Mxode's picture
upload
f6519b4
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 9480,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010548523206751054,
"grad_norm": 1.2868106365203857,
"learning_rate": 0.00015822784810126583,
"loss": 7.4625,
"step": 10
},
{
"epoch": 0.002109704641350211,
"grad_norm": 1.081502914428711,
"learning_rate": 0.00031645569620253165,
"loss": 6.7982,
"step": 20
},
{
"epoch": 0.0031645569620253164,
"grad_norm": 0.7845546007156372,
"learning_rate": 0.00047468354430379745,
"loss": 6.1976,
"step": 30
},
{
"epoch": 0.004219409282700422,
"grad_norm": 1.0056753158569336,
"learning_rate": 0.0006329113924050633,
"loss": 5.7558,
"step": 40
},
{
"epoch": 0.005274261603375527,
"grad_norm": 0.6206520795822144,
"learning_rate": 0.0007911392405063291,
"loss": 5.3613,
"step": 50
},
{
"epoch": 0.006329113924050633,
"grad_norm": 0.6065470576286316,
"learning_rate": 0.0009493670886075949,
"loss": 4.9295,
"step": 60
},
{
"epoch": 0.007383966244725738,
"grad_norm": 1.051611065864563,
"learning_rate": 0.0011075949367088608,
"loss": 4.5799,
"step": 70
},
{
"epoch": 0.008438818565400843,
"grad_norm": 0.678774893283844,
"learning_rate": 0.0012658227848101266,
"loss": 4.3343,
"step": 80
},
{
"epoch": 0.00949367088607595,
"grad_norm": 0.6915785074234009,
"learning_rate": 0.0014240506329113926,
"loss": 4.1424,
"step": 90
},
{
"epoch": 0.010548523206751054,
"grad_norm": 0.4927213788032532,
"learning_rate": 0.0015,
"loss": 4.0223,
"step": 100
},
{
"epoch": 0.011603375527426161,
"grad_norm": 0.4323209524154663,
"learning_rate": 0.0015,
"loss": 3.8911,
"step": 110
},
{
"epoch": 0.012658227848101266,
"grad_norm": 0.5385138988494873,
"learning_rate": 0.0015,
"loss": 3.8258,
"step": 120
},
{
"epoch": 0.013713080168776372,
"grad_norm": 0.4882274866104126,
"learning_rate": 0.0015,
"loss": 3.7476,
"step": 130
},
{
"epoch": 0.014767932489451477,
"grad_norm": 1.0732852220535278,
"learning_rate": 0.0015,
"loss": 3.6807,
"step": 140
},
{
"epoch": 0.015822784810126583,
"grad_norm": 0.6531216502189636,
"learning_rate": 0.0015,
"loss": 3.631,
"step": 150
},
{
"epoch": 0.016877637130801686,
"grad_norm": 0.6538905501365662,
"learning_rate": 0.0015,
"loss": 3.5829,
"step": 160
},
{
"epoch": 0.017932489451476793,
"grad_norm": 0.5252573490142822,
"learning_rate": 0.0015,
"loss": 3.5372,
"step": 170
},
{
"epoch": 0.0189873417721519,
"grad_norm": 0.5157744288444519,
"learning_rate": 0.0015,
"loss": 3.5138,
"step": 180
},
{
"epoch": 0.020042194092827006,
"grad_norm": 0.7058530449867249,
"learning_rate": 0.0015,
"loss": 3.4513,
"step": 190
},
{
"epoch": 0.02109704641350211,
"grad_norm": 0.4870591461658478,
"learning_rate": 0.0015,
"loss": 3.4176,
"step": 200
},
{
"epoch": 0.022151898734177215,
"grad_norm": 0.5080909132957458,
"learning_rate": 0.0015,
"loss": 3.3788,
"step": 210
},
{
"epoch": 0.023206751054852322,
"grad_norm": 0.6322933435440063,
"learning_rate": 0.0015,
"loss": 3.3503,
"step": 220
},
{
"epoch": 0.024261603375527425,
"grad_norm": 0.5261993408203125,
"learning_rate": 0.0015,
"loss": 3.3152,
"step": 230
},
{
"epoch": 0.02531645569620253,
"grad_norm": 0.5326963663101196,
"learning_rate": 0.0015,
"loss": 3.2764,
"step": 240
},
{
"epoch": 0.026371308016877638,
"grad_norm": 0.5141389966011047,
"learning_rate": 0.0015,
"loss": 3.2478,
"step": 250
},
{
"epoch": 0.027426160337552744,
"grad_norm": 0.5861570239067078,
"learning_rate": 0.0015,
"loss": 3.213,
"step": 260
},
{
"epoch": 0.028481012658227847,
"grad_norm": 0.5424656867980957,
"learning_rate": 0.0015,
"loss": 3.2038,
"step": 270
},
{
"epoch": 0.029535864978902954,
"grad_norm": 0.5439494252204895,
"learning_rate": 0.0015,
"loss": 3.1589,
"step": 280
},
{
"epoch": 0.03059071729957806,
"grad_norm": 0.48905304074287415,
"learning_rate": 0.0015,
"loss": 3.1448,
"step": 290
},
{
"epoch": 0.03164556962025317,
"grad_norm": 0.5016011595726013,
"learning_rate": 0.0015,
"loss": 3.1269,
"step": 300
},
{
"epoch": 0.03270042194092827,
"grad_norm": 0.5477200150489807,
"learning_rate": 0.0015,
"loss": 3.0937,
"step": 310
},
{
"epoch": 0.03375527426160337,
"grad_norm": 0.4754057824611664,
"learning_rate": 0.0015,
"loss": 3.072,
"step": 320
},
{
"epoch": 0.03481012658227848,
"grad_norm": 0.5856311917304993,
"learning_rate": 0.0015,
"loss": 3.0638,
"step": 330
},
{
"epoch": 0.035864978902953586,
"grad_norm": 0.5928363800048828,
"learning_rate": 0.0015,
"loss": 3.0494,
"step": 340
},
{
"epoch": 0.03691983122362869,
"grad_norm": 0.48620685935020447,
"learning_rate": 0.0015,
"loss": 3.0092,
"step": 350
},
{
"epoch": 0.0379746835443038,
"grad_norm": 0.6079722046852112,
"learning_rate": 0.0015,
"loss": 2.9939,
"step": 360
},
{
"epoch": 0.039029535864978905,
"grad_norm": 0.6120694279670715,
"learning_rate": 0.0015,
"loss": 2.9895,
"step": 370
},
{
"epoch": 0.04008438818565401,
"grad_norm": 0.6466580033302307,
"learning_rate": 0.0015,
"loss": 2.9753,
"step": 380
},
{
"epoch": 0.04113924050632911,
"grad_norm": 0.6039373874664307,
"learning_rate": 0.0015,
"loss": 2.961,
"step": 390
},
{
"epoch": 0.04219409282700422,
"grad_norm": 0.4796377420425415,
"learning_rate": 0.0015,
"loss": 2.9316,
"step": 400
},
{
"epoch": 0.043248945147679324,
"grad_norm": 0.6468100547790527,
"learning_rate": 0.0015,
"loss": 2.9278,
"step": 410
},
{
"epoch": 0.04430379746835443,
"grad_norm": 0.7683860063552856,
"learning_rate": 0.0015,
"loss": 2.9088,
"step": 420
},
{
"epoch": 0.04535864978902954,
"grad_norm": 0.5952186584472656,
"learning_rate": 0.0015,
"loss": 2.886,
"step": 430
},
{
"epoch": 0.046413502109704644,
"grad_norm": 0.5001847147941589,
"learning_rate": 0.0015,
"loss": 2.8733,
"step": 440
},
{
"epoch": 0.04746835443037975,
"grad_norm": 0.6091427803039551,
"learning_rate": 0.0015,
"loss": 2.8715,
"step": 450
},
{
"epoch": 0.04852320675105485,
"grad_norm": 0.5184801816940308,
"learning_rate": 0.0015,
"loss": 2.864,
"step": 460
},
{
"epoch": 0.049578059071729956,
"grad_norm": 0.5876625776290894,
"learning_rate": 0.0015,
"loss": 2.8365,
"step": 470
},
{
"epoch": 0.05063291139240506,
"grad_norm": 0.6384932994842529,
"learning_rate": 0.0015,
"loss": 2.835,
"step": 480
},
{
"epoch": 0.05168776371308017,
"grad_norm": 0.6414058804512024,
"learning_rate": 0.0015,
"loss": 2.8105,
"step": 490
},
{
"epoch": 0.052742616033755275,
"grad_norm": 0.5449652671813965,
"learning_rate": 0.0015,
"loss": 2.8079,
"step": 500
},
{
"epoch": 0.05379746835443038,
"grad_norm": 0.5134292840957642,
"learning_rate": 0.0015,
"loss": 2.7963,
"step": 510
},
{
"epoch": 0.05485232067510549,
"grad_norm": 0.5442935824394226,
"learning_rate": 0.0015,
"loss": 2.7782,
"step": 520
},
{
"epoch": 0.05590717299578059,
"grad_norm": 0.48734530806541443,
"learning_rate": 0.0015,
"loss": 2.7682,
"step": 530
},
{
"epoch": 0.056962025316455694,
"grad_norm": 0.5748555064201355,
"learning_rate": 0.0015,
"loss": 2.7495,
"step": 540
},
{
"epoch": 0.0580168776371308,
"grad_norm": 0.6285437941551208,
"learning_rate": 0.0015,
"loss": 2.7557,
"step": 550
},
{
"epoch": 0.05907172995780591,
"grad_norm": 0.6870784759521484,
"learning_rate": 0.0015,
"loss": 2.7206,
"step": 560
},
{
"epoch": 0.060126582278481014,
"grad_norm": 0.6807124018669128,
"learning_rate": 0.0015,
"loss": 2.7254,
"step": 570
},
{
"epoch": 0.06118143459915612,
"grad_norm": 0.6352162957191467,
"learning_rate": 0.0015,
"loss": 2.7221,
"step": 580
},
{
"epoch": 0.06223628691983123,
"grad_norm": 0.8213168382644653,
"learning_rate": 0.0015,
"loss": 2.7018,
"step": 590
},
{
"epoch": 0.06329113924050633,
"grad_norm": 0.8000382781028748,
"learning_rate": 0.0015,
"loss": 2.6865,
"step": 600
},
{
"epoch": 0.06434599156118144,
"grad_norm": 0.5520359873771667,
"learning_rate": 0.0015,
"loss": 2.6886,
"step": 610
},
{
"epoch": 0.06540084388185655,
"grad_norm": 0.5529119968414307,
"learning_rate": 0.0015,
"loss": 2.6749,
"step": 620
},
{
"epoch": 0.06645569620253164,
"grad_norm": 0.5327541828155518,
"learning_rate": 0.0015,
"loss": 2.6622,
"step": 630
},
{
"epoch": 0.06751054852320675,
"grad_norm": 0.6201050281524658,
"learning_rate": 0.0015,
"loss": 2.6688,
"step": 640
},
{
"epoch": 0.06856540084388185,
"grad_norm": 0.612938404083252,
"learning_rate": 0.0015,
"loss": 2.6624,
"step": 650
},
{
"epoch": 0.06962025316455696,
"grad_norm": 0.5672504901885986,
"learning_rate": 0.0015,
"loss": 2.6479,
"step": 660
},
{
"epoch": 0.07067510548523206,
"grad_norm": 0.5667068958282471,
"learning_rate": 0.0015,
"loss": 2.6356,
"step": 670
},
{
"epoch": 0.07172995780590717,
"grad_norm": 0.6461451649665833,
"learning_rate": 0.0015,
"loss": 2.6288,
"step": 680
},
{
"epoch": 0.07278481012658228,
"grad_norm": 0.879687488079071,
"learning_rate": 0.0015,
"loss": 2.6352,
"step": 690
},
{
"epoch": 0.07383966244725738,
"grad_norm": 0.6234734058380127,
"learning_rate": 0.0015,
"loss": 2.6209,
"step": 700
},
{
"epoch": 0.07489451476793249,
"grad_norm": 0.5032368898391724,
"learning_rate": 0.0015,
"loss": 2.6052,
"step": 710
},
{
"epoch": 0.0759493670886076,
"grad_norm": 0.5342044830322266,
"learning_rate": 0.0015,
"loss": 2.5859,
"step": 720
},
{
"epoch": 0.0770042194092827,
"grad_norm": 0.6016132235527039,
"learning_rate": 0.0015,
"loss": 2.5785,
"step": 730
},
{
"epoch": 0.07805907172995781,
"grad_norm": 0.5635907649993896,
"learning_rate": 0.0015,
"loss": 2.5911,
"step": 740
},
{
"epoch": 0.07911392405063292,
"grad_norm": 0.6573704481124878,
"learning_rate": 0.0015,
"loss": 2.5732,
"step": 750
},
{
"epoch": 0.08016877637130802,
"grad_norm": 0.725891649723053,
"learning_rate": 0.0015,
"loss": 2.5707,
"step": 760
},
{
"epoch": 0.08122362869198312,
"grad_norm": 0.5711265206336975,
"learning_rate": 0.0015,
"loss": 2.5593,
"step": 770
},
{
"epoch": 0.08227848101265822,
"grad_norm": 0.785830020904541,
"learning_rate": 0.0015,
"loss": 2.5554,
"step": 780
},
{
"epoch": 0.08333333333333333,
"grad_norm": 0.6236185431480408,
"learning_rate": 0.0015,
"loss": 2.5627,
"step": 790
},
{
"epoch": 0.08438818565400844,
"grad_norm": 0.5514404773712158,
"learning_rate": 0.0015,
"loss": 2.5346,
"step": 800
},
{
"epoch": 0.08544303797468354,
"grad_norm": 0.5685513019561768,
"learning_rate": 0.0015,
"loss": 2.5257,
"step": 810
},
{
"epoch": 0.08649789029535865,
"grad_norm": 0.571617603302002,
"learning_rate": 0.0015,
"loss": 2.5345,
"step": 820
},
{
"epoch": 0.08755274261603375,
"grad_norm": 0.6274116635322571,
"learning_rate": 0.0015,
"loss": 2.5334,
"step": 830
},
{
"epoch": 0.08860759493670886,
"grad_norm": 0.6448284387588501,
"learning_rate": 0.0015,
"loss": 2.5129,
"step": 840
},
{
"epoch": 0.08966244725738397,
"grad_norm": 0.550652265548706,
"learning_rate": 0.0015,
"loss": 2.5105,
"step": 850
},
{
"epoch": 0.09071729957805907,
"grad_norm": 0.4846270978450775,
"learning_rate": 0.0015,
"loss": 2.5089,
"step": 860
},
{
"epoch": 0.09177215189873418,
"grad_norm": 0.5789787769317627,
"learning_rate": 0.0015,
"loss": 2.5095,
"step": 870
},
{
"epoch": 0.09282700421940929,
"grad_norm": 0.5167219638824463,
"learning_rate": 0.0015,
"loss": 2.4978,
"step": 880
},
{
"epoch": 0.0938818565400844,
"grad_norm": 0.5709961652755737,
"learning_rate": 0.0015,
"loss": 2.4925,
"step": 890
},
{
"epoch": 0.0949367088607595,
"grad_norm": 0.6126795411109924,
"learning_rate": 0.0015,
"loss": 2.4843,
"step": 900
},
{
"epoch": 0.09599156118143459,
"grad_norm": 0.5470844507217407,
"learning_rate": 0.0015,
"loss": 2.4912,
"step": 910
},
{
"epoch": 0.0970464135021097,
"grad_norm": 0.487741082906723,
"learning_rate": 0.0015,
"loss": 2.4809,
"step": 920
},
{
"epoch": 0.0981012658227848,
"grad_norm": 0.5449468493461609,
"learning_rate": 0.0015,
"loss": 2.4667,
"step": 930
},
{
"epoch": 0.09915611814345991,
"grad_norm": 0.5165682435035706,
"learning_rate": 0.0015,
"loss": 2.4695,
"step": 940
},
{
"epoch": 0.10021097046413502,
"grad_norm": 0.523166298866272,
"learning_rate": 0.0015,
"loss": 2.4533,
"step": 950
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.5307929515838623,
"learning_rate": 0.0015,
"loss": 2.4599,
"step": 960
},
{
"epoch": 0.10232067510548523,
"grad_norm": 0.5590803623199463,
"learning_rate": 0.0015,
"loss": 2.4603,
"step": 970
},
{
"epoch": 0.10337552742616034,
"grad_norm": 0.586479127407074,
"learning_rate": 0.0015,
"loss": 2.4378,
"step": 980
},
{
"epoch": 0.10443037974683544,
"grad_norm": 0.6927133202552795,
"learning_rate": 0.0015,
"loss": 2.4365,
"step": 990
},
{
"epoch": 0.10548523206751055,
"grad_norm": 0.7192505598068237,
"learning_rate": 0.0015,
"loss": 2.4458,
"step": 1000
},
{
"epoch": 0.10654008438818566,
"grad_norm": 0.5297435522079468,
"learning_rate": 0.0015,
"loss": 2.4368,
"step": 1010
},
{
"epoch": 0.10759493670886076,
"grad_norm": 0.5197161436080933,
"learning_rate": 0.0015,
"loss": 2.431,
"step": 1020
},
{
"epoch": 0.10864978902953587,
"grad_norm": 0.7736219763755798,
"learning_rate": 0.0015,
"loss": 2.4302,
"step": 1030
},
{
"epoch": 0.10970464135021098,
"grad_norm": 0.5818594694137573,
"learning_rate": 0.0015,
"loss": 2.4228,
"step": 1040
},
{
"epoch": 0.11075949367088607,
"grad_norm": 0.5666146874427795,
"learning_rate": 0.0015,
"loss": 2.4246,
"step": 1050
},
{
"epoch": 0.11181434599156118,
"grad_norm": 0.5254496335983276,
"learning_rate": 0.0015,
"loss": 2.4131,
"step": 1060
},
{
"epoch": 0.11286919831223628,
"grad_norm": 0.5331057906150818,
"learning_rate": 0.0015,
"loss": 2.4027,
"step": 1070
},
{
"epoch": 0.11392405063291139,
"grad_norm": 0.5084558129310608,
"learning_rate": 0.0015,
"loss": 2.4073,
"step": 1080
},
{
"epoch": 0.1149789029535865,
"grad_norm": 0.5524882078170776,
"learning_rate": 0.0015,
"loss": 2.4098,
"step": 1090
},
{
"epoch": 0.1160337552742616,
"grad_norm": 0.5621328353881836,
"learning_rate": 0.0015,
"loss": 2.3943,
"step": 1100
},
{
"epoch": 0.11708860759493671,
"grad_norm": 0.4843035042285919,
"learning_rate": 0.0015,
"loss": 2.3961,
"step": 1110
},
{
"epoch": 0.11814345991561181,
"grad_norm": 0.528351366519928,
"learning_rate": 0.0015,
"loss": 2.3914,
"step": 1120
},
{
"epoch": 0.11919831223628692,
"grad_norm": 0.599197506904602,
"learning_rate": 0.0015,
"loss": 2.3768,
"step": 1130
},
{
"epoch": 0.12025316455696203,
"grad_norm": 0.513840913772583,
"learning_rate": 0.0015,
"loss": 2.3961,
"step": 1140
},
{
"epoch": 0.12130801687763713,
"grad_norm": 0.6017542481422424,
"learning_rate": 0.0015,
"loss": 2.3786,
"step": 1150
},
{
"epoch": 0.12236286919831224,
"grad_norm": 0.509437620639801,
"learning_rate": 0.0015,
"loss": 2.3859,
"step": 1160
},
{
"epoch": 0.12341772151898735,
"grad_norm": 0.5826842784881592,
"learning_rate": 0.0015,
"loss": 2.3702,
"step": 1170
},
{
"epoch": 0.12447257383966245,
"grad_norm": 0.529461145401001,
"learning_rate": 0.0015,
"loss": 2.3571,
"step": 1180
},
{
"epoch": 0.12552742616033755,
"grad_norm": 0.5379661321640015,
"learning_rate": 0.0015,
"loss": 2.3604,
"step": 1190
},
{
"epoch": 0.12658227848101267,
"grad_norm": 0.5306044220924377,
"learning_rate": 0.0015,
"loss": 2.369,
"step": 1200
},
{
"epoch": 0.12763713080168776,
"grad_norm": 0.5906445980072021,
"learning_rate": 0.0015,
"loss": 2.3662,
"step": 1210
},
{
"epoch": 0.12869198312236288,
"grad_norm": 0.5382705926895142,
"learning_rate": 0.0015,
"loss": 2.3502,
"step": 1220
},
{
