MultimodalAgent's picture
Upload multi_student_3/checkpoint-3600/trainer_state.json with huggingface_hub
775649e verified
Raw
History Blame Contribute Delete
63.3 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.7975126311698406,
"eval_steps": 500,
"global_step": 3600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00777302759424796,
"grad_norm": 19.018803349985753,
"learning_rate": 8.737864077669904e-07,
"loss": 8.2473,
"step": 10
},
{
"epoch": 0.01554605518849592,
"grad_norm": 17.967208179671328,
"learning_rate": 1.8446601941747572e-06,
"loss": 8.1304,
"step": 20
},
{
"epoch": 0.023319082782743878,
"grad_norm": 9.21576906708124,
"learning_rate": 2.8155339805825245e-06,
"loss": 7.4913,
"step": 30
},
{
"epoch": 0.03109211037699184,
"grad_norm": 4.8422984662740465,
"learning_rate": 3.7864077669902915e-06,
"loss": 6.7578,
"step": 40
},
{
"epoch": 0.0388651379712398,
"grad_norm": 3.0883237528757603,
"learning_rate": 4.7572815533980585e-06,
"loss": 6.2789,
"step": 50
},
{
"epoch": 0.046638165565487756,
"grad_norm": 3.0641013238554837,
"learning_rate": 5.728155339805825e-06,
"loss": 5.9381,
"step": 60
},
{
"epoch": 0.054411193159735714,
"grad_norm": 3.6989178160715803,
"learning_rate": 6.699029126213592e-06,
"loss": 5.547,
"step": 70
},
{
"epoch": 0.06218422075398368,
"grad_norm": 3.8926133061231023,
"learning_rate": 7.66990291262136e-06,
"loss": 4.9449,
"step": 80
},
{
"epoch": 0.06995724834823164,
"grad_norm": 6.005297193380389,
"learning_rate": 8.640776699029127e-06,
"loss": 4.2741,
"step": 90
},
{
"epoch": 0.0777302759424796,
"grad_norm": 4.634097942418493,
"learning_rate": 9.611650485436892e-06,
"loss": 3.692,
"step": 100
},
{
"epoch": 0.08550330353672755,
"grad_norm": 3.9161281136802915,
"learning_rate": 1.0582524271844662e-05,
"loss": 3.3021,
"step": 110
},
{
"epoch": 0.09327633113097551,
"grad_norm": 2.668188197386523,
"learning_rate": 1.1553398058252427e-05,
"loss": 3.0699,
"step": 120
},
{
"epoch": 0.10104935872522347,
"grad_norm": 5.67542909747789,
"learning_rate": 1.2524271844660197e-05,
"loss": 2.9232,
"step": 130
},
{
"epoch": 0.10882238631947143,
"grad_norm": 8.358891491079559,
"learning_rate": 1.3495145631067962e-05,
"loss": 2.8142,
"step": 140
},
{
"epoch": 0.11659541391371939,
"grad_norm": 7.521330978238392,
"learning_rate": 1.446601941747573e-05,
"loss": 2.736,
"step": 150
},
{
"epoch": 0.12436844150796736,
"grad_norm": 16.178722459983547,
"learning_rate": 1.5436893203883496e-05,
"loss": 2.6651,
"step": 160
},
{
"epoch": 0.1321414691022153,
"grad_norm": 8.335965189770683,
"learning_rate": 1.6407766990291263e-05,
"loss": 2.6021,
"step": 170
},
{
"epoch": 0.13991449669646328,
"grad_norm": 8.022474097021126,
"learning_rate": 1.737864077669903e-05,
"loss": 2.5458,
"step": 180
},
{
"epoch": 0.14768752429071122,
"grad_norm": 11.647238033711355,
"learning_rate": 1.8349514563106798e-05,
"loss": 2.4889,
"step": 190
},
{
"epoch": 0.1554605518849592,
"grad_norm": 11.910883080289608,
"learning_rate": 1.9320388349514565e-05,
"loss": 2.4326,
"step": 200
},
{
"epoch": 0.16323357947920716,
"grad_norm": 11.410643498860912,
"learning_rate": 2.0291262135922333e-05,
"loss": 2.3697,
"step": 210
},
{
"epoch": 0.1710066070734551,
"grad_norm": 11.108752154756692,
"learning_rate": 2.1262135922330097e-05,
"loss": 2.309,
"step": 220
},
{
"epoch": 0.17877963466770308,
"grad_norm": 19.998718279627422,
"learning_rate": 2.2233009708737864e-05,
"loss": 2.241,
"step": 230
},
{
"epoch": 0.18655266226195102,
"grad_norm": 11.758179967806427,
"learning_rate": 2.3203883495145632e-05,
"loss": 2.1862,
"step": 240
},
{
"epoch": 0.194325689856199,
"grad_norm": 13.686852090688575,
"learning_rate": 2.41747572815534e-05,
"loss": 2.1354,
"step": 250
},
{
"epoch": 0.20209871745044694,
"grad_norm": 7.097958211758118,
"learning_rate": 2.5145631067961167e-05,
"loss": 2.0252,
"step": 260
},
{
"epoch": 0.2098717450446949,
"grad_norm": 12.553623894402433,
"learning_rate": 2.6116504854368934e-05,
"loss": 1.92,
"step": 270
},
{
"epoch": 0.21764477263894286,
"grad_norm": 7.033016298250838,
"learning_rate": 2.7087378640776702e-05,
"loss": 1.832,
"step": 280
},
{
"epoch": 0.22541780023319083,
"grad_norm": 8.05454776441266,
"learning_rate": 2.8058252427184466e-05,
"loss": 1.7505,
"step": 290
},
{
"epoch": 0.23319082782743877,
"grad_norm": 6.32342513670527,
"learning_rate": 2.9029126213592233e-05,
"loss": 1.6596,
"step": 300
},
{
"epoch": 0.24096385542168675,
"grad_norm": 6.0607588085936435,
"learning_rate": 3e-05,
"loss": 1.5836,
"step": 310
},
{
"epoch": 0.24873688301593472,
"grad_norm": 7.205879956336063,
"learning_rate": 3.097087378640777e-05,
"loss": 1.5148,
"step": 320
},
{
"epoch": 0.2565099106101827,
"grad_norm": 5.233158820350977,
"learning_rate": 3.194174757281553e-05,
"loss": 1.4655,
"step": 330
},
{
"epoch": 0.2642829382044306,
"grad_norm": 6.5668167184777495,
"learning_rate": 3.29126213592233e-05,
"loss": 1.3996,
"step": 340
},
{
"epoch": 0.2720559657986786,
"grad_norm": 5.866172867997477,
"learning_rate": 3.3883495145631074e-05,
"loss": 1.3546,
"step": 350
},
{
"epoch": 0.27982899339292655,
"grad_norm": 7.078627326366601,
"learning_rate": 3.485436893203884e-05,
"loss": 1.2957,
"step": 360
},
{
"epoch": 0.2876020209871745,
"grad_norm": 7.013906915714377,
"learning_rate": 3.58252427184466e-05,
"loss": 1.265,
"step": 370
},
{
"epoch": 0.29537504858142244,
"grad_norm": 5.721220928268079,
"learning_rate": 3.679611650485437e-05,
"loss": 1.2163,
"step": 380
},
{
"epoch": 0.3031480761756704,
"grad_norm": 5.571852622479908,
"learning_rate": 3.776699029126214e-05,
"loss": 1.163,
"step": 390
},
{
"epoch": 0.3109211037699184,
"grad_norm": 5.4843565963743295,
"learning_rate": 3.873786407766991e-05,
"loss": 1.1277,
"step": 400
},
{
"epoch": 0.31869413136416636,
"grad_norm": 7.069970365890535,
"learning_rate": 3.970873786407767e-05,
"loss": 1.1094,
"step": 410
},
{
"epoch": 0.3264671589584143,
"grad_norm": 5.425993491783754,
"learning_rate": 4.0679611650485436e-05,
"loss": 1.069,
"step": 420
},
{
"epoch": 0.33424018655266224,
"grad_norm": 5.03054433074538,
"learning_rate": 4.165048543689321e-05,
"loss": 1.0382,
"step": 430
},
{
"epoch": 0.3420132141469102,
"grad_norm": 7.103802226586121,
