Jinwu01's picture
Upload folder using huggingface_hub
c3dc34b verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002,
"grad_norm": 0.1880519837141037,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.5832,
"step": 10
},
{
"epoch": 0.004,
"grad_norm": 0.4404986798763275,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.834,
"step": 20
},
{
"epoch": 0.006,
"grad_norm": 0.8119221925735474,
"learning_rate": 3e-06,
"loss": 2.3087,
"step": 30
},
{
"epoch": 0.008,
"grad_norm": 1.038776159286499,
"learning_rate": 4.000000000000001e-06,
"loss": 1.8821,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 0.41819652915000916,
"learning_rate": 5e-06,
"loss": 4.2502,
"step": 50
},
{
"epoch": 0.012,
"grad_norm": 0.6414370536804199,
"learning_rate": 6e-06,
"loss": 3.6424,
"step": 60
},
{
"epoch": 0.014,
"grad_norm": 1.6446870565414429,
"learning_rate": 7.000000000000001e-06,
"loss": 3.1319,
"step": 70
},
{
"epoch": 0.016,
"grad_norm": 3.257716655731201,
"learning_rate": 8.000000000000001e-06,
"loss": 1.7977,
"step": 80
},
{
"epoch": 0.018,
"grad_norm": 0.3212912678718567,
"learning_rate": 9e-06,
"loss": 2.3069,
"step": 90
},
{
"epoch": 0.02,
"grad_norm": 0.817084014415741,
"learning_rate": 1e-05,
"loss": 1.9409,
"step": 100
},
{
"epoch": 0.022,
"grad_norm": 6.956598281860352,
"learning_rate": 1.1000000000000001e-05,
"loss": 3.9241,
"step": 110
},
{
"epoch": 0.024,
"grad_norm": 0.4231731593608856,
"learning_rate": 1.2e-05,
"loss": 1.6482,
"step": 120
},
{
"epoch": 0.026,
"grad_norm": 2.118069887161255,
"learning_rate": 1.3000000000000001e-05,
"loss": 3.1564,
"step": 130
},
{
"epoch": 0.028,
"grad_norm": 0.23121468722820282,
"learning_rate": 1.4000000000000001e-05,
"loss": 2.426,
"step": 140
},
{
"epoch": 0.03,
"grad_norm": 0.7299257516860962,
"learning_rate": 1.5e-05,
"loss": 1.3026,
"step": 150
},
{
"epoch": 0.032,
"grad_norm": 2.510737895965576,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.1178,
"step": 160
},
{
"epoch": 0.034,
"grad_norm": 2.180232524871826,
"learning_rate": 1.7000000000000003e-05,
"loss": 4.0788,
"step": 170
},
{
"epoch": 0.036,
"grad_norm": 1.0466824769973755,
"learning_rate": 1.8e-05,
"loss": 1.6576,
"step": 180
},
{
"epoch": 0.038,
"grad_norm": 0.0,
"learning_rate": 1.9e-05,
"loss": 1.4615,
"step": 190
},
{
"epoch": 0.04,
"grad_norm": 0.35433244705200195,
"learning_rate": 2e-05,
"loss": 2.2587,
"step": 200
},
{
"epoch": 0.042,
"grad_norm": 0.0,
"learning_rate": 2.1e-05,
"loss": 2.3845,
"step": 210
},
{
"epoch": 0.044,
"grad_norm": 1.116017460823059,
"learning_rate": 2.2000000000000003e-05,
"loss": 1.3799,
"step": 220
},
{
"epoch": 0.046,
"grad_norm": 0.6432040333747864,
"learning_rate": 2.3000000000000003e-05,
"loss": 1.4828,
"step": 230
},
{
"epoch": 0.048,
"grad_norm": 0.34233808517456055,
"learning_rate": 2.4e-05,
"loss": 2.1998,
"step": 240
},
{
"epoch": 0.05,
"grad_norm": 0.6910992860794067,
"learning_rate": 2.5e-05,
"loss": 2.2553,
"step": 250
},
{
"epoch": 0.052,
"grad_norm": 0.4735572338104248,
"learning_rate": 2.6000000000000002e-05,
"loss": 1.2649,
"step": 260
},
{
"epoch": 0.054,
"grad_norm": 0.5137017965316772,
"learning_rate": 2.7000000000000002e-05,
"loss": 0.943,
"step": 270
},
{
"epoch": 0.056,
"grad_norm": 1.4285893440246582,
"learning_rate": 2.8000000000000003e-05,
"loss": 0.8752,
"step": 280
},
{
"epoch": 0.058,
"grad_norm": 0.0,
"learning_rate": 2.9e-05,
"loss": 1.1963,
"step": 290
},
{
"epoch": 0.06,
"grad_norm": 0.4973921477794647,
"learning_rate": 3e-05,
"loss": 1.115,
"step": 300
},
{
"epoch": 0.062,
"grad_norm": 0.6414042115211487,
"learning_rate": 3.1e-05,
"loss": 2.2816,
"step": 310
},
{
"epoch": 0.064,
"grad_norm": 0.0,
"learning_rate": 3.2000000000000005e-05,
"loss": 0.4115,
"step": 320
},
{
"epoch": 0.066,
"grad_norm": 0.5332188010215759,
"learning_rate": 3.3e-05,
"loss": 1.3494,
"step": 330
},
{
"epoch": 0.068,
"grad_norm": 3.6028225421905518,
"learning_rate": 3.4000000000000007e-05,
"loss": 1.6002,
"step": 340
},
{
"epoch": 0.07,
"grad_norm": 0.7303054928779602,
"learning_rate": 3.5e-05,
"loss": 1.2186,
"step": 350
},
{
"epoch": 0.072,
"grad_norm": 0.922789990901947,
"learning_rate": 3.6e-05,
"loss": 0.792,
"step": 360
},
{
"epoch": 0.074,
"grad_norm": 1.9681700468063354,
"learning_rate": 3.7e-05,
"loss": 2.0036,
"step": 370
},
{
"epoch": 0.076,
"grad_norm": 0.6458554863929749,
"learning_rate": 3.8e-05,
"loss": 1.4322,
"step": 380
},
{
"epoch": 0.078,
"grad_norm": 11.577519416809082,
"learning_rate": 3.9000000000000006e-05,
"loss": 3.1319,
"step": 390
},
{
"epoch": 0.08,
"grad_norm": 4.925251483917236,
"learning_rate": 4e-05,
"loss": 1.1777,
"step": 400
},
{
"epoch": 0.082,
"grad_norm": 0.41002562642097473,
"learning_rate": 4.1e-05,
"loss": 0.7938,
"step": 410
},
{
"epoch": 0.084,
"grad_norm": 1.1357100009918213,
"learning_rate": 4.2e-05,
"loss": 1.0942,
"step": 420
},
{
"epoch": 0.086,
"grad_norm": 0.2831510603427887,
"learning_rate": 4.3e-05,
"loss": 1.276,
"step": 430
},
{
"epoch": 0.088,
"grad_norm": 0.47275394201278687,
"learning_rate": 4.4000000000000006e-05,
"loss": 2.1237,
"step": 440
},
{
"epoch": 0.09,
"grad_norm": 0.25966623425483704,
"learning_rate": 4.5e-05,
"loss": 0.9086,
"step": 450
},
{
"epoch": 0.092,
"grad_norm": 3.9267749786376953,
"learning_rate": 4.600000000000001e-05,
"loss": 2.0123,
"step": 460
},
{
"epoch": 0.094,
"grad_norm": 0.0,
"learning_rate": 4.7e-05,
"loss": 0.4836,
"step": 470
},
{
"epoch": 0.096,