"epoch": 0.12974683544303797,
"grad_norm": 0.49285948276519775,
"learning_rate": 0.0015,
"loss": 2.3365,
"step": 1230
},
{
"epoch": 0.1308016877637131,
"grad_norm": 0.5219690799713135,
"learning_rate": 0.0015,
"loss": 2.354,
"step": 1240
},
{
"epoch": 0.13185654008438819,
"grad_norm": 0.5424527525901794,
"learning_rate": 0.0015,
"loss": 2.3568,
"step": 1250
},
{
"epoch": 0.13291139240506328,
"grad_norm": 0.5400310158729553,
"learning_rate": 0.0015,
"loss": 2.336,
"step": 1260
},
{
"epoch": 0.1339662447257384,
"grad_norm": 0.5023423433303833,
"learning_rate": 0.0015,
"loss": 2.3316,
"step": 1270
},
{
"epoch": 0.1350210970464135,
"grad_norm": 0.5966672301292419,
"learning_rate": 0.0015,
"loss": 2.3487,
"step": 1280
},
{
"epoch": 0.1360759493670886,
"grad_norm": 0.4899304509162903,
"learning_rate": 0.0015,
"loss": 2.3344,
"step": 1290
},
{
"epoch": 0.1371308016877637,
"grad_norm": 0.4942777752876282,
"learning_rate": 0.0015,
"loss": 2.3203,
"step": 1300
},
{
"epoch": 0.13818565400843882,
"grad_norm": 0.5217072367668152,
"learning_rate": 0.0015,
"loss": 2.325,
"step": 1310
},
{
"epoch": 0.13924050632911392,
"grad_norm": 0.5340680480003357,
"learning_rate": 0.0015,
"loss": 2.3268,
"step": 1320
},
{
"epoch": 0.14029535864978904,
"grad_norm": 0.5798514485359192,
"learning_rate": 0.0015,
"loss": 2.3177,
"step": 1330
},
{
"epoch": 0.14135021097046413,
"grad_norm": 0.5663682818412781,
"learning_rate": 0.0015,
"loss": 2.3212,
"step": 1340
},
{
"epoch": 0.14240506329113925,
"grad_norm": 0.5789469480514526,
"learning_rate": 0.0015,
"loss": 2.3274,
"step": 1350
},
{
"epoch": 0.14345991561181434,
"grad_norm": 0.5599437952041626,
"learning_rate": 0.0015,
"loss": 2.3133,
"step": 1360
},
{
"epoch": 0.14451476793248946,
"grad_norm": 0.5702401995658875,
"learning_rate": 0.0015,
"loss": 2.3123,
"step": 1370
},
{
"epoch": 0.14556962025316456,
"grad_norm": 0.5608999133110046,
"learning_rate": 0.0015,
"loss": 2.2989,
"step": 1380
},
{
"epoch": 0.14662447257383968,
"grad_norm": 0.5066065192222595,
"learning_rate": 0.0015,
"loss": 2.3107,
"step": 1390
},
{
"epoch": 0.14767932489451477,
"grad_norm": 0.6196108460426331,
"learning_rate": 0.0015,
"loss": 2.2991,
"step": 1400
},
{
"epoch": 0.14873417721518986,
"grad_norm": 0.5460635423660278,
"learning_rate": 0.0015,
"loss": 2.2974,
"step": 1410
},
{
"epoch": 0.14978902953586498,
"grad_norm": 0.4733991026878357,
"learning_rate": 0.0015,
"loss": 2.3018,
"step": 1420
},
{
"epoch": 0.15084388185654007,
"grad_norm": 0.5400242805480957,
"learning_rate": 0.0015,
"loss": 2.2962,
"step": 1430
},
{
"epoch": 0.1518987341772152,
"grad_norm": 0.5293214917182922,
"learning_rate": 0.0015,
"loss": 2.2908,
"step": 1440
},
{
"epoch": 0.1529535864978903,
"grad_norm": 0.45472586154937744,
"learning_rate": 0.0015,
"loss": 2.2902,
"step": 1450
},
{
"epoch": 0.1540084388185654,
"grad_norm": 0.5257254838943481,
"learning_rate": 0.0015,
"loss": 2.2861,
"step": 1460
},
{
"epoch": 0.1550632911392405,
"grad_norm": 0.5383533835411072,
"learning_rate": 0.0015,
"loss": 2.2832,
"step": 1470
},
{
"epoch": 0.15611814345991562,
"grad_norm": 0.5611907243728638,
"learning_rate": 0.0015,
"loss": 2.2883,
"step": 1480
},
{
"epoch": 0.1571729957805907,
"grad_norm": 0.5457258224487305,
"learning_rate": 0.0015,
"loss": 2.2838,
"step": 1490
},
{
"epoch": 0.15822784810126583,
"grad_norm": 0.49772271513938904,
"learning_rate": 0.0015,
"loss": 2.2817,
"step": 1500
},
{
"epoch": 0.15928270042194093,
"grad_norm": 0.5408310294151306,
"learning_rate": 0.0015,
"loss": 2.2784,
"step": 1510
},
{
"epoch": 0.16033755274261605,
"grad_norm": 0.5458889603614807,
"learning_rate": 0.0015,
"loss": 2.2805,
"step": 1520
},
{
"epoch": 0.16139240506329114,
"grad_norm": 0.4789125323295593,
"learning_rate": 0.0015,
"loss": 2.2612,
"step": 1530
},
{
"epoch": 0.16244725738396623,
"grad_norm": 0.4473031759262085,
"learning_rate": 0.0015,
"loss": 2.252,
"step": 1540
},
{
"epoch": 0.16350210970464135,
"grad_norm": 0.5544670820236206,
"learning_rate": 0.0015,
"loss": 2.2557,
"step": 1550
},
{
"epoch": 0.16455696202531644,
"grad_norm": 0.5125646591186523,
"learning_rate": 0.0015,
"loss": 2.2613,
"step": 1560
},
{
"epoch": 0.16561181434599156,
"grad_norm": 0.49753326177597046,
"learning_rate": 0.0015,
"loss": 2.2521,
"step": 1570
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.5130689740180969,
"learning_rate": 0.0015,
"loss": 2.2621,
"step": 1580
},
{
"epoch": 0.16772151898734178,
"grad_norm": 0.5498607158660889,
"learning_rate": 0.0015,
"loss": 2.2588,
"step": 1590
},
{
"epoch": 0.16877637130801687,
"grad_norm": 0.4919903874397278,
"learning_rate": 0.0015,
"loss": 2.2635,
"step": 1600
},
{
"epoch": 0.169831223628692,
"grad_norm": 0.4785551130771637,
"learning_rate": 0.0015,
"loss": 2.2627,
"step": 1610
},
{
"epoch": 0.17088607594936708,
"grad_norm": 0.487488716840744,
"learning_rate": 0.0015,
"loss": 2.2331,
"step": 1620
},
{
"epoch": 0.1719409282700422,
"grad_norm": 0.7447993755340576,
"learning_rate": 0.0015,
"loss": 2.2502,
"step": 1630
},
{
"epoch": 0.1729957805907173,
"grad_norm": 0.517065703868866,
"learning_rate": 0.0015,
"loss": 2.2445,
"step": 1640
},
{
"epoch": 0.17405063291139242,
"grad_norm": 0.48035067319869995,
"learning_rate": 0.0015,
"loss": 2.2383,
"step": 1650
},
{
"epoch": 0.1751054852320675,
"grad_norm": 0.4717473089694977,
"learning_rate": 0.0015,
"loss": 2.2332,
"step": 1660
},
{
"epoch": 0.17616033755274263,
"grad_norm": 0.5340759754180908,
"learning_rate": 0.0015,
"loss": 2.245,
"step": 1670
},
{
"epoch": 0.17721518987341772,
"grad_norm": 0.5645149350166321,
"learning_rate": 0.0015,
"loss": 2.2369,
"step": 1680
},
{
"epoch": 0.17827004219409281,
"grad_norm": 0.479165256023407,
"learning_rate": 0.0015,
"loss": 2.2194,
"step": 1690
},
{
"epoch": 0.17932489451476794,
"grad_norm": 0.5128156542778015,
"learning_rate": 0.0015,
"loss": 2.2411,
"step": 1700
},
{
"epoch": 0.18037974683544303,
"grad_norm": 0.5018277764320374,
"learning_rate": 0.0015,
"loss": 2.2197,
"step": 1710
},
{
"epoch": 0.18143459915611815,
"grad_norm": 0.48069244623184204,
"learning_rate": 0.0015,
"loss": 2.2192,
"step": 1720
},
{
"epoch": 0.18248945147679324,
"grad_norm": 0.5087079405784607,
"learning_rate": 0.0015,
"loss": 2.223,
"step": 1730
},
{
"epoch": 0.18354430379746836,
"grad_norm": 0.5914036631584167,
"learning_rate": 0.0015,
"loss": 2.2284,
"step": 1740
},
{
"epoch": 0.18459915611814345,
"grad_norm": 0.5060463547706604,
"learning_rate": 0.0015,
"loss": 2.2485,
"step": 1750
},
{
"epoch": 0.18565400843881857,
"grad_norm": 0.5524370670318604,
"learning_rate": 0.0015,
"loss": 2.2249,
"step": 1760
},
{
"epoch": 0.18670886075949367,
"grad_norm": 0.5056757926940918,
"learning_rate": 0.0015,
"loss": 2.2237,
"step": 1770
},
{
"epoch": 0.1877637130801688,
"grad_norm": 0.4828547239303589,
"learning_rate": 0.0015,
"loss": 2.2252,
"step": 1780
},
{
"epoch": 0.18881856540084388,
"grad_norm": 0.5432929396629333,
"learning_rate": 0.0015,
"loss": 2.2214,
"step": 1790
},
{
"epoch": 0.189873417721519,
"grad_norm": 0.51959228515625,
"learning_rate": 0.0015,
"loss": 2.2189,
"step": 1800
},
{
"epoch": 0.1909282700421941,
"grad_norm": 0.513494074344635,
"learning_rate": 0.0015,
"loss": 2.2098,
"step": 1810
},
{
"epoch": 0.19198312236286919,
"grad_norm": 0.4728807508945465,
"learning_rate": 0.0015,
"loss": 2.2085,
"step": 1820
},
{
"epoch": 0.1930379746835443,
"grad_norm": 0.4876849055290222,
"learning_rate": 0.0015,
"loss": 2.2015,
"step": 1830
},
{
"epoch": 0.1940928270042194,
"grad_norm": 0.487371027469635,
"learning_rate": 0.0015,
"loss": 2.2029,
"step": 1840
},
{
"epoch": 0.19514767932489452,
"grad_norm": 0.5107097029685974,
"learning_rate": 0.0015,
"loss": 2.2053,
"step": 1850
},
{
"epoch": 0.1962025316455696,
"grad_norm": 0.4802800416946411,
"learning_rate": 0.0015,
"loss": 2.2095,
"step": 1860
},
{
"epoch": 0.19725738396624473,
"grad_norm": 0.5234653949737549,
"learning_rate": 0.0015,
"loss": 2.2067,
"step": 1870
},
{
"epoch": 0.19831223628691982,
"grad_norm": 0.5987045764923096,
"learning_rate": 0.0015,
"loss": 2.1995,
"step": 1880
},
{
"epoch": 0.19936708860759494,
"grad_norm": 0.5111141204833984,
"learning_rate": 0.0015,
"loss": 2.2054,
"step": 1890
},
{
"epoch": 0.20042194092827004,
"grad_norm": 0.516825258731842,
"learning_rate": 0.0015,
"loss": 2.1997,
"step": 1900
},
{
"epoch": 0.20147679324894516,
"grad_norm": 0.5078732371330261,
"learning_rate": 0.0015,
"loss": 2.201,
"step": 1910
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.4825908839702606,
"learning_rate": 0.0015,
"loss": 2.195,
"step": 1920
},
{
"epoch": 0.20358649789029537,
"grad_norm": 0.5081301331520081,
"learning_rate": 0.0015,
"loss": 2.1994,
"step": 1930
},
{
"epoch": 0.20464135021097046,
"grad_norm": 0.4621557593345642,
"learning_rate": 0.0015,
"loss": 2.1866,
"step": 1940
},
{
"epoch": 0.20569620253164558,
"grad_norm": 0.49421605467796326,
"learning_rate": 0.0015,
"loss": 2.1891,
"step": 1950
},
{
"epoch": 0.20675105485232068,
"grad_norm": 0.446591854095459,
"learning_rate": 0.0015,
"loss": 2.1931,
"step": 1960
},
{
"epoch": 0.20780590717299577,
"grad_norm": 0.47465038299560547,
"learning_rate": 0.0015,
"loss": 2.198,
"step": 1970
},
{
"epoch": 0.2088607594936709,
"grad_norm": 0.47602617740631104,
"learning_rate": 0.0015,
"loss": 2.1803,
"step": 1980
},
{
"epoch": 0.20991561181434598,
"grad_norm": 0.4852505028247833,
"learning_rate": 0.0015,
"loss": 2.1811,
"step": 1990
},
{
"epoch": 0.2109704641350211,
"grad_norm": 0.495806485414505,
"learning_rate": 0.0015,
"loss": 2.179,
"step": 2000
},
{
"epoch": 0.2120253164556962,
"grad_norm": 0.4961065948009491,
"learning_rate": 0.0015,
"loss": 2.1852,
"step": 2010
},
{
"epoch": 0.21308016877637131,
"grad_norm": 0.5539078116416931,
"learning_rate": 0.0015,
"loss": 2.1965,
"step": 2020
},
{
"epoch": 0.2141350210970464,
"grad_norm": 0.5105553269386292,
"learning_rate": 0.0015,
"loss": 2.1847,
"step": 2030
},
{
"epoch": 0.21518987341772153,
"grad_norm": 0.47178715467453003,
"learning_rate": 0.0015,
"loss": 2.1823,
"step": 2040
},
{
"epoch": 0.21624472573839662,
"grad_norm": 0.47604134678840637,
"learning_rate": 0.0015,
"loss": 2.1838,
"step": 2050
},
{
"epoch": 0.21729957805907174,
"grad_norm": 0.5011481046676636,
"learning_rate": 0.0015,
"loss": 2.1696,
"step": 2060
},
{
"epoch": 0.21835443037974683,
"grad_norm": 0.454166978597641,
"learning_rate": 0.0015,
"loss": 2.1681,
"step": 2070
},
{
"epoch": 0.21940928270042195,
"grad_norm": 0.5561933517456055,
"learning_rate": 0.0015,
"loss": 2.1811,
"step": 2080
},
{
"epoch": 0.22046413502109705,
"grad_norm": 0.47614896297454834,
"learning_rate": 0.0015,
"loss": 2.1668,
"step": 2090
},
{
"epoch": 0.22151898734177214,
"grad_norm": 0.6641237139701843,
"learning_rate": 0.0015,
"loss": 2.1758,
"step": 2100
},
{
"epoch": 0.22257383966244726,
"grad_norm": 0.5254981517791748,
"learning_rate": 0.0015,
"loss": 2.1685,
"step": 2110
},
{
"epoch": 0.22362869198312235,
"grad_norm": 0.4500657618045807,
"learning_rate": 0.0015,
"loss": 2.164,
"step": 2120
},
{
"epoch": 0.22468354430379747,
"grad_norm": 0.4955187737941742,
"learning_rate": 0.0015,
"loss": 2.1673,
"step": 2130
},
{
"epoch": 0.22573839662447256,
"grad_norm": 0.48642462491989136,
"learning_rate": 0.0015,
"loss": 2.1589,
"step": 2140
},
{
"epoch": 0.22679324894514769,
"grad_norm": 0.4678893983364105,
"learning_rate": 0.0015,
"loss": 2.1645,
"step": 2150
},
{
"epoch": 0.22784810126582278,
"grad_norm": 0.5280480980873108,
"learning_rate": 0.0015,
"loss": 2.1805,
"step": 2160
},
{
"epoch": 0.2289029535864979,
"grad_norm": 0.5271260142326355,
"learning_rate": 0.0015,
"loss": 2.1607,
"step": 2170
},
{
"epoch": 0.229957805907173,
"grad_norm": 0.45616209506988525,
"learning_rate": 0.0015,
"loss": 2.1469,
"step": 2180
},
{
"epoch": 0.2310126582278481,
"grad_norm": 0.5835659503936768,
"learning_rate": 0.0015,
"loss": 2.1636,
"step": 2190
},
{
"epoch": 0.2320675105485232,
"grad_norm": 0.5413613319396973,
"learning_rate": 0.0015,
"loss": 2.1651,
"step": 2200
},
{
"epoch": 0.23312236286919832,
"grad_norm": 0.44273021817207336,
"learning_rate": 0.0015,
"loss": 2.1505,
"step": 2210
},
{
"epoch": 0.23417721518987342,
"grad_norm": 0.536941409111023,
"learning_rate": 0.0015,
"loss": 2.1682,
"step": 2220
},
{
"epoch": 0.23523206751054854,
"grad_norm": 0.45308029651641846,
"learning_rate": 0.0015,
"loss": 2.1552,
"step": 2230
},
{
"epoch": 0.23628691983122363,
"grad_norm": 0.46434181928634644,
"learning_rate": 0.0015,
"loss": 2.1493,
"step": 2240
},
{
"epoch": 0.23734177215189872,
"grad_norm": 0.5993618369102478,
"learning_rate": 0.0015,
"loss": 2.1494,
"step": 2250
},
{
"epoch": 0.23839662447257384,
"grad_norm": 0.4938041865825653,
"learning_rate": 0.0015,
"loss": 2.1387,
"step": 2260
},
{
"epoch": 0.23945147679324894,
"grad_norm": 0.4831969141960144,
"learning_rate": 0.0015,
"loss": 2.1735,
"step": 2270
},
{
"epoch": 0.24050632911392406,
"grad_norm": 0.5053781867027283,
"learning_rate": 0.0015,
"loss": 2.1568,
"step": 2280
},
{
"epoch": 0.24156118143459915,
"grad_norm": 0.6139144897460938,
"learning_rate": 0.0015,
"loss": 2.1437,
"step": 2290
},
{
"epoch": 0.24261603375527427,
"grad_norm": 0.45216429233551025,
"learning_rate": 0.0015,
"loss": 2.141,
"step": 2300
},
{
"epoch": 0.24367088607594936,
"grad_norm": 0.5123635530471802,
"learning_rate": 0.0015,
"loss": 2.1475,
"step": 2310
},
{
"epoch": 0.24472573839662448,
"grad_norm": 0.49576327204704285,
"learning_rate": 0.0015,
"loss": 2.1526,
"step": 2320
},
{
"epoch": 0.24578059071729957,
"grad_norm": 0.47483232617378235,
"learning_rate": 0.0015,
"loss": 2.1349,
"step": 2330
},
{
"epoch": 0.2468354430379747,
"grad_norm": 0.47471559047698975,
"learning_rate": 0.0015,
"loss": 2.1336,
"step": 2340
},
{
"epoch": 0.2478902953586498,
"grad_norm": 0.49153319001197815,
"learning_rate": 0.0015,
"loss": 2.1316,
"step": 2350
},
{
"epoch": 0.2489451476793249,
"grad_norm": 0.5166982412338257,
"learning_rate": 0.0015,
"loss": 2.1433,
"step": 2360
},
{
"epoch": 0.25,
"grad_norm": 0.5839778780937195,
"learning_rate": 0.0015,
"loss": 2.1414,
"step": 2370
},
{
"epoch": 0.2510548523206751,
"grad_norm": 0.5124769806861877,
"learning_rate": 0.0015,
"loss": 2.1426,
"step": 2380
},
{
"epoch": 0.2521097046413502,
"grad_norm": 0.5506088137626648,
"learning_rate": 0.0015,
"loss": 2.1383,
"step": 2390
},
{
"epoch": 0.25316455696202533,
"grad_norm": 0.4894935190677643,
"learning_rate": 0.0015,
"loss": 2.1348,
"step": 2400
},
{
"epoch": 0.2542194092827004,
"grad_norm": 0.5619579553604126,
"learning_rate": 0.0015,
"loss": 2.1337,
"step": 2410
},
{
"epoch": 0.2552742616033755,
"grad_norm": 0.5544256567955017,
"learning_rate": 0.0015,
"loss": 2.1305,
"step": 2420
},
{
"epoch": 0.2563291139240506,
"grad_norm": 0.47035881876945496,
"learning_rate": 0.0015,
"loss": 2.1299,
"step": 2430
},
{
"epoch": 0.25738396624472576,
"grad_norm": 0.5639828443527222,
"learning_rate": 0.0015,
"loss": 2.1393,
"step": 2440
},
{
"epoch": 0.25843881856540085,
"grad_norm": 0.5048669576644897,
"learning_rate": 0.0015,
"loss": 2.1196,
"step": 2450
},
{
"epoch": 0.25949367088607594,
"grad_norm": 0.4745074510574341,
"learning_rate": 0.0015,
"loss": 2.1368,
"step": 2460
},
