 8510
},
{
"epoch": 30.75812274368231,
"grad_norm": 0.1977151781320572,
"learning_rate": 5.877658143096265e-06,
"loss": 0.018,
"step": 8520
},
{
"epoch": 30.794223826714802,
"grad_norm": 0.17285385727882385,
"learning_rate": 5.800118315080661e-06,
"loss": 0.0175,
"step": 8530
},
{
"epoch": 30.830324909747294,
"grad_norm": 0.18532536923885345,
"learning_rate": 5.723061849889716e-06,
"loss": 0.0147,
"step": 8540
},
{
"epoch": 30.866425992779785,
"grad_norm": 0.24875099956989288,
"learning_rate": 5.646489590200604e-06,
"loss": 0.0187,
"step": 8550
},
{
"epoch": 30.902527075812273,
"grad_norm": 0.23656758666038513,
"learning_rate": 5.570402373395256e-06,
"loss": 0.0186,
"step": 8560
},
{
"epoch": 30.938628158844764,
"grad_norm": 0.26919370889663696,
"learning_rate": 5.494801031551305e-06,
"loss": 0.0172,
"step": 8570
},
{
"epoch": 30.974729241877256,
"grad_norm": 0.20205961167812347,
"learning_rate": 5.41968639143291e-06,
"loss": 0.0184,
"step": 8580
},
{
"epoch": 31.010830324909747,
"grad_norm": 0.251251220703125,
"learning_rate": 5.345059274481751e-06,
"loss": 0.0169,
"step": 8590
},
{
"epoch": 31.04693140794224,
"grad_norm": 0.2158970832824707,
"learning_rate": 5.270920496808002e-06,
"loss": 0.0196,
"step": 8600
},
{
"epoch": 31.08303249097473,
"grad_norm": 0.26518523693084717,
"learning_rate": 5.1972708691814695e-06,
"loss": 0.0173,
"step": 8610
},
{
"epoch": 31.119133574007222,
"grad_norm": 0.21849223971366882,
"learning_rate": 5.124111197022674e-06,
"loss": 0.0209,
"step": 8620
},
{
"epoch": 31.15523465703971,
"grad_norm": 0.16551697254180908,
"learning_rate": 5.051442280394081e-06,
"loss": 0.0153,
"step": 8630
},
{
"epoch": 31.1913357400722,
"grad_norm": 0.1864350438117981,
"learning_rate": 4.979264913991322e-06,
"loss": 0.0152,
"step": 8640
},
{
"epoch": 31.227436823104693,
"grad_norm": 0.13301555812358856,
"learning_rate": 4.907579887134489e-06,
"loss": 0.0173,
"step": 8650
},
{
"epoch": 31.263537906137184,
"grad_norm": 0.13425064086914062,
"learning_rate": 4.836387983759572e-06,
"loss": 0.0184,
"step": 8660
},
{
"epoch": 31.299638989169676,
"grad_norm": 0.1871993988752365,
"learning_rate": 4.765689982409816e-06,
"loss": 0.0165,
"step": 8670
},
{
"epoch": 31.335740072202167,
"grad_norm": 0.10523172467947006,
"learning_rate": 4.695486656227233e-06,
"loss": 0.0167,
"step": 8680
},
{
"epoch": 31.37184115523466,
"grad_norm": 0.19608139991760254,
"learning_rate": 4.625778772944156e-06,
"loss": 0.0185,
"step": 8690
},
{
"epoch": 31.407942238267147,
"grad_norm": 0.137054905295372,
"learning_rate": 4.556567094874825e-06,
"loss": 0.0203,
"step": 8700
},
{
"epoch": 31.444043321299638,
"grad_norm": 0.21377761662006378,
"learning_rate": 4.487852378907059e-06,
"loss": 0.0155,
"step": 8710
},
{
"epoch": 31.48014440433213,
"grad_norm": 0.17693573236465454,
"learning_rate": 4.419635376493986e-06,
"loss": 0.0187,
"step": 8720
},
{
"epoch": 31.51624548736462,
"grad_norm": 0.22736315429210663,