"grad_norm": 3.281909942626953,
"learning_rate": 4.8e-05,
"loss": 1.2907,
"step": 480
},
{
"epoch": 0.098,
"grad_norm": 16.264896392822266,
"learning_rate": 4.9e-05,
"loss": 1.8895,
"step": 490
},
{
"epoch": 0.1,
"grad_norm": 1.467863917350769,
"learning_rate": 5e-05,
"loss": 1.4207,
"step": 500
},
{
"epoch": 0.102,
"grad_norm": 4.497130870819092,
"learning_rate": 4.999939076763487e-05,
"loss": 1.5309,
"step": 510
},
{
"epoch": 0.104,
"grad_norm": 0.6231054663658142,
"learning_rate": 4.999756310023261e-05,
"loss": 2.412,
"step": 520
},
{
"epoch": 0.106,
"grad_norm": 2.1034438610076904,
"learning_rate": 4.999451708687114e-05,
"loss": 1.599,
"step": 530
},
{
"epoch": 0.108,
"grad_norm": 1.5209999084472656,
"learning_rate": 4.999025287600886e-05,
"loss": 0.9927,
"step": 540
},
{
"epoch": 0.11,
"grad_norm": 0.0,
"learning_rate": 4.99847706754774e-05,
"loss": 0.9149,
"step": 550
},
{
"epoch": 0.112,
"grad_norm": 2.1328773498535156,
"learning_rate": 4.997807075247146e-05,
"loss": 1.8017,
"step": 560
},
{
"epoch": 0.114,
"grad_norm": 1.0151857137680054,
"learning_rate": 4.997015343353585e-05,
"loss": 0.948,
"step": 570
},
{
"epoch": 0.116,
"grad_norm": 0.2761138081550598,
"learning_rate": 4.996101910454953e-05,
"loss": 0.6686,
"step": 580
},
{
"epoch": 0.118,
"grad_norm": 4.962229251861572,
"learning_rate": 4.995066821070679e-05,
"loss": 1.2988,
"step": 590
},
{
"epoch": 0.12,
"grad_norm": 1.0377299785614014,
"learning_rate": 4.993910125649561e-05,
"loss": 1.6317,
"step": 600
},
{
"epoch": 0.122,
"grad_norm": 3.0478384494781494,
"learning_rate": 4.992631880567301e-05,
"loss": 1.2735,
"step": 610
},
{
"epoch": 0.124,
"grad_norm": 1.3416221141815186,
"learning_rate": 4.991232148123761e-05,
"loss": 0.9292,
"step": 620
},
{
"epoch": 0.126,
"grad_norm": 0.15793165564537048,
"learning_rate": 4.989710996539926e-05,
"loss": 1.7753,
"step": 630
},
{
"epoch": 0.128,
"grad_norm": 8.417360305786133,
"learning_rate": 4.988068499954578e-05,
"loss": 1.3691,
"step": 640
},
{
"epoch": 0.13,
"grad_norm": 2.0238218307495117,
"learning_rate": 4.9863047384206835e-05,
"loss": 1.1858,
"step": 650
},
{
"epoch": 0.132,
"grad_norm": 0.5965118408203125,
"learning_rate": 4.984419797901491e-05,
"loss": 0.4702,
"step": 660
},
{
"epoch": 0.134,
"grad_norm": 0.0,
"learning_rate": 4.982413770266342e-05,
"loss": 1.8529,
"step": 670
},
{
"epoch": 0.136,
"grad_norm": 5.714442729949951,
"learning_rate": 4.980286753286195e-05,
"loss": 1.3363,
"step": 680
},
{
"epoch": 0.138,
"grad_norm": 1.3500193357467651,
"learning_rate": 4.978038850628854e-05,
"loss": 1.3565,
"step": 690
},
{
"epoch": 0.14,
"grad_norm": 12.879951477050781,
"learning_rate": 4.975670171853926e-05,
"loss": 1.7252,
"step": 700
},
{
"epoch": 0.142,
"grad_norm": 4.59855318069458,
"learning_rate": 4.9731808324074717e-05,
"loss": 1.3579,
"step": 710
},
{
"epoch": 0.144,
"grad_norm": 2.8700144290924072,
"learning_rate": 4.9705709536163824e-05,
"loss": 1.9292,
"step": 720
},
{
"epoch": 0.146,
"grad_norm": 3.4623522758483887,
"learning_rate": 4.96784066268247e-05,
"loss": 0.6254,
"step": 730
},
{
"epoch": 0.148,
"grad_norm": 8.593720436096191,
"learning_rate": 4.964990092676263e-05,
"loss": 1.6293,
"step": 740
},
{
"epoch": 0.15,
"grad_norm": 9.260454177856445,
"learning_rate": 4.962019382530521e-05,
"loss": 1.3849,
"step": 750
},
{
"epoch": 0.152,
"grad_norm": 0.13983963429927826,
"learning_rate": 4.9589286770334654e-05,
"loss": 0.8336,
"step": 760
},
{
"epoch": 0.154,
"grad_norm": 0.0,
"learning_rate": 4.9557181268217227e-05,
"loss": 0.9091,
"step": 770
},
{
"epoch": 0.156,
"grad_norm": 0.8305982351303101,
"learning_rate": 4.952387888372979e-05,
"loss": 2.1102,
"step": 780
},
{
"epoch": 0.158,
"grad_norm": 1.0168547630310059,
"learning_rate": 4.94893812399836e-05,
"loss": 1.3355,
"step": 790
},
{
"epoch": 0.16,
"grad_norm": 0.6084606647491455,
"learning_rate": 4.9453690018345144e-05,
"loss": 1.7654,
"step": 800
},
{
"epoch": 0.162,
"grad_norm": 0.0,
"learning_rate": 4.94168069583542e-05,
"loss": 2.7706,
"step": 810
},
{
"epoch": 0.164,
"grad_norm": 1.6526423692703247,
"learning_rate": 4.937873385763908e-05,
"loss": 1.1375,
"step": 820
},
{
"epoch": 0.166,
"grad_norm": 0.6791658401489258,
"learning_rate": 4.933947257182901e-05,
"loss": 2.3548,
"step": 830
},
{
"epoch": 0.168,
"grad_norm": 0.8403846025466919,
"learning_rate": 4.929902501446366e-05,
"loss": 1.6854,
"step": 840
},
{
"epoch": 0.17,
"grad_norm": 9.99562931060791,
"learning_rate": 4.925739315689991e-05,
"loss": 1.7509,
"step": 850
},
{
"epoch": 0.172,
"grad_norm": 5.381862163543701,
"learning_rate": 4.9214579028215776e-05,
"loss": 1.382,
"step": 860
},
{
"epoch": 0.174,
"grad_norm": 1.614503026008606,
"learning_rate": 4.917058471511149e-05,
"loss": 2.0779,
"step": 870
},
{
"epoch": 0.176,
"grad_norm": 1.4542772769927979,
"learning_rate": 4.912541236180779e-05,
"loss": 4.77,
"step": 880
},
{
"epoch": 0.178,
"grad_norm": 12.149937629699707,
"learning_rate": 4.907906416994146e-05,
"loss": 1.5973,
"step": 890
},
{
"epoch": 0.18,
"grad_norm": 0.8259605169296265,
"learning_rate": 4.9031542398457974e-05,
"loss": 1.5671,
"step": 900
},
{
"epoch": 0.182,
"grad_norm": 0.0,
"learning_rate": 4.898284936350144e-05,
"loss": 0.7493,
"step": 910
},
{
"epoch": 0.184,
"grad_norm": 1.1340575218200684,
"learning_rate": 4.893298743830168e-05,
"loss": 0.9516,
"step": 920
},
{
"epoch": 0.186,
"grad_norm": 0.0,