{
"epoch": 0.26054852320675104,
"grad_norm": 0.4397202730178833,
"learning_rate": 0.0015,
"loss": 2.1272,
"step": 2470
},
{
"epoch": 0.2616033755274262,
"grad_norm": 0.4900398850440979,
"learning_rate": 0.0015,
"loss": 2.1307,
"step": 2480
},
{
"epoch": 0.2626582278481013,
"grad_norm": 0.46921002864837646,
"learning_rate": 0.0015,
"loss": 2.122,
"step": 2490
},
{
"epoch": 0.26371308016877637,
"grad_norm": 0.4493680000305176,
"learning_rate": 0.0015,
"loss": 2.1253,
"step": 2500
},
{
"epoch": 0.26476793248945146,
"grad_norm": 0.4484681785106659,
"learning_rate": 0.0015,
"loss": 2.1222,
"step": 2510
},
{
"epoch": 0.26582278481012656,
"grad_norm": 0.45130497217178345,
"learning_rate": 0.0015,
"loss": 2.1284,
"step": 2520
},
{
"epoch": 0.2668776371308017,
"grad_norm": 0.4377133548259735,
"learning_rate": 0.0015,
"loss": 2.1191,
"step": 2530
},
{
"epoch": 0.2679324894514768,
"grad_norm": 0.49162405729293823,
"learning_rate": 0.0015,
"loss": 2.1242,
"step": 2540
},
{
"epoch": 0.2689873417721519,
"grad_norm": 0.46707892417907715,
"learning_rate": 0.0015,
"loss": 2.1122,
"step": 2550
},
{
"epoch": 0.270042194092827,
"grad_norm": 0.4644486606121063,
"learning_rate": 0.0015,
"loss": 2.1259,
"step": 2560
},
{
"epoch": 0.27109704641350213,
"grad_norm": 0.5053285956382751,
"learning_rate": 0.0015,
"loss": 2.1178,
"step": 2570
},
{
"epoch": 0.2721518987341772,
"grad_norm": 0.48815426230430603,
"learning_rate": 0.0015,
"loss": 2.1179,
"step": 2580
},
{
"epoch": 0.2732067510548523,
"grad_norm": 0.4503813683986664,
"learning_rate": 0.0015,
"loss": 2.1237,
"step": 2590
},
{
"epoch": 0.2742616033755274,
"grad_norm": 0.4196797013282776,
"learning_rate": 0.0015,
"loss": 2.1218,
"step": 2600
},
{
"epoch": 0.27531645569620256,
"grad_norm": 0.5734556317329407,
"learning_rate": 0.0015,
"loss": 2.1233,
"step": 2610
},
{
"epoch": 0.27637130801687765,
"grad_norm": 0.487870991230011,
"learning_rate": 0.0015,
"loss": 2.1204,
"step": 2620
},
{
"epoch": 0.27742616033755274,
"grad_norm": 0.4825158417224884,
"learning_rate": 0.0015,
"loss": 2.1064,
"step": 2630
},
{
"epoch": 0.27848101265822783,
"grad_norm": 0.498384028673172,
"learning_rate": 0.0015,
"loss": 2.1209,
"step": 2640
},
{
"epoch": 0.2795358649789029,
"grad_norm": 0.5111238360404968,
"learning_rate": 0.0015,
"loss": 2.1117,
"step": 2650
},
{
"epoch": 0.2805907172995781,
"grad_norm": 0.5499308705329895,
"learning_rate": 0.0015,
"loss": 2.1076,
"step": 2660
},
{
"epoch": 0.28164556962025317,
"grad_norm": 0.5574318170547485,
"learning_rate": 0.0015,
"loss": 2.105,
"step": 2670
},
{
"epoch": 0.28270042194092826,
"grad_norm": 0.4500906765460968,
"learning_rate": 0.0015,
"loss": 2.1053,
"step": 2680
},
{
"epoch": 0.28375527426160335,
"grad_norm": 0.4742302894592285,
"learning_rate": 0.0015,
"loss": 2.1083,
"step": 2690
},
{
"epoch": 0.2848101265822785,
"grad_norm": 0.47450193762779236,
"learning_rate": 0.0015,
"loss": 2.1058,
"step": 2700
},
{
"epoch": 0.2858649789029536,
"grad_norm": 0.4674990475177765,
"learning_rate": 0.0015,
"loss": 2.1062,
"step": 2710
},
{
"epoch": 0.2869198312236287,
"grad_norm": 0.4899941682815552,
"learning_rate": 0.0015,
"loss": 2.1095,
"step": 2720
},
{
"epoch": 0.2879746835443038,
"grad_norm": 0.4762454628944397,
"learning_rate": 0.0015,
"loss": 2.0897,
"step": 2730
},
{
"epoch": 0.2890295358649789,
"grad_norm": 0.49402663111686707,
"learning_rate": 0.0015,
"loss": 2.1016,
"step": 2740
},
{
"epoch": 0.290084388185654,
"grad_norm": 0.43774598836898804,
"learning_rate": 0.0015,
"loss": 2.1069,
"step": 2750
},
{
"epoch": 0.2911392405063291,
"grad_norm": 0.4219803512096405,
"learning_rate": 0.0015,
"loss": 2.1115,
"step": 2760
},
{
"epoch": 0.2921940928270042,
"grad_norm": 0.42314285039901733,
"learning_rate": 0.0015,
"loss": 2.1059,
"step": 2770
},
{
"epoch": 0.29324894514767935,
"grad_norm": 0.5017038583755493,
"learning_rate": 0.0015,
"loss": 2.1045,
"step": 2780
},
{
"epoch": 0.29430379746835444,
"grad_norm": 0.5033962726593018,
"learning_rate": 0.0015,
"loss": 2.1004,
"step": 2790
},
{
"epoch": 0.29535864978902954,
"grad_norm": 0.411020427942276,
"learning_rate": 0.0015,
"loss": 2.0992,
"step": 2800
},
{
"epoch": 0.29641350210970463,
"grad_norm": 0.4837932288646698,
"learning_rate": 0.0015,
"loss": 2.1011,
"step": 2810
},
{
"epoch": 0.2974683544303797,
"grad_norm": 0.48591822385787964,
"learning_rate": 0.0015,
"loss": 2.1056,
"step": 2820
},
{
"epoch": 0.29852320675105487,
"grad_norm": 0.5090926885604858,
"learning_rate": 0.0015,
"loss": 2.0751,
"step": 2830
},
{
"epoch": 0.29957805907172996,
"grad_norm": 0.4709317088127136,
"learning_rate": 0.0015,
"loss": 2.078,
"step": 2840
},
{
"epoch": 0.30063291139240506,
"grad_norm": 0.4561120867729187,
"learning_rate": 0.0015,
"loss": 2.093,
"step": 2850
},
{
"epoch": 0.30168776371308015,
"grad_norm": 0.4594179689884186,
"learning_rate": 0.0015,
"loss": 2.0942,
"step": 2860
},
{
"epoch": 0.3027426160337553,
"grad_norm": 0.5431883931159973,
"learning_rate": 0.0015,
"loss": 2.094,
"step": 2870
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.4916505515575409,
"learning_rate": 0.0015,
"loss": 2.079,
"step": 2880
},
{
"epoch": 0.3048523206751055,
"grad_norm": 0.43344438076019287,
"learning_rate": 0.0015,
"loss": 2.0862,
"step": 2890
},
{
"epoch": 0.3059071729957806,
"grad_norm": 0.4944959580898285,
"learning_rate": 0.0015,
"loss": 2.093,
"step": 2900
},
{
"epoch": 0.3069620253164557,
"grad_norm": 0.4984077215194702,
"learning_rate": 0.0015,
"loss": 2.0932,
"step": 2910
},
{
"epoch": 0.3080168776371308,
"grad_norm": 0.4800587594509125,
"learning_rate": 0.0015,
"loss": 2.086,
"step": 2920
},
{
"epoch": 0.3090717299578059,
"grad_norm": 0.45485249161720276,
"learning_rate": 0.0015,
"loss": 2.0843,
"step": 2930
},
{
"epoch": 0.310126582278481,
"grad_norm": 0.45084336400032043,
"learning_rate": 0.0015,
"loss": 2.0896,
"step": 2940
},
{
"epoch": 0.3111814345991561,
"grad_norm": 0.42323753237724304,
"learning_rate": 0.0015,
"loss": 2.0951,
"step": 2950
},
{
"epoch": 0.31223628691983124,
"grad_norm": 0.5326976180076599,
"learning_rate": 0.0015,
"loss": 2.0801,
"step": 2960
},
{
"epoch": 0.31329113924050633,
"grad_norm": 0.46457424759864807,
"learning_rate": 0.0015,
"loss": 2.0727,
"step": 2970
},
{
"epoch": 0.3143459915611814,
"grad_norm": 0.47872284054756165,
"learning_rate": 0.0015,
"loss": 2.0803,
"step": 2980
},
{
"epoch": 0.3154008438818565,
"grad_norm": 0.4702615439891815,
"learning_rate": 0.0015,
"loss": 2.0967,
"step": 2990
},
{
"epoch": 0.31645569620253167,
"grad_norm": 0.45781195163726807,
"learning_rate": 0.0015,
"loss": 2.0685,
"step": 3000
},
{
"epoch": 0.31751054852320676,
"grad_norm": 0.44454801082611084,
"learning_rate": 0.0015,
"loss": 2.0806,
"step": 3010
},
{
"epoch": 0.31856540084388185,
"grad_norm": 0.479475200176239,
"learning_rate": 0.0015,
"loss": 2.0787,
"step": 3020
},
{
"epoch": 0.31962025316455694,
"grad_norm": 0.4887007474899292,
"learning_rate": 0.0015,
"loss": 2.0717,
"step": 3030
},
{
"epoch": 0.3206751054852321,
"grad_norm": 0.46655696630477905,
"learning_rate": 0.0015,
"loss": 2.0776,
"step": 3040
},
{
"epoch": 0.3217299578059072,
"grad_norm": 0.48713260889053345,
"learning_rate": 0.0015,
"loss": 2.0746,
"step": 3050
},
{
"epoch": 0.3227848101265823,
"grad_norm": 0.5423433780670166,
"learning_rate": 0.0015,
"loss": 2.0836,
"step": 3060
},
{
"epoch": 0.32383966244725737,
"grad_norm": 0.49016839265823364,
"learning_rate": 0.0015,
"loss": 2.0772,
"step": 3070
},
{
"epoch": 0.32489451476793246,
"grad_norm": 0.4809034466743469,
"learning_rate": 0.0015,
"loss": 2.0638,
"step": 3080
},
{
"epoch": 0.3259493670886076,
"grad_norm": 0.4437301754951477,
"learning_rate": 0.0015,
"loss": 2.0813,
"step": 3090
},
{
"epoch": 0.3270042194092827,
"grad_norm": 0.42194899916648865,
"learning_rate": 0.0015,
"loss": 2.0676,
"step": 3100
},
{
"epoch": 0.3280590717299578,
"grad_norm": 0.47773346304893494,
"learning_rate": 0.0015,
"loss": 2.0682,
"step": 3110
},
{
"epoch": 0.3291139240506329,
"grad_norm": 0.47165659070014954,
"learning_rate": 0.0015,
"loss": 2.0726,
"step": 3120
},
{
"epoch": 0.33016877637130804,
"grad_norm": 0.5155400037765503,
"learning_rate": 0.0015,
"loss": 2.0847,
"step": 3130
},
{
"epoch": 0.33122362869198313,
"grad_norm": 0.5024751424789429,
"learning_rate": 0.0015,
"loss": 2.0605,
"step": 3140
},
{
"epoch": 0.3322784810126582,
"grad_norm": 0.47432848811149597,
"learning_rate": 0.0015,
"loss": 2.0699,
"step": 3150
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.4940382242202759,
"learning_rate": 0.0015,
"loss": 2.0705,
"step": 3160
},
{
"epoch": 0.33438818565400846,
"grad_norm": 0.4658326804637909,
"learning_rate": 0.0015,
"loss": 2.0717,
"step": 3170
},
{
"epoch": 0.33544303797468356,
"grad_norm": 0.4399244487285614,
"learning_rate": 0.0015,
"loss": 2.067,
"step": 3180
},
{
"epoch": 0.33649789029535865,
"grad_norm": 0.41705337166786194,
"learning_rate": 0.0015,
"loss": 2.0647,
"step": 3190
},
{
"epoch": 0.33755274261603374,
"grad_norm": 0.4723767936229706,
"learning_rate": 0.0015,
"loss": 2.0601,
"step": 3200
},
{
"epoch": 0.33860759493670883,
"grad_norm": 0.4406121075153351,
"learning_rate": 0.0015,
"loss": 2.0763,
"step": 3210
},
{
"epoch": 0.339662447257384,
"grad_norm": 0.4701217710971832,
"learning_rate": 0.0015,
"loss": 2.0646,
"step": 3220
},
{
"epoch": 0.3407172995780591,
"grad_norm": 0.4767507314682007,
"learning_rate": 0.0015,
"loss": 2.0624,
"step": 3230
},
{
"epoch": 0.34177215189873417,
"grad_norm": 0.45047277212142944,
"learning_rate": 0.0015,
"loss": 2.042,
"step": 3240
},
{
"epoch": 0.34282700421940926,
"grad_norm": 0.45256108045578003,
"learning_rate": 0.0015,
"loss": 2.0642,
"step": 3250
},
{
"epoch": 0.3438818565400844,
"grad_norm": 0.46273359656333923,
"learning_rate": 0.0015,
"loss": 2.0556,
"step": 3260
},
{
"epoch": 0.3449367088607595,
"grad_norm": 0.4902341067790985,
"learning_rate": 0.0015,
"loss": 2.0697,
"step": 3270
},
{
"epoch": 0.3459915611814346,
"grad_norm": 0.47359898686408997,
"learning_rate": 0.0015,
"loss": 2.0715,
"step": 3280
},
{
"epoch": 0.3470464135021097,
"grad_norm": 0.5300987958908081,
"learning_rate": 0.0015,
"loss": 2.0537,
"step": 3290
},
{
"epoch": 0.34810126582278483,
"grad_norm": 0.4852631688117981,
"learning_rate": 0.0015,
"loss": 2.0675,
"step": 3300
},
{
"epoch": 0.3491561181434599,
"grad_norm": 0.49096518754959106,
"learning_rate": 0.0015,
"loss": 2.0699,
"step": 3310
},
{
"epoch": 0.350210970464135,
"grad_norm": 0.4786725342273712,
"learning_rate": 0.0015,
"loss": 2.0695,
"step": 3320
},
{
"epoch": 0.3512658227848101,
"grad_norm": 0.4786069095134735,
"learning_rate": 0.0015,
"loss": 2.0587,
"step": 3330
},
{
"epoch": 0.35232067510548526,
"grad_norm": 0.529323399066925,
"learning_rate": 0.0015,
"loss": 2.058,
"step": 3340
},
{
"epoch": 0.35337552742616035,
"grad_norm": 0.45865610241889954,
"learning_rate": 0.0015,
"loss": 2.0528,
"step": 3350
},
{
"epoch": 0.35443037974683544,
"grad_norm": 0.45312967896461487,
"learning_rate": 0.0015,
"loss": 2.0501,
"step": 3360
},
{
"epoch": 0.35548523206751054,
"grad_norm": 0.48587530851364136,
"learning_rate": 0.0015,
"loss": 2.0594,
"step": 3370
},
{
"epoch": 0.35654008438818563,
"grad_norm": 0.4497227966785431,
"learning_rate": 0.0015,
"loss": 2.0552,
"step": 3380
},
{
"epoch": 0.3575949367088608,
"grad_norm": 0.4384129047393799,
"learning_rate": 0.0015,
"loss": 2.0482,
"step": 3390
},
{
"epoch": 0.35864978902953587,
"grad_norm": 0.47989708185195923,
"learning_rate": 0.0015,
"loss": 2.0617,
"step": 3400
},
{
"epoch": 0.35970464135021096,
"grad_norm": 0.5777666568756104,
"learning_rate": 0.0015,
"loss": 2.0587,
"step": 3410
},
{
"epoch": 0.36075949367088606,
"grad_norm": 0.5106038451194763,
"learning_rate": 0.0015,
"loss": 2.0394,
"step": 3420
},
{
"epoch": 0.3618143459915612,
"grad_norm": 0.4221823215484619,
"learning_rate": 0.0015,
"loss": 2.0559,
"step": 3430
},
{
"epoch": 0.3628691983122363,
"grad_norm": 0.4602856934070587,
"learning_rate": 0.0015,
"loss": 2.0448,
"step": 3440
},
{
"epoch": 0.3639240506329114,
"grad_norm": 0.45464810729026794,
"learning_rate": 0.0015,
"loss": 2.0508,
"step": 3450
},
{
"epoch": 0.3649789029535865,
"grad_norm": 0.4791298508644104,
"learning_rate": 0.0015,
"loss": 2.0377,
"step": 3460
},
{
"epoch": 0.36603375527426163,
"grad_norm": 0.4933064877986908,
"learning_rate": 0.0015,
"loss": 2.0569,
"step": 3470
},
{
"epoch": 0.3670886075949367,
"grad_norm": 0.4469650983810425,
"learning_rate": 0.0015,
"loss": 2.0604,
"step": 3480
},
{
"epoch": 0.3681434599156118,
"grad_norm": 0.49007025361061096,
"learning_rate": 0.0015,
"loss": 2.057,
"step": 3490
},
{
"epoch": 0.3691983122362869,
"grad_norm": 0.49333617091178894,
"learning_rate": 0.0015,
"loss": 2.0383,
"step": 3500
},
{
"epoch": 0.370253164556962,
"grad_norm": 0.4251687228679657,
"learning_rate": 0.0015,
"loss": 2.0538,
"step": 3510
},
{
"epoch": 0.37130801687763715,
"grad_norm": 0.4521111249923706,
"learning_rate": 0.0015,
"loss": 2.0465,
"step": 3520
},
{
"epoch": 0.37236286919831224,
"grad_norm": 0.46906667947769165,
"learning_rate": 0.0015,
"loss": 2.0612,
"step": 3530
},
{
"epoch": 0.37341772151898733,
"grad_norm": 0.4442053437232971,
"learning_rate": 0.0015,
"loss": 2.0575,
"step": 3540
},
{
"epoch": 0.3744725738396624,
"grad_norm": 0.4792494475841522,
"learning_rate": 0.0015,
"loss": 2.0446,
"step": 3550
},
{
"epoch": 0.3755274261603376,
"grad_norm": 0.4594126343727112,
"learning_rate": 0.0015,
"loss": 2.0446,
"step": 3560
},
{
"epoch": 0.37658227848101267,
"grad_norm": 0.4464714229106903,
"learning_rate": 0.0015,
"loss": 2.0402,
"step": 3570
},
{
"epoch": 0.37763713080168776,
"grad_norm": 0.5364466309547424,
"learning_rate": 0.0015,
"loss": 2.041,
"step": 3580
},
{
"epoch": 0.37869198312236285,
"grad_norm": 0.4744153320789337,
"learning_rate": 0.0015,
"loss": 2.0515,
"step": 3590
},
{
"epoch": 0.379746835443038,
"grad_norm": 0.4536152184009552,
"learning_rate": 0.0015,
"loss": 2.0367,
"step": 3600
},
{
"epoch": 0.3808016877637131,
"grad_norm": 0.514091968536377,
"learning_rate": 0.0015,
"loss": 2.0391,
"step": 3610
},
{
"epoch": 0.3818565400843882,
"grad_norm": 0.4596957564353943,
"learning_rate": 0.0015,
"loss": 2.0433,
"step": 3620
},
{
"epoch": 0.3829113924050633,
"grad_norm": 0.43287795782089233,
"learning_rate": 0.0015,
"loss": 2.0429,
"step": 3630
},
{
"epoch": 0.38396624472573837,
"grad_norm": 0.4937414526939392,
"learning_rate": 0.0015,
"loss": 2.0449,
"step": 3640
},
{
"epoch": 0.3850210970464135,
"grad_norm": 0.4595278203487396,
"learning_rate": 0.0015,
"loss": 2.0459,
"step": 3650
},
{
"epoch": 0.3860759493670886,
"grad_norm": 0.4414646029472351,
"learning_rate": 0.0015,
"loss": 2.0383,
"step": 3660
},
{
"epoch": 0.3871308016877637,
"grad_norm": 0.4963432550430298,
"learning_rate": 0.0015,
"loss": 2.0354,
"step": 3670
},
{
"epoch": 0.3881856540084388,
"grad_norm": 0.4459683895111084,
"learning_rate": 0.0015,
"loss": 2.0356,
"step": 3680
},
{
"epoch": 0.38924050632911394,
"grad_norm": 0.5404839515686035,
"learning_rate": 0.0015,
"loss": 2.0467,
"step": 3690
},
{
"epoch": 0.39029535864978904,
"grad_norm": 0.4792265295982361,
"learning_rate": 0.0015,
"loss": 2.053,