"learning_rate": 4.262135922330097e-05,
"loss": 0.9988,
"step": 440
},
{
"epoch": 0.3497862417411582,
"grad_norm": 5.180871741408765,
"learning_rate": 4.3592233009708735e-05,
"loss": 0.979,
"step": 450
},
{
"epoch": 0.35755926933540616,
"grad_norm": 5.168730423055745,
"learning_rate": 4.4563106796116506e-05,
"loss": 0.9442,
"step": 460
},
{
"epoch": 0.3653322969296541,
"grad_norm": 4.509772926215358,
"learning_rate": 4.553398058252427e-05,
"loss": 0.9304,
"step": 470
},
{
"epoch": 0.37310532452390205,
"grad_norm": 5.497064414551351,
"learning_rate": 4.650485436893204e-05,
"loss": 0.8971,
"step": 480
},
{
"epoch": 0.38087835211815,
"grad_norm": 4.2878368849547375,
"learning_rate": 4.747572815533981e-05,
"loss": 0.8754,
"step": 490
},
{
"epoch": 0.388651379712398,
"grad_norm": 5.023778493195039,
"learning_rate": 4.8446601941747576e-05,
"loss": 0.8535,
"step": 500
},
{
"epoch": 0.39642440730664597,
"grad_norm": 4.85429288760384,
"learning_rate": 4.941747572815535e-05,
"loss": 0.8176,
"step": 510
},
{
"epoch": 0.4041974349008939,
"grad_norm": 4.78285229843812,
"learning_rate": 5.0388349514563104e-05,
"loss": 0.8114,
"step": 520
},
{
"epoch": 0.41197046249514185,
"grad_norm": 5.101751542601933,
"learning_rate": 5.135922330097087e-05,
"loss": 0.7752,
"step": 530
},
{
"epoch": 0.4197434900893898,
"grad_norm": 5.448054773232898,
"learning_rate": 5.233009708737864e-05,
"loss": 0.755,
"step": 540
},
{
"epoch": 0.4275165176836378,
"grad_norm": 4.141934961389859,
"learning_rate": 5.3300970873786417e-05,
"loss": 0.7273,
"step": 550
},
{
"epoch": 0.4352895452778857,
"grad_norm": 4.514745713617805,
"learning_rate": 5.427184466019418e-05,
"loss": 0.7143,
"step": 560
},
{
"epoch": 0.4430625728721337,
"grad_norm": 3.852074915498448,
"learning_rate": 5.524271844660195e-05,
"loss": 0.6954,
"step": 570
},
{
"epoch": 0.45083560046638166,
"grad_norm": 4.549795432019635,
"learning_rate": 5.6213592233009716e-05,
"loss": 0.6695,
"step": 580
},
{
"epoch": 0.45860862806062963,
"grad_norm": 3.920314510813715,
"learning_rate": 5.718446601941748e-05,
"loss": 0.6503,
"step": 590
},
{
"epoch": 0.46638165565487755,
"grad_norm": 5.518859711149864,
"learning_rate": 5.815533980582525e-05,
"loss": 0.6249,
"step": 600
},
{
"epoch": 0.4741546832491255,
"grad_norm": 4.876638883083892,
"learning_rate": 5.9126213592233015e-05,
"loss": 0.6064,
"step": 610
},
{
"epoch": 0.4819277108433735,
"grad_norm": 5.2941697252063955,
"learning_rate": 6.009708737864078e-05,
"loss": 0.5911,
"step": 620
},
{
"epoch": 0.48970073843762146,
"grad_norm": 4.245229081646686,
"learning_rate": 6.106796116504854e-05,
"loss": 0.5704,
"step": 630
},
{
"epoch": 0.49747376603186944,
"grad_norm": 4.127680025681991,
"learning_rate": 6.203883495145631e-05,
"loss": 0.5589,
"step": 640
},
{
"epoch": 0.5052467936261174,
"grad_norm": 3.5437231540377616,
"learning_rate": 6.300970873786408e-05,
"loss": 0.5378,
"step": 650
},
{
"epoch": 0.5130198212203654,
"grad_norm": 4.180991087743002,
"learning_rate": 6.398058252427184e-05,
"loss": 0.5056,
"step": 660
},
{
"epoch": 0.5207928488146133,
"grad_norm": 3.7905222943864034,
"learning_rate": 6.495145631067961e-05,
"loss": 0.4944,
"step": 670
},
{
"epoch": 0.5285658764088612,
"grad_norm": 2.829726998778342,
"learning_rate": 6.592233009708738e-05,
"loss": 0.4706,
"step": 680
},
{
"epoch": 0.5363389040031092,
"grad_norm": 2.9270824536877456,
"learning_rate": 6.689320388349515e-05,
"loss": 0.4539,
"step": 690
},
{
"epoch": 0.5441119315973572,
"grad_norm": 2.742603316305724,
"learning_rate": 6.786407766990291e-05,
"loss": 0.4376,
"step": 700
},
{
"epoch": 0.5518849591916052,
"grad_norm": 2.896538805797215,
"learning_rate": 6.883495145631068e-05,
"loss": 0.4113,
"step": 710
},
{
"epoch": 0.5596579867858531,
"grad_norm": 3.4815145675652013,
"learning_rate": 6.980582524271845e-05,
"loss": 0.4005,
"step": 720
},
{
"epoch": 0.567431014380101,
"grad_norm": 2.270621987351258,
"learning_rate": 7.077669902912621e-05,
"loss": 0.3886,
"step": 730
},
{
"epoch": 0.575204041974349,
"grad_norm": 2.7187166457028344,
"learning_rate": 7.174757281553398e-05,
"loss": 0.3765,
"step": 740
},
{
"epoch": 0.582977069568597,
"grad_norm": 3.184495607766821,
"learning_rate": 7.271844660194175e-05,
"loss": 0.3514,
"step": 750
},
{
"epoch": 0.5907500971628449,
"grad_norm": 2.379059028542611,
"learning_rate": 7.368932038834951e-05,
"loss": 0.341,
"step": 760
},
{
"epoch": 0.5985231247570929,
"grad_norm": 2.4310101684060705,
"learning_rate": 7.466019417475728e-05,
"loss": 0.3263,
"step": 770
},
{
"epoch": 0.6062961523513408,
"grad_norm": 2.5710567898956005,
"learning_rate": 7.563106796116505e-05,
"loss": 0.3117,
"step": 780
},
{
"epoch": 0.6140691799455888,
"grad_norm": 2.416082850923419,
"learning_rate": 7.660194174757281e-05,
"loss": 0.2944,
"step": 790
},
{
"epoch": 0.6218422075398368,
"grad_norm": 2.109447352556029,
"learning_rate": 7.757281553398059e-05,
"loss": 0.2832,
"step": 800
},
{
"epoch": 0.6296152351340847,
"grad_norm": 2.2994104123752988,
"learning_rate": 7.854368932038836e-05,
"loss": 0.2726,
"step": 810
},
{
"epoch": 0.6373882627283327,
"grad_norm": 2.9267817958414524,
"learning_rate": 7.951456310679612e-05,
"loss": 0.2606,
"step": 820
},
{
"epoch": 0.6451612903225806,
"grad_norm": 2.1699995560860197,
"learning_rate": 8.048543689320389e-05,
"loss": 0.2471,
"step": 830
},
{
"epoch": 0.6529343179168287,
"grad_norm": 1.680681697729957,
"learning_rate": 8.145631067961166e-05,
"loss": 0.2354,
"step": 840
},
{
"epoch": 0.6607073455110766,
"grad_norm": 1.88393966323043,
"learning_rate": 8.242718446601942e-05,
"loss": 0.2263,
"step": 850
},
{
"epoch": 0.6684803731053245,
"grad_norm": 2.7385999995385957,
"learning_rate": 8.339805825242719e-05,
"loss": 0.2109,
"step": 860
},
{
"epoch": 0.6762534006995725,
"grad_norm": 1.71125390353767,
"learning_rate": 8.436893203883496e-05,
"loss": 0.2034,
"step": 870
},
{
"epoch": 0.6840264282938204,
"grad_norm": 1.8856456915119455,
"learning_rate": 8.533980582524272e-05,
"loss": 0.1928,
"step": 880
},
{
"epoch": 0.6917994558880683,
"grad_norm": 1.54640226468132,
"learning_rate": 8.631067961165049e-05,
"loss": 0.1797,
"step": 890
},
{
"epoch": 0.6995724834823164,
"grad_norm": 1.3309710354478357,