"learning_rate": 4.888195905305859e-05,
"loss": 1.2972,
"step": 930
},
{
"epoch": 0.188,
"grad_norm": 0.9685637950897217,
"learning_rate": 4.882976669482367e-05,
"loss": 0.561,
"step": 940
},
{
"epoch": 0.19,
"grad_norm": 0.0,
"learning_rate": 4.877641290737884e-05,
"loss": 0.4192,
"step": 950
},
{
"epoch": 0.192,
"grad_norm": 0.41207355260849,
"learning_rate": 4.8721900291112415e-05,
"loss": 0.8683,
"step": 960
},
{
"epoch": 0.194,
"grad_norm": 1.2403852939605713,
"learning_rate": 4.8666231502892415e-05,
"loss": 0.8114,
"step": 970
},
{
"epoch": 0.196,
"grad_norm": 3.257436990737915,
"learning_rate": 4.860940925593703e-05,
"loss": 0.8031,
"step": 980
},
{
"epoch": 0.198,
"grad_norm": 0.0,
"learning_rate": 4.855143631968242e-05,
"loss": 1.1436,
"step": 990
},
{
"epoch": 0.2,
"grad_norm": 3.756723642349243,
"learning_rate": 4.849231551964771e-05,
"loss": 1.4053,
"step": 1000
},
{
"epoch": 0.202,
"grad_norm": 2.7545039653778076,
"learning_rate": 4.843204973729729e-05,
"loss": 1.2775,
"step": 1010
},
{
"epoch": 0.204,
"grad_norm": 15.873637199401855,
"learning_rate": 4.837064190990036e-05,
"loss": 1.4629,
"step": 1020
},
{
"epoch": 0.206,
"grad_norm": 1.0889697074890137,
"learning_rate": 4.830809503038781e-05,
"loss": 2.1119,
"step": 1030
},
{
"epoch": 0.208,
"grad_norm": 3.773542881011963,
"learning_rate": 4.8244412147206284e-05,
"loss": 1.5775,
"step": 1040
},
{
"epoch": 0.21,
"grad_norm": 0.398959219455719,
"learning_rate": 4.817959636416969e-05,
"loss": 1.2064,
"step": 1050
},
{
"epoch": 0.212,
"grad_norm": 3.363379955291748,
"learning_rate": 4.8113650840307834e-05,
"loss": 1.1176,
"step": 1060
},
{
"epoch": 0.214,
"grad_norm": 7.889254093170166,
"learning_rate": 4.8046578789712515e-05,
"loss": 1.2473,
"step": 1070
},
{
"epoch": 0.216,
"grad_norm": 0.2269558608531952,
"learning_rate": 4.797838348138086e-05,
"loss": 1.6337,
"step": 1080
},
{
"epoch": 0.218,
"grad_norm": 14.9801607131958,
"learning_rate": 4.790906823905599e-05,
"loss": 2.1006,
"step": 1090
},
{
"epoch": 0.22,
"grad_norm": 5.640057563781738,
"learning_rate": 4.783863644106502e-05,
"loss": 1.538,
"step": 1100
},
{
"epoch": 0.222,
"grad_norm": 1.0203982591629028,
"learning_rate": 4.776709152015443e-05,
"loss": 0.9892,
"step": 1110
},
{
"epoch": 0.224,
"grad_norm": 0.0,
"learning_rate": 4.769443696332272e-05,
"loss": 1.1776,
"step": 1120
},
{
"epoch": 0.226,
"grad_norm": 10.02802562713623,
"learning_rate": 4.762067631165049e-05,
"loss": 1.1521,
"step": 1130
},
{
"epoch": 0.228,
"grad_norm": 7.899293899536133,
"learning_rate": 4.754581316012785e-05,
"loss": 1.8144,
"step": 1140
},
{
"epoch": 0.23,
"grad_norm": 2.8550174236297607,
"learning_rate": 4.7469851157479177e-05,
"loss": 1.8925,
"step": 1150
},
{
"epoch": 0.232,
"grad_norm": 1.191503643989563,
"learning_rate": 4.7392794005985326e-05,
"loss": 1.9765,
"step": 1160
},
{
"epoch": 0.234,
"grad_norm": 0.0,
"learning_rate": 4.731464546130314e-05,
"loss": 1.3849,
"step": 1170
},
{
"epoch": 0.236,
"grad_norm": 7.479146480560303,
"learning_rate": 4.723540933228244e-05,
"loss": 1.0749,
"step": 1180
},
{
"epoch": 0.238,
"grad_norm": 3.3298346996307373,
"learning_rate": 4.715508948078037e-05,
"loss": 1.2765,
"step": 1190
},
{
"epoch": 0.24,
"grad_norm": 2.338376998901367,
"learning_rate": 4.707368982147318e-05,
"loss": 1.3458,
"step": 1200
},
{
"epoch": 0.242,
"grad_norm": 4.633726596832275,
"learning_rate": 4.6991214321665414e-05,
"loss": 2.4307,
"step": 1210
},
{
"epoch": 0.244,
"grad_norm": 3.250617027282715,
"learning_rate": 4.690766700109659e-05,
"loss": 1.2513,
"step": 1220
},
{
"epoch": 0.246,
"grad_norm": 4.539650917053223,
"learning_rate": 4.682305193174524e-05,
"loss": 1.5237,
"step": 1230
},
{
"epoch": 0.248,
"grad_norm": 0.0,
"learning_rate": 4.6737373237630476e-05,
"loss": 0.5587,
"step": 1240
},
{
"epoch": 0.25,
"grad_norm": 0.0,
"learning_rate": 4.665063509461097e-05,
"loss": 1.2764,
"step": 1250
},
{
"epoch": 0.252,
"grad_norm": 2.6641793251037598,
"learning_rate": 4.656284173018144e-05,
"loss": 1.3583,
"step": 1260
},
{
"epoch": 0.254,
"grad_norm": 0.6401162147521973,
"learning_rate": 4.6473997423266614e-05,
"loss": 1.0476,
"step": 1270
},
{
"epoch": 0.256,
"grad_norm": 17.269447326660156,
"learning_rate": 4.638410650401267e-05,
"loss": 1.2224,
"step": 1280
},
{
"epoch": 0.258,
"grad_norm": 5.4362993240356445,
"learning_rate": 4.629317335357619e-05,
"loss": 1.2511,
"step": 1290
},
{
"epoch": 0.26,
"grad_norm": 9.284676551818848,
"learning_rate": 4.620120240391065e-05,
"loss": 0.9118,
"step": 1300
},
{
"epoch": 0.262,
"grad_norm": 5.196897029876709,
"learning_rate": 4.610819813755038e-05,
"loss": 1.6257,
"step": 1310
},
{
"epoch": 0.264,
"grad_norm": 3.060595750808716,
"learning_rate": 4.601416508739211e-05,
"loss": 1.3779,
"step": 1320
},
{
"epoch": 0.266,
"grad_norm": 4.4460673332214355,
"learning_rate": 4.591910783647404e-05,
"loss": 0.7513,
"step": 1330
},
{
"epoch": 0.268,
"grad_norm": 3.9478137493133545,
"learning_rate": 4.5823031017752485e-05,
"loss": 0.9175,
"step": 1340
},
{
"epoch": 0.27,
"grad_norm": 0.0,
"learning_rate": 4.572593931387604e-05,
"loss": 0.5732,
"step": 1350
},
{
"epoch": 0.272,
"grad_norm": 0.0,
"learning_rate": 4.562783745695738e-05,
"loss": 2.0532,
"step": 1360
},
{
"epoch": 0.274,
"grad_norm": 12.004858016967773,
"learning_rate": 4.5528730228342605e-05,
"loss": 1.1836,
"step": 1370
},
{
"epoch": 0.276,
"grad_norm": 1.7193065881729126,
"learning_rate": 4.542862245837821e-05,