"learning_rate": 4.351916833645825e-06,
"loss": 0.0172,
"step": 8730
},
{
"epoch": 31.552346570397113,
"grad_norm": 0.12939462065696716,
"learning_rate": 4.284697490921691e-06,
"loss": 0.0157,
"step": 8740
},
{
"epoch": 31.588447653429604,
"grad_norm": 0.12454534322023392,
"learning_rate": 4.2179780834215585e-06,
"loss": 0.0185,
"step": 8750
},
{
"epoch": 31.624548736462096,
"grad_norm": 0.1691960096359253,
"learning_rate": 4.151759340778178e-06,
"loss": 0.018,
"step": 8760
},
{
"epoch": 31.660649819494584,
"grad_norm": 0.18643519282341003,
"learning_rate": 4.086041987149109e-06,
"loss": 0.0172,
"step": 8770
},
{
"epoch": 31.696750902527075,
"grad_norm": 0.19169756770133972,
"learning_rate": 4.020826741208811e-06,
"loss": 0.0171,
"step": 8780
},
{
"epoch": 31.732851985559567,
"grad_norm": 0.1198309063911438,
"learning_rate": 3.956114316140746e-06,
"loss": 0.02,
"step": 8790
},
{
"epoch": 31.768953068592058,
"grad_norm": 0.22697773575782776,
"learning_rate": 3.891905419629643e-06,
"loss": 0.0184,
"step": 8800
},
{
"epoch": 31.80505415162455,
"grad_norm": 0.16717450320720673,
"learning_rate": 3.8282007538536946e-06,
"loss": 0.0202,
"step": 8810
},
{
"epoch": 31.84115523465704,
"grad_norm": 0.3170683979988098,
"learning_rate": 3.7650010154769265e-06,
"loss": 0.0142,
"step": 8820
},
{
"epoch": 31.87725631768953,
"grad_norm": 0.13118068873882294,
"learning_rate": 3.7023068956415608e-06,
"loss": 0.0193,
"step": 8830
},
{
"epoch": 31.91335740072202,
"grad_norm": 0.18915139138698578,
"learning_rate": 3.6401190799604303e-06,
"loss": 0.0149,
"step": 8840
},
{
"epoch": 31.949458483754512,
"grad_norm": 0.20172129571437836,
"learning_rate": 3.578438248509536e-06,
"loss": 0.0178,
"step": 8850
},
{
"epoch": 31.985559566787003,
"grad_norm": 0.188405379652977,
"learning_rate": 3.5172650758205583e-06,
"loss": 0.02,
"step": 8860
},
{
"epoch": 32.021660649819495,
"grad_norm": 0.15788529813289642,
"learning_rate": 3.45660023087353e-06,
"loss": 0.0147,
"step": 8870
},
{
"epoch": 32.057761732851986,
"grad_norm": 0.15065881609916687,
"learning_rate": 3.3964443770894528e-06,
"loss": 0.0146,
"step": 8880
},
{
"epoch": 32.09386281588448,
"grad_norm": 0.18288753926753998,
"learning_rate": 3.3367981723231245e-06,
"loss": 0.0172,
"step": 8890
},
{
"epoch": 32.12996389891697,
"grad_norm": 0.17745760083198547,
"learning_rate": 3.2776622688558746e-06,
"loss": 0.014,
"step": 8900
},
{
"epoch": 32.16606498194946,
"grad_norm": 0.17741569876670837,
"learning_rate": 3.2190373133884677e-06,
"loss": 0.0171,
"step": 8910
},
{
"epoch": 32.20216606498195,
"grad_norm": 0.15399572253227234,
"learning_rate": 3.1609239470340446e-06,
"loss": 0.0148,
"step": 8920
},
{
"epoch": 32.238267148014444,
"grad_norm": 0.1344560980796814,
"learning_rate": 3.1033228053110373e-06,
"loss": 0.0164,
"step": 8930
},
{
"epoch": 32.27436823104693,
"grad_norm": 0.263134628534317,
"learning_rate": 3.0462345181363314e-06,