"learning_rate": 8.728155339805826e-05,
"loss": 0.1712,
"step": 900
},
{
"epoch": 0.7073455110765643,
"grad_norm": 1.2897601826602498,
"learning_rate": 8.825242718446602e-05,
"loss": 0.1624,
"step": 910
},
{
"epoch": 0.7151185386708123,
"grad_norm": 1.1022252566548159,
"learning_rate": 8.922330097087379e-05,
"loss": 0.1509,
"step": 920
},
{
"epoch": 0.7228915662650602,
"grad_norm": 1.4797407171136665,
"learning_rate": 9.019417475728156e-05,
"loss": 0.1495,
"step": 930
},
{
"epoch": 0.7306645938593082,
"grad_norm": 1.4708443604696282,
"learning_rate": 9.116504854368933e-05,
"loss": 0.1407,
"step": 940
},
{
"epoch": 0.7384376214535562,
"grad_norm": 1.2062811965627402,
"learning_rate": 9.213592233009709e-05,
"loss": 0.1409,
"step": 950
},
{
"epoch": 0.7462106490478041,
"grad_norm": 0.991340571431122,
"learning_rate": 9.310679611650486e-05,
"loss": 0.1267,
"step": 960
},
{
"epoch": 0.7539836766420521,
"grad_norm": 1.0799976482859923,
"learning_rate": 9.407766990291263e-05,
"loss": 0.1294,
"step": 970
},
{
"epoch": 0.7617567042363,
"grad_norm": 1.0055612260520772,
"learning_rate": 9.504854368932039e-05,
"loss": 0.1167,
"step": 980
},
{
"epoch": 0.769529731830548,
"grad_norm": 0.99809772380525,
"learning_rate": 9.601941747572816e-05,
"loss": 0.1104,
"step": 990
},
{
"epoch": 0.777302759424796,
"grad_norm": 0.8341966391057081,
"learning_rate": 9.699029126213593e-05,
"loss": 0.104,
"step": 1000
},
{
"epoch": 0.7850757870190439,
"grad_norm": 0.7440675821604489,
"learning_rate": 9.796116504854369e-05,
"loss": 0.1,
"step": 1010
},
{
"epoch": 0.7928488146132919,
"grad_norm": 1.0089217362175902,
"learning_rate": 9.893203883495146e-05,
"loss": 0.0926,
"step": 1020
},
{
"epoch": 0.8006218422075398,
"grad_norm": 0.669475479205569,
"learning_rate": 9.990291262135923e-05,
"loss": 0.0855,
"step": 1030
},
{
"epoch": 0.8083948698017878,
"grad_norm": 0.7508700947786369,
"learning_rate": 9.999976722311928e-05,
"loss": 0.0824,
"step": 1040
},
{
"epoch": 0.8161678973960358,
"grad_norm": 0.5833904781662174,
"learning_rate": 9.999896256507963e-05,
"loss": 0.0794,
"step": 1050
},
{
"epoch": 0.8239409249902837,
"grad_norm": 0.6285661168460489,
"learning_rate": 9.99975831613399e-05,
"loss": 0.0737,
"step": 1060
},
{
"epoch": 0.8317139525845316,
"grad_norm": 0.574291945207083,
"learning_rate": 9.99956290277565e-05,
"loss": 0.0689,
"step": 1070
},
{
"epoch": 0.8394869801787797,
"grad_norm": 0.4876511188851532,
"learning_rate": 9.999310018679253e-05,
"loss": 0.0659,
"step": 1080
},
{
"epoch": 0.8472600077730276,
"grad_norm": 0.553792085705227,
"learning_rate": 9.998999666751738e-05,
"loss": 0.064,
"step": 1090
},
{
"epoch": 0.8550330353672756,
"grad_norm": 0.4302541179456994,
"learning_rate": 9.998631850560647e-05,
"loss": 0.0612,
"step": 1100
},
{
"epoch": 0.8628060629615235,
"grad_norm": 0.47710435973221127,
"learning_rate": 9.998206574334089e-05,
"loss": 0.0572,
"step": 1110
},
{
"epoch": 0.8705790905557714,
"grad_norm": 0.6437234961571133,
"learning_rate": 9.997723842960671e-05,
"loss": 0.0556,
"step": 1120
},
{
"epoch": 0.8783521181500195,
"grad_norm": 0.44297654610119586,
"learning_rate": 9.997183661989468e-05,
"loss": 0.0551,
"step": 1130
},
{
"epoch": 0.8861251457442674,
"grad_norm": 0.40789308058272417,
"learning_rate": 9.996586037629944e-05,
"loss": 0.052,
"step": 1140
},
{
"epoch": 0.8938981733385154,
"grad_norm": 0.412683747214122,
"learning_rate": 9.99593097675188e-05,
"loss": 0.0503,
"step": 1150
},
{
"epoch": 0.9016712009327633,
"grad_norm": 0.4042307583726218,
"learning_rate": 9.995218486885302e-05,
"loss": 0.0491,
"step": 1160
},
{
"epoch": 0.9094442285270112,
"grad_norm": 0.3917357425568075,
"learning_rate": 9.994448576220388e-05,
"loss": 0.0472,
"step": 1170
},
{
"epoch": 0.9172172561212593,
"grad_norm": 0.42847580686284464,
"learning_rate": 9.99362125360738e-05,
"loss": 0.0461,
"step": 1180
},
{
"epoch": 0.9249902837155072,
"grad_norm": 0.36275912435050334,
"learning_rate": 9.992736528556474e-05,
"loss": 0.045,
"step": 1190
},
{
"epoch": 0.9327633113097551,
"grad_norm": 0.29922320469821373,
"learning_rate": 9.991794411237725e-05,
"loss": 0.0429,
"step": 1200
},
{
"epoch": 0.9405363389040031,
"grad_norm": 0.3914552553052167,
"learning_rate": 9.990794912480913e-05,
"loss": 0.0426,
"step": 1210
},
{
"epoch": 0.948309366498251,
"grad_norm": 0.38509264588619113,
"learning_rate": 9.989738043775428e-05,
"loss": 0.0423,
"step": 1220
},
{
"epoch": 0.9560823940924991,
"grad_norm": 0.3259823714473278,
"learning_rate": 9.988623817270137e-05,
"loss": 0.0406,
"step": 1230
},
{
"epoch": 0.963855421686747,
"grad_norm": 0.31452752346688945,
"learning_rate": 9.987452245773243e-05,
"loss": 0.0398,
"step": 1240
},
{
"epoch": 0.9716284492809949,
"grad_norm": 0.338353347661472,
"learning_rate": 9.986223342752141e-05,
"loss": 0.0386,
"step": 1250
},
{
"epoch": 0.9794014768752429,
"grad_norm": 0.3002478510776732,
"learning_rate": 9.984937122333256e-05,
"loss": 0.0382,
"step": 1260
},
{
"epoch": 0.9871745044694908,
"grad_norm": 0.40670865355319846,
"learning_rate": 9.983593599301891e-05,
"loss": 0.037,
"step": 1270
},
{
"epoch": 0.9949475320637389,
"grad_norm": 0.6429283417769656,
"learning_rate": 9.982192789102047e-05,
"loss": 0.0373,
"step": 1280
},
{
"epoch": 1.0023319082782745,
"grad_norm": 0.2801176127856459,
"learning_rate": 9.980734707836248e-05,
"loss": 0.0354,
"step": 1290
},
{
"epoch": 1.0101049358725223,
"grad_norm": 0.3724656663427553,
"learning_rate": 9.979219372265365e-05,
"loss": 0.0344,
"step": 1300
},
{
"epoch": 1.0178779634667703,
"grad_norm": 0.2573842211275922,
"learning_rate": 9.97764679980841e-05,
"loss": 0.0333,
"step": 1310
},
{
"epoch": 1.0256509910610183,
"grad_norm": 0.24713329711260812,
"learning_rate": 9.97601700854234e-05,
"loss": 0.0336,
"step": 1320
},
{
"epoch": 1.0334240186552661,
"grad_norm": 0.23958650404178872,
"learning_rate": 9.974330017201859e-05,
"loss": 0.0328,
"step": 1330
},
{
"epoch": 1.0411970462495141,
"grad_norm": 0.22856953483452255,
"learning_rate": 9.972585845179188e-05,
"loss": 0.0316,
"step": 1340
},
{
"epoch": 1.0489700738437622,
"grad_norm": 0.25194563553211635,
"learning_rate": 9.970784512523848e-05,
"loss": 0.0311,
"step": 1350
},
{