"loss": 0.876,
"step": 1380
},
{
"epoch": 0.278,
"grad_norm": 8.586918830871582,
"learning_rate": 4.532751902617569e-05,
"loss": 1.036,
"step": 1390
},
{
"epoch": 0.28,
"grad_norm": 7.1235198974609375,
"learning_rate": 4.522542485937369e-05,
"loss": 1.2111,
"step": 1400
},
{
"epoch": 0.282,
"grad_norm": 4.701413631439209,
"learning_rate": 4.512234493389785e-05,
"loss": 1.6001,
"step": 1410
},
{
"epoch": 0.284,
"grad_norm": 11.681370735168457,
"learning_rate": 4.5018284273718336e-05,
"loss": 1.4541,
"step": 1420
},
{
"epoch": 0.286,
"grad_norm": 1.9920321702957153,
"learning_rate": 4.491324795060491e-05,
"loss": 1.9069,
"step": 1430
},
{
"epoch": 0.288,
"grad_norm": 17.279251098632812,
"learning_rate": 4.480724108387977e-05,
"loss": 1.6968,
"step": 1440
},
{
"epoch": 0.29,
"grad_norm": 23.974655151367188,
"learning_rate": 4.4700268840168045e-05,
"loss": 2.8442,
"step": 1450
},
{
"epoch": 0.292,
"grad_norm": 3.4535789489746094,
"learning_rate": 4.4592336433146e-05,
"loss": 1.5812,
"step": 1460
},
{
"epoch": 0.294,
"grad_norm": 14.773252487182617,
"learning_rate": 4.448344912328686e-05,
"loss": 2.3539,
"step": 1470
},
{
"epoch": 0.296,
"grad_norm": 1.3638978004455566,
"learning_rate": 4.4373612217604496e-05,
"loss": 1.0708,
"step": 1480
},
{
"epoch": 0.298,
"grad_norm": 1.5786101818084717,
"learning_rate": 4.426283106939474e-05,
"loss": 1.464,
"step": 1490
},
{
"epoch": 0.3,
"grad_norm": 1.6620932817459106,
"learning_rate": 4.415111107797445e-05,
"loss": 0.9687,
"step": 1500
},
{
"epoch": 0.302,
"grad_norm": 4.670433521270752,
"learning_rate": 4.403845768841842e-05,
"loss": 1.822,
"step": 1510
},
{
"epoch": 0.304,
"grad_norm": 0.0,
"learning_rate": 4.3924876391293915e-05,
"loss": 1.1844,
"step": 1520
},
{
"epoch": 0.306,
"grad_norm": 13.988329887390137,
"learning_rate": 4.381037272239311e-05,
"loss": 1.8112,
"step": 1530
},
{
"epoch": 0.308,
"grad_norm": 0.3172394037246704,
"learning_rate": 4.36949522624633e-05,
"loss": 0.8056,
"step": 1540
},
{
"epoch": 0.31,
"grad_norm": 0.0,
"learning_rate": 4.357862063693486e-05,
"loss": 2.317,
"step": 1550
},
{
"epoch": 0.312,
"grad_norm": 5.500033855438232,
"learning_rate": 4.3461383515647106e-05,
"loss": 1.1789,
"step": 1560
},
{
"epoch": 0.314,
"grad_norm": 3.911036729812622,
"learning_rate": 4.334324661257191e-05,
"loss": 1.6664,
"step": 1570
},
{
"epoch": 0.316,
"grad_norm": 3.9234087467193604,
"learning_rate": 4.3224215685535294e-05,
"loss": 0.8933,
"step": 1580
},
{
"epoch": 0.318,
"grad_norm": 0.0,
"learning_rate": 4.3104296535936695e-05,
"loss": 2.5625,
"step": 1590
},
{
"epoch": 0.32,
"grad_norm": 2.117269515991211,
"learning_rate": 4.2983495008466276e-05,
"loss": 1.6555,
"step": 1600
},
{
"epoch": 0.322,
"grad_norm": 9.193464279174805,
"learning_rate": 4.2861816990820084e-05,
"loss": 1.1192,
"step": 1610
},
{
"epoch": 0.324,
"grad_norm": 4.3306565284729,
"learning_rate": 4.273926841341302e-05,
"loss": 1.3019,
"step": 1620
},
{
"epoch": 0.326,
"grad_norm": 7.020684242248535,
"learning_rate": 4.261585524908987e-05,
"loss": 1.3233,
"step": 1630
},
{
"epoch": 0.328,
"grad_norm": 2.99652099609375,
"learning_rate": 4.249158351283414e-05,
"loss": 0.8989,
"step": 1640
},
{
"epoch": 0.33,
"grad_norm": 19.782791137695312,
"learning_rate": 4.2366459261474933e-05,
"loss": 1.1397,
"step": 1650
},
{
"epoch": 0.332,
"grad_norm": 2.0220930576324463,
"learning_rate": 4.224048859339175e-05,
"loss": 1.099,
"step": 1660
},
{
"epoch": 0.334,
"grad_norm": 0.0,
"learning_rate": 4.211367764821722e-05,
"loss": 0.7078,
"step": 1670
},
{
"epoch": 0.336,
"grad_norm": 9.233697891235352,
"learning_rate": 4.198603260653792e-05,
"loss": 1.958,
"step": 1680
},
{
"epoch": 0.338,
"grad_norm": 14.23305606842041,
"learning_rate": 4.185755968959308e-05,
"loss": 4.0017,
"step": 1690
},
{
"epoch": 0.34,
"grad_norm": 1.316114068031311,
"learning_rate": 4.172826515897146e-05,
"loss": 0.9028,
"step": 1700
},
{
"epoch": 0.342,
"grad_norm": 2.208270311355591,
"learning_rate": 4.1598155316306044e-05,
"loss": 1.3906,
"step": 1710
},
{
"epoch": 0.344,
"grad_norm": 0.7679595947265625,
"learning_rate": 4.146723650296701e-05,
"loss": 1.1884,
"step": 1720
},
{
"epoch": 0.346,
"grad_norm": 0.0,
"learning_rate": 4.133551509975264e-05,
"loss": 0.7797,
"step": 1730
},
{
"epoch": 0.348,
"grad_norm": 0.0,
"learning_rate": 4.1202997526578276e-05,
"loss": 1.9353,
"step": 1740
},
{
"epoch": 0.35,
"grad_norm": 0.4529210925102234,
"learning_rate": 4.1069690242163484e-05,
"loss": 1.5543,
"step": 1750
},
{
"epoch": 0.352,
"grad_norm": 1.7549803256988525,
"learning_rate": 4.093559974371725e-05,
"loss": 1.1649,
"step": 1760
},
{
"epoch": 0.354,
"grad_norm": 0.4439053237438202,
"learning_rate": 4.080073256662127e-05,
"loss": 1.1743,
"step": 1770
},
{
"epoch": 0.356,
"grad_norm": 1.59803307056427,
"learning_rate": 4.066509528411152e-05,
"loss": 1.5422,
"step": 1780
},
{
"epoch": 0.358,
"grad_norm": 1.4424983263015747,
"learning_rate": 4.052869450695776e-05,
"loss": 0.9694,
"step": 1790
},
{
"epoch": 0.36,
"grad_norm": 3.5582339763641357,
"learning_rate": 4.039153688314145e-05,
"loss": 1.37,
"step": 1800
},
{
"epoch": 0.362,
"grad_norm": 1.2062020301818848,
"learning_rate": 4.02536290975317e-05,
"loss": 1.0036,
"step": 1810
},
{
"epoch": 0.364,
"grad_norm": 0.0,
"learning_rate": 4.011497787155938e-05,
"loss": 1.8302,
"step": 1820
},
{
"epoch": 0.366,
"grad_norm": 0.0,
"learning_rate": 3.997558996288965e-05,