"loss": 0.0159,
"step": 8940
},
{
"epoch": 32.31046931407942,
"grad_norm": 0.17395150661468506,
"learning_rate": 2.9896597098182654e-06,
"loss": 0.0181,
"step": 8950
},
{
"epoch": 32.34657039711191,
"grad_norm": 0.23743000626564026,
"learning_rate": 2.933598999049891e-06,
"loss": 0.0158,
"step": 8960
},
{
"epoch": 32.3826714801444,
"grad_norm": 0.1234792023897171,
"learning_rate": 2.8780529989021697e-06,
"loss": 0.0155,
"step": 8970
},
{
"epoch": 32.418772563176894,
"grad_norm": 0.1488943099975586,
"learning_rate": 2.823022316817242e-06,
"loss": 0.0191,
"step": 8980
},
{
"epoch": 32.454873646209386,
"grad_norm": 0.20114587247371674,
"learning_rate": 2.7685075546018456e-06,
"loss": 0.0207,
"step": 8990
},
{
"epoch": 32.49097472924188,
"grad_norm": 0.15291468799114227,
"learning_rate": 2.7145093084206598e-06,
"loss": 0.0214,
"step": 9000
},
{
"epoch": 32.52707581227437,
"grad_norm": 0.16634592413902283,
"learning_rate": 2.661028168789892e-06,
"loss": 0.0166,
"step": 9010
},
{
"epoch": 32.56317689530686,
"grad_norm": 0.13826407492160797,
"learning_rate": 2.6080647205706855e-06,
"loss": 0.0133,
"step": 9020
},
{
"epoch": 32.59927797833935,
"grad_norm": 0.17645834386348724,
"learning_rate": 2.555619542962834e-06,
"loss": 0.0181,
"step": 9030
},
{
"epoch": 32.63537906137184,
"grad_norm": 0.16048942506313324,
"learning_rate": 2.503693209498409e-06,
"loss": 0.0139,
"step": 9040
},
{
"epoch": 32.671480144404335,
"grad_norm": 0.19027777016162872,
"learning_rate": 2.452286288035449e-06,
"loss": 0.0156,
"step": 9050
},
{
"epoch": 32.707581227436826,
"grad_norm": 0.14160841703414917,
"learning_rate": 2.4013993407518363e-06,
"loss": 0.0176,
"step": 9060
},
{
"epoch": 32.74368231046932,
"grad_norm": 0.11572880297899246,
"learning_rate": 2.351032924139063e-06,
"loss": 0.0174,
"step": 9070
},
{
"epoch": 32.7797833935018,
"grad_norm": 0.18979032337665558,
"learning_rate": 2.30118758899619e-06,
"loss": 0.0191,
"step": 9080
},
{
"epoch": 32.81588447653429,
"grad_norm": 0.18322326242923737,
"learning_rate": 2.2518638804238157e-06,
"loss": 0.0172,
"step": 9090
},
{
"epoch": 32.851985559566785,
"grad_norm": 0.18762503564357758,
"learning_rate": 2.203062337818118e-06,
"loss": 0.018,
"step": 9100
},
{
"epoch": 32.888086642599276,
"grad_norm": 0.1830272078514099,
"learning_rate": 2.1547834948649483e-06,
"loss": 0.0161,
"step": 9110
},
{
"epoch": 32.92418772563177,
"grad_norm": 0.23481222987174988,
"learning_rate": 2.1070278795340017e-06,
"loss": 0.0173,
"step": 9120
},
{
"epoch": 32.96028880866426,
"grad_norm": 0.18570080399513245,
"learning_rate": 2.059796014073029e-06,
"loss": 0.0145,
"step": 9130
},
{
"epoch": 32.99638989169675,
"grad_norm": 0.15050573647022247,
"learning_rate": 2.01308841500214e-06,
"loss": 0.0144,
"step": 9140
},
{
"epoch": 33.03249097472924,
"grad_norm": 0.21648050844669342,
"learning_rate": 1.9669055931081704e-06,
"loss": 0.0179,
"step": 9150
},
{
"epoch": 33.068592057761734,