"step": 3700
},
{
"epoch": 0.39135021097046413,
"grad_norm": 0.4621272683143616,
"learning_rate": 0.0015,
"loss": 2.0392,
"step": 3710
},
{
"epoch": 0.3924050632911392,
"grad_norm": 0.3897222578525543,
"learning_rate": 0.0015,
"loss": 2.0163,
"step": 3720
},
{
"epoch": 0.39345991561181437,
"grad_norm": 0.5289087891578674,
"learning_rate": 0.0015,
"loss": 2.0355,
"step": 3730
},
{
"epoch": 0.39451476793248946,
"grad_norm": 0.497032105922699,
"learning_rate": 0.0015,
"loss": 2.0388,
"step": 3740
},
{
"epoch": 0.39556962025316456,
"grad_norm": 0.464529424905777,
"learning_rate": 0.0015,
"loss": 2.0368,
"step": 3750
},
{
"epoch": 0.39662447257383965,
"grad_norm": 0.5141547918319702,
"learning_rate": 0.0015,
"loss": 2.0313,
"step": 3760
},
{
"epoch": 0.39767932489451474,
"grad_norm": 0.5320308208465576,
"learning_rate": 0.0015,
"loss": 2.0457,
"step": 3770
},
{
"epoch": 0.3987341772151899,
"grad_norm": 0.4653412699699402,
"learning_rate": 0.0015,
"loss": 2.0315,
"step": 3780
},
{
"epoch": 0.399789029535865,
"grad_norm": 0.4444263279438019,
"learning_rate": 0.0015,
"loss": 2.0343,
"step": 3790
},
{
"epoch": 0.4008438818565401,
"grad_norm": 0.444671094417572,
"learning_rate": 0.0015,
"loss": 2.0378,
"step": 3800
},
{
"epoch": 0.40189873417721517,
"grad_norm": 0.4838058054447174,
"learning_rate": 0.0015,
"loss": 2.0343,
"step": 3810
},
{
"epoch": 0.4029535864978903,
"grad_norm": 0.47420215606689453,
"learning_rate": 0.0015,
"loss": 2.0185,
"step": 3820
},
{
"epoch": 0.4040084388185654,
"grad_norm": 0.48603472113609314,
"learning_rate": 0.0015,
"loss": 2.0268,
"step": 3830
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.5206816792488098,
"learning_rate": 0.0015,
"loss": 2.0435,
"step": 3840
},
{
"epoch": 0.4061181434599156,
"grad_norm": 0.5574729442596436,
"learning_rate": 0.0015,
"loss": 2.0265,
"step": 3850
},
{
"epoch": 0.40717299578059074,
"grad_norm": 0.48063263297080994,
"learning_rate": 0.0015,
"loss": 2.0411,
"step": 3860
},
{
"epoch": 0.40822784810126583,
"grad_norm": 0.4758067727088928,
"learning_rate": 0.0015,
"loss": 2.0247,
"step": 3870
},
{
"epoch": 0.4092827004219409,
"grad_norm": 0.4571913778781891,
"learning_rate": 0.0015,
"loss": 2.0172,
"step": 3880
},
{
"epoch": 0.410337552742616,
"grad_norm": 0.4651147723197937,
"learning_rate": 0.0015,
"loss": 2.0175,
"step": 3890
},
{
"epoch": 0.41139240506329117,
"grad_norm": 0.4823998212814331,
"learning_rate": 0.0015,
"loss": 2.0225,
"step": 3900
},
{
"epoch": 0.41244725738396626,
"grad_norm": 0.4863017797470093,
"learning_rate": 0.0015,
"loss": 2.0234,
"step": 3910
},
{
"epoch": 0.41350210970464135,
"grad_norm": 0.5648422241210938,
"learning_rate": 0.0015,
"loss": 2.0218,
"step": 3920
},
{
"epoch": 0.41455696202531644,
"grad_norm": 0.4649427533149719,
"learning_rate": 0.0015,
"loss": 2.0249,
"step": 3930
},
{
"epoch": 0.41561181434599154,
"grad_norm": 0.55061274766922,
"learning_rate": 0.0015,
"loss": 2.0203,
"step": 3940
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.4301551282405853,
"learning_rate": 0.0015,
"loss": 2.0268,
"step": 3950
},
{
"epoch": 0.4177215189873418,
"grad_norm": 0.4754408299922943,
"learning_rate": 0.0015,
"loss": 2.0229,
"step": 3960
},
{
"epoch": 0.41877637130801687,
"grad_norm": 0.43963631987571716,
"learning_rate": 0.0015,
"loss": 2.0149,
"step": 3970
},
{
"epoch": 0.41983122362869196,
"grad_norm": 0.4324744641780853,
"learning_rate": 0.0015,
"loss": 2.0223,
"step": 3980
},
{
"epoch": 0.4208860759493671,
"grad_norm": 0.40844643115997314,
"learning_rate": 0.0015,
"loss": 2.0142,
"step": 3990
},
{
"epoch": 0.4219409282700422,
"grad_norm": 0.5064060688018799,
"learning_rate": 0.0015,
"loss": 2.0341,
"step": 4000
},
{
"epoch": 0.4229957805907173,
"grad_norm": 0.463920921087265,
"learning_rate": 0.0015,
"loss": 2.0165,
"step": 4010
},
{
"epoch": 0.4240506329113924,
"grad_norm": 0.41492554545402527,
"learning_rate": 0.0015,
"loss": 2.0221,
"step": 4020
},
{
"epoch": 0.42510548523206754,
"grad_norm": 0.490600049495697,
"learning_rate": 0.0015,
"loss": 2.012,
"step": 4030
},
{
"epoch": 0.42616033755274263,
"grad_norm": 0.44256430864334106,
"learning_rate": 0.0015,
"loss": 2.0225,
"step": 4040
},
{
"epoch": 0.4272151898734177,
"grad_norm": 0.5451088547706604,
"learning_rate": 0.0015,
"loss": 2.0237,
"step": 4050
},
{
"epoch": 0.4282700421940928,
"grad_norm": 0.4461345076560974,
"learning_rate": 0.0015,
"loss": 2.0311,
"step": 4060
},
{
"epoch": 0.4293248945147679,
"grad_norm": 0.4550500512123108,
"learning_rate": 0.0015,
"loss": 2.0216,
"step": 4070
},
{
"epoch": 0.43037974683544306,
"grad_norm": 0.5397707223892212,
"learning_rate": 0.0015,
"loss": 2.0264,
"step": 4080
},
{
"epoch": 0.43143459915611815,
"grad_norm": 0.4498850405216217,
"learning_rate": 0.0015,
"loss": 2.0201,
"step": 4090
},
{
"epoch": 0.43248945147679324,
"grad_norm": 0.45243409276008606,
"learning_rate": 0.0015,
"loss": 2.0145,
"step": 4100
},
{
"epoch": 0.43354430379746833,
"grad_norm": 0.46291717886924744,
"learning_rate": 0.0015,
"loss": 2.0125,
"step": 4110
},
{
"epoch": 0.4345991561181435,
"grad_norm": 0.4428139328956604,
"learning_rate": 0.0015,
"loss": 2.0166,
"step": 4120
},
{
"epoch": 0.4356540084388186,
"grad_norm": 0.5002286434173584,
"learning_rate": 0.0015,
"loss": 2.0118,
"step": 4130
},
{
"epoch": 0.43670886075949367,
"grad_norm": 0.4523918926715851,
"learning_rate": 0.0015,
"loss": 2.012,
"step": 4140
},
{
"epoch": 0.43776371308016876,
"grad_norm": 0.4352469742298126,
"learning_rate": 0.0015,
"loss": 2.0215,
"step": 4150
},
{
"epoch": 0.4388185654008439,
"grad_norm": 0.502491295337677,
"learning_rate": 0.0015,
"loss": 2.0121,
"step": 4160
},
{
"epoch": 0.439873417721519,
"grad_norm": 0.4340950846672058,
"learning_rate": 0.0015,
"loss": 2.0134,
"step": 4170
},
{
"epoch": 0.4409282700421941,
"grad_norm": 0.42705026268959045,
"learning_rate": 0.0015,
"loss": 2.0095,
"step": 4180
},
{
"epoch": 0.4419831223628692,
"grad_norm": 0.44456130266189575,
"learning_rate": 0.0015,
"loss": 2.0157,
"step": 4190
},
{
"epoch": 0.4430379746835443,
"grad_norm": 0.44883906841278076,
"learning_rate": 0.0015,
"loss": 2.012,
"step": 4200
},
{
"epoch": 0.4440928270042194,
"grad_norm": 0.43857628107070923,
"learning_rate": 0.0015,
"loss": 2.0005,
"step": 4210
},
{
"epoch": 0.4451476793248945,
"grad_norm": 0.46458226442337036,
"learning_rate": 0.0015,
"loss": 2.0051,
"step": 4220
},
{
"epoch": 0.4462025316455696,
"grad_norm": 0.5564199686050415,
"learning_rate": 0.0015,
"loss": 2.0069,
"step": 4230
},
{
"epoch": 0.4472573839662447,
"grad_norm": 0.43587273359298706,
"learning_rate": 0.0015,
"loss": 2.0094,
"step": 4240
},
{
"epoch": 0.44831223628691985,
"grad_norm": 0.4776957929134369,
"learning_rate": 0.0015,
"loss": 2.0142,
"step": 4250
},
{
"epoch": 0.44936708860759494,
"grad_norm": 0.5066412091255188,
"learning_rate": 0.0015,
"loss": 2.0189,
"step": 4260
},
{
"epoch": 0.45042194092827004,
"grad_norm": 0.46694788336753845,
"learning_rate": 0.0015,
"loss": 2.0051,
"step": 4270
},
{
"epoch": 0.45147679324894513,
"grad_norm": 0.5150187015533447,
"learning_rate": 0.0015,
"loss": 2.0017,
"step": 4280
},
{
"epoch": 0.4525316455696203,
"grad_norm": 0.5508468747138977,
"learning_rate": 0.0015,
"loss": 2.0142,
"step": 4290
},
{
"epoch": 0.45358649789029537,
"grad_norm": 0.49552634358406067,
"learning_rate": 0.0015,
"loss": 2.013,
"step": 4300
},
{
"epoch": 0.45464135021097046,
"grad_norm": 0.4847050905227661,
"learning_rate": 0.0015,
"loss": 2.0094,
"step": 4310
},
{
"epoch": 0.45569620253164556,
"grad_norm": 0.4787251949310303,
"learning_rate": 0.0015,
"loss": 2.001,
"step": 4320
},
{
"epoch": 0.45675105485232065,
"grad_norm": 0.48696401715278625,
"learning_rate": 0.0015,
"loss": 2.0134,
"step": 4330
},
{
"epoch": 0.4578059071729958,
"grad_norm": 0.42618006467819214,
"learning_rate": 0.0015,
"loss": 2.0014,
"step": 4340
},
{
"epoch": 0.4588607594936709,
"grad_norm": 0.44857847690582275,
"learning_rate": 0.0015,
"loss": 2.0068,
"step": 4350
},
{
"epoch": 0.459915611814346,
"grad_norm": 0.47533655166625977,
"learning_rate": 0.0015,
"loss": 2.0225,
"step": 4360
},
{
"epoch": 0.4609704641350211,
"grad_norm": 0.45596417784690857,
"learning_rate": 0.0015,
"loss": 1.9968,
"step": 4370
},
{
"epoch": 0.4620253164556962,
"grad_norm": 0.42831170558929443,
"learning_rate": 0.0015,
"loss": 2.0036,
"step": 4380
},
{
"epoch": 0.4630801687763713,
"grad_norm": 0.4378739893436432,
"learning_rate": 0.0015,
"loss": 1.9962,
"step": 4390
},
{
"epoch": 0.4641350210970464,
"grad_norm": 0.5626639723777771,
"learning_rate": 0.0015,
"loss": 2.0053,
"step": 4400
},
{
"epoch": 0.4651898734177215,
"grad_norm": 0.46488156914711,
"learning_rate": 0.0015,
"loss": 2.015,
"step": 4410
},
{
"epoch": 0.46624472573839665,
"grad_norm": 0.45077210664749146,
"learning_rate": 0.0015,
"loss": 2.0003,
"step": 4420
},
{
"epoch": 0.46729957805907174,
"grad_norm": 0.4862898290157318,
"learning_rate": 0.0015,
"loss": 2.0038,
"step": 4430
},
{
"epoch": 0.46835443037974683,
"grad_norm": 0.4480973482131958,
"learning_rate": 0.0015,
"loss": 2.006,
"step": 4440
},
{
"epoch": 0.4694092827004219,
"grad_norm": 0.44025543332099915,
"learning_rate": 0.0015,
"loss": 2.0102,
"step": 4450
},
{
"epoch": 0.4704641350210971,
"grad_norm": 0.45903876423835754,
"learning_rate": 0.0015,
"loss": 1.9943,
"step": 4460
},
{
"epoch": 0.47151898734177217,
"grad_norm": 0.4817822277545929,
"learning_rate": 0.0015,
"loss": 2.0085,
"step": 4470
},
{
"epoch": 0.47257383966244726,
"grad_norm": 0.4522876441478729,
"learning_rate": 0.0015,
"loss": 2.0136,
"step": 4480
},
{
"epoch": 0.47362869198312235,
"grad_norm": 0.48608750104904175,
"learning_rate": 0.0015,
"loss": 1.9993,
"step": 4490
},
{
"epoch": 0.47468354430379744,
"grad_norm": 0.45206281542778015,
"learning_rate": 0.0015,
"loss": 1.9961,
"step": 4500
},
{
"epoch": 0.4757383966244726,
"grad_norm": 0.4605315327644348,
"learning_rate": 0.0015,
"loss": 2.0039,
"step": 4510
},
{
"epoch": 0.4767932489451477,
"grad_norm": 0.45295825600624084,
"learning_rate": 0.0015,
"loss": 1.9876,
"step": 4520
},
{
"epoch": 0.4778481012658228,
"grad_norm": 0.48463356494903564,
"learning_rate": 0.0015,
"loss": 2.0063,
"step": 4530
},
{
"epoch": 0.47890295358649787,
"grad_norm": 0.4252591133117676,
"learning_rate": 0.0015,
"loss": 2.0067,
"step": 4540
},
{
"epoch": 0.479957805907173,
"grad_norm": 0.5178452134132385,
"learning_rate": 0.0015,
"loss": 2.003,
"step": 4550
},
{
"epoch": 0.4810126582278481,
"grad_norm": 0.4417160749435425,
"learning_rate": 0.0015,
"loss": 1.9941,
"step": 4560
},
{
"epoch": 0.4820675105485232,
"grad_norm": 0.46367692947387695,
"learning_rate": 0.0015,
"loss": 1.9731,
"step": 4570
},
{
"epoch": 0.4831223628691983,
"grad_norm": 0.4723823666572571,
"learning_rate": 0.0015,
"loss": 1.9998,
"step": 4580
},
{
"epoch": 0.48417721518987344,
"grad_norm": 0.46017590165138245,
"learning_rate": 0.0015,
"loss": 1.9865,
"step": 4590
},
{
"epoch": 0.48523206751054854,
"grad_norm": 0.5154760479927063,
"learning_rate": 0.0015,
"loss": 2.0025,
"step": 4600
},
{
"epoch": 0.48628691983122363,
"grad_norm": 0.4670974612236023,
"learning_rate": 0.0015,
"loss": 2.0056,
"step": 4610
},
{
"epoch": 0.4873417721518987,
"grad_norm": 0.4248136878013611,
"learning_rate": 0.0015,
"loss": 1.9816,
"step": 4620
},
{
"epoch": 0.4883966244725738,
"grad_norm": 0.4667222797870636,
"learning_rate": 0.0015,
"loss": 2.0021,
"step": 4630
},
{
"epoch": 0.48945147679324896,
"grad_norm": 0.4552115499973297,
"learning_rate": 0.0015,
"loss": 2.0007,
"step": 4640
},
{
"epoch": 0.49050632911392406,
"grad_norm": 0.46759188175201416,
"learning_rate": 0.0015,
"loss": 2.0032,
"step": 4650
},
{
"epoch": 0.49156118143459915,
"grad_norm": 0.505689263343811,
"learning_rate": 0.0015,
"loss": 1.9887,
"step": 4660
},
{
"epoch": 0.49261603375527424,
"grad_norm": 0.49472060799598694,
"learning_rate": 0.0015,
"loss": 1.9916,
"step": 4670
},
{
"epoch": 0.4936708860759494,
"grad_norm": 0.4831889867782593,
"learning_rate": 0.0015,
"loss": 1.9867,
"step": 4680
},
{
"epoch": 0.4947257383966245,
"grad_norm": 0.4475608170032501,
"learning_rate": 0.0015,
"loss": 1.9846,
"step": 4690
},
{
"epoch": 0.4957805907172996,
"grad_norm": 0.45800697803497314,
"learning_rate": 0.0015,
"loss": 1.998,
"step": 4700
},
{
"epoch": 0.49683544303797467,
"grad_norm": 0.463772177696228,
"learning_rate": 0.0015,
"loss": 1.9962,
"step": 4710
},
{
"epoch": 0.4978902953586498,
"grad_norm": 0.4545440971851349,
"learning_rate": 0.0015,
"loss": 1.9925,
"step": 4720
},
{
"epoch": 0.4989451476793249,
"grad_norm": 0.4118850529193878,
"learning_rate": 0.0015,
"loss": 1.9913,
"step": 4730
},
{
"epoch": 0.5,
"grad_norm": 0.4960225522518158,
"learning_rate": 0.0015,
"loss": 1.9887,
"step": 4740
},
{
"epoch": 0.5010548523206751,
"grad_norm": 0.48252198100090027,
"learning_rate": 0.0015,
"loss": 2.0057,
"step": 4750
},
{
"epoch": 0.5021097046413502,
"grad_norm": 0.4500682055950165,
"learning_rate": 0.0015,
"loss": 1.9894,
"step": 4760
},
{
"epoch": 0.5031645569620253,
"grad_norm": 0.4634728729724884,
"learning_rate": 0.0015,
"loss": 1.9879,
"step": 4770
},
{
"epoch": 0.5042194092827004,
"grad_norm": 0.46008265018463135,
"learning_rate": 0.0015,
"loss": 1.996,
"step": 4780
},
{
"epoch": 0.5052742616033755,
"grad_norm": 0.489211767911911,
"learning_rate": 0.0015,
"loss": 1.9842,
"step": 4790
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.445319801568985,
"learning_rate": 0.0015,
"loss": 1.9871,
"step": 4800
},
{
"epoch": 0.5073839662447257,
"grad_norm": 0.4555240869522095,
"learning_rate": 0.0015,
"loss": 1.9885,
"step": 4810
},
{
"epoch": 0.5084388185654009,
"grad_norm": 0.4640772044658661,
"learning_rate": 0.0015,
"loss": 2.0086,
"step": 4820
},
{
"epoch": 0.509493670886076,
"grad_norm": 0.4786563813686371,
"learning_rate": 0.0015,
"loss": 1.9898,
"step": 4830
},
{
"epoch": 0.510548523206751,
"grad_norm": 0.44662174582481384,
"learning_rate": 0.0015,
"loss": 1.9836,
"step": 4840
},
{
"epoch": 0.5116033755274262,
"grad_norm": 0.4463147521018982,
"learning_rate": 0.0015,
"loss": 1.998,
"step": 4850
},
{
"epoch": 0.5126582278481012,
"grad_norm": 0.4866524338722229,
"learning_rate": 0.0015,
"loss": 1.982,
"step": 4860
},
{
"epoch": 0.5137130801687764,
"grad_norm": 0.4643232226371765,
"learning_rate": 0.0015,
"loss": 1.9781,
"step": 4870
},
{
"epoch": 0.5147679324894515,
"grad_norm": 0.49099764227867126,
"learning_rate": 0.0015,
"loss": 1.9965,
"step": 4880
},
{
"epoch": 0.5158227848101266,
"grad_norm": 0.4760585427284241,
"learning_rate": 0.0015,
"loss": 1.9889,
"step": 4890
},
{
"epoch": 0.5168776371308017,
"grad_norm": 0.4203405976295471,
"learning_rate": 0.0015,
"loss": 1.994,
"step": 4900
},
{
"epoch": 0.5179324894514767,
"grad_norm": 0.4211159348487854,
"learning_rate": 0.0015,
"loss": 1.9888,
"step": 4910
},
{
"epoch": 0.5189873417721519,
"grad_norm": 0.46969079971313477,
"learning_rate": 0.0015,
"loss": 1.9813,
"step": 4920
},
{
"epoch": 0.520042194092827,
"grad_norm": 0.4826894998550415,
"learning_rate": 0.0015,
"loss": 1.9888,
"step": 4930
},
{
"epoch": 0.5210970464135021,
"grad_norm": 0.4152940511703491,
"learning_rate": 0.0015,
"loss": 1.9859,