"loss": 1.5304,
"step": 1830
},
{
"epoch": 0.368,
"grad_norm": 0.0,
"learning_rate": 3.983547216509254e-05,
"loss": 0.8427,
"step": 1840
},
{
"epoch": 0.37,
"grad_norm": 0.9621694087982178,
"learning_rate": 3.969463130731183e-05,
"loss": 0.9384,
"step": 1850
},
{
"epoch": 0.372,
"grad_norm": 0.9074615240097046,
"learning_rate": 3.955307425393224e-05,
"loss": 0.956,
"step": 1860
},
{
"epoch": 0.374,
"grad_norm": 94.73406219482422,
"learning_rate": 3.941080790424484e-05,
"loss": 2.5897,
"step": 1870
},
{
"epoch": 0.376,
"grad_norm": 1.4673329591751099,
"learning_rate": 3.92678391921108e-05,
"loss": 1.0552,
"step": 1880
},
{
"epoch": 0.378,
"grad_norm": 2.7422103881835938,
"learning_rate": 3.912417508562345e-05,
"loss": 1.314,
"step": 1890
},
{
"epoch": 0.38,
"grad_norm": 8.64188003540039,
"learning_rate": 3.897982258676867e-05,
"loss": 2.4508,
"step": 1900
},
{
"epoch": 0.382,
"grad_norm": 3.8510308265686035,
"learning_rate": 3.883478873108361e-05,
"loss": 1.5485,
"step": 1910
},
{
"epoch": 0.384,
"grad_norm": 0.0,
"learning_rate": 3.868908058731376e-05,
"loss": 0.6966,
"step": 1920
},
{
"epoch": 0.386,
"grad_norm": 0.0,
"learning_rate": 3.85427052570685e-05,
"loss": 0.8407,
"step": 1930
},
{
"epoch": 0.388,
"grad_norm": 1.5290547609329224,
"learning_rate": 3.8395669874474915e-05,
"loss": 1.1045,
"step": 1940
},
{
"epoch": 0.39,
"grad_norm": 0.0,
"learning_rate": 3.824798160583012e-05,
"loss": 0.7849,
"step": 1950
},
{
"epoch": 0.392,
"grad_norm": 2.1444456577301025,
"learning_rate": 3.8099647649251986e-05,
"loss": 1.8512,
"step": 1960
},
{
"epoch": 0.394,
"grad_norm": 1.4146127700805664,
"learning_rate": 3.795067523432826e-05,
"loss": 0.9294,
"step": 1970
},
{
"epoch": 0.396,
"grad_norm": 8.307282447814941,
"learning_rate": 3.780107162176429e-05,
"loss": 1.0097,
"step": 1980
},
{
"epoch": 0.398,
"grad_norm": 17.269071578979492,
"learning_rate": 3.765084410302909e-05,
"loss": 2.4489,
"step": 1990
},
{
"epoch": 0.4,
"grad_norm": 7.6245198249816895,
"learning_rate": 3.7500000000000003e-05,
"loss": 1.441,
"step": 2000
},
{
"epoch": 0.402,
"grad_norm": 0.9176368713378906,
"learning_rate": 3.7348546664605777e-05,
"loss": 1.1639,
"step": 2010
},
{
"epoch": 0.404,
"grad_norm": 1.319966435432434,
"learning_rate": 3.719649147846832e-05,
"loss": 1.0539,
"step": 2020
},
{
"epoch": 0.406,
"grad_norm": 3.70522403717041,
"learning_rate": 3.704384185254288e-05,
"loss": 0.9333,
"step": 2030
},
{
"epoch": 0.408,
"grad_norm": 45.268592834472656,
"learning_rate": 3.689060522675689e-05,
"loss": 1.8345,
"step": 2040
},
{
"epoch": 0.41,
"grad_norm": 3.782158136367798,
"learning_rate": 3.673678906964727e-05,
"loss": 2.0458,
"step": 2050
},
{
"epoch": 0.412,
"grad_norm": 7.8290791511535645,
"learning_rate": 3.6582400877996546e-05,
"loss": 1.3346,
"step": 2060
},
{
"epoch": 0.414,
"grad_norm": 2.4729154109954834,
"learning_rate": 3.642744817646736e-05,
"loss": 1.5323,
"step": 2070
},
{
"epoch": 0.416,
"grad_norm": 0.838581383228302,
"learning_rate": 3.627193851723577e-05,
"loss": 0.8685,
"step": 2080
},
{
"epoch": 0.418,
"grad_norm": 16.806278228759766,
"learning_rate": 3.611587947962319e-05,
"loss": 1.3805,
"step": 2090
},
{
"epoch": 0.42,
"grad_norm": 0.0,
"learning_rate": 3.5959278669726935e-05,
"loss": 3.2155,
"step": 2100
},
{
"epoch": 0.422,
"grad_norm": 3.4624030590057373,
"learning_rate": 3.580214372004956e-05,
"loss": 1.3376,
"step": 2110
},
{
"epoch": 0.424,
"grad_norm": 3.215987205505371,
"learning_rate": 3.564448228912682e-05,
"loss": 2.3062,
"step": 2120
},
{
"epoch": 0.426,
"grad_norm": 24.706693649291992,
"learning_rate": 3.548630206115443e-05,
"loss": 0.8054,
"step": 2130
},
{
"epoch": 0.428,
"grad_norm": 28.45231056213379,
"learning_rate": 3.532761074561355e-05,
"loss": 1.1863,
"step": 2140
},
{
"epoch": 0.43,
"grad_norm": 3.422032356262207,
"learning_rate": 3.516841607689501e-05,
"loss": 1.2644,
"step": 2150
},
{
"epoch": 0.432,
"grad_norm": 3.279686450958252,
"learning_rate": 3.5008725813922386e-05,
"loss": 1.1396,
"step": 2160
},
{
"epoch": 0.434,
"grad_norm": 6.026493549346924,
"learning_rate": 3.484854773977378e-05,
"loss": 1.6754,
"step": 2170
},
{
"epoch": 0.436,
"grad_norm": 1.911526083946228,
"learning_rate": 3.4687889661302576e-05,
"loss": 0.7079,
"step": 2180
},
{
"epoch": 0.438,
"grad_norm": 14.799755096435547,
"learning_rate": 3.452675940875686e-05,
"loss": 1.4206,
"step": 2190
},
{
"epoch": 0.44,
"grad_norm": 6.455738544464111,
"learning_rate": 3.436516483539781e-05,
"loss": 1.3386,
"step": 2200
},
{
"epoch": 0.442,
"grad_norm": 1.276485800743103,
"learning_rate": 3.4203113817116957e-05,
"loss": 0.8194,
"step": 2210
},
{
"epoch": 0.444,
"grad_norm": 6.4939799308776855,
"learning_rate": 3.4040614252052305e-05,
"loss": 1.8051,
"step": 2220
},
{
"epoch": 0.446,
"grad_norm": 1.6123688220977783,
"learning_rate": 3.387767406020343e-05,
"loss": 1.8429,
"step": 2230
},
{
"epoch": 0.448,
"grad_norm": 3.2084081172943115,
"learning_rate": 3.3714301183045385e-05,
"loss": 1.7987,
"step": 2240
},
{
"epoch": 0.45,
"grad_norm": 2.2829909324645996,
"learning_rate": 3.355050358314172e-05,
"loss": 1.9843,
"step": 2250
},
{
"epoch": 0.452,
"grad_norm": 4.479115009307861,
"learning_rate": 3.338628924375638e-05,
"loss": 0.6176,
"step": 2260
},
{
"epoch": 0.454,
"grad_norm": 1.7939817905426025,
"learning_rate": 3.322166616846458e-05,
"loss": 1.2847,
"step": 2270
},
{
"epoch": 0.456,