"grad_norm": 0.15288197994232178,
"learning_rate": 1.9212480534390507e-06,
"loss": 0.0174,
"step": 9160
},
{
"epoch": 33.104693140794225,
"grad_norm": 0.17982974648475647,
"learning_rate": 1.8761162952983246e-06,
"loss": 0.0146,
"step": 9170
},
{
"epoch": 33.14079422382672,
"grad_norm": 0.19630658626556396,
"learning_rate": 1.8315108122396618e-06,
"loss": 0.0194,
"step": 9180
},
{
"epoch": 33.17689530685921,
"grad_norm": 0.13119301199913025,
"learning_rate": 1.787432092061475e-06,
"loss": 0.0156,
"step": 9190
},
{
"epoch": 33.2129963898917,
"grad_norm": 0.17562414705753326,
"learning_rate": 1.743880616801602e-06,
"loss": 0.0163,
"step": 9200
},
{
"epoch": 33.24909747292419,
"grad_norm": 0.21962830424308777,
"learning_rate": 1.7008568627319865e-06,
"loss": 0.0191,
"step": 9210
},
{
"epoch": 33.285198555956676,
"grad_norm": 0.20217278599739075,
"learning_rate": 1.6583613003535226e-06,
"loss": 0.0164,
"step": 9220
},
{
"epoch": 33.32129963898917,
"grad_norm": 0.1368408203125,
"learning_rate": 1.6163943943908522e-06,
"loss": 0.0193,
"step": 9230
},
{
"epoch": 33.35740072202166,
"grad_norm": 0.1425715684890747,
"learning_rate": 1.5749566037873476e-06,
"loss": 0.0166,
"step": 9240
},
{
"epoch": 33.39350180505415,
"grad_norm": 0.17141273617744446,
"learning_rate": 1.5340483817000428e-06,
"loss": 0.0197,
"step": 9250
},
{
"epoch": 33.42960288808664,
"grad_norm": 0.1339203268289566,
"learning_rate": 1.4936701754947101e-06,
"loss": 0.0181,
"step": 9260
},
{
"epoch": 33.46570397111913,
"grad_norm": 0.1516958326101303,
"learning_rate": 1.4538224267409361e-06,
"loss": 0.0183,
"step": 9270
},
{
"epoch": 33.501805054151625,
"grad_norm": 0.17948590219020844,
"learning_rate": 1.414505571207314e-06,
"loss": 0.0174,
"step": 9280
},
{
"epoch": 33.537906137184116,
"grad_norm": 0.0888703390955925,
"learning_rate": 1.3757200388566816e-06,
"loss": 0.0132,
"step": 9290
},
{
"epoch": 33.57400722021661,
"grad_norm": 0.2608892619609833,
"learning_rate": 1.3374662538414074e-06,
"loss": 0.0174,
"step": 9300
},
{
"epoch": 33.6101083032491,
"grad_norm": 0.1301056444644928,
"learning_rate": 1.2997446344987617e-06,
"loss": 0.0156,
"step": 9310
},
{
"epoch": 33.64620938628159,
"grad_norm": 0.1602219045162201,
"learning_rate": 1.262555593346315e-06,
"loss": 0.0153,
"step": 9320
},
{
"epoch": 33.68231046931408,
"grad_norm": 0.17319612205028534,
"learning_rate": 1.2258995370774685e-06,
"loss": 0.0172,
"step": 9330
},
{
"epoch": 33.71841155234657,
"grad_norm": 0.17861145734786987,
"learning_rate": 1.1897768665569798e-06,
"loss": 0.0182,
"step": 9340
},
{
"epoch": 33.75451263537906,
"grad_norm": 0.13086837530136108,
"learning_rate": 1.1541879768165954e-06,
"loss": 0.0141,
"step": 9350
},
{
"epoch": 33.79061371841155,
"grad_norm": 0.15529681742191315,
"learning_rate": 1.1191332570507085e-06,
"loss": 0.0145,
"step": 9360
},
{
"epoch": 33.82671480144404,
"grad_norm": 0.16968750953674316,