"epoch": 1.0567431014380102,
"grad_norm": 0.23842502807724422,
"learning_rate": 9.968926039942435e-05,
"loss": 0.0306,
"step": 1360
},
{
"epoch": 1.064516129032258,
"grad_norm": 0.2701262980707877,
"learning_rate": 9.967010448798375e-05,
"loss": 0.0303,
"step": 1370
},
{
"epoch": 1.072289156626506,
"grad_norm": 0.25970806872919033,
"learning_rate": 9.965037761111681e-05,
"loss": 0.0295,
"step": 1380
},
{
"epoch": 1.080062184220754,
"grad_norm": 0.24331306995976767,
"learning_rate": 9.963007999558696e-05,
"loss": 0.0291,
"step": 1390
},
{
"epoch": 1.0878352118150019,
"grad_norm": 0.24100978412084842,
"learning_rate": 9.96092118747184e-05,
"loss": 0.0287,
"step": 1400
},
{
"epoch": 1.09560823940925,
"grad_norm": 0.2184207676934967,
"learning_rate": 9.958777348839336e-05,
"loss": 0.0277,
"step": 1410
},
{
"epoch": 1.103381267003498,
"grad_norm": 0.21996045383271537,
"learning_rate": 9.956576508304938e-05,
"loss": 0.0272,
"step": 1420
},
{
"epoch": 1.1111542945977457,
"grad_norm": 0.21628224837569832,
"learning_rate": 9.954318691167641e-05,
"loss": 0.0273,
"step": 1430
},
{
"epoch": 1.1189273221919938,
"grad_norm": 0.22298883167746006,
"learning_rate": 9.952003923381399e-05,
"loss": 0.0264,
"step": 1440
},
{
"epoch": 1.1267003497862418,
"grad_norm": 0.2246150118171153,
"learning_rate": 9.949632231554819e-05,
"loss": 0.0265,
"step": 1450
},
{
"epoch": 1.1344733773804898,
"grad_norm": 0.23186180017701422,
"learning_rate": 9.947203642950862e-05,
"loss": 0.0262,
"step": 1460
},
{
"epoch": 1.1422464049747376,
"grad_norm": 0.24648849817769855,
"learning_rate": 9.944718185486523e-05,
"loss": 0.0256,
"step": 1470
},
{
"epoch": 1.1500194325689856,
"grad_norm": 0.30413187225809407,
"learning_rate": 9.942175887732517e-05,
"loss": 0.025,
"step": 1480
},
{
"epoch": 1.1577924601632337,
"grad_norm": 0.23138441255979614,
"learning_rate": 9.93957677891294e-05,
"loss": 0.0251,
"step": 1490
},
{
"epoch": 1.1655654877574815,
"grad_norm": 0.24336635292704567,
"learning_rate": 9.93692088890495e-05,
"loss": 0.0244,
"step": 1500
},
{
"epoch": 1.1733385153517295,
"grad_norm": 0.21185931343408212,
"learning_rate": 9.934208248238403e-05,
"loss": 0.0243,
"step": 1510
},
{
"epoch": 1.1811115429459775,
"grad_norm": 0.19638440276738353,
"learning_rate": 9.93143888809552e-05,
"loss": 0.0239,
"step": 1520
},
{
"epoch": 1.1888845705402253,
"grad_norm": 0.21756730427462398,
"learning_rate": 9.928612840310514e-05,
"loss": 0.0232,
"step": 1530
},
{
"epoch": 1.1966575981344734,
"grad_norm": 0.21305245786411112,
"learning_rate": 9.925730137369242e-05,
"loss": 0.023,
"step": 1540
},
{
"epoch": 1.2044306257287214,
"grad_norm": 0.22828714542453502,
"learning_rate": 9.922790812408808e-05,
"loss": 0.0226,
"step": 1550
},
{
"epoch": 1.2122036533229692,
"grad_norm": 0.19930203197693303,
"learning_rate": 9.919794899217203e-05,
"loss": 0.0224,
"step": 1560
},
{
"epoch": 1.2199766809172172,
"grad_norm": 0.20433748074735467,
"learning_rate": 9.916742432232909e-05,
"loss": 0.0222,
"step": 1570
},
{
"epoch": 1.2277497085114653,
"grad_norm": 0.20184151603700423,
"learning_rate": 9.913633446544493e-05,
"loss": 0.0215,
"step": 1580
},
{
"epoch": 1.2355227361057133,
"grad_norm": 0.2081529716929725,
"learning_rate": 9.910467977890225e-05,
"loss": 0.0216,
"step": 1590
},
{
"epoch": 1.243295763699961,
"grad_norm": 0.27882658958977674,
"learning_rate": 9.907246062657648e-05,
"loss": 0.0234,
"step": 1600
},
{
"epoch": 1.2510687912942091,
"grad_norm": 0.30986570505669037,
"learning_rate": 9.903967737883168e-05,
"loss": 0.0248,
"step": 1610
},
{
"epoch": 1.258841818888457,
"grad_norm": 0.2205524563203657,
"learning_rate": 9.900633041251631e-05,
"loss": 0.0222,
"step": 1620
},
{
"epoch": 1.266614846482705,
"grad_norm": 0.19645809273796086,
"learning_rate": 9.897242011095883e-05,
"loss": 0.0208,
"step": 1630
},
{
"epoch": 1.274387874076953,
"grad_norm": 0.20220041423279006,
"learning_rate": 9.893794686396329e-05,
"loss": 0.0203,
"step": 1640
},
{
"epoch": 1.282160901671201,
"grad_norm": 0.19701852522226898,
"learning_rate": 9.890291106780495e-05,
"loss": 0.0199,
"step": 1650
},
{
"epoch": 1.289933929265449,
"grad_norm": 0.19055264078653414,
"learning_rate": 9.886731312522563e-05,
"loss": 0.0196,
"step": 1660
},
{
"epoch": 1.2977069568596968,
"grad_norm": 0.17498611562411595,
"learning_rate": 9.883115344542911e-05,
"loss": 0.0189,
"step": 1670
},
{
"epoch": 1.3054799844539449,
"grad_norm": 0.18664040460319856,
"learning_rate": 9.879443244407643e-05,
"loss": 0.019,
"step": 1680
},
{
"epoch": 1.3132530120481927,
"grad_norm": 0.186899833087349,
"learning_rate": 9.875715054328108e-05,
"loss": 0.0186,
"step": 1690
},
{
"epoch": 1.3210260396424407,
"grad_norm": 0.18960588899249967,
"learning_rate": 9.871930817160422e-05,
"loss": 0.018,
"step": 1700
},
{
"epoch": 1.3287990672366887,
"grad_norm": 0.2246998119833756,
"learning_rate": 9.868090576404966e-05,
"loss": 0.0185,
"step": 1710
},
{
"epoch": 1.3365720948309368,
"grad_norm": 0.18612335330028418,
"learning_rate": 9.864194376205895e-05,
"loss": 0.0181,
"step": 1720
},
{
"epoch": 1.3443451224251846,
"grad_norm": 0.19504144084562025,
"learning_rate": 9.860242261350623e-05,
"loss": 0.0174,
"step": 1730
},
{
"epoch": 1.3521181500194326,
"grad_norm": 0.1940460381388096,
"learning_rate": 9.856234277269313e-05,
"loss": 0.0173,
"step": 1740
},
{
"epoch": 1.3598911776136804,
"grad_norm": 0.1882305169943444,
"learning_rate": 9.852170470034356e-05,
"loss": 0.0173,
"step": 1750
},
{
"epoch": 1.3676642052079284,
"grad_norm": 0.1844609631309131,
"learning_rate": 9.848050886359835e-05,
"loss": 0.0171,
"step": 1760
},
{
"epoch": 1.3754372328021764,
"grad_norm": 0.16755276537647784,
"learning_rate": 9.843875573600989e-05,
"loss": 0.0169,
"step": 1770
},
{
"epoch": 1.3832102603964245,
"grad_norm": 0.16696100318285242,
"learning_rate": 9.839644579753677e-05,
"loss": 0.0164,
"step": 1780
},
{
"epoch": 1.3909832879906725,
"grad_norm": 0.1818918991872044,
"learning_rate": 9.835357953453821e-05,
"loss": 0.0163,
"step": 1790
},
{
"epoch": 1.3987563155849203,
"grad_norm": 0.18963889299339048,
"learning_rate": 9.831015743976843e-05,
"loss": 0.0158,
"step": 1800
},
{
"epoch": 1.4065293431791683,