"step": 4940
},
{
"epoch": 0.5221518987341772,
"grad_norm": 0.4372897744178772,
"learning_rate": 0.0015,
"loss": 1.998,
"step": 4950
},
{
"epoch": 0.5232067510548524,
"grad_norm": 0.47398361563682556,
"learning_rate": 0.0015,
"loss": 1.9737,
"step": 4960
},
{
"epoch": 0.5242616033755274,
"grad_norm": 0.5493957996368408,
"learning_rate": 0.0015,
"loss": 1.9861,
"step": 4970
},
{
"epoch": 0.5253164556962026,
"grad_norm": 0.4829331338405609,
"learning_rate": 0.0015,
"loss": 1.9797,
"step": 4980
},
{
"epoch": 0.5263713080168776,
"grad_norm": 0.43320485949516296,
"learning_rate": 0.0015,
"loss": 1.9816,
"step": 4990
},
{
"epoch": 0.5274261603375527,
"grad_norm": 0.4184596538543701,
"learning_rate": 0.0015,
"loss": 1.9937,
"step": 5000
},
{
"epoch": 0.5284810126582279,
"grad_norm": 0.44668543338775635,
"learning_rate": 0.0015,
"loss": 1.9817,
"step": 5010
},
{
"epoch": 0.5295358649789029,
"grad_norm": 0.5183205604553223,
"learning_rate": 0.0015,
"loss": 1.9779,
"step": 5020
},
{
"epoch": 0.5305907172995781,
"grad_norm": 0.4478100538253784,
"learning_rate": 0.0015,
"loss": 1.9669,
"step": 5030
},
{
"epoch": 0.5316455696202531,
"grad_norm": 0.4046509265899658,
"learning_rate": 0.0015,
"loss": 1.9882,
"step": 5040
},
{
"epoch": 0.5327004219409283,
"grad_norm": 0.4287264943122864,
"learning_rate": 0.0015,
"loss": 1.9743,
"step": 5050
},
{
"epoch": 0.5337552742616034,
"grad_norm": 0.46131548285484314,
"learning_rate": 0.0015,
"loss": 1.9925,
"step": 5060
},
{
"epoch": 0.5348101265822784,
"grad_norm": 0.4695143401622772,
"learning_rate": 0.0015,
"loss": 1.9816,
"step": 5070
},
{
"epoch": 0.5358649789029536,
"grad_norm": 0.433845579624176,
"learning_rate": 0.0015,
"loss": 1.9909,
"step": 5080
},
{
"epoch": 0.5369198312236287,
"grad_norm": 0.4523727297782898,
"learning_rate": 0.0015,
"loss": 1.9869,
"step": 5090
},
{
"epoch": 0.5379746835443038,
"grad_norm": 0.4685346782207489,
"learning_rate": 0.0015,
"loss": 1.9675,
"step": 5100
},
{
"epoch": 0.5390295358649789,
"grad_norm": 0.4254748821258545,
"learning_rate": 0.0015,
"loss": 1.9816,
"step": 5110
},
{
"epoch": 0.540084388185654,
"grad_norm": 0.46817123889923096,
"learning_rate": 0.0015,
"loss": 1.9789,
"step": 5120
},
{
"epoch": 0.5411392405063291,
"grad_norm": 0.43347081542015076,
"learning_rate": 0.0015,
"loss": 1.9776,
"step": 5130
},
{
"epoch": 0.5421940928270043,
"grad_norm": 0.4349901080131531,
"learning_rate": 0.0015,
"loss": 1.991,
"step": 5140
},
{
"epoch": 0.5432489451476793,
"grad_norm": 0.43023401498794556,
"learning_rate": 0.0015,
"loss": 1.9843,
"step": 5150
},
{
"epoch": 0.5443037974683544,
"grad_norm": 0.434885174036026,
"learning_rate": 0.0015,
"loss": 1.9818,
"step": 5160
},
{
"epoch": 0.5453586497890295,
"grad_norm": 0.43523484468460083,
"learning_rate": 0.0015,
"loss": 1.9804,
"step": 5170
},
{
"epoch": 0.5464135021097046,
"grad_norm": 0.43468087911605835,
"learning_rate": 0.0015,
"loss": 1.9945,
"step": 5180
},
{
"epoch": 0.5474683544303798,
"grad_norm": 0.4559604525566101,
"learning_rate": 0.0015,
"loss": 1.9718,
"step": 5190
},
{
"epoch": 0.5485232067510548,
"grad_norm": 0.45978057384490967,
"learning_rate": 0.0015,
"loss": 1.9846,
"step": 5200
},
{
"epoch": 0.54957805907173,
"grad_norm": 0.3937625586986542,
"learning_rate": 0.0015,
"loss": 1.9651,
"step": 5210
},
{
"epoch": 0.5506329113924051,
"grad_norm": 0.4718064069747925,
"learning_rate": 0.0015,
"loss": 1.9792,
"step": 5220
},
{
"epoch": 0.5516877637130801,
"grad_norm": 0.4999498426914215,
"learning_rate": 0.0015,
"loss": 1.9754,
"step": 5230
},
{
"epoch": 0.5527426160337553,
"grad_norm": 0.47621315717697144,
"learning_rate": 0.0015,
"loss": 1.98,
"step": 5240
},
{
"epoch": 0.5537974683544303,
"grad_norm": 0.44971024990081787,
"learning_rate": 0.0015,
"loss": 1.971,
"step": 5250
},
{
"epoch": 0.5548523206751055,
"grad_norm": 0.4371758699417114,
"learning_rate": 0.0015,
"loss": 1.9741,
"step": 5260
},
{
"epoch": 0.5559071729957806,
"grad_norm": 0.4647037982940674,
"learning_rate": 0.0015,
"loss": 1.9827,
"step": 5270
},
{
"epoch": 0.5569620253164557,
"grad_norm": 0.5894455909729004,
"learning_rate": 0.0015,
"loss": 1.9752,
"step": 5280
},
{
"epoch": 0.5580168776371308,
"grad_norm": 0.4337363839149475,
"learning_rate": 0.0015,
"loss": 1.976,
"step": 5290
},
{
"epoch": 0.5590717299578059,
"grad_norm": 0.4197293519973755,
"learning_rate": 0.0015,
"loss": 1.9784,
"step": 5300
},
{
"epoch": 0.560126582278481,
"grad_norm": 0.46126967668533325,
"learning_rate": 0.0015,
"loss": 1.9671,
"step": 5310
},
{
"epoch": 0.5611814345991561,
"grad_norm": 0.4594750702381134,
"learning_rate": 0.0015,
"loss": 1.9685,
"step": 5320
},
{
"epoch": 0.5622362869198312,
"grad_norm": 0.41427746415138245,
"learning_rate": 0.0015,
"loss": 1.9812,
"step": 5330
},
{
"epoch": 0.5632911392405063,
"grad_norm": 0.41871240735054016,
"learning_rate": 0.0015,
"loss": 1.9702,
"step": 5340
},
{
"epoch": 0.5643459915611815,
"grad_norm": 0.4757276177406311,
"learning_rate": 0.0015,
"loss": 1.9606,
"step": 5350
},
{
"epoch": 0.5654008438818565,
"grad_norm": 0.5399381518363953,
"learning_rate": 0.0015,
"loss": 1.9752,
"step": 5360
},
{
"epoch": 0.5664556962025317,
"grad_norm": 0.43040159344673157,
"learning_rate": 0.0015,
"loss": 1.9782,
"step": 5370
},
{
"epoch": 0.5675105485232067,
"grad_norm": 0.46205800771713257,
"learning_rate": 0.0015,
"loss": 1.9658,
"step": 5380
},
{
"epoch": 0.5685654008438819,
"grad_norm": 0.4588663578033447,
"learning_rate": 0.0015,
"loss": 1.9691,
"step": 5390
},
{
"epoch": 0.569620253164557,
"grad_norm": 0.44170957803726196,
"learning_rate": 0.0015,
"loss": 1.9674,
"step": 5400
},
{
"epoch": 0.570675105485232,
"grad_norm": 0.42715734243392944,
"learning_rate": 0.0015,
"loss": 1.9688,
"step": 5410
},
{
"epoch": 0.5717299578059072,
"grad_norm": 0.4816579818725586,
"learning_rate": 0.0015,
"loss": 1.9638,
"step": 5420
},
{
"epoch": 0.5727848101265823,
"grad_norm": 0.46428218483924866,
"learning_rate": 0.0015,
"loss": 1.9637,
"step": 5430
},
{
"epoch": 0.5738396624472574,
"grad_norm": 0.43607380986213684,
"learning_rate": 0.0015,
"loss": 1.9593,
"step": 5440
},
{
"epoch": 0.5748945147679325,
"grad_norm": 0.4220513701438904,
"learning_rate": 0.0015,
"loss": 1.9736,
"step": 5450
},
{
"epoch": 0.5759493670886076,
"grad_norm": 0.39790844917297363,
"learning_rate": 0.0015,
"loss": 1.971,
"step": 5460
},
{
"epoch": 0.5770042194092827,
"grad_norm": 0.4477349817752838,
"learning_rate": 0.0015,
"loss": 1.9656,
"step": 5470
},
{
"epoch": 0.5780590717299579,
"grad_norm": 0.44955015182495117,
"learning_rate": 0.0015,
"loss": 1.974,
"step": 5480
},
{
"epoch": 0.5791139240506329,
"grad_norm": 0.44701093435287476,
"learning_rate": 0.0015,
"loss": 1.9717,
"step": 5490
},
{
"epoch": 0.580168776371308,
"grad_norm": 0.4384399950504303,
"learning_rate": 0.0015,
"loss": 1.9659,
"step": 5500
},
{
"epoch": 0.5812236286919831,
"grad_norm": 0.4545252323150635,
"learning_rate": 0.0015,
"loss": 1.9755,
"step": 5510
},
{
"epoch": 0.5822784810126582,
"grad_norm": 0.46129509806632996,
"learning_rate": 0.0015,
"loss": 1.9656,
"step": 5520
},
{
"epoch": 0.5833333333333334,
"grad_norm": 0.4120708703994751,
"learning_rate": 0.0015,
"loss": 1.9609,
"step": 5530
},
{
"epoch": 0.5843881856540084,
"grad_norm": 0.45053553581237793,
"learning_rate": 0.0015,
"loss": 1.9585,
"step": 5540
},
{
"epoch": 0.5854430379746836,
"grad_norm": 0.4825860559940338,
"learning_rate": 0.0015,
"loss": 1.9737,
"step": 5550
},
{
"epoch": 0.5864978902953587,
"grad_norm": 0.42986005544662476,
"learning_rate": 0.0015,
"loss": 1.9796,
"step": 5560
},
{
"epoch": 0.5875527426160337,
"grad_norm": 0.45834633708000183,
"learning_rate": 0.0015,
"loss": 1.9632,
"step": 5570
},
{
"epoch": 0.5886075949367089,
"grad_norm": 0.43091991543769836,
"learning_rate": 0.0015,
"loss": 1.9597,
"step": 5580
},
{
"epoch": 0.5896624472573839,
"grad_norm": 0.4816896319389343,
"learning_rate": 0.0015,
"loss": 1.9655,
"step": 5590
},
{
"epoch": 0.5907172995780591,
"grad_norm": 0.44736018776893616,
"learning_rate": 0.0015,
"loss": 1.9636,
"step": 5600
},
{
"epoch": 0.5917721518987342,
"grad_norm": 0.46847572922706604,
"learning_rate": 0.0015,
"loss": 1.9637,
"step": 5610
},
{
"epoch": 0.5928270042194093,
"grad_norm": 0.4232116639614105,
"learning_rate": 0.0015,
"loss": 1.9726,
"step": 5620
},
{
"epoch": 0.5938818565400844,
"grad_norm": 0.4705914556980133,
"learning_rate": 0.0015,
"loss": 1.9672,
"step": 5630
},
{
"epoch": 0.5949367088607594,
"grad_norm": 0.450400173664093,
"learning_rate": 0.0015,
"loss": 1.9768,
"step": 5640
},
{
"epoch": 0.5959915611814346,
"grad_norm": 0.4197351038455963,
"learning_rate": 0.0015,
"loss": 1.9741,
"step": 5650
},
{
"epoch": 0.5970464135021097,
"grad_norm": 0.4076090157032013,
"learning_rate": 0.0015,
"loss": 1.9676,
"step": 5660
},
{
"epoch": 0.5981012658227848,
"grad_norm": 0.4663560688495636,
"learning_rate": 0.0015,
"loss": 1.9779,
"step": 5670
},
{
"epoch": 0.5991561181434599,
"grad_norm": 0.4360770285129547,
"learning_rate": 0.0015,
"loss": 1.9764,
"step": 5680
},
{
"epoch": 0.6002109704641351,
"grad_norm": 0.48296862840652466,
"learning_rate": 0.0015,
"loss": 1.9544,
"step": 5690
},
{
"epoch": 0.6012658227848101,
"grad_norm": 0.5105125308036804,
"learning_rate": 0.0015,
"loss": 1.9578,
"step": 5700
},
{
"epoch": 0.6023206751054853,
"grad_norm": 0.465209037065506,
"learning_rate": 0.0015,
"loss": 1.9654,
"step": 5710
},
{
"epoch": 0.6033755274261603,
"grad_norm": 0.4125451147556305,
"learning_rate": 0.0015,
"loss": 1.9658,
"step": 5720
},
{
"epoch": 0.6044303797468354,
"grad_norm": 0.44935154914855957,
"learning_rate": 0.0015,
"loss": 1.9567,
"step": 5730
},
{
"epoch": 0.6054852320675106,
"grad_norm": 0.4448432922363281,
"learning_rate": 0.0015,
"loss": 1.9608,
"step": 5740
},
{
"epoch": 0.6065400843881856,
"grad_norm": 0.44799625873565674,
"learning_rate": 0.0015,
"loss": 1.9359,
"step": 5750
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.44629913568496704,
"learning_rate": 0.0015,
"loss": 1.9477,
"step": 5760
},
{
"epoch": 0.6086497890295358,
"grad_norm": 0.4163765609264374,
"learning_rate": 0.0015,
"loss": 1.9526,
"step": 5770
},
{
"epoch": 0.609704641350211,
"grad_norm": 0.4490468204021454,
"learning_rate": 0.0015,
"loss": 1.9567,
"step": 5780
},
{
"epoch": 0.6107594936708861,
"grad_norm": 0.45518097281455994,
"learning_rate": 0.0015,
"loss": 1.952,
"step": 5790
},
{
"epoch": 0.6118143459915611,
"grad_norm": 0.4716593325138092,
"learning_rate": 0.0015,
"loss": 1.9624,
"step": 5800
},
{
"epoch": 0.6128691983122363,
"grad_norm": 0.4235859215259552,
"learning_rate": 0.0015,
"loss": 1.9624,
"step": 5810
},
{
"epoch": 0.6139240506329114,
"grad_norm": 0.4250667989253998,
"learning_rate": 0.0015,
"loss": 1.973,
"step": 5820
},
{
"epoch": 0.6149789029535865,
"grad_norm": 0.4351049065589905,
"learning_rate": 0.0015,
"loss": 1.9551,
"step": 5830
},
{
"epoch": 0.6160337552742616,
"grad_norm": 0.4573591947555542,
"learning_rate": 0.0015,
"loss": 1.9703,
"step": 5840
},
{
"epoch": 0.6170886075949367,
"grad_norm": 0.5216925740242004,
"learning_rate": 0.0015,
"loss": 1.9746,
"step": 5850
},
{
"epoch": 0.6181434599156118,
"grad_norm": 0.48470115661621094,
"learning_rate": 0.0015,
"loss": 1.9637,
"step": 5860
},
{
"epoch": 0.619198312236287,
"grad_norm": 0.451644629240036,
"learning_rate": 0.0015,
"loss": 1.9556,
"step": 5870
},
{
"epoch": 0.620253164556962,
"grad_norm": 0.4359816014766693,
"learning_rate": 0.0015,
"loss": 1.9662,
"step": 5880
},
{
"epoch": 0.6213080168776371,
"grad_norm": 0.5675661563873291,
"learning_rate": 0.0015,
"loss": 1.9608,
"step": 5890
},
{
"epoch": 0.6223628691983122,
"grad_norm": 0.3987058103084564,
"learning_rate": 0.0015,
"loss": 1.9614,
"step": 5900
},
{
"epoch": 0.6234177215189873,
"grad_norm": 0.4844250977039337,
"learning_rate": 0.0015,
"loss": 1.9527,
"step": 5910
},
{
"epoch": 0.6244725738396625,
"grad_norm": 0.39574798941612244,
"learning_rate": 0.0015,
"loss": 1.9496,
"step": 5920
},
{
"epoch": 0.6255274261603375,
"grad_norm": 0.6383304595947266,
"learning_rate": 0.0015,
"loss": 1.9576,
"step": 5930
},
{
"epoch": 0.6265822784810127,
"grad_norm": 0.4334365725517273,
"learning_rate": 0.0015,
"loss": 1.9622,
"step": 5940
},
{
"epoch": 0.6276371308016878,
"grad_norm": 0.4857499301433563,
"learning_rate": 0.0015,
"loss": 1.9487,
"step": 5950
},
{
"epoch": 0.6286919831223629,
"grad_norm": 0.46174830198287964,
"learning_rate": 0.0015,
"loss": 1.9569,
"step": 5960
},
{
"epoch": 0.629746835443038,
"grad_norm": 0.44857001304626465,
"learning_rate": 0.0015,
"loss": 1.9579,
"step": 5970
},
{
"epoch": 0.630801687763713,
"grad_norm": 0.4137052297592163,
"learning_rate": 0.0015,
"loss": 1.9404,
"step": 5980
},
{
"epoch": 0.6318565400843882,
"grad_norm": 0.4595361649990082,
"learning_rate": 0.0015,
"loss": 1.9581,
"step": 5990
},
{
"epoch": 0.6329113924050633,
"grad_norm": 0.47880202531814575,
"learning_rate": 0.0015,
"loss": 1.9679,
"step": 6000
},
{
"epoch": 0.6339662447257384,
"grad_norm": 0.47764644026756287,
"learning_rate": 0.0015,
"loss": 1.9593,
"step": 6010
},
{
"epoch": 0.6350210970464135,
"grad_norm": 0.43710753321647644,
"learning_rate": 0.0015,
"loss": 1.9655,
"step": 6020
},
{
"epoch": 0.6360759493670886,
"grad_norm": 0.4158395230770111,
"learning_rate": 0.0015,
"loss": 1.9437,
"step": 6030
},
{
"epoch": 0.6371308016877637,
"grad_norm": 0.4571756422519684,
"learning_rate": 0.0015,
"loss": 1.9627,
"step": 6040
},
{
"epoch": 0.6381856540084389,
"grad_norm": 0.5196099281311035,
"learning_rate": 0.0015,
"loss": 1.9435,
"step": 6050
},
{
"epoch": 0.6392405063291139,
"grad_norm": 0.4487728476524353,
"learning_rate": 0.0015,
"loss": 1.9611,
"step": 6060
},
{
"epoch": 0.640295358649789,
"grad_norm": 0.5082179307937622,
"learning_rate": 0.0015,
"loss": 1.9533,
"step": 6070
},
{
"epoch": 0.6413502109704642,
"grad_norm": 0.4590562880039215,
"learning_rate": 0.0015,
"loss": 1.9481,
"step": 6080
},
{
"epoch": 0.6424050632911392,
"grad_norm": 0.42560580372810364,
"learning_rate": 0.0015,
"loss": 1.9488,
"step": 6090
},
{
"epoch": 0.6434599156118144,
"grad_norm": 0.44798460602760315,
"learning_rate": 0.0015,
"loss": 1.9419,
"step": 6100
},
{
"epoch": 0.6445147679324894,
"grad_norm": 0.4697492718696594,
"learning_rate": 0.0015,
"loss": 1.9605,
"step": 6110
},
{
"epoch": 0.6455696202531646,
"grad_norm": 0.4113924205303192,
"learning_rate": 0.0015,
"loss": 1.9505,
"step": 6120
},
{
"epoch": 0.6466244725738397,
"grad_norm": 0.4310709834098816,
"learning_rate": 0.0015,
"loss": 1.9475,
"step": 6130
},
{
"epoch": 0.6476793248945147,
"grad_norm": 0.5099350214004517,
"learning_rate": 0.0015,
"loss": 1.9516,
"step": 6140
},
{
"epoch": 0.6487341772151899,
"grad_norm": 0.41777700185775757,
"learning_rate": 0.0015,
"loss": 1.9598,
"step": 6150
},
{
"epoch": 0.6497890295358649,
"grad_norm": 0.4158155024051666,
"learning_rate": 0.0015,
"loss": 1.9517,
"step": 6160
},
{
"epoch": 0.6508438818565401,
"grad_norm": 0.4126010835170746,
"learning_rate": 0.0014834368975312174,
"loss": 1.9332,
"step": 6170
},
{
"epoch": 0.6518987341772152,
"grad_norm": 0.4793664813041687,
"learning_rate": 0.0014629899726345957,
"loss": 1.9533,
"step": 6180