"learning_rate": 1.0846130906121132e-06,
"loss": 0.0232,
"step": 9370
},
{
"epoch": 33.86281588447653,
"grad_norm": 0.19526614248752594,
"learning_rate": 1.0506278550078131e-06,
"loss": 0.0178,
"step": 9380
},
{
"epoch": 33.898916967509024,
"grad_norm": 0.2040262073278427,
"learning_rate": 1.0171779218949185e-06,
"loss": 0.019,
"step": 9390
},
{
"epoch": 33.935018050541515,
"grad_norm": 0.20133711397647858,
"learning_rate": 9.842636570765174e-07,
"loss": 0.0163,
"step": 9400
},
{
"epoch": 33.97111913357401,
"grad_norm": 0.14701826870441437,
"learning_rate": 9.518854204977612e-07,
"loss": 0.0179,
"step": 9410
},
{
"epoch": 34.0072202166065,
"grad_norm": 0.18041811883449554,
"learning_rate": 9.200435662418349e-07,
"loss": 0.017,
"step": 9420
},
{
"epoch": 34.04332129963899,
"grad_norm": 0.22241787612438202,
"learning_rate": 8.887384425261658e-07,
"loss": 0.0168,
"step": 9430
},
{
"epoch": 34.07942238267148,
"grad_norm": 0.20928716659545898,
"learning_rate": 8.579703916985648e-07,
"loss": 0.0139,
"step": 9440
},
{
"epoch": 34.11552346570397,
"grad_norm": 0.19215941429138184,
"learning_rate": 8.277397502335194e-07,
"loss": 0.0209,
"step": 9450
},
{
"epoch": 34.151624548736464,
"grad_norm": 0.19478574395179749,
"learning_rate": 7.980468487284675e-07,
"loss": 0.0146,
"step": 9460
},
{
"epoch": 34.187725631768956,
"grad_norm": 0.12717410922050476,
"learning_rate": 7.688920119002297e-07,
"loss": 0.0161,
"step": 9470
},
{
"epoch": 34.22382671480145,
"grad_norm": 0.24346430599689484,
"learning_rate": 7.402755585814269e-07,
"loss": 0.0153,
"step": 9480
},
{
"epoch": 34.25992779783393,
"grad_norm": 0.18664444983005524,
"learning_rate": 7.121978017170073e-07,
"loss": 0.0168,
"step": 9490
},
{
"epoch": 34.29602888086642,
"grad_norm": 0.18650779128074646,
"learning_rate": 6.846590483608306e-07,
"loss": 0.0166,
"step": 9500
},
{
"epoch": 34.332129963898915,
"grad_norm": 0.13057780265808105,
"learning_rate": 6.576595996722834e-07,
"loss": 0.0216,
"step": 9510
},
{
"epoch": 34.368231046931406,
"grad_norm": 0.11951325833797455,
"learning_rate": 6.311997509130141e-07,
"loss": 0.0182,
"step": 9520
},
{
"epoch": 34.4043321299639,
"grad_norm": 0.24225543439388275,
"learning_rate": 6.052797914436803e-07,
"loss": 0.0161,
"step": 9530
},
{
"epoch": 34.44043321299639,
"grad_norm": 0.15939517319202423,
"learning_rate": 5.799000047208181e-07,
"loss": 0.0192,
"step": 9540
},
{
"epoch": 34.47653429602888,
"grad_norm": 0.23033608496189117,
"learning_rate": 5.550606682937054e-07,
"loss": 0.0192,
"step": 9550
},
{
"epoch": 34.51263537906137,
"grad_norm": 0.24206401407718658,
"learning_rate": 5.307620538013481e-07,
"loss": 0.0175,
"step": 9560
},
{
"epoch": 34.54873646209386,
"grad_norm": 0.13644522428512573,
"learning_rate": 5.070044269694874e-07,
"loss": 0.0171,
"step": 9570
},
{
"epoch": 34.584837545126355,
"grad_norm": 0.25693774223327637,
"learning_rate": 4.837880476077417e-07,
"loss": 0.0188,
"step": 9580