"grad_norm": 0.1634818998306774,
"learning_rate": 9.8266180012371e-05,
"loss": 0.0154,
"step": 1810
},
{
"epoch": 1.4143023707734161,
"grad_norm": 0.197090617938791,
"learning_rate": 9.82216477578732e-05,
"loss": 0.0152,
"step": 1820
},
{
"epoch": 1.4220753983676642,
"grad_norm": 0.15850580732892314,
"learning_rate": 9.817656118818006e-05,
"loss": 0.0148,
"step": 1830
},
{
"epoch": 1.4298484259619122,
"grad_norm": 0.15186976089409565,
"learning_rate": 9.813092082156858e-05,
"loss": 0.0146,
"step": 1840
},
{
"epoch": 1.4376214535561602,
"grad_norm": 0.17352561780495993,
"learning_rate": 9.808472718268174e-05,
"loss": 0.0146,
"step": 1850
},
{
"epoch": 1.445394481150408,
"grad_norm": 0.16575574275890378,
"learning_rate": 9.803798080252245e-05,
"loss": 0.014,
"step": 1860
},
{
"epoch": 1.453167508744656,
"grad_norm": 0.16700742476995556,
"learning_rate": 9.79906822184475e-05,
"loss": 0.0139,
"step": 1870
},
{
"epoch": 1.460940536338904,
"grad_norm": 0.15966930327908196,
"learning_rate": 9.79428319741613e-05,
"loss": 0.0136,
"step": 1880
},
{
"epoch": 1.4687135639331519,
"grad_norm": 0.16653090767039122,
"learning_rate": 9.789443061970973e-05,
"loss": 0.0134,
"step": 1890
},
{
"epoch": 1.4764865915274,
"grad_norm": 0.190626948029195,
"learning_rate": 9.784547871147374e-05,
"loss": 0.0129,
"step": 1900
},
{
"epoch": 1.484259619121648,
"grad_norm": 0.1763435524084487,
"learning_rate": 9.779597681216297e-05,
"loss": 0.0127,
"step": 1910
},
{
"epoch": 1.492032646715896,
"grad_norm": 0.1672345466206139,
"learning_rate": 9.774592549080932e-05,
"loss": 0.0126,
"step": 1920
},
{
"epoch": 1.4998056743101438,
"grad_norm": 0.15678075113388967,
"learning_rate": 9.769532532276034e-05,
"loss": 0.0123,
"step": 1930
},
{
"epoch": 1.5075787019043918,
"grad_norm": 0.14761977546971236,
"learning_rate": 9.764417688967268e-05,
"loss": 0.0118,
"step": 1940
},
{
"epoch": 1.5153517294986396,
"grad_norm": 0.17063805371316892,
"learning_rate": 9.75924807795054e-05,
"loss": 0.0118,
"step": 1950
},
{
"epoch": 1.5231247570928876,
"grad_norm": 0.14571436924477155,
"learning_rate": 9.754023758651314e-05,
"loss": 0.0115,
"step": 1960
},
{
"epoch": 1.5308977846871357,
"grad_norm": 0.15118080081237256,
"learning_rate": 9.74874479112394e-05,
"loss": 0.0112,
"step": 1970
},
{
"epoch": 1.5386708122813837,
"grad_norm": 0.1524883133010242,
"learning_rate": 9.743411236050953e-05,
"loss": 0.0112,
"step": 1980
},
{
"epoch": 1.5464438398756317,
"grad_norm": 0.15389453717185902,
"learning_rate": 9.738023154742383e-05,
"loss": 0.0114,
"step": 1990
},
{
"epoch": 1.5542168674698795,
"grad_norm": 0.14981753056593858,
"learning_rate": 9.732580609135047e-05,
"loss": 0.0108,
"step": 2000
},
{
"epoch": 1.5619898950641273,
"grad_norm": 1.9198802022850896,
"learning_rate": 9.727083661791836e-05,
"loss": 0.0165,
"step": 2010
},
{
"epoch": 1.5697629226583754,
"grad_norm": 0.26855012872950995,
"learning_rate": 9.721532375900997e-05,
"loss": 0.0144,
"step": 2020
},
{
"epoch": 1.5775359502526234,
"grad_norm": 0.2071283206940206,
"learning_rate": 9.71592681527541e-05,
"loss": 0.0117,
"step": 2030
},
{
"epoch": 1.5853089778468714,
"grad_norm": 0.14770534677344455,
"learning_rate": 9.710267044351846e-05,
"loss": 0.0109,
"step": 2040
},
{
"epoch": 1.5930820054411194,
"grad_norm": 0.14368540945251085,
"learning_rate": 9.704553128190239e-05,
"loss": 0.0104,
"step": 2050
},
{
"epoch": 1.6008550330353672,
"grad_norm": 0.13894193794675433,
"learning_rate": 9.698785132472926e-05,
"loss": 0.01,
"step": 2060
},
{
"epoch": 1.6086280606296153,
"grad_norm": 0.13441313900735385,
"learning_rate": 9.692963123503894e-05,
"loss": 0.0094,
"step": 2070
},
{
"epoch": 1.616401088223863,
"grad_norm": 0.12841760115701012,
"learning_rate": 9.687087168208026e-05,
"loss": 0.0094,
"step": 2080
},
{
"epoch": 1.624174115818111,
"grad_norm": 0.1378414358921271,
"learning_rate": 9.681157334130326e-05,
"loss": 0.0092,
"step": 2090
},
{
"epoch": 1.6319471434123591,
"grad_norm": 0.13875742382402598,
"learning_rate": 9.67517368943514e-05,
"loss": 0.0093,
"step": 2100
},
{
"epoch": 1.6397201710066072,
"grad_norm": 0.12150978897988295,
"learning_rate": 9.669136302905374e-05,
"loss": 0.0094,
"step": 2110
},
{
"epoch": 1.6474931986008552,
"grad_norm": 0.13386154431399386,
"learning_rate": 9.663045243941711e-05,
"loss": 0.009,
"step": 2120
},
{
"epoch": 1.655266226195103,
"grad_norm": 0.14353227081750533,
"learning_rate": 9.656900582561801e-05,
"loss": 0.009,
"step": 2130
},
{
"epoch": 1.6630392537893508,
"grad_norm": 0.13602583881387148,
"learning_rate": 9.650702389399461e-05,
"loss": 0.0088,
"step": 2140
},
{
"epoch": 1.6708122813835988,
"grad_norm": 0.13201111610838254,
"learning_rate": 9.644450735703872e-05,
"loss": 0.0087,
"step": 2150
},
{
"epoch": 1.6785853089778469,
"grad_norm": 0.11982302853617144,
"learning_rate": 9.638145693338743e-05,
"loss": 0.0087,
"step": 2160
},
{
"epoch": 1.6863583365720949,
"grad_norm": 0.11312867242927008,
"learning_rate": 9.631787334781497e-05,
"loss": 0.0087,
"step": 2170
},
{
"epoch": 1.694131364166343,
"grad_norm": 0.11767848177751597,
"learning_rate": 9.625375733122439e-05,
"loss": 0.0085,
"step": 2180
},
{
"epoch": 1.7019043917605907,
"grad_norm": 0.11041145208342112,
"learning_rate": 9.618910962063903e-05,
"loss": 0.0082,
"step": 2190
},
{
"epoch": 1.7096774193548387,
"grad_norm": 0.10799469734884444,
"learning_rate": 9.612393095919423e-05,
"loss": 0.0078,
"step": 2200
},
{
"epoch": 1.7174504469490866,
"grad_norm": 0.11477500796146126,
"learning_rate": 9.605822209612862e-05,
"loss": 0.0078,
"step": 2210
},
{
"epoch": 1.7252234745433346,
"grad_norm": 0.17423934102739813,
"learning_rate": 9.599198378677559e-05,
"loss": 0.0079,
"step": 2220
},
{
"epoch": 1.7329965021375826,
"grad_norm": 0.11632697553641431,
"learning_rate": 9.592521679255464e-05,
"loss": 0.008,
"step": 2230
},
{
"epoch": 1.7407695297318306,
"grad_norm": 0.12288031599865465,
"learning_rate": 9.585792188096252e-05,
"loss": 0.008,
"step": 2240
},
{
"epoch": 1.7485425573260787,
"grad_norm": 0.15984027115213267,
"learning_rate": 9.579009982556452e-05,
"loss": 0.0079,
"step": 2250
},
{
"epoch": 1.7563155849203265,
"grad_norm": 0.11306629622107851,
"learning_rate": 9.572175140598553e-05,