"grad_norm": 2.055887460708618,
"learning_rate": 3.305664238076278e-05,
"loss": 1.7299,
"step": 2280
},
{
"epoch": 0.458,
"grad_norm": 0.5422760248184204,
"learning_rate": 3.289122592367757e-05,
"loss": 0.7908,
"step": 2290
},
{
"epoch": 0.46,
"grad_norm": 0.0,
"learning_rate": 3.272542485937369e-05,
"loss": 1.0017,
"step": 2300
},
{
"epoch": 0.462,
"grad_norm": 0.32280176877975464,
"learning_rate": 3.2559247268761115e-05,
"loss": 0.8984,
"step": 2310
},
{
"epoch": 0.464,
"grad_norm": 18.99048614501953,
"learning_rate": 3.239270125110117e-05,
"loss": 1.7084,
"step": 2320
},
{
"epoch": 0.466,
"grad_norm": 5.556630611419678,
"learning_rate": 3.222579492361179e-05,
"loss": 0.6562,
"step": 2330
},
{
"epoch": 0.468,
"grad_norm": 21.25049591064453,
"learning_rate": 3.205853642107192e-05,
"loss": 1.1915,
"step": 2340
},
{
"epoch": 0.47,
"grad_norm": 0.0,
"learning_rate": 3.1890933895424976e-05,
"loss": 1.3534,
"step": 2350
},
{
"epoch": 0.472,
"grad_norm": 1.8797976970672607,
"learning_rate": 3.172299551538164e-05,
"loss": 1.4973,
"step": 2360
},
{
"epoch": 0.474,
"grad_norm": 3.0614185333251953,
"learning_rate": 3.155472946602162e-05,
"loss": 2.4774,
"step": 2370
},
{
"epoch": 0.476,
"grad_norm": 6.993570804595947,
"learning_rate": 3.138614394839476e-05,
"loss": 1.5159,
"step": 2380
},
{
"epoch": 0.478,
"grad_norm": 0.0,
"learning_rate": 3.121724717912138e-05,
"loss": 1.0125,
"step": 2390
},
{
"epoch": 0.48,
"grad_norm": 2.762420892715454,
"learning_rate": 3.104804738999169e-05,
"loss": 0.5848,
"step": 2400
},
{
"epoch": 0.482,
"grad_norm": 24.377647399902344,
"learning_rate": 3.087855282756475e-05,
"loss": 1.3007,
"step": 2410
},
{
"epoch": 0.484,
"grad_norm": 15.386418342590332,
"learning_rate": 3.0708771752766394e-05,
"loss": 1.6124,
"step": 2420
},
{
"epoch": 0.486,
"grad_norm": 4.332566261291504,
"learning_rate": 3.053871244048669e-05,
"loss": 1.0171,
"step": 2430
},
{
"epoch": 0.488,
"grad_norm": 1.804961919784546,
"learning_rate": 3.0368383179176585e-05,
"loss": 0.5442,
"step": 2440
},
{
"epoch": 0.49,
"grad_norm": 1.1589062213897705,
"learning_rate": 3.0197792270443982e-05,
"loss": 0.824,
"step": 2450
},
{
"epoch": 0.492,
"grad_norm": 1.190585732460022,
"learning_rate": 3.002694802864912e-05,
"loss": 1.5214,
"step": 2460
},
{
"epoch": 0.494,
"grad_norm": 7.825082302093506,
"learning_rate": 2.98558587804993e-05,
"loss": 1.9628,
"step": 2470
},
{
"epoch": 0.496,
"grad_norm": 7.856967449188232,
"learning_rate": 2.9684532864643122e-05,
"loss": 1.6987,
"step": 2480
},
{
"epoch": 0.498,
"grad_norm": 11.930267333984375,
"learning_rate": 2.9512978631264006e-05,
"loss": 1.3293,
"step": 2490
},
{
"epoch": 0.5,
"grad_norm": 1.154645562171936,
"learning_rate": 2.9341204441673266e-05,
"loss": 1.4358,
"step": 2500
},
{
"epoch": 0.502,
"grad_norm": 0.0,
"learning_rate": 2.916921866790256e-05,
"loss": 0.9238,
"step": 2510
},
{
"epoch": 0.504,
"grad_norm": 2.808257579803467,
"learning_rate": 2.8997029692295874e-05,
"loss": 0.6767,
"step": 2520
},
{
"epoch": 0.506,
"grad_norm": 7.462977886199951,
"learning_rate": 2.8824645907100954e-05,
"loss": 1.5141,
"step": 2530
},
{
"epoch": 0.508,
"grad_norm": 4.376765727996826,
"learning_rate": 2.8652075714060295e-05,
"loss": 0.7271,
"step": 2540
},
{
"epoch": 0.51,
"grad_norm": 0.0,
"learning_rate": 2.8479327524001636e-05,
"loss": 1.1601,
"step": 2550
},
{
"epoch": 0.512,
"grad_norm": 2.4635355472564697,
"learning_rate": 2.8306409756428064e-05,
"loss": 1.1031,
"step": 2560
},
{
"epoch": 0.514,
"grad_norm": 16.775001525878906,
"learning_rate": 2.8133330839107608e-05,
"loss": 1.9696,
"step": 2570
},
{
"epoch": 0.516,
"grad_norm": 0.4682444632053375,
"learning_rate": 2.7960099207662532e-05,
"loss": 1.1183,
"step": 2580
},
{
"epoch": 0.518,
"grad_norm": 4.674540042877197,
"learning_rate": 2.7786723305158136e-05,
"loss": 0.8266,
"step": 2590
},
{
"epoch": 0.52,
"grad_norm": 3.9463658332824707,
"learning_rate": 2.761321158169134e-05,
"loss": 1.5703,
"step": 2600
},
{
"epoch": 0.522,
"grad_norm": 9.975082397460938,
"learning_rate": 2.7439572493978736e-05,
"loss": 1.6023,
"step": 2610
},
{
"epoch": 0.524,
"grad_norm": 5.6594462394714355,
"learning_rate": 2.726581450494451e-05,
"loss": 0.9304,
"step": 2620
},
{
"epoch": 0.526,
"grad_norm": 9.62289047241211,
"learning_rate": 2.7091946083307896e-05,
"loss": 1.5715,
"step": 2630
},
{
"epoch": 0.528,
"grad_norm": 15.356546401977539,
"learning_rate": 2.6917975703170466e-05,
"loss": 1.8024,
"step": 2640
},
{
"epoch": 0.53,
"grad_norm": 1.7599657773971558,
"learning_rate": 2.674391184360313e-05,
"loss": 0.8039,
"step": 2650
},
{
"epoch": 0.532,
"grad_norm": 1.521369218826294,
"learning_rate": 2.656976298823284e-05,
"loss": 0.6528,
"step": 2660
},
{
"epoch": 0.534,
"grad_norm": 0.0,
"learning_rate": 2.6395537624829096e-05,
"loss": 1.7035,
"step": 2670
},
{
"epoch": 0.536,
"grad_norm": 2.2401411533355713,
"learning_rate": 2.6221244244890336e-05,
"loss": 1.2272,
"step": 2680
},
{
"epoch": 0.538,
"grad_norm": 1.1615192890167236,
"learning_rate": 2.604689134322999e-05,
"loss": 0.5353,
"step": 2690
},
{
"epoch": 0.54,
"grad_norm": 15.130184173583984,
"learning_rate": 2.587248741756253e-05,
"loss": 1.1314,
"step": 2700
},
{
"epoch": 0.542,
"grad_norm": 0.6911365389823914,
"learning_rate": 2.5698040968089225e-05,
"loss": 1.8837,
"step": 2710
},
{
"epoch": 0.544,
"grad_norm": 29.572574615478516,
"learning_rate": 2.5523560497083926e-05,
"loss": 1.7309,