},
{
"epoch": 0.6529535864978903,
"grad_norm": 0.4547656178474426,
"learning_rate": 0.0014428248775471316,
"loss": 1.9602,
"step": 6190
},
{
"epoch": 0.6540084388185654,
"grad_norm": 0.46379294991493225,
"learning_rate": 0.00142293772767289,
"loss": 1.95,
"step": 6200
},
{
"epoch": 0.6550632911392406,
"grad_norm": 0.4061027765274048,
"learning_rate": 0.001403324691959192,
"loss": 1.9455,
"step": 6210
},
{
"epoch": 0.6561181434599156,
"grad_norm": 0.4377910792827606,
"learning_rate": 0.0013839819921586025,
"loss": 1.9519,
"step": 6220
},
{
"epoch": 0.6571729957805907,
"grad_norm": 0.44470494985580444,
"learning_rate": 0.0013649059021010894,
"loss": 1.9404,
"step": 6230
},
{
"epoch": 0.6582278481012658,
"grad_norm": 0.394380658864975,
"learning_rate": 0.0013460927469762154,
"loss": 1.9402,
"step": 6240
},
{
"epoch": 0.6592827004219409,
"grad_norm": 0.3865656554698944,
"learning_rate": 0.0013275389026252255,
"loss": 1.9495,
"step": 6250
},
{
"epoch": 0.6603375527426161,
"grad_norm": 0.41206902265548706,
"learning_rate": 0.0013092407948428887,
"loss": 1.9417,
"step": 6260
},
{
"epoch": 0.6613924050632911,
"grad_norm": 0.40238043665885925,
"learning_rate": 0.001291194898688966,
"loss": 1.9467,
"step": 6270
},
{
"epoch": 0.6624472573839663,
"grad_norm": 0.4450398087501526,
"learning_rate": 0.001273397737809166,
"loss": 1.9452,
"step": 6280
},
{
"epoch": 0.6635021097046413,
"grad_norm": 0.4000122547149658,
"learning_rate": 0.001255845883765463,
"loss": 1.9376,
"step": 6290
},
{
"epoch": 0.6645569620253164,
"grad_norm": 0.5313910245895386,
"learning_rate": 0.001238535955375642,
"loss": 1.9362,
"step": 6300
},
{
"epoch": 0.6656118143459916,
"grad_norm": 0.43466663360595703,
"learning_rate": 0.0012214646180619506,
"loss": 1.9318,
"step": 6310
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.4663276672363281,
"learning_rate": 0.001204628583208727,
"loss": 1.925,
"step": 6320
},
{
"epoch": 0.6677215189873418,
"grad_norm": 0.42209333181381226,
"learning_rate": 0.0011880246075288827,
"loss": 1.9321,
"step": 6330
},
{
"epoch": 0.6687763713080169,
"grad_norm": 0.4251249134540558,
"learning_rate": 0.001171649492439115,
"loss": 1.9279,
"step": 6340
},
{
"epoch": 0.669831223628692,
"grad_norm": 0.4062013328075409,
"learning_rate": 0.0011555000834437364,
"loss": 1.928,
"step": 6350
},
{
"epoch": 0.6708860759493671,
"grad_norm": 0.4061965048313141,
"learning_rate": 0.0011395732695269908,
"loss": 1.9243,
"step": 6360
},
{
"epoch": 0.6719409282700421,
"grad_norm": 0.4409763514995575,
"learning_rate": 0.0011238659825537505,
"loss": 1.9109,
"step": 6370
},
{
"epoch": 0.6729957805907173,
"grad_norm": 0.41895362734794617,
"learning_rate": 0.0011083751966784717,
"loss": 1.9137,
"step": 6380
},
{
"epoch": 0.6740506329113924,
"grad_norm": 0.42582714557647705,
"learning_rate": 0.0010930979277622953,
"loss": 1.9278,
"step": 6390
},
{
"epoch": 0.6751054852320675,
"grad_norm": 0.4710099995136261,
"learning_rate": 0.0010780312327981854,
"loss": 1.9272,
"step": 6400
},
{
"epoch": 0.6761603375527426,
"grad_norm": 0.4434272050857544,
"learning_rate": 0.0010631722093439888,
"loss": 1.925,
"step": 6410
},
{
"epoch": 0.6772151898734177,
"grad_norm": 0.42562806606292725,
"learning_rate": 0.00104851799496331,
"loss": 1.9119,
"step": 6420
},
{
"epoch": 0.6782700421940928,
"grad_norm": 0.4150448143482208,
"learning_rate": 0.0010340657666740914,
"loss": 1.9191,
"step": 6430
},
{
"epoch": 0.679324894514768,
"grad_norm": 0.4621260166168213,
"learning_rate": 0.0010198127404047975,
"loss": 1.9031,
"step": 6440
},
{
"epoch": 0.680379746835443,
"grad_norm": 0.3927471339702606,
"learning_rate": 0.0010057561704580897,
"loss": 1.9063,
"step": 6450
},
{
"epoch": 0.6814345991561181,
"grad_norm": 0.3816894292831421,
"learning_rate": 0.0009918933489818985,
"loss": 1.919,
"step": 6460
},
{
"epoch": 0.6824894514767933,
"grad_norm": 0.4114357531070709,
"learning_rate": 0.0009782216054477827,
"loss": 1.9112,
"step": 6470
},
{
"epoch": 0.6835443037974683,
"grad_norm": 0.4401596784591675,
"learning_rate": 0.0009647383061364801,
"loss": 1.9177,
"step": 6480
},
{
"epoch": 0.6845991561181435,
"grad_norm": 0.4112340211868286,
"learning_rate": 0.0009514408536305495,
"loss": 1.9135,
"step": 6490
},
{
"epoch": 0.6856540084388185,
"grad_norm": 0.42365604639053345,
"learning_rate": 0.0009383266863140042,
"loss": 1.9279,
"step": 6500
},
{
"epoch": 0.6867088607594937,
"grad_norm": 0.3973536491394043,
"learning_rate": 0.000925393277878844,
"loss": 1.9232,
"step": 6510
},
{
"epoch": 0.6877637130801688,
"grad_norm": 0.4078630805015564,
"learning_rate": 0.0009126381368383879,
"loss": 1.908,
"step": 6520
},
{
"epoch": 0.6888185654008439,
"grad_norm": 0.4189813435077667,
"learning_rate": 0.0009000588060473156,
"loss": 1.9038,
"step": 6530
},
{
"epoch": 0.689873417721519,
"grad_norm": 0.3901522159576416,
"learning_rate": 0.0008876528622283235,
"loss": 1.9061,
"step": 6540
},
{
"epoch": 0.6909282700421941,
"grad_norm": 0.39721715450286865,
"learning_rate": 0.0008754179155053053,
"loss": 1.9038,
"step": 6550
},
{
"epoch": 0.6919831223628692,
"grad_norm": 0.5176498293876648,
"learning_rate": 0.0008633516089429683,
"loss": 1.9048,
"step": 6560
},
{
"epoch": 0.6930379746835443,
"grad_norm": 0.38577041029930115,
"learning_rate": 0.0008514516180927928,
"loss": 1.9032,
"step": 6570
},
{
"epoch": 0.6940928270042194,
"grad_norm": 0.40588951110839844,
"learning_rate": 0.0008397156505452524,
"loss": 1.8961,
"step": 6580
},
{
"epoch": 0.6951476793248945,
"grad_norm": 0.42049214243888855,
"learning_rate": 0.0008281414454882051,
"loss": 1.9051,
"step": 6590
},
{
"epoch": 0.6962025316455697,
"grad_norm": 0.4043624699115753,
"learning_rate": 0.0008167267732713704,
"loss": 1.9076,
"step": 6600
},
{
"epoch": 0.6972573839662447,
"grad_norm": 0.3989948332309723,
"learning_rate": 0.0008054694349768117,
"loss": 1.8948,
"step": 6610
},
{
"epoch": 0.6983122362869199,
"grad_norm": 0.389932781457901,
"learning_rate": 0.0007943672619953359,
"loss": 1.8984,
"step": 6620
},
{
"epoch": 0.6993670886075949,
"grad_norm": 0.4338935613632202,
"learning_rate": 0.0007834181156087356,
"loss": 1.8912,
"step": 6630
},
{
"epoch": 0.70042194092827,
"grad_norm": 0.4143203794956207,
"learning_rate": 0.0007726198865777852,
"loss": 1.8992,
"step": 6640
},
{
"epoch": 0.7014767932489452,
"grad_norm": 0.4350084364414215,
"learning_rate": 0.0007619704947359191,
"loss": 1.8883,
"step": 6650
},
{
"epoch": 0.7025316455696202,
"grad_norm": 0.4138067662715912,
"learning_rate": 0.0007514678885885087,
"loss": 1.8949,
"step": 6660
},
{
"epoch": 0.7035864978902954,
"grad_norm": 0.36295846104621887,
"learning_rate": 0.0007411100449176633,
"loss": 1.8896,
"step": 6670
},
{
"epoch": 0.7046413502109705,
"grad_norm": 0.39966461062431335,
"learning_rate": 0.0007308949683924791,
"loss": 1.8926,
"step": 6680
},
{
"epoch": 0.7056962025316456,
"grad_norm": 0.396759569644928,
"learning_rate": 0.000720820691184658,
"loss": 1.8905,
"step": 6690
},
{
"epoch": 0.7067510548523207,
"grad_norm": 0.40356534719467163,
"learning_rate": 0.0007108852725894269,
"loss": 1.8842,
"step": 6700
},
{
"epoch": 0.7078059071729957,
"grad_norm": 0.43995505571365356,
"learning_rate": 0.000701086798651681,
"loss": 1.8876,
"step": 6710
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.39463213086128235,
"learning_rate": 0.0006914233817972798,
"loss": 1.8817,
"step": 6720
},
{
"epoch": 0.709915611814346,
"grad_norm": 0.37957167625427246,
"learning_rate": 0.0006818931604694261,
"loss": 1.8875,
"step": 6730
},
{
"epoch": 0.7109704641350211,
"grad_norm": 0.37148046493530273,
"learning_rate": 0.0006724942987700563,
"loss": 1.8937,
"step": 6740
},
{
"epoch": 0.7120253164556962,
"grad_norm": 0.3996550440788269,
"learning_rate": 0.0006632249861061732,
"loss": 1.8938,
"step": 6750
},
{
"epoch": 0.7130801687763713,
"grad_norm": 0.3949614465236664,
"learning_rate": 0.0006540834368410549,
"loss": 1.888,
"step": 6760
},
{
"epoch": 0.7141350210970464,
"grad_norm": 0.3813718855381012,
"learning_rate": 0.0006450678899502701,
"loss": 1.8907,
"step": 6770
},
{
"epoch": 0.7151898734177216,
"grad_norm": 0.37598955631256104,
"learning_rate": 0.0006361766086824345,
"loss": 1.8861,
"step": 6780
},
{
"epoch": 0.7162447257383966,
"grad_norm": 0.39398065209388733,
"learning_rate": 0.000627407880224645,
"loss": 1.8921,
"step": 6790
},
{
"epoch": 0.7172995780590717,
"grad_norm": 0.37036997079849243,
"learning_rate": 0.0006187600153725225,
"loss": 1.8747,
"step": 6800
},
{
"epoch": 0.7183544303797469,
"grad_norm": 0.4019232392311096,
"learning_rate": 0.0006102313482048055,
"loss": 1.881,
"step": 6810
},
{
"epoch": 0.7194092827004219,
"grad_norm": 0.41694897413253784,
"learning_rate": 0.0006018202357624274,
"loss": 1.8829,
"step": 6820
},
{
"epoch": 0.7204641350210971,
"grad_norm": 0.40881526470184326,
"learning_rate": 0.0005935250577320168,
"loss": 1.8787,
"step": 6830
},
{
"epoch": 0.7215189873417721,
"grad_norm": 0.4087632894515991,
"learning_rate": 0.0005853442161337618,
"loss": 1.8696,
"step": 6840
},
{
"epoch": 0.7225738396624473,
"grad_norm": 0.37924131751060486,
"learning_rate": 0.0005772761350135759,
"loss": 1.8772,
"step": 6850
},
{
"epoch": 0.7236286919831224,
"grad_norm": 0.38739898800849915,
"learning_rate": 0.0005693192601395058,
"loss": 1.8742,
"step": 6860
},
{
"epoch": 0.7246835443037974,
"grad_norm": 0.3744981288909912,
"learning_rate": 0.000561472058702326,
"loss": 1.8695,
"step": 6870
},
{
"epoch": 0.7257383966244726,
"grad_norm": 0.3714418113231659,
"learning_rate": 0.000553733019020258,
"loss": 1.8793,
"step": 6880
},
{
"epoch": 0.7267932489451476,
"grad_norm": 0.3763202726840973,
"learning_rate": 0.0005461006502477612,
"loss": 1.8629,
"step": 6890
},
{
"epoch": 0.7278481012658228,
"grad_norm": 0.38656339049339294,
"learning_rate": 0.0005385734820883369,
"loss": 1.8729,
"step": 6900
},
{
"epoch": 0.7289029535864979,
"grad_norm": 0.36681076884269714,
"learning_rate": 0.0005311500645112907,
"loss": 1.8944,
"step": 6910
},
{
"epoch": 0.729957805907173,
"grad_norm": 0.3906095623970032,
"learning_rate": 0.0005238289674723993,
"loss": 1.8737,
"step": 6920
},
{
"epoch": 0.7310126582278481,
"grad_norm": 0.3659605383872986,
"learning_rate": 0.0005166087806384274,
"loss": 1.8813,
"step": 6930
},
{
"epoch": 0.7320675105485233,
"grad_norm": 0.4291110336780548,
"learning_rate": 0.0005094881131154418,
"loss": 1.8803,
"step": 6940
},
{
"epoch": 0.7331223628691983,
"grad_norm": 0.39232853055000305,
"learning_rate": 0.0005024655931808696,
"loss": 1.8777,
"step": 6950
},
{
"epoch": 0.7341772151898734,
"grad_norm": 0.39790189266204834,
"learning_rate": 0.0004955398680192508,
"loss": 1.8676,
"step": 6960
},
{
"epoch": 0.7352320675105485,
"grad_norm": 0.40521377325057983,
"learning_rate": 0.000488709603461632,
"loss": 1.8615,
"step": 6970
},
{
"epoch": 0.7362869198312236,
"grad_norm": 0.41293227672576904,
"learning_rate": 0.000481973483728553,
"loss": 1.8663,
"step": 6980
},
{
"epoch": 0.7373417721518988,
"grad_norm": 0.43020930886268616,
"learning_rate": 0.0004753302111765748,
"loss": 1.8671,
"step": 6990
},
{
"epoch": 0.7383966244725738,
"grad_norm": 0.3803190588951111,
"learning_rate": 0.0004687785060483032,
"loss": 1.8765,
"step": 7000
},
{
"epoch": 0.739451476793249,
"grad_norm": 0.3787606954574585,
"learning_rate": 0.0004623171062258558,
"loss": 1.8469,
"step": 7010
},
{
"epoch": 0.740506329113924,
"grad_norm": 0.38570114970207214,
"learning_rate": 0.0004559447669877288,
"loss": 1.8631,
"step": 7020
},
{
"epoch": 0.7415611814345991,
"grad_norm": 0.3946021795272827,
"learning_rate": 0.00044966026076901413,
"loss": 1.864,
"step": 7030
},
{
"epoch": 0.7426160337552743,
"grad_norm": 0.4139018654823303,
"learning_rate": 0.00044346237692492177,
"loss": 1.8711,
"step": 7040
},
{
"epoch": 0.7436708860759493,
"grad_norm": 0.3821570873260498,
"learning_rate": 0.0004373499214975615,
"loss": 1.8613,
"step": 7050
},
{
"epoch": 0.7447257383966245,
"grad_norm": 0.3973364531993866,
"learning_rate": 0.0004313217169859396,
"loss": 1.8656,
"step": 7060
},
{
"epoch": 0.7457805907172996,
"grad_norm": 0.40247097611427307,
"learning_rate": 0.0004253766021191256,
"loss": 1.8697,
"step": 7070
},
{
"epoch": 0.7468354430379747,
"grad_norm": 0.36728376150131226,
"learning_rate": 0.00041951343163254497,
"loss": 1.8667,
"step": 7080
},
{
"epoch": 0.7478902953586498,
"grad_norm": 0.37559038400650024,
"learning_rate": 0.00041373107604735626,
"loss": 1.8658,
"step": 7090
},
{
"epoch": 0.7489451476793249,
"grad_norm": 0.37511202692985535,
"learning_rate": 0.0004080284214528687,
"loss": 1.8595,
"step": 7100
},
{
"epoch": 0.75,
"grad_norm": 0.41109174489974976,
"learning_rate": 0.0004024043692919589,
"loss": 1.871,
"step": 7110
},
{
"epoch": 0.7510548523206751,
"grad_norm": 0.39778023958206177,
"learning_rate": 0.0003968578361494449,
"loss": 1.8691,
"step": 7120
},
{
"epoch": 0.7521097046413502,
"grad_norm": 0.3845960795879364,
"learning_rate": 0.000391387753543378,
"loss": 1.8756,
"step": 7130
},
{
"epoch": 0.7531645569620253,
"grad_norm": 0.3910766839981079,
"learning_rate": 0.00038599306771921023,
"loss": 1.8543,
"step": 7140
},
{
"epoch": 0.7542194092827004,
"grad_norm": 0.39525917172431946,
"learning_rate": 0.0003806727394468004,
"loss": 1.8514,
"step": 7150
},
{
"epoch": 0.7552742616033755,
"grad_norm": 0.36833056807518005,
"learning_rate": 0.0003754257438202162,
"loss": 1.8615,
"step": 7160
},
{
"epoch": 0.7563291139240507,
"grad_norm": 0.3827219605445862,
"learning_rate": 0.0003702510700602974,
"loss": 1.8718,
"step": 7170
},
{
"epoch": 0.7573839662447257,
"grad_norm": 0.388357013463974,
"learning_rate": 0.0003651477213199393,
"loss": 1.8462,
"step": 7180
},
{
"epoch": 0.7584388185654009,
"grad_norm": 0.39934271574020386,
"learning_rate": 0.000360114714492061,
"loss": 1.8507,
"step": 7190
},
{
"epoch": 0.759493670886076,
"grad_norm": 0.41285282373428345,
"learning_rate": 0.0003551510800202195,
"loss": 1.8544,
"step": 7200
},
{
"epoch": 0.760548523206751,
"grad_norm": 0.3969578742980957,
"learning_rate": 0.0003502558617118353,
"loss": 1.8598,
"step": 7210
},
{
"epoch": 0.7616033755274262,
"grad_norm": 0.3763035237789154,
"learning_rate": 0.0003454281165539914,
"loss": 1.8752,
"step": 7220
},
{
"epoch": 0.7626582278481012,
"grad_norm": 0.4243927597999573,
"learning_rate": 0.00034066691453177176,
"loss": 1.8671,
"step": 7230
},
{
"epoch": 0.7637130801687764,
"grad_norm": 0.3984666168689728,
"learning_rate": 0.0003359713384491037,
"loss": 1.8618,
"step": 7240
},
{
"epoch": 0.7647679324894515,
"grad_norm": 0.39229220151901245,
"learning_rate": 0.00033134048375206944,
"loss": 1.8598,
"step": 7250
},
{
"epoch": 0.7658227848101266,
"grad_norm": 0.359428733587265,
"learning_rate": 0.0003267734583546536,
"loss": 1.8541,
"step": 7260
},
{
"epoch": 0.7668776371308017,
"grad_norm": 0.3993035852909088,
"learning_rate": 0.00032226938246689157,
"loss": 1.8583,
"step": 7270
},
{
"epoch": 0.7679324894514767,
"grad_norm": 0.39195823669433594,
"learning_rate": 0.0003178273884253874,
"loss": 1.855,
"step": 7280
},
{
"epoch": 0.7689873417721519,
"grad_norm": 0.37241217494010925,
"learning_rate": 0.0003134466205261674,