"loss": 0.0076,
"step": 2260
},
{
"epoch": 1.7640886125145743,
"grad_norm": 0.1278172766126942,
"learning_rate": 9.565287740790101e-05,
"loss": 0.0077,
"step": 2270
},
{
"epoch": 1.7718616401088223,
"grad_norm": 0.1251067207191811,
"learning_rate": 9.55834786230281e-05,
"loss": 0.0074,
"step": 2280
},
{
"epoch": 1.7796346677030703,
"grad_norm": 0.12072123910224757,
"learning_rate": 9.551355584911641e-05,
"loss": 0.0077,
"step": 2290
},
{
"epoch": 1.7874076952973184,
"grad_norm": 0.13420830427564445,
"learning_rate": 9.544310988993889e-05,
"loss": 0.0076,
"step": 2300
},
{
"epoch": 1.7951807228915664,
"grad_norm": 0.11275927835211168,
"learning_rate": 9.53721415552826e-05,
"loss": 0.0072,
"step": 2310
},
{
"epoch": 1.8029537504858142,
"grad_norm": 0.11663410063710995,
"learning_rate": 9.530065166093936e-05,
"loss": 0.007,
"step": 2320
},
{
"epoch": 1.8107267780800622,
"grad_norm": 0.11515297587501505,
"learning_rate": 9.52286410286964e-05,
"loss": 0.007,
"step": 2330
},
{
"epoch": 1.81849980567431,
"grad_norm": 0.12034539299056037,
"learning_rate": 9.515611048632692e-05,
"loss": 0.0071,
"step": 2340
},
{
"epoch": 1.826272833268558,
"grad_norm": 0.13525219378529196,
"learning_rate": 9.508306086758059e-05,
"loss": 0.0069,
"step": 2350
},
{
"epoch": 1.834045860862806,
"grad_norm": 0.11355987672855966,
"learning_rate": 9.500949301217388e-05,
"loss": 0.0066,
"step": 2360
},
{
"epoch": 1.841818888457054,
"grad_norm": 0.10553177998244642,
"learning_rate": 9.493540776578053e-05,
"loss": 0.0067,
"step": 2370
},
{
"epoch": 1.8495919160513021,
"grad_norm": 0.10733467006153838,
"learning_rate": 9.486080598002176e-05,
"loss": 0.0066,
"step": 2380
},
{
"epoch": 1.85736494364555,
"grad_norm": 0.11088682968161348,
"learning_rate": 9.478568851245643e-05,
"loss": 0.0064,
"step": 2390
},
{
"epoch": 1.8651379712397977,
"grad_norm": 0.11788349609407499,
"learning_rate": 9.471005622657134e-05,
"loss": 0.0064,
"step": 2400
},
{
"epoch": 1.8729109988340458,
"grad_norm": 0.1382091673100836,
"learning_rate": 9.463390999177113e-05,
"loss": 0.0066,
"step": 2410
},
{
"epoch": 1.8806840264282938,
"grad_norm": 0.11080598206299531,
"learning_rate": 9.455725068336837e-05,
"loss": 0.0067,
"step": 2420
},
{
"epoch": 1.8884570540225418,
"grad_norm": 0.11345029521361591,
"learning_rate": 9.44800791825735e-05,
"loss": 0.0063,
"step": 2430
},
{
"epoch": 1.8962300816167899,
"grad_norm": 0.12000127996735861,
"learning_rate": 9.440239637648473e-05,
"loss": 0.0064,
"step": 2440
},
{
"epoch": 1.9040031092110377,
"grad_norm": 0.11742977407222017,
"learning_rate": 9.432420315807771e-05,
"loss": 0.0063,
"step": 2450
},
{
"epoch": 1.9117761368052857,
"grad_norm": 0.10552545531273427,
"learning_rate": 9.424550042619544e-05,
"loss": 0.0062,
"step": 2460
},
{
"epoch": 1.9195491643995335,
"grad_norm": 0.0996572096889176,
"learning_rate": 9.416628908553784e-05,
"loss": 0.0059,
"step": 2470
},
{
"epoch": 1.9273221919937815,
"grad_norm": 0.09626416174818198,
"learning_rate": 9.408657004665131e-05,
"loss": 0.0058,
"step": 2480
},
{
"epoch": 1.9350952195880295,
"grad_norm": 0.09783061431201936,
"learning_rate": 9.400634422591838e-05,
"loss": 0.0058,
"step": 2490
},
{
"epoch": 1.9428682471822776,
"grad_norm": 0.11102429789832584,
"learning_rate": 9.392561254554713e-05,
"loss": 0.0059,
"step": 2500
},
{
"epoch": 1.9506412747765256,
"grad_norm": 0.10011364069857738,
"learning_rate": 9.384437593356048e-05,
"loss": 0.0059,
"step": 2510
},
{
"epoch": 1.9584143023707734,
"grad_norm": 0.10584225350181801,
"learning_rate": 9.37626353237857e-05,
"loss": 0.0061,
"step": 2520
},
{
"epoch": 1.9661873299650212,
"grad_norm": 0.10444281925647403,
"learning_rate": 9.368039165584359e-05,
"loss": 0.0061,
"step": 2530
},
{
"epoch": 1.9739603575592692,
"grad_norm": 0.10604191503186843,
"learning_rate": 9.359764587513761e-05,
"loss": 0.0057,
"step": 2540
},
{
"epoch": 1.9817333851535173,
"grad_norm": 0.09211496249761421,
"learning_rate": 9.351439893284317e-05,
"loss": 0.0056,
"step": 2550
},
{
"epoch": 1.9895064127477653,
"grad_norm": 0.14501544032188865,
"learning_rate": 9.343065178589654e-05,
"loss": 0.0056,
"step": 2560
},
{
"epoch": 1.9972794403420133,
"grad_norm": 0.11456289818413888,
"learning_rate": 9.334640539698396e-05,
"loss": 0.0058,
"step": 2570
},
{
"epoch": 2.004663816556549,
"grad_norm": 0.08068921187905599,
"learning_rate": 9.326166073453051e-05,
"loss": 0.0052,
"step": 2580
},
{
"epoch": 2.012436844150797,
"grad_norm": 0.10032295518621465,
"learning_rate": 9.317641877268901e-05,
"loss": 0.0045,
"step": 2590
},
{
"epoch": 2.0202098717450445,
"grad_norm": 0.0906748520799375,
"learning_rate": 9.309068049132882e-05,
"loss": 0.0044,
"step": 2600
},
{
"epoch": 2.0279828993392925,
"grad_norm": 0.07982904566291607,
"learning_rate": 9.300444687602457e-05,
"loss": 0.0041,
"step": 2610
},
{
"epoch": 2.0357559269335406,
"grad_norm": 0.1253437061299033,
"learning_rate": 9.291771891804478e-05,
"loss": 0.004,
"step": 2620
},
{
"epoch": 2.0435289545277886,
"grad_norm": 0.09180533228491213,
"learning_rate": 9.283049761434059e-05,
"loss": 0.0041,
"step": 2630
},
{
"epoch": 2.0513019821220366,
"grad_norm": 0.09238927797437002,
"learning_rate": 9.27427839675342e-05,
"loss": 0.004,
"step": 2640
},
{
"epoch": 2.0590750097162847,
"grad_norm": 0.08085544322712232,
"learning_rate": 9.26545789859073e-05,
"loss": 0.004,
"step": 2650
},
{
"epoch": 2.0668480373105322,
"grad_norm": 0.08278933444675111,
"learning_rate": 9.256588368338968e-05,
"loss": 0.0039,
"step": 2660
},
{
"epoch": 2.0746210649047803,
"grad_norm": 0.0821491544443607,
"learning_rate": 9.24766990795473e-05,
"loss": 0.004,
"step": 2670
},
{
"epoch": 2.0823940924990283,
"grad_norm": 0.0820536970017854,
"learning_rate": 9.238702619957083e-05,
"loss": 0.0039,
"step": 2680
},
{
"epoch": 2.0901671200932763,
"grad_norm": 0.07783393489314461,
"learning_rate": 9.229686607426368e-05,
"loss": 0.0038,
"step": 2690
},
{
"epoch": 2.0979401476875243,
"grad_norm": 0.08972295811517289,
"learning_rate": 9.220621974003025e-05,
"loss": 0.0038,
"step": 2700
},
{
"epoch": 2.1057131752817724,
"grad_norm": 0.10850037565865495,
"learning_rate": 9.211508823886398e-05,
"loss": 0.0038,
"step": 2710