"step": 2720
},
{
"epoch": 0.546,
"grad_norm": 2.712412118911743,
"learning_rate": 2.5349054508478637e-05,
"loss": 1.5027,
"step": 2730
},
{
"epoch": 0.548,
"grad_norm": 2.607976198196411,
"learning_rate": 2.517453150744904e-05,
"loss": 0.7212,
"step": 2740
},
{
"epoch": 0.55,
"grad_norm": 0.0,
"learning_rate": 2.5e-05,
"loss": 0.9073,
"step": 2750
},
{
"epoch": 0.552,
"grad_norm": 4.585177898406982,
"learning_rate": 2.4825468492550964e-05,
"loss": 1.3612,
"step": 2760
},
{
"epoch": 0.554,
"grad_norm": 38.40209197998047,
"learning_rate": 2.4650945491521372e-05,
"loss": 2.5574,
"step": 2770
},
{
"epoch": 0.556,
"grad_norm": 0.7733972668647766,
"learning_rate": 2.447643950291608e-05,
"loss": 1.5628,
"step": 2780
},
{
"epoch": 0.558,
"grad_norm": 10.307479858398438,
"learning_rate": 2.4301959031910784e-05,
"loss": 1.4349,
"step": 2790
},
{
"epoch": 0.56,
"grad_norm": 0.0,
"learning_rate": 2.4127512582437485e-05,
"loss": 1.2743,
"step": 2800
},
{
"epoch": 0.562,
"grad_norm": 17.315427780151367,
"learning_rate": 2.3953108656770016e-05,
"loss": 1.4333,
"step": 2810
},
{
"epoch": 0.564,
"grad_norm": 0.3390060365200043,
"learning_rate": 2.377875575510967e-05,
"loss": 0.7756,
"step": 2820
},
{
"epoch": 0.566,
"grad_norm": 0.7559804916381836,
"learning_rate": 2.3604462375170906e-05,
"loss": 1.8001,
"step": 2830
},
{
"epoch": 0.568,
"grad_norm": 6.139041423797607,
"learning_rate": 2.3430237011767167e-05,
"loss": 1.5661,
"step": 2840
},
{
"epoch": 0.57,
"grad_norm": 0.0,
"learning_rate": 2.3256088156396868e-05,
"loss": 0.801,
"step": 2850
},
{
"epoch": 0.572,
"grad_norm": 0.6441702246665955,
"learning_rate": 2.3082024296829536e-05,
"loss": 1.4609,
"step": 2860
},
{
"epoch": 0.574,
"grad_norm": 11.885534286499023,
"learning_rate": 2.2908053916692117e-05,
"loss": 1.174,
"step": 2870
},
{
"epoch": 0.576,
"grad_norm": 0.4849330186843872,
"learning_rate": 2.2734185495055503e-05,
"loss": 1.5587,
"step": 2880
},
{
"epoch": 0.578,
"grad_norm": 2.4908790588378906,
"learning_rate": 2.2560427506021266e-05,
"loss": 1.1499,
"step": 2890
},
{
"epoch": 0.58,
"grad_norm": 11.29605484008789,
"learning_rate": 2.238678841830867e-05,
"loss": 1.0868,
"step": 2900
},
{
"epoch": 0.582,
"grad_norm": 8.866591453552246,
"learning_rate": 2.2213276694841866e-05,
"loss": 1.2244,
"step": 2910
},
{
"epoch": 0.584,
"grad_norm": 0.6604860424995422,
"learning_rate": 2.2039900792337474e-05,
"loss": 1.1256,
"step": 2920
},
{
"epoch": 0.586,
"grad_norm": 6.171046733856201,
"learning_rate": 2.186666916089239e-05,
"loss": 0.4934,
"step": 2930
},
{
"epoch": 0.588,
"grad_norm": 12.340161323547363,
"learning_rate": 2.1693590243571938e-05,
"loss": 1.223,
"step": 2940
},
{
"epoch": 0.59,
"grad_norm": 1.2349804639816284,
"learning_rate": 2.1520672475998373e-05,
"loss": 0.6751,
"step": 2950
},
{
"epoch": 0.592,
"grad_norm": 22.765419006347656,
"learning_rate": 2.1347924285939714e-05,
"loss": 2.4113,
"step": 2960
},
{
"epoch": 0.594,
"grad_norm": 7.209038257598877,
"learning_rate": 2.117535409289905e-05,
"loss": 1.3446,
"step": 2970
},
{
"epoch": 0.596,
"grad_norm": 1.229001522064209,
"learning_rate": 2.1002970307704132e-05,
"loss": 1.0472,
"step": 2980
},
{
"epoch": 0.598,
"grad_norm": 2.125377893447876,
"learning_rate": 2.0830781332097446e-05,
"loss": 0.9695,
"step": 2990
},
{
"epoch": 0.6,
"grad_norm": 3.410578489303589,
"learning_rate": 2.0658795558326743e-05,
"loss": 1.1868,
"step": 3000
},
{
"epoch": 0.602,
"grad_norm": 1.8903169631958008,
"learning_rate": 2.0487021368736003e-05,
"loss": 1.3204,
"step": 3010
},
{
"epoch": 0.604,
"grad_norm": 7.726503849029541,
"learning_rate": 2.031546713535688e-05,
"loss": 0.8136,
"step": 3020
},
{
"epoch": 0.606,
"grad_norm": 2.2459115982055664,
"learning_rate": 2.0144141219500705e-05,
"loss": 1.2284,
"step": 3030
},
{
"epoch": 0.608,
"grad_norm": 18.302974700927734,
"learning_rate": 1.9973051971350888e-05,
"loss": 2.5085,
"step": 3040
},
{
"epoch": 0.61,
"grad_norm": 0.5879172682762146,
"learning_rate": 1.980220772955602e-05,
"loss": 0.8985,
"step": 3050
},
{
"epoch": 0.612,
"grad_norm": 1.7284860610961914,
"learning_rate": 1.963161682082342e-05,
"loss": 1.7433,
"step": 3060
},
{
"epoch": 0.614,
"grad_norm": 12.673727035522461,
"learning_rate": 1.946128755951332e-05,
"loss": 1.393,
"step": 3070
},
{
"epoch": 0.616,
"grad_norm": 1.9405444860458374,
"learning_rate": 1.9291228247233605e-05,
"loss": 1.4175,
"step": 3080
},
{
"epoch": 0.618,
"grad_norm": 10.212315559387207,
"learning_rate": 1.912144717243525e-05,
"loss": 1.9178,
"step": 3090
},
{
"epoch": 0.62,
"grad_norm": 1.4168413877487183,
"learning_rate": 1.895195261000831e-05,
"loss": 0.9371,
"step": 3100
},
{
"epoch": 0.622,
"grad_norm": 0.0,
"learning_rate": 1.8782752820878634e-05,
"loss": 0.7683,
"step": 3110
},
{
"epoch": 0.624,
"grad_norm": 8.363224983215332,
"learning_rate": 1.8613856051605243e-05,
"loss": 1.173,
"step": 3120
},
{
"epoch": 0.626,
"grad_norm": 4.870674133300781,
"learning_rate": 1.8445270533978388e-05,
"loss": 1.2886,
"step": 3130
},
{
"epoch": 0.628,
"grad_norm": 1.6593077182769775,
"learning_rate": 1.827700448461836e-05,
"loss": 0.7759,
"step": 3140
},
{
"epoch": 0.63,
"grad_norm": 1.97660231590271,
"learning_rate": 1.8109066104575023e-05,
"loss": 1.0199,
"step": 3150
},
{
"epoch": 0.632,
"grad_norm": 10.020951271057129,
"learning_rate": 1.7941463578928086e-05,
"loss": 1.3115,
"step": 3160
},
{
"epoch": 0.634,