"loss": 1.8653,
"step": 7290
},
{
"epoch": 0.770042194092827,
"grad_norm": 0.3893034756183624,
"learning_rate": 0.0003091262348598378,
"loss": 1.8726,
"step": 7300
},
{
"epoch": 0.7710970464135021,
"grad_norm": 0.3817334473133087,
"learning_rate": 0.0003048653991490141,
"loss": 1.8498,
"step": 7310
},
{
"epoch": 0.7721518987341772,
"grad_norm": 0.3646955192089081,
"learning_rate": 0.00030066329258799187,
"loss": 1.8465,
"step": 7320
},
{
"epoch": 0.7732067510548524,
"grad_norm": 0.4202854335308075,
"learning_rate": 0.0002965191056846266,
"loss": 1.8541,
"step": 7330
},
{
"epoch": 0.7742616033755274,
"grad_norm": 0.3694995641708374,
"learning_rate": 0.000292432040104394,
"loss": 1.8509,
"step": 7340
},
{
"epoch": 0.7753164556962026,
"grad_norm": 0.35707733035087585,
"learning_rate": 0.00028840130851659853,
"loss": 1.8467,
"step": 7350
},
{
"epoch": 0.7763713080168776,
"grad_norm": 0.38870304822921753,
"learning_rate": 0.0002844261344427028,
"loss": 1.8572,
"step": 7360
},
{
"epoch": 0.7774261603375527,
"grad_norm": 0.3943558633327484,
"learning_rate": 0.0002805057521067471,
"loss": 1.8527,
"step": 7370
},
{
"epoch": 0.7784810126582279,
"grad_norm": 0.3791944980621338,
"learning_rate": 0.00027663940628783017,
"loss": 1.8454,
"step": 7380
},
{
"epoch": 0.7795358649789029,
"grad_norm": 0.37139755487442017,
"learning_rate": 0.00027282635217462393,
"loss": 1.8519,
"step": 7390
},
{
"epoch": 0.7805907172995781,
"grad_norm": 0.3750312030315399,
"learning_rate": 0.0002690658552218937,
"loss": 1.8623,
"step": 7400
},
{
"epoch": 0.7816455696202531,
"grad_norm": 0.3591516315937042,
"learning_rate": 0.00026535719100899516,
"loss": 1.8381,
"step": 7410
},
{
"epoch": 0.7827004219409283,
"grad_norm": 0.3792029321193695,
"learning_rate": 0.00026169964510032245,
"loss": 1.8462,
"step": 7420
},
{
"epoch": 0.7837552742616034,
"grad_norm": 0.3520404100418091,
"learning_rate": 0.00025809251290767984,
"loss": 1.8338,
"step": 7430
},
{
"epoch": 0.7848101265822784,
"grad_norm": 0.3666081130504608,
"learning_rate": 0.00025453509955454957,
"loss": 1.8392,
"step": 7440
},
{
"epoch": 0.7858649789029536,
"grad_norm": 0.38259756565093994,
"learning_rate": 0.00025102671974223175,
"loss": 1.8421,
"step": 7450
},
{
"epoch": 0.7869198312236287,
"grad_norm": 0.3695935010910034,
"learning_rate": 0.00024756669761782815,
"loss": 1.8536,
"step": 7460
},
{
"epoch": 0.7879746835443038,
"grad_norm": 0.36817803978919983,
"learning_rate": 0.0002441543666440464,
"loss": 1.8405,
"step": 7470
},
{
"epoch": 0.7890295358649789,
"grad_norm": 0.38346537947654724,
"learning_rate": 0.00024078906947079878,
"loss": 1.844,
"step": 7480
},
{
"epoch": 0.790084388185654,
"grad_norm": 0.37116846442222595,
"learning_rate": 0.00023747015780857005,
"loss": 1.8563,
"step": 7490
},
{
"epoch": 0.7911392405063291,
"grad_norm": 0.37044385075569153,
"learning_rate": 0.00023419699230353144,
"loss": 1.85,
"step": 7500
},
{
"epoch": 0.7921940928270043,
"grad_norm": 0.38595202565193176,
"learning_rate": 0.00023096894241437586,
"loss": 1.8571,
"step": 7510
},
{
"epoch": 0.7932489451476793,
"grad_norm": 0.36819988489151,
"learning_rate": 0.00022778538629085056,
"loss": 1.8393,
"step": 7520
},
{
"epoch": 0.7943037974683544,
"grad_norm": 0.36314669251441956,
"learning_rate": 0.00022464571065396427,
"loss": 1.8443,
"step": 7530
},
{
"epoch": 0.7953586497890295,
"grad_norm": 0.39035850763320923,
"learning_rate": 0.00022154931067784521,
"loss": 1.8413,
"step": 7540
},
{
"epoch": 0.7964135021097046,
"grad_norm": 0.3708403408527374,
"learning_rate": 0.00021849558987322782,
"loss": 1.8374,
"step": 7550
},
{
"epoch": 0.7974683544303798,
"grad_norm": 0.38795605301856995,
"learning_rate": 0.0002154839599725452,
"loss": 1.837,
"step": 7560
},
{
"epoch": 0.7985232067510548,
"grad_norm": 0.38620778918266296,
"learning_rate": 0.00021251384081660544,
"loss": 1.8519,
"step": 7570
},
{
"epoch": 0.79957805907173,
"grad_norm": 0.36514854431152344,
"learning_rate": 0.0002095846602428303,
"loss": 1.8473,
"step": 7580
},
{
"epoch": 0.8006329113924051,
"grad_norm": 0.3578909635543823,
"learning_rate": 0.00020669585397503358,
"loss": 1.8404,
"step": 7590
},
{
"epoch": 0.8016877637130801,
"grad_norm": 0.37365859746932983,
"learning_rate": 0.0002038468655147195,
"loss": 1.8403,
"step": 7600
},
{
"epoch": 0.8027426160337553,
"grad_norm": 0.38331690430641174,
"learning_rate": 0.00020103714603387894,
"loss": 1.8533,
"step": 7610
},
{
"epoch": 0.8037974683544303,
"grad_norm": 0.3489489257335663,
"learning_rate": 0.00019826615426926338,
"loss": 1.8245,
"step": 7620
},
{
"epoch": 0.8048523206751055,
"grad_norm": 0.3655901253223419,
"learning_rate": 0.00019553335641811625,
"loss": 1.8473,
"step": 7630
},
{
"epoch": 0.8059071729957806,
"grad_norm": 0.38881802558898926,
"learning_rate": 0.0001928382260353415,
"loss": 1.8409,
"step": 7640
},
{
"epoch": 0.8069620253164557,
"grad_norm": 0.3771028518676758,
"learning_rate": 0.00019018024393208902,
"loss": 1.8537,
"step": 7650
},
{
"epoch": 0.8080168776371308,
"grad_norm": 0.38920238614082336,
"learning_rate": 0.00018755889807573872,
"loss": 1.8341,
"step": 7660
},
{
"epoch": 0.8090717299578059,
"grad_norm": 0.35742661356925964,
"learning_rate": 0.00018497368349126262,
"loss": 1.8412,
"step": 7670
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.4079872667789459,
"learning_rate": 0.00018242410216394648,
"loss": 1.8581,
"step": 7680
},
{
"epoch": 0.8111814345991561,
"grad_norm": 0.37757664918899536,
"learning_rate": 0.0001799096629434529,
"loss": 1.8299,
"step": 7690
},
{
"epoch": 0.8122362869198312,
"grad_norm": 0.36545130610466003,
"learning_rate": 0.00017742988144920578,
"loss": 1.84,
"step": 7700
},
{
"epoch": 0.8132911392405063,
"grad_norm": 0.38178738951683044,
"learning_rate": 0.00017498427997707976,
"loss": 1.8369,
"step": 7710
},
{
"epoch": 0.8143459915611815,
"grad_norm": 0.36409515142440796,
"learning_rate": 0.00017257238740737548,
"loss": 1.8373,
"step": 7720
},
{
"epoch": 0.8154008438818565,
"grad_norm": 0.39177459478378296,
"learning_rate": 0.00017019373911406307,
"loss": 1.8498,
"step": 7730
},
{
"epoch": 0.8164556962025317,
"grad_norm": 0.3647187650203705,
"learning_rate": 0.000167847876875277,
"loss": 1.8464,
"step": 7740
},
{
"epoch": 0.8175105485232067,
"grad_norm": 0.39089101552963257,
"learning_rate": 0.00016553434878504428,
"loss": 1.8306,
"step": 7750
},
{
"epoch": 0.8185654008438819,
"grad_norm": 0.38606345653533936,
"learning_rate": 0.00016325270916622947,
"loss": 1.8328,
"step": 7760
},
{
"epoch": 0.819620253164557,
"grad_norm": 0.36788392066955566,
"learning_rate": 0.00016100251848467966,
"loss": 1.8427,
"step": 7770
},
{
"epoch": 0.820675105485232,
"grad_norm": 0.35645923018455505,
"learning_rate": 0.0001587833432645528,
"loss": 1.8318,
"step": 7780
},
{
"epoch": 0.8217299578059072,
"grad_norm": 0.3621348738670349,
"learning_rate": 0.00015659475600481292,
"loss": 1.8535,
"step": 7790
},
{
"epoch": 0.8227848101265823,
"grad_norm": 0.35121673345565796,
"learning_rate": 0.00015443633509687688,
"loss": 1.8381,
"step": 7800
},
{
"epoch": 0.8238396624472574,
"grad_norm": 0.3815164566040039,
"learning_rate": 0.00015230766474339536,
"loss": 1.836,
"step": 7810
},
{
"epoch": 0.8248945147679325,
"grad_norm": 0.37245428562164307,
"learning_rate": 0.00015020833487815416,
"loss": 1.8467,
"step": 7820
},
{
"epoch": 0.8259493670886076,
"grad_norm": 0.3836027979850769,
"learning_rate": 0.0001481379410870792,
"loss": 1.8344,
"step": 7830
},
{
"epoch": 0.8270042194092827,
"grad_norm": 0.3576463460922241,
"learning_rate": 0.00014609608453033013,
"loss": 1.8205,
"step": 7840
},
{
"epoch": 0.8280590717299579,
"grad_norm": 0.35346224904060364,
"learning_rate": 0.00014408237186546807,
"loss": 1.8307,
"step": 7850
},
{
"epoch": 0.8291139240506329,
"grad_norm": 0.378290593624115,
"learning_rate": 0.00014209641517168273,
"loss": 1.8216,
"step": 7860
},
{
"epoch": 0.830168776371308,
"grad_norm": 0.35224640369415283,
"learning_rate": 0.00014013783187506265,
"loss": 1.8318,
"step": 7870
},
{
"epoch": 0.8312236286919831,
"grad_norm": 0.37513265013694763,
"learning_rate": 0.00013820624467489697,
"loss": 1.85,
"step": 7880
},
{
"epoch": 0.8322784810126582,
"grad_norm": 0.35514310002326965,
"learning_rate": 0.00013630128147099213,
"loss": 1.8478,
"step": 7890
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.3997390568256378,
"learning_rate": 0.00013442257529199068,
"loss": 1.83,
"step": 7900
},
{
"epoch": 0.8343881856540084,
"grad_norm": 0.3484932780265808,
"learning_rate": 0.00013256976422467803,
"loss": 1.8357,
"step": 7910
},
{
"epoch": 0.8354430379746836,
"grad_norm": 0.35105887055397034,
"learning_rate": 0.00013074249134426366,
"loss": 1.8379,
"step": 7920
},
{
"epoch": 0.8364978902953587,
"grad_norm": 0.3573038876056671,
"learning_rate": 0.0001289404046456233,
"loss": 1.8476,
"step": 7930
},
{
"epoch": 0.8375527426160337,
"grad_norm": 0.36516422033309937,
"learning_rate": 0.0001271631569754887,
"loss": 1.8416,
"step": 7940
},
{
"epoch": 0.8386075949367089,
"grad_norm": 0.3589839041233063,
"learning_rate": 0.0001254104059655723,
"loss": 1.8402,
"step": 7950
},
{
"epoch": 0.8396624472573839,
"grad_norm": 0.3751431703567505,
"learning_rate": 0.00012368181396661337,
"loss": 1.8229,
"step": 7960
},
{
"epoch": 0.8407172995780591,
"grad_norm": 0.362739622592926,
"learning_rate": 0.00012197704798333364,
"loss": 1.8323,
"step": 7970
},
{
"epoch": 0.8417721518987342,
"grad_norm": 0.36905282735824585,
"learning_rate": 0.00012029577961028894,
"loss": 1.8251,
"step": 7980
},
{
"epoch": 0.8428270042194093,
"grad_norm": 0.3701053559780121,
"learning_rate": 0.00011863768496860542,
"loss": 1.8409,
"step": 7990
},
{
"epoch": 0.8438818565400844,
"grad_norm": 0.3674405813217163,
"learning_rate": 0.00011700244464358777,
"loss": 1.8328,
"step": 8000
},
{
"epoch": 0.8449367088607594,
"grad_norm": 0.36590641736984253,
"learning_rate": 0.00011538974362318715,
"loss": 1.8391,
"step": 8010
},
{
"epoch": 0.8459915611814346,
"grad_norm": 0.3725239932537079,
"learning_rate": 0.00011379927123731737,
"loss": 1.8305,
"step": 8020
},
{
"epoch": 0.8470464135021097,
"grad_norm": 0.37020644545555115,
"learning_rate": 0.0001122307210980077,
"loss": 1.829,
"step": 8030
},
{
"epoch": 0.8481012658227848,
"grad_norm": 0.3549184799194336,
"learning_rate": 0.00011068379104038026,
"loss": 1.8424,
"step": 8040
},
{
"epoch": 0.8491561181434599,
"grad_norm": 0.3790867030620575,
"learning_rate": 0.00010915818306444116,
"loss": 1.8258,
"step": 8050
},
{
"epoch": 0.8502109704641351,
"grad_norm": 0.3635125160217285,
"learning_rate": 0.00010765360327767384,
"loss": 1.8301,
"step": 8060
},
{
"epoch": 0.8512658227848101,
"grad_norm": 0.3487062454223633,
"learning_rate": 0.00010616976183842376,
"loss": 1.8365,
"step": 8070
},
{
"epoch": 0.8523206751054853,
"grad_norm": 0.35935455560684204,
"learning_rate": 0.00010470637290006365,
"loss": 1.8393,
"step": 8080
},
{
"epoch": 0.8533755274261603,
"grad_norm": 0.34557873010635376,
"learning_rate": 0.00010326315455592764,
"loss": 1.8289,
"step": 8090
},
{
"epoch": 0.8544303797468354,
"grad_norm": 0.359907865524292,
"learning_rate": 0.0001018398287850053,
"loss": 1.8204,
"step": 8100
},
{
"epoch": 0.8554852320675106,
"grad_norm": 0.38513684272766113,
"learning_rate": 0.00010043612139838357,
"loss": 1.8426,
"step": 8110
},
{
"epoch": 0.8565400843881856,
"grad_norm": 0.3592938184738159,
"learning_rate": 9.905176198642719e-05,
"loss": 1.8298,
"step": 8120
},
{
"epoch": 0.8575949367088608,
"grad_norm": 0.37656906247138977,
"learning_rate": 9.76864838666871e-05,
"loss": 1.8326,
"step": 8130
},
{
"epoch": 0.8586497890295358,
"grad_norm": 0.3696509301662445,
"learning_rate": 9.634002403252676e-05,
"loss": 1.8277,
"step": 8140
},
{
"epoch": 0.859704641350211,
"grad_norm": 0.36902931332588196,
"learning_rate": 9.501212310245681e-05,
"loss": 1.8259,
"step": 8150
},
{
"epoch": 0.8607594936708861,
"grad_norm": 0.35556498169898987,
"learning_rate": 9.370252527016777e-05,
"loss": 1.8357,
"step": 8160
},
{
"epoch": 0.8618143459915611,
"grad_norm": 0.3559355139732361,
"learning_rate": 9.241097825525163e-05,
"loss": 1.8289,
"step": 8170
},
{
"epoch": 0.8628691983122363,
"grad_norm": 0.37157970666885376,
"learning_rate": 9.113723325460276e-05,
"loss": 1.8367,
"step": 8180
},
{
"epoch": 0.8639240506329114,
"grad_norm": 0.355340838432312,
"learning_rate": 8.988104489448849e-05,
"loss": 1.8299,
"step": 8190
},
{
"epoch": 0.8649789029535865,
"grad_norm": 0.3648318946361542,
"learning_rate": 8.864217118328042e-05,
"loss": 1.8382,
"step": 8200
},
{
"epoch": 0.8660337552742616,
"grad_norm": 0.3711951971054077,
"learning_rate": 8.742037346483729e-05,
"loss": 1.8308,
"step": 8210
},
{
"epoch": 0.8670886075949367,
"grad_norm": 0.3465905785560608,
"learning_rate": 8.62154163725303e-05,
"loss": 1.84,
"step": 8220
},
{
"epoch": 0.8681434599156118,
"grad_norm": 0.3531224727630615,
"learning_rate": 8.502706778390219e-05,
"loss": 1.8361,
"step": 8230
},
{
"epoch": 0.869198312236287,
"grad_norm": 0.3585963547229767,
"learning_rate": 8.38550987759513e-05,
"loss": 1.8324,
"step": 8240
},
{
"epoch": 0.870253164556962,
"grad_norm": 0.3620217740535736,
"learning_rate": 8.269928358103191e-05,
"loss": 1.8472,
"step": 8250
},
{
"epoch": 0.8713080168776371,
"grad_norm": 0.36264878511428833,
"learning_rate": 8.155939954336243e-05,
"loss": 1.8382,
"step": 8260
},
{
"epoch": 0.8723628691983122,
"grad_norm": 0.3570832908153534,
"learning_rate": 8.043522707613312e-05,
"loss": 1.8293,
"step": 8270
},
{
"epoch": 0.8734177215189873,
"grad_norm": 0.3508296608924866,
"learning_rate": 7.932654961920486e-05,
"loss": 1.8209,
"step": 8280
},
{
"epoch": 0.8744725738396625,
"grad_norm": 0.3548789620399475,
"learning_rate": 7.823315359739135e-05,
"loss": 1.818,
"step": 8290
},
{
"epoch": 0.8755274261603375,
"grad_norm": 0.350575715303421,
"learning_rate": 7.715482837931577e-05,
"loss": 1.8457,
"step": 8300
},
{
"epoch": 0.8765822784810127,
"grad_norm": 0.34618890285491943,
"learning_rate": 7.6091366236835e-05,
"loss": 1.8207,
"step": 8310
},
{
"epoch": 0.8776371308016878,
"grad_norm": 0.36400192975997925,
"learning_rate": 7.504256230502289e-05,
"loss": 1.8395,
"step": 8320
},
{
"epoch": 0.8786919831223629,
"grad_norm": 0.3459380269050598,
"learning_rate": 7.400821454270524e-05,
"loss": 1.8339,
"step": 8330
},
{
"epoch": 0.879746835443038,
"grad_norm": 0.34697189927101135,
"learning_rate": 7.29881236935386e-05,
"loss": 1.8194,
"step": 8340
},
{
"epoch": 0.880801687763713,
"grad_norm": 0.3628111779689789,
"learning_rate": 7.198209324762562e-05,
"loss": 1.8248,
"step": 8350
},
{
"epoch": 0.8818565400843882,
"grad_norm": 0.3612930178642273,
"learning_rate": 7.098992940365946e-05,
"loss": 1.8225,
"step": 8360
},
{
"epoch": 0.8829113924050633,
"grad_norm": 0.37480759620666504,
"learning_rate": 7.001144103159e-05,
"loss": 1.832,
"step": 8370
},
{
"epoch": 0.8839662447257384,
"grad_norm": 0.3713092505931854,
"learning_rate": 6.904643963580461e-05,
"loss": 1.8362,
"step": 8380
},
{
"epoch": 0.8850210970464135,
"grad_norm": 0.36273178458213806,
"learning_rate": 6.809473931881644e-05,
"loss": 1.8276,