},
{
"epoch": 2.1134862028760204,
"grad_norm": 0.1165050619921317,
"learning_rate": 9.202347261833536e-05,
"loss": 0.004,
"step": 2720
},
{
"epoch": 2.121259230470268,
"grad_norm": 0.09613835587740172,
"learning_rate": 9.193137393157994e-05,
"loss": 0.004,
"step": 2730
},
{
"epoch": 2.129032258064516,
"grad_norm": 0.08222637851192974,
"learning_rate": 9.183879323728613e-05,
"loss": 0.0041,
"step": 2740
},
{
"epoch": 2.136805285658764,
"grad_norm": 0.09364974494864589,
"learning_rate": 9.174573159968319e-05,
"loss": 0.0041,
"step": 2750
},
{
"epoch": 2.144578313253012,
"grad_norm": 0.09079410215789208,
"learning_rate": 9.165219008852879e-05,
"loss": 0.0037,
"step": 2760
},
{
"epoch": 2.15235134084726,
"grad_norm": 0.08693641953192356,
"learning_rate": 9.155816977909688e-05,
"loss": 0.0038,
"step": 2770
},
{
"epoch": 2.160124368441508,
"grad_norm": 0.08407058222727168,
"learning_rate": 9.146367175216527e-05,
"loss": 0.0038,
"step": 2780
},
{
"epoch": 2.167897396035756,
"grad_norm": 0.08864727152361798,
"learning_rate": 9.136869709400318e-05,
"loss": 0.0037,
"step": 2790
},
{
"epoch": 2.1756704236300037,
"grad_norm": 0.08585633309380411,
"learning_rate": 9.127324689635879e-05,
"loss": 0.0038,
"step": 2800
},
{
"epoch": 2.1834434512242518,
"grad_norm": 0.09079038707160886,
"learning_rate": 9.117732225644668e-05,
"loss": 0.0037,
"step": 2810
},
{
"epoch": 2.1912164788185,
"grad_norm": 0.08945120074693964,
"learning_rate": 9.108092427693521e-05,
"loss": 0.0039,
"step": 2820
},
{
"epoch": 2.198989506412748,
"grad_norm": 0.09181352080010166,
"learning_rate": 9.098405406593387e-05,
"loss": 0.0038,
"step": 2830
},
{
"epoch": 2.206762534006996,
"grad_norm": 0.08585652120551285,
"learning_rate": 9.088671273698049e-05,
"loss": 0.0038,
"step": 2840
},
{
"epoch": 2.214535561601244,
"grad_norm": 0.08716662155761687,
"learning_rate": 9.078890140902851e-05,
"loss": 0.0037,
"step": 2850
},
{
"epoch": 2.2223085891954915,
"grad_norm": 0.08274857092352111,
"learning_rate": 9.069062120643406e-05,
"loss": 0.0036,
"step": 2860
},
{
"epoch": 2.2300816167897395,
"grad_norm": 0.08054936545881469,
"learning_rate": 9.059187325894305e-05,
"loss": 0.0037,
"step": 2870
},
{
"epoch": 2.2378546443839875,
"grad_norm": 0.08928250602229335,
"learning_rate": 9.04926587016782e-05,
"loss": 0.0037,
"step": 2880
},
{
"epoch": 2.2456276719782355,
"grad_norm": 0.14580475649103322,
"learning_rate": 9.039297867512598e-05,
"loss": 0.0039,
"step": 2890
},
{
"epoch": 2.2534006995724836,
"grad_norm": 0.09684044113116728,
"learning_rate": 9.02928343251235e-05,
"loss": 0.0036,
"step": 2900
},
{
"epoch": 2.2611737271667316,
"grad_norm": 4.610227312049106,
"learning_rate": 9.019222680284531e-05,
"loss": 0.0177,
"step": 2910
},
{
"epoch": 2.2689467547609796,
"grad_norm": 3.9441098766180716,
"learning_rate": 9.009115726479028e-05,
"loss": 0.0845,
"step": 2920
},
{
"epoch": 2.276719782355227,
"grad_norm": 0.30362597840942473,
"learning_rate": 8.998962687276809e-05,
"loss": 0.0336,
"step": 2930
},
{
"epoch": 2.2844928099494752,
"grad_norm": 0.17872746957322608,
"learning_rate": 8.988763679388614e-05,
"loss": 0.0106,
"step": 2940
},
{
"epoch": 2.2922658375437233,
"grad_norm": 0.11891620122269164,
"learning_rate": 8.97851882005359e-05,
"loss": 0.0069,
"step": 2950
},
{
"epoch": 2.3000388651379713,
"grad_norm": 0.11931811737114671,
"learning_rate": 8.968228227037961e-05,
"loss": 0.0054,
"step": 2960
},
{
"epoch": 2.3078118927322193,
"grad_norm": 0.11297205800033525,
"learning_rate": 8.957892018633662e-05,
"loss": 0.005,
"step": 2970
},
{
"epoch": 2.3155849203264673,
"grad_norm": 0.08607374863983304,
"learning_rate": 8.947510313656983e-05,
"loss": 0.0046,
"step": 2980
},
{
"epoch": 2.323357947920715,
"grad_norm": 0.08646791211129752,
"learning_rate": 8.93708323144721e-05,
"loss": 0.0043,
"step": 2990
},
{
"epoch": 2.331130975514963,
"grad_norm": 0.09633865197527709,
"learning_rate": 8.926610891865239e-05,
"loss": 0.0042,
"step": 3000
},
{
"epoch": 2.338904003109211,
"grad_norm": 0.07797950909850374,
"learning_rate": 8.916093415292212e-05,
"loss": 0.004,
"step": 3010
},
{
"epoch": 2.346677030703459,
"grad_norm": 0.08803265774925108,
"learning_rate": 8.905530922628129e-05,
"loss": 0.0039,
"step": 3020
},
{
"epoch": 2.354450058297707,
"grad_norm": 0.08578035815680632,
"learning_rate": 8.894923535290452e-05,
"loss": 0.0038,
"step": 3030
},
{
"epoch": 2.362223085891955,
"grad_norm": 0.0723027097191447,
"learning_rate": 8.884271375212713e-05,
"loss": 0.0037,
"step": 3040
},
{
"epoch": 2.369996113486203,
"grad_norm": 0.07933890872564231,
"learning_rate": 8.873574564843123e-05,
"loss": 0.0037,
"step": 3050
},
{
"epoch": 2.3777691410804507,
"grad_norm": 0.08309597358808056,
"learning_rate": 8.862833227143143e-05,
"loss": 0.0037,
"step": 3060
},
{
"epoch": 2.3855421686746987,
"grad_norm": 0.0796136928178535,
"learning_rate": 8.852047485586093e-05,
"loss": 0.0034,
"step": 3070
},
{
"epoch": 2.3933151962689467,
"grad_norm": 0.0906501106792225,
"learning_rate": 8.841217464155714e-05,
"loss": 0.0036,
"step": 3080
},
{
"epoch": 2.4010882238631948,
"grad_norm": 0.09232238343629308,
"learning_rate": 8.830343287344753e-05,
"loss": 0.0036,
"step": 3090
},
{
"epoch": 2.408861251457443,
"grad_norm": 0.07677160549185906,
"learning_rate": 8.819425080153533e-05,
"loss": 0.0036,
"step": 3100
},
{
"epoch": 2.416634279051691,
"grad_norm": 0.07965459624459517,
"learning_rate": 8.808462968088512e-05,
"loss": 0.0036,
"step": 3110
},
{
"epoch": 2.4244073066459384,
"grad_norm": 0.08152518289735053,
"learning_rate": 8.797457077160836e-05,
"loss": 0.0036,
"step": 3120
},
{
"epoch": 2.4321803342401864,
"grad_norm": 0.07840524642090889,
"learning_rate": 8.786407533884902e-05,
"loss": 0.0035,
"step": 3130
},
{
"epoch": 2.4399533618344345,
"grad_norm": 0.0753906754915079,
"learning_rate": 8.775314465276896e-05,
"loss": 0.0034,
"step": 3140
},
{
"epoch": 2.4477263894286825,
"grad_norm": 0.07553857495501508,
"learning_rate": 8.764177998853334e-05,
"loss": 0.0034,
"step": 3150
},
{
"epoch": 2.4554994170229305,
"grad_norm": 0.07989895676116511,
"learning_rate": 8.752998262629593e-05,
"loss": 0.0034,
"step": 3160
},
{
"epoch": 2.4632724446171785,