},
{
"epoch": 34.62093862815885,
"grad_norm": 0.17360013723373413,
"learning_rate": 4.6111316960670835e-07,
"loss": 0.0195,
"step": 9590
},
{
"epoch": 34.65703971119134,
"grad_norm": 0.14875541627407074,
"learning_rate": 4.389800409352218e-07,
"loss": 0.0153,
"step": 9600
},
{
"epoch": 34.69314079422383,
"grad_norm": 0.1965004801750183,
"learning_rate": 4.173889036376277e-07,
"loss": 0.0201,
"step": 9610
},
{
"epoch": 34.72924187725632,
"grad_norm": 0.20799556374549866,
"learning_rate": 3.963399938311463e-07,
"loss": 0.0191,
"step": 9620
},
{
"epoch": 34.765342960288805,
"grad_norm": 0.14635834097862244,
"learning_rate": 3.7583354170328545e-07,
"loss": 0.0222,
"step": 9630
},
{
"epoch": 34.8014440433213,
"grad_norm": 0.19555388391017914,
"learning_rate": 3.558697715093207e-07,
"loss": 0.0136,
"step": 9640
},
{
"epoch": 34.83754512635379,
"grad_norm": 0.1937207281589508,
"learning_rate": 3.3644890156983576e-07,
"loss": 0.0168,
"step": 9650
},
{
"epoch": 34.87364620938628,
"grad_norm": 0.10331624001264572,
"learning_rate": 3.175711442683638e-07,
"loss": 0.0165,
"step": 9660
},
{
"epoch": 34.90974729241877,
"grad_norm": 0.20205675065517426,
"learning_rate": 2.9923670604902197e-07,
"loss": 0.0171,
"step": 9670
},
{
"epoch": 34.94584837545126,
"grad_norm": 0.1662183403968811,
"learning_rate": 2.814457874143028e-07,
"loss": 0.0202,
"step": 9680
},
{
"epoch": 34.981949458483754,
"grad_norm": 0.11857236176729202,
"learning_rate": 2.641985829228366e-07,
"loss": 0.0147,
"step": 9690
},
{
"epoch": 35.018050541516246,
"grad_norm": 0.11744869500398636,
"learning_rate": 2.474952811872877e-07,
"loss": 0.0158,
"step": 9700
},
{
"epoch": 35.05415162454874,
"grad_norm": 0.16060422360897064,
"learning_rate": 2.3133606487228397e-07,
"loss": 0.0152,
"step": 9710
},
{
"epoch": 35.09025270758123,
"grad_norm": 0.10732783377170563,
"learning_rate": 2.157211106924295e-07,
"loss": 0.0168,
"step": 9720
},
{
"epoch": 35.12635379061372,
"grad_norm": 0.19921289384365082,
"learning_rate": 2.006505894103672e-07,
"loss": 0.0146,
"step": 9730
},
{
"epoch": 35.16245487364621,
"grad_norm": 0.16379214823246002,
"learning_rate": 1.8612466583489696e-07,
"loss": 0.0177,
"step": 9740
},
{
"epoch": 35.1985559566787,
"grad_norm": 0.20076480507850647,
"learning_rate": 1.7214349881918834e-07,
"loss": 0.0183,
"step": 9750
},
{
"epoch": 35.234657039711195,
"grad_norm": 0.14486785233020782,
"learning_rate": 1.5870724125904845e-07,
"loss": 0.014,
"step": 9760
},
{
"epoch": 35.27075812274368,
"grad_norm": 0.1228007823228836,
"learning_rate": 1.4581604009124006e-07,
"loss": 0.0164,
"step": 9770
},
{
"epoch": 35.30685920577617,
"grad_norm": 0.18470965325832367,
"learning_rate": 1.334700362918717e-07,
"loss": 0.0136,
"step": 9780
},
{
"epoch": 35.34296028880866,
"grad_norm": 0.2023715376853943,
"learning_rate": 1.2166936487486015e-07,
"loss": 0.0131,
"step": 9790
},
{
"epoch": 35.37906137184115,
"grad_norm": 0.1507968306541443,