"step": 8390
},
{
"epoch": 0.8860759493670886,
"grad_norm": 0.35557547211647034,
"learning_rate": 6.71561567454532e-05,
"loss": 1.8295,
"step": 8400
},
{
"epoch": 0.8871308016877637,
"grad_norm": 0.35513368248939514,
"learning_rate": 6.623051110753948e-05,
"loss": 1.8483,
"step": 8410
},
{
"epoch": 0.8881856540084389,
"grad_norm": 0.35310035943984985,
"learning_rate": 6.531762408906607e-05,
"loss": 1.8331,
"step": 8420
},
{
"epoch": 0.8892405063291139,
"grad_norm": 0.36005258560180664,
"learning_rate": 6.441731983183912e-05,
"loss": 1.8305,
"step": 8430
},
{
"epoch": 0.890295358649789,
"grad_norm": 0.356106162071228,
"learning_rate": 6.352942490160292e-05,
"loss": 1.8234,
"step": 8440
},
{
"epoch": 0.8913502109704642,
"grad_norm": 0.37876901030540466,
"learning_rate": 6.265376825462966e-05,
"loss": 1.8346,
"step": 8450
},
{
"epoch": 0.8924050632911392,
"grad_norm": 0.3676668405532837,
"learning_rate": 6.179018120476945e-05,
"loss": 1.8302,
"step": 8460
},
{
"epoch": 0.8934599156118144,
"grad_norm": 0.3509625792503357,
"learning_rate": 6.0938497390954946e-05,
"loss": 1.8309,
"step": 8470
},
{
"epoch": 0.8945147679324894,
"grad_norm": 0.3570344150066376,
"learning_rate": 6.009855274515339e-05,
"loss": 1.8234,
"step": 8480
},
{
"epoch": 0.8955696202531646,
"grad_norm": 0.3523993194103241,
"learning_rate": 5.9270185460760735e-05,
"loss": 1.833,
"step": 8490
},
{
"epoch": 0.8966244725738397,
"grad_norm": 0.35637491941452026,
"learning_rate": 5.8453235961431225e-05,
"loss": 1.8195,
"step": 8500
},
{
"epoch": 0.8976793248945147,
"grad_norm": 0.34806060791015625,
"learning_rate": 5.764754687033678e-05,
"loss": 1.8327,
"step": 8510
},
{
"epoch": 0.8987341772151899,
"grad_norm": 0.3617207705974579,
"learning_rate": 5.6852962979849836e-05,
"loss": 1.8257,
"step": 8520
},
{
"epoch": 0.8997890295358649,
"grad_norm": 0.3532322645187378,
"learning_rate": 5.6069331221644284e-05,
"loss": 1.8388,
"step": 8530
},
{
"epoch": 0.9008438818565401,
"grad_norm": 0.35625216364860535,
"learning_rate": 5.529650063720842e-05,
"loss": 1.8383,
"step": 8540
},
{
"epoch": 0.9018987341772152,
"grad_norm": 0.3520863652229309,
"learning_rate": 5.453432234876445e-05,
"loss": 1.8293,
"step": 8550
},
{
"epoch": 0.9029535864978903,
"grad_norm": 0.34727564454078674,
"learning_rate": 5.37826495305886e-05,
"loss": 1.8185,
"step": 8560
},
{
"epoch": 0.9040084388185654,
"grad_norm": 0.35971036553382874,
"learning_rate": 5.304133738072674e-05,
"loss": 1.8376,
"step": 8570
},
{
"epoch": 0.9050632911392406,
"grad_norm": 0.3352327346801758,
"learning_rate": 5.2310243093099814e-05,
"loss": 1.831,
"step": 8580
},
{
"epoch": 0.9061181434599156,
"grad_norm": 0.3646147549152374,
"learning_rate": 5.158922582999368e-05,
"loss": 1.8354,
"step": 8590
},
{
"epoch": 0.9071729957805907,
"grad_norm": 0.33844876289367676,
"learning_rate": 5.087814669492819e-05,
"loss": 1.8275,
"step": 8600
},
{
"epoch": 0.9082278481012658,
"grad_norm": 0.3586468994617462,
"learning_rate": 5.017686870590028e-05,
"loss": 1.8258,
"step": 8610
},
{
"epoch": 0.9092827004219409,
"grad_norm": 0.36400601267814636,
"learning_rate": 4.948525676899577e-05,
"loss": 1.8222,
"step": 8620
},
{
"epoch": 0.9103375527426161,
"grad_norm": 0.35185590386390686,
"learning_rate": 4.880317765236493e-05,
"loss": 1.834,
"step": 8630
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.36754703521728516,
"learning_rate": 4.8130499960556755e-05,
"loss": 1.8111,
"step": 8640
},
{
"epoch": 0.9124472573839663,
"grad_norm": 0.3563450872898102,
"learning_rate": 4.746709410920699e-05,
"loss": 1.8212,
"step": 8650
},
{
"epoch": 0.9135021097046413,
"grad_norm": 0.35486289858818054,
"learning_rate": 4.681283230007507e-05,
"loss": 1.8219,
"step": 8660
},
{
"epoch": 0.9145569620253164,
"grad_norm": 0.35444745421409607,
"learning_rate": 4.616758849642509e-05,
"loss": 1.8267,
"step": 8670
},
{
"epoch": 0.9156118143459916,
"grad_norm": 0.35601454973220825,
"learning_rate": 4.553123839874615e-05,
"loss": 1.8359,
"step": 8680
},
{
"epoch": 0.9166666666666666,
"grad_norm": 0.35390567779541016,
"learning_rate": 4.490365942080736e-05,
"loss": 1.8325,
"step": 8690
},
{
"epoch": 0.9177215189873418,
"grad_norm": 0.3538806736469269,
"learning_rate": 4.428473066604285e-05,
"loss": 1.8294,
"step": 8700
},
{
"epoch": 0.9187763713080169,
"grad_norm": 0.3609935939311981,
"learning_rate": 4.367433290426233e-05,
"loss": 1.8301,
"step": 8710
},
{
"epoch": 0.919831223628692,
"grad_norm": 0.3562791049480438,
"learning_rate": 4.3072348548682595e-05,
"loss": 1.8191,
"step": 8720
},
{
"epoch": 0.9208860759493671,
"grad_norm": 0.35799530148506165,
"learning_rate": 4.247866163327575e-05,
"loss": 1.8328,
"step": 8730
},
{
"epoch": 0.9219409282700421,
"grad_norm": 0.35550785064697266,
"learning_rate": 4.1893157790429404e-05,
"loss": 1.8192,
"step": 8740
},
{
"epoch": 0.9229957805907173,
"grad_norm": 0.354793906211853,
"learning_rate": 4.1315724228915066e-05,
"loss": 1.8205,
"step": 8750
},
{
"epoch": 0.9240506329113924,
"grad_norm": 0.35129454731941223,
"learning_rate": 4.074624971216005e-05,
"loss": 1.8088,
"step": 8760
},
{
"epoch": 0.9251054852320675,
"grad_norm": 0.34905755519866943,
"learning_rate": 4.018462453681889e-05,
"loss": 1.8224,
"step": 8770
},
{
"epoch": 0.9261603375527426,
"grad_norm": 0.3482454717159271,
"learning_rate": 3.963074051164014e-05,
"loss": 1.8312,
"step": 8780
},
{
"epoch": 0.9272151898734177,
"grad_norm": 0.3517247438430786,
"learning_rate": 3.908449093662446e-05,
"loss": 1.8254,
"step": 8790
},
{
"epoch": 0.9282700421940928,
"grad_norm": 0.3577110171318054,
"learning_rate": 3.854577058246998e-05,
"loss": 1.8213,
"step": 8800
},
{
"epoch": 0.929324894514768,
"grad_norm": 0.3506202697753906,
"learning_rate": 3.801447567030094e-05,
"loss": 1.8372,
"step": 8810
},
{
"epoch": 0.930379746835443,
"grad_norm": 0.349944531917572,
"learning_rate": 3.7490503851675777e-05,
"loss": 1.8253,
"step": 8820
},
{
"epoch": 0.9314345991561181,
"grad_norm": 0.3520268201828003,
"learning_rate": 3.6973754188870806e-05,
"loss": 1.8422,
"step": 8830
},
{
"epoch": 0.9324894514767933,
"grad_norm": 0.3507753610610962,
"learning_rate": 3.6464127135435536e-05,
"loss": 1.8241,
"step": 8840
},
{
"epoch": 0.9335443037974683,
"grad_norm": 0.35455137491226196,
"learning_rate": 3.596152451701616e-05,
"loss": 1.8224,
"step": 8850
},
{
"epoch": 0.9345991561181435,
"grad_norm": 0.3520313501358032,
"learning_rate": 3.5465849512443226e-05,
"loss": 1.8255,
"step": 8860
},
{
"epoch": 0.9356540084388185,
"grad_norm": 0.346407413482666,
"learning_rate": 3.4977006635080086e-05,
"loss": 1.8293,
"step": 8870
},
{
"epoch": 0.9367088607594937,
"grad_norm": 0.3564207851886749,
"learning_rate": 3.449490171442838e-05,
"loss": 1.8308,
"step": 8880
},
{
"epoch": 0.9377637130801688,
"grad_norm": 0.3574276566505432,
"learning_rate": 3.401944187798702e-05,
"loss": 1.8308,
"step": 8890
},
{
"epoch": 0.9388185654008439,
"grad_norm": 0.3549087643623352,
"learning_rate": 3.355053553336137e-05,
"loss": 1.821,
"step": 8900
},
{
"epoch": 0.939873417721519,
"grad_norm": 0.37319085001945496,
"learning_rate": 3.308809235061882e-05,
"loss": 1.8203,
"step": 8910
},
{
"epoch": 0.9409282700421941,
"grad_norm": 0.34253865480422974,
"learning_rate": 3.263202324488772e-05,
"loss": 1.8236,
"step": 8920
},
{
"epoch": 0.9419831223628692,
"grad_norm": 0.3503594696521759,
"learning_rate": 3.218224035919609e-05,
"loss": 1.8276,
"step": 8930
},
{
"epoch": 0.9430379746835443,
"grad_norm": 0.35104867815971375,
"learning_rate": 3.173865704754688e-05,
"loss": 1.8252,
"step": 8940
},
{
"epoch": 0.9440928270042194,
"grad_norm": 0.34539034962654114,
"learning_rate": 3.130118785822657e-05,
"loss": 1.8284,
"step": 8950
},
{
"epoch": 0.9451476793248945,
"grad_norm": 0.3551845848560333,
"learning_rate": 3.08697485173437e-05,
"loss": 1.8309,
"step": 8960
},
{
"epoch": 0.9462025316455697,
"grad_norm": 0.3424545228481293,
"learning_rate": 3.0444255912594442e-05,
"loss": 1.8306,
"step": 8970
},
{
"epoch": 0.9472573839662447,
"grad_norm": 0.35456904768943787,
"learning_rate": 3.002462807725185e-05,
"loss": 1.8216,
"step": 8980
},
{
"epoch": 0.9483122362869199,
"grad_norm": 0.35064905881881714,
"learning_rate": 2.9610784174375868e-05,
"loss": 1.8312,
"step": 8990
},
{
"epoch": 0.9493670886075949,
"grad_norm": 0.34413984417915344,
"learning_rate": 2.920264448124087e-05,
"loss": 1.8245,
"step": 9000
},
{
"epoch": 0.95042194092827,
"grad_norm": 0.34440872073173523,
"learning_rate": 2.8800130373977936e-05,
"loss": 1.8232,
"step": 9010
},
{
"epoch": 0.9514767932489452,
"grad_norm": 0.366588830947876,
"learning_rate": 2.84031643124288e-05,
"loss": 1.8241,
"step": 9020
},
{
"epoch": 0.9525316455696202,
"grad_norm": 0.3599245250225067,
"learning_rate": 2.8011669825208517e-05,
"loss": 1.8427,
"step": 9030
},
{
"epoch": 0.9535864978902954,
"grad_norm": 0.34534797072410583,
"learning_rate": 2.762557149497405e-05,
"loss": 1.8167,
"step": 9040
},
{
"epoch": 0.9546413502109705,
"grad_norm": 0.34453675150871277,
"learning_rate": 2.724479494389592e-05,
"loss": 1.8323,
"step": 9050
},
{
"epoch": 0.9556962025316456,
"grad_norm": 0.35049787163734436,
"learning_rate": 2.6869266819330058e-05,
"loss": 1.8328,
"step": 9060
},
{
"epoch": 0.9567510548523207,
"grad_norm": 0.34922167658805847,
"learning_rate": 2.6498914779687228e-05,
"loss": 1.8298,
"step": 9070
},
{
"epoch": 0.9578059071729957,
"grad_norm": 0.34359559416770935,
"learning_rate": 2.6133667480497115e-05,
"loss": 1.8293,
"step": 9080
},
{
"epoch": 0.9588607594936709,
"grad_norm": 0.3608853816986084,
"learning_rate": 2.5773454560664597e-05,
"loss": 1.8354,
"step": 9090
},
{
"epoch": 0.959915611814346,
"grad_norm": 0.3448154628276825,
"learning_rate": 2.541820662891541e-05,
"loss": 1.8088,
"step": 9100
},
{
"epoch": 0.9609704641350211,
"grad_norm": 0.35206976532936096,
"learning_rate": 2.5067855250428616e-05,
"loss": 1.8256,
"step": 9110
},
{
"epoch": 0.9620253164556962,
"grad_norm": 0.3525342047214508,
"learning_rate": 2.472233293365335e-05,
"loss": 1.8293,
"step": 9120
},
{
"epoch": 0.9630801687763713,
"grad_norm": 0.3429815173149109,
"learning_rate": 2.4381573117307307e-05,
"loss": 1.8172,
"step": 9130
},
{
"epoch": 0.9641350210970464,
"grad_norm": 0.34919753670692444,
"learning_rate": 2.4045510157554362e-05,
"loss": 1.8359,
"step": 9140
},
{
"epoch": 0.9651898734177216,
"grad_norm": 0.3607841432094574,
"learning_rate": 2.3714079315358985e-05,
"loss": 1.8267,
"step": 9150
},
{
"epoch": 0.9662447257383966,
"grad_norm": 0.3457028269767761,
"learning_rate": 2.338721674401494e-05,
"loss": 1.8331,
"step": 9160
},
{
"epoch": 0.9672995780590717,
"grad_norm": 0.3522280752658844,
"learning_rate": 2.30648594768459e-05,
"loss": 1.8313,
"step": 9170
},
{
"epoch": 0.9683544303797469,
"grad_norm": 0.3416561186313629,
"learning_rate": 2.2746945415075523e-05,
"loss": 1.827,
"step": 9180
},
{
"epoch": 0.9694092827004219,
"grad_norm": 0.35568854212760925,
"learning_rate": 2.2433413315864803e-05,
"loss": 1.8216,
"step": 9190
},
{
"epoch": 0.9704641350210971,
"grad_norm": 0.3549978733062744,
"learning_rate": 2.2124202780514277e-05,
"loss": 1.8219,
"step": 9200
},
{
"epoch": 0.9715189873417721,
"grad_norm": 0.3559131920337677,
"learning_rate": 2.1819254242828815e-05,
"loss": 1.8254,
"step": 9210
},
{
"epoch": 0.9725738396624473,
"grad_norm": 0.3483375906944275,
"learning_rate": 2.151850895764285e-05,
"loss": 1.8242,
"step": 9220
},
{
"epoch": 0.9736286919831224,
"grad_norm": 0.3538123667240143,
"learning_rate": 2.12219089895037e-05,
"loss": 1.8114,
"step": 9230
},
{
"epoch": 0.9746835443037974,
"grad_norm": 0.3543187975883484,
"learning_rate": 2.092939720151092e-05,
"loss": 1.8096,
"step": 9240
},
{
"epoch": 0.9757383966244726,
"grad_norm": 0.34672802686691284,
"learning_rate": 2.064091724430947e-05,
"loss": 1.8195,
"step": 9250
},
{
"epoch": 0.9767932489451476,
"grad_norm": 0.3511864244937897,
"learning_rate": 2.0356413545234603e-05,
"loss": 1.8189,
"step": 9260
},
{
"epoch": 0.9778481012658228,
"grad_norm": 0.33919981122016907,
"learning_rate": 2.0075831297606357e-05,
"loss": 1.8304,
"step": 9270
},
{
"epoch": 0.9789029535864979,
"grad_norm": 0.34800052642822266,
"learning_rate": 1.9799116450171627e-05,
"loss": 1.8168,
"step": 9280
},
{
"epoch": 0.979957805907173,
"grad_norm": 0.35135072469711304,
"learning_rate": 1.952621569669175e-05,
"loss": 1.8247,
"step": 9290
},
{
"epoch": 0.9810126582278481,
"grad_norm": 0.35404956340789795,
"learning_rate": 1.9257076465673605e-05,
"loss": 1.8213,
"step": 9300
},
{
"epoch": 0.9820675105485233,
"grad_norm": 0.3485180735588074,
"learning_rate": 1.899164691024229e-05,
"loss": 1.8222,
"step": 9310
},
{
"epoch": 0.9831223628691983,
"grad_norm": 0.35099634528160095,
"learning_rate": 1.872987589815331e-05,
"loss": 1.8157,
"step": 9320
},
{
"epoch": 0.9841772151898734,
"grad_norm": 0.3508138060569763,
"learning_rate": 1.8471713001942538e-05,
"loss": 1.8287,
"step": 9330
},
{
"epoch": 0.9852320675105485,
"grad_norm": 0.3524887263774872,
"learning_rate": 1.8217108489211845e-05,
"loss": 1.8299,
"step": 9340
},
{
"epoch": 0.9862869198312236,
"grad_norm": 0.34672585129737854,
"learning_rate": 1.7966013313048696e-05,
"loss": 1.8415,
"step": 9350
},
{
"epoch": 0.9873417721518988,
"grad_norm": 0.35972487926483154,
"learning_rate": 1.7718379102577752e-05,
"loss": 1.8344,
"step": 9360
},
{
"epoch": 0.9883966244725738,
"grad_norm": 0.3603551685810089,
"learning_rate": 1.7474158153642745e-05,
"loss": 1.8302,
"step": 9370
},
{
"epoch": 0.989451476793249,
"grad_norm": 0.35453343391418457,
"learning_rate": 1.7233303419616745e-05,
"loss": 1.8158,
"step": 9380
},
{
"epoch": 0.990506329113924,
"grad_norm": 0.3581453561782837,
"learning_rate": 1.699576850233916e-05,
"loss": 1.8205,
"step": 9390
},
{
"epoch": 0.9915611814345991,
"grad_norm": 0.34704574942588806,
"learning_rate": 1.6761507643177553e-05,
"loss": 1.8312,
"step": 9400
},
{
"epoch": 0.9926160337552743,
"grad_norm": 0.353828489780426,
"learning_rate": 1.6530475714212752e-05,
"loss": 1.8063,
"step": 9410
},
{
"epoch": 0.9936708860759493,
"grad_norm": 0.3566914498806,
"learning_rate": 1.6302628209545423e-05,
"loss": 1.826,
"step": 9420
},
{
"epoch": 0.9947257383966245,
"grad_norm": 0.3434826731681824,
"learning_rate": 1.6077921236722464e-05,
"loss": 1.8289,
"step": 9430
},
{
"epoch": 0.9957805907172996,
"grad_norm": 0.34118619561195374,
"learning_rate": 1.5856311508281594e-05,
"loss": 1.8211,
"step": 9440
},
{
"epoch": 0.9968354430379747,
"grad_norm": 0.35303235054016113,
"learning_rate": 1.5637756333412454e-05,
"loss": 1.8344,
"step": 9450
},
{
"epoch": 0.9978902953586498,
"grad_norm": 0.35012632608413696,
"learning_rate": 1.542221360973268e-05,
"loss": 1.8168,
"step": 9460
},
{
"epoch": 0.9989451476793249,
"grad_norm": 0.3457571864128113,
"learning_rate": 1.5209641815177312e-05,
"loss": 1.832,
"step": 9470
},
{
"epoch": 1.0,
"grad_norm": 1.0240792036056519,
"learning_rate": 1.5e-05,
"loss": 1.819,
"step": 9480
}
],
"logging_steps": 10,
"max_steps": 9480,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.8391618477891584e+16,
"train_batch_size": 1024,
"trial_name": null,
"trial_params": null
}