"grad_norm": 0.08604538052038023,
"learning_rate": 8.741775385118447e-05,
"loss": 0.0035,
"step": 3170
},
{
"epoch": 2.4710454722114266,
"grad_norm": 0.07301243329178161,
"learning_rate": 8.730509495328587e-05,
"loss": 0.0034,
"step": 3180
},
{
"epoch": 2.478818499805674,
"grad_norm": 0.07831526315789164,
"learning_rate": 8.71920072276313e-05,
"loss": 0.0034,
"step": 3190
},
{
"epoch": 2.486591527399922,
"grad_norm": 0.10194734979674533,
"learning_rate": 8.707849197418147e-05,
"loss": 0.0034,
"step": 3200
},
{
"epoch": 2.49436455499417,
"grad_norm": 0.07574798451139898,
"learning_rate": 8.696455049781147e-05,
"loss": 0.0035,
"step": 3210
},
{
"epoch": 2.5021375825884182,
"grad_norm": 0.07681960051751799,
"learning_rate": 8.6850184108296e-05,
"loss": 0.0033,
"step": 3220
},
{
"epoch": 2.5099106101826663,
"grad_norm": 0.07836534928409306,
"learning_rate": 8.673539412029413e-05,
"loss": 0.0033,
"step": 3230
},
{
"epoch": 2.517683637776914,
"grad_norm": 0.08548600777071955,
"learning_rate": 8.662018185333425e-05,
"loss": 0.0034,
"step": 3240
},
{
"epoch": 2.525456665371162,
"grad_norm": 0.0873567409210325,
"learning_rate": 8.650454863179899e-05,
"loss": 0.0032,
"step": 3250
},
{
"epoch": 2.53322969296541,
"grad_norm": 0.08119732193506358,
"learning_rate": 8.638849578490983e-05,
"loss": 0.0032,
"step": 3260
},
{
"epoch": 2.541002720559658,
"grad_norm": 0.07579199273478096,
"learning_rate": 8.627202464671196e-05,
"loss": 0.0031,
"step": 3270
},
{
"epoch": 2.548775748153906,
"grad_norm": 0.07826696555904925,
"learning_rate": 8.615513655605888e-05,
"loss": 0.0031,
"step": 3280
},
{
"epoch": 2.556548775748154,
"grad_norm": 0.07982815545083491,
"learning_rate": 8.6037832856597e-05,
"loss": 0.0031,
"step": 3290
},
{
"epoch": 2.564321803342402,
"grad_norm": 0.08481539873916785,
"learning_rate": 8.592011489675025e-05,
"loss": 0.0031,
"step": 3300
},
{
"epoch": 2.57209483093665,
"grad_norm": 0.07916097799005735,
"learning_rate": 8.580198402970456e-05,
"loss": 0.003,
"step": 3310
},
{
"epoch": 2.579867858530898,
"grad_norm": 0.07473474197283404,
"learning_rate": 8.56834416133922e-05,
"loss": 0.003,
"step": 3320
},
{
"epoch": 2.5876408861251456,
"grad_norm": 0.09130974953799981,
"learning_rate": 8.556448901047638e-05,
"loss": 0.0031,
"step": 3330
},
{
"epoch": 2.5954139137193937,
"grad_norm": 0.1047401105445018,
"learning_rate": 8.544512758833537e-05,
"loss": 0.003,
"step": 3340
},
{
"epoch": 2.6031869413136417,
"grad_norm": 0.06673230506522639,
"learning_rate": 8.532535871904695e-05,
"loss": 0.003,
"step": 3350
},
{
"epoch": 2.6109599689078897,
"grad_norm": 0.07876904924601609,
"learning_rate": 8.520518377937253e-05,
"loss": 0.0031,
"step": 3360
},
{
"epoch": 2.6187329965021373,
"grad_norm": 0.07009255329786931,
"learning_rate": 8.508460415074137e-05,
"loss": 0.0029,
"step": 3370
},
{
"epoch": 2.6265060240963853,
"grad_norm": 0.07519024159981848,
"learning_rate": 8.49636212192347e-05,
"loss": 0.003,
"step": 3380
},
{
"epoch": 2.6342790516906334,
"grad_norm": 0.08094145785537127,
"learning_rate": 8.484223637556974e-05,
"loss": 0.0031,
"step": 3390
},
{
"epoch": 2.6420520792848814,
"grad_norm": 0.07736222204990759,
"learning_rate": 8.472045101508382e-05,
"loss": 0.003,
"step": 3400
},
{
"epoch": 2.6498251068791294,
"grad_norm": 0.0776215180930073,
"learning_rate": 8.459826653771821e-05,
"loss": 0.003,
"step": 3410
},
{
"epoch": 2.6575981344733774,
"grad_norm": 0.06742420975546093,
"learning_rate": 8.44756843480021e-05,
"loss": 0.0028,
"step": 3420
},
{
"epoch": 2.6653711620676255,
"grad_norm": 0.0698399904847748,
"learning_rate": 8.435270585503647e-05,
"loss": 0.0029,
"step": 3430
},
{
"epoch": 2.6731441896618735,
"grad_norm": 0.07591621006160705,
"learning_rate": 8.422933247247786e-05,
"loss": 0.0029,
"step": 3440
},
{
"epoch": 2.6809172172561215,
"grad_norm": 0.06930363699655918,
"learning_rate": 8.410556561852212e-05,
"loss": 0.0029,
"step": 3450
},
{
"epoch": 2.688690244850369,
"grad_norm": 0.07067953717607145,
"learning_rate": 8.39814067158881e-05,
"loss": 0.0028,
"step": 3460
},
{
"epoch": 2.696463272444617,
"grad_norm": 0.077677315745107,
"learning_rate": 8.385685719180132e-05,
"loss": 0.0029,
"step": 3470
},
{
"epoch": 2.704236300038865,
"grad_norm": 0.07570167649828695,
"learning_rate": 8.373191847797754e-05,
"loss": 0.0028,
"step": 3480
},
{
"epoch": 2.712009327633113,
"grad_norm": 0.06955310603007175,
"learning_rate": 8.360659201060639e-05,
"loss": 0.0029,
"step": 3490
},
{
"epoch": 2.719782355227361,
"grad_norm": 0.0659051677019647,
"learning_rate": 8.348087923033465e-05,
"loss": 0.0029,
"step": 3500
},
{
"epoch": 2.727555382821609,
"grad_norm": 0.07181363122453262,
"learning_rate": 8.335478158224996e-05,
"loss": 0.0027,
"step": 3510
},
{
"epoch": 2.735328410415857,
"grad_norm": 0.06585918912737933,
"learning_rate": 8.3228300515864e-05,
"loss": 0.0027,
"step": 3520
},
{
"epoch": 2.743101438010105,
"grad_norm": 0.08664632265940041,
"learning_rate": 8.310143748509588e-05,
"loss": 0.0029,
"step": 3530
},
{
"epoch": 2.750874465604353,
"grad_norm": 0.0773656650662347,
"learning_rate": 8.297419394825554e-05,
"loss": 0.0028,
"step": 3540
},
{
"epoch": 2.758647493198601,
"grad_norm": 0.07477928292429042,
"learning_rate": 8.28465713680268e-05,
"loss": 0.0029,
"step": 3550
},
{
"epoch": 2.766420520792849,
"grad_norm": 0.06614018293609181,
"learning_rate": 8.271857121145073e-05,
"loss": 0.0028,
"step": 3560
},
{
"epoch": 2.774193548387097,
"grad_norm": 0.07953850815353782,
"learning_rate": 8.25901949499086e-05,
"loss": 0.0028,
"step": 3570
},
{
"epoch": 2.781966575981345,
"grad_norm": 0.06857404401820201,
"learning_rate": 8.246144405910517e-05,
"loss": 0.0028,
"step": 3580
},
{
"epoch": 2.7897396035755926,
"grad_norm": 0.06706667623803336,
"learning_rate": 8.233232001905151e-05,
"loss": 0.0029,
"step": 3590
},
{
"epoch": 2.7975126311698406,
"grad_norm": 0.06548223623701806,
"learning_rate": 8.220282431404818e-05,
"loss": 0.0027,
"step": 3600
}
],
"logging_steps": 10,
"max_steps": 10296,
"num_input_tokens_seen": 0,
"num_train_epochs": 8,
"save_steps": 1800,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 265508976328704.0,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}