"learning_rate": 1.1041415489045914e-07,
"loss": 0.0183,
"step": 9800
},
{
"epoch": 35.415162454873645,
"grad_norm": 0.16795922815799713,
"learning_rate": 9.970452942384412e-08,
"loss": 0.013,
"step": 9810
},
{
"epoch": 35.45126353790614,
"grad_norm": 0.1879134327173233,
"learning_rate": 8.954060559375754e-08,
"loss": 0.0126,
"step": 9820
},
{
"epoch": 35.48736462093863,
"grad_norm": 0.1272624135017395,
"learning_rate": 7.99224945512489e-08,
"loss": 0.0191,
"step": 9830
},
{
"epoch": 35.52346570397112,
"grad_norm": 0.23458939790725708,
"learning_rate": 7.085030147843675e-08,
"loss": 0.0188,
"step": 9840
},
{
"epoch": 35.55956678700361,
"grad_norm": 0.31009164452552795,
"learning_rate": 6.232412558736523e-08,
"loss": 0.0187,
"step": 9850
},
{
"epoch": 35.5956678700361,
"grad_norm": 0.19827063381671906,
"learning_rate": 5.434406011893822e-08,
"loss": 0.0195,
"step": 9860
},
{
"epoch": 35.631768953068594,
"grad_norm": 0.1300552785396576,
"learning_rate": 4.6910192341864664e-08,
"loss": 0.0152,
"step": 9870
},
{
"epoch": 35.667870036101085,
"grad_norm": 0.11563283205032349,
"learning_rate": 4.0022603551737035e-08,
"loss": 0.0146,
"step": 9880
},
{
"epoch": 35.70397111913358,
"grad_norm": 0.24187107384204865,
"learning_rate": 3.3681369070120985e-08,
"loss": 0.0187,
"step": 9890
},
{
"epoch": 35.74007220216607,
"grad_norm": 0.18035714328289032,
"learning_rate": 2.7886558243744866e-08,
"loss": 0.0211,
"step": 9900
},
{
"epoch": 35.77617328519855,
"grad_norm": 0.21188323199748993,
"learning_rate": 2.2638234443722596e-08,
"loss": 0.015,
"step": 9910
},
{
"epoch": 35.812274368231044,
"grad_norm": 0.21406260132789612,
"learning_rate": 1.7936455064887504e-08,
"loss": 0.0206,
"step": 9920
},
{
"epoch": 35.848375451263536,
"grad_norm": 0.1407003402709961,
"learning_rate": 1.378127152514841e-08,
"loss": 0.0132,
"step": 9930
},
{
"epoch": 35.88447653429603,
"grad_norm": 0.20881906151771545,
"learning_rate": 1.0172729264917857e-08,
"loss": 0.0177,
"step": 9940
},
{
"epoch": 35.92057761732852,
"grad_norm": 0.1569938063621521,
"learning_rate": 7.1108677466458215e-09,
"loss": 0.0202,
"step": 9950
},
{
"epoch": 35.95667870036101,
"grad_norm": 0.12689387798309326,
"learning_rate": 4.595720454353414e-09,
"loss": 0.0183,
"step": 9960
},
{
"epoch": 35.9927797833935,
"grad_norm": 0.1784266084432602,
"learning_rate": 2.627314893294264e-09,
"loss": 0.0127,
"step": 9970
},
{
"epoch": 36.02888086642599,
"grad_norm": 0.25251060724258423,
"learning_rate": 1.2056725896270048e-09,
"loss": 0.0176,
"step": 9980
},
{
"epoch": 36.064981949458485,
"grad_norm": 0.22773143649101257,
"learning_rate": 3.308090902098826e-10,
"loss": 0.0202,
"step": 9990
},
{
"epoch": 36.101083032490976,
"grad_norm": 0.27981123328208923,
"learning_rate": 2.7339624120159555e-12,
"loss": 0.019,
"step": 10000
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 37,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}