"grad_norm": 2.821995973587036,
"learning_rate": 1.7774205076388206e-05,
"loss": 1.3926,
"step": 3170
},
{
"epoch": 0.636,
"grad_norm": 24.37348747253418,
"learning_rate": 1.7607298748898842e-05,
"loss": 2.0498,
"step": 3180
},
{
"epoch": 0.638,
"grad_norm": 5.457019329071045,
"learning_rate": 1.744075273123889e-05,
"loss": 0.5073,
"step": 3190
},
{
"epoch": 0.64,
"grad_norm": 0.0,
"learning_rate": 1.7274575140626318e-05,
"loss": 1.159,
"step": 3200
},
{
"epoch": 0.642,
"grad_norm": 0.691412091255188,
"learning_rate": 1.7108774076322443e-05,
"loss": 0.5308,
"step": 3210
},
{
"epoch": 0.644,
"grad_norm": 1.4707207679748535,
"learning_rate": 1.6943357619237226e-05,
"loss": 0.8804,
"step": 3220
},
{
"epoch": 0.646,
"grad_norm": 1.5302475690841675,
"learning_rate": 1.677833383153542e-05,
"loss": 0.562,
"step": 3230
},
{
"epoch": 0.648,
"grad_norm": 1.94536292552948,
"learning_rate": 1.6613710756243626e-05,
"loss": 1.184,
"step": 3240
},
{
"epoch": 0.65,
"grad_norm": 3.658015251159668,
"learning_rate": 1.6449496416858284e-05,
"loss": 1.5939,
"step": 3250
},
{
"epoch": 0.652,
"grad_norm": 13.352316856384277,
"learning_rate": 1.6285698816954624e-05,
"loss": 0.6112,
"step": 3260
},
{
"epoch": 0.654,
"grad_norm": 9.652981758117676,
"learning_rate": 1.612232593979658e-05,
"loss": 1.3099,
"step": 3270
},
{
"epoch": 0.656,
"grad_norm": 0.0,
"learning_rate": 1.5959385747947698e-05,
"loss": 2.5607,
"step": 3280
},
{
"epoch": 0.658,
"grad_norm": 0.0,
"learning_rate": 1.5796886182883053e-05,
"loss": 1.0217,
"step": 3290
},
{
"epoch": 0.66,
"grad_norm": 3.3009743690490723,
"learning_rate": 1.56348351646022e-05,
"loss": 0.9401,
"step": 3300
},
{
"epoch": 0.662,
"grad_norm": 0.0,
"learning_rate": 1.547324059124315e-05,
"loss": 1.0247,
"step": 3310
},
{
"epoch": 0.664,
"grad_norm": 0.5582336783409119,
"learning_rate": 1.5312110338697426e-05,
"loss": 1.9519,
"step": 3320
},
{
"epoch": 0.666,
"grad_norm": 5.420542240142822,
"learning_rate": 1.5151452260226224e-05,
"loss": 0.7734,
"step": 3330
},
{
"epoch": 0.668,
"grad_norm": 3.886523962020874,
"learning_rate": 1.4991274186077632e-05,
"loss": 1.0947,
"step": 3340
},
{
"epoch": 0.67,
"grad_norm": 1.0134683847427368,
"learning_rate": 1.4831583923104999e-05,
"loss": 0.5691,
"step": 3350
},
{
"epoch": 0.672,
"grad_norm": 2.0535247325897217,
"learning_rate": 1.467238925438646e-05,
"loss": 1.966,
"step": 3360
},
{
"epoch": 0.674,
"grad_norm": 1.5086647272109985,
"learning_rate": 1.4513697938845572e-05,
"loss": 0.8153,
"step": 3370
},
{
"epoch": 0.676,
"grad_norm": 0.9438655972480774,
"learning_rate": 1.4355517710873184e-05,
"loss": 0.4211,
"step": 3380
},
{
"epoch": 0.678,
"grad_norm": 9.679282188415527,
"learning_rate": 1.4197856279950438e-05,
"loss": 1.4362,
"step": 3390
},
{
"epoch": 0.68,
"grad_norm": 3.571068048477173,
"learning_rate": 1.4040721330273062e-05,
"loss": 0.7938,
"step": 3400
},
{
"epoch": 0.682,
"grad_norm": 0.3649686872959137,
"learning_rate": 1.388412052037682e-05,
"loss": 0.7605,
"step": 3410
},
{
"epoch": 0.684,
"grad_norm": 6.745936870574951,
"learning_rate": 1.3728061482764238e-05,
"loss": 2.1024,
"step": 3420
},
{
"epoch": 0.686,
"grad_norm": 0.0,
"learning_rate": 1.3572551823532654e-05,
"loss": 1.7508,
"step": 3430
},
{
"epoch": 0.688,
"grad_norm": 10.104940414428711,
"learning_rate": 1.3417599122003464e-05,
"loss": 1.4949,
"step": 3440
},
{
"epoch": 0.69,
"grad_norm": 2.5721874237060547,
"learning_rate": 1.3263210930352737e-05,
"loss": 1.3493,
"step": 3450
},
{
"epoch": 0.692,
"grad_norm": 1.289473533630371,
"learning_rate": 1.3109394773243117e-05,
"loss": 1.7492,
"step": 3460
},
{
"epoch": 0.694,
"grad_norm": 0.9646878242492676,
"learning_rate": 1.2956158147457115e-05,
"loss": 1.2414,
"step": 3470
},
{
"epoch": 0.696,
"grad_norm": 1.2052338123321533,
"learning_rate": 1.280350852153168e-05,
"loss": 0.978,
"step": 3480
},
{
"epoch": 0.698,
"grad_norm": 1.5356535911560059,
"learning_rate": 1.2651453335394231e-05,
"loss": 0.6575,
"step": 3490
},
{
"epoch": 0.7,
"grad_norm": 0.6956253051757812,
"learning_rate": 1.2500000000000006e-05,
"loss": 1.6792,
"step": 3500
},
{
"epoch": 0.702,
"grad_norm": 2.102616548538208,
"learning_rate": 1.234915589697091e-05,
"loss": 0.6078,
"step": 3510
},
{
"epoch": 0.704,
"grad_norm": 0.0,
"learning_rate": 1.2198928378235716e-05,
"loss": 1.0625,
"step": 3520
},
{
"epoch": 0.706,
"grad_norm": 7.8556294441223145,
"learning_rate": 1.2049324765671749e-05,
"loss": 1.7064,
"step": 3530
},
{
"epoch": 0.708,
"grad_norm": 2.802849292755127,
"learning_rate": 1.1900352350748026e-05,
"loss": 0.8152,
"step": 3540
},
{
"epoch": 0.71,
"grad_norm": 0.0,
"learning_rate": 1.175201839416988e-05,
"loss": 1.1932,
"step": 3550
},
{
"epoch": 0.712,
"grad_norm": 18.025968551635742,
"learning_rate": 1.1604330125525079e-05,
"loss": 1.2326,
"step": 3560
},
{
"epoch": 0.714,
"grad_norm": 1.438110113143921,
"learning_rate": 1.1457294742931507e-05,
"loss": 1.2993,
"step": 3570
},
{
"epoch": 0.716,
"grad_norm": 0.0,
"learning_rate": 1.1310919412686247e-05,
"loss": 1.4532,
"step": 3580
},
{
"epoch": 0.718,
"grad_norm": 0.0,
"learning_rate": 1.11652112689164e-05,
"loss": 1.545,
"step": 3590
},
{
"epoch": 0.72,
"grad_norm": 0.0,
"learning_rate": 1.1020177413231334e-05,
"loss": 1.3031,
"step": 3600
},
{
"epoch": 0.722,
"grad_norm": 0.0,
"learning_rate": 1.0875824914376553e-05,
"loss": 1.0103,
"step": 3610
},
{
"epoch": 0.724,
"grad_norm": 7.254970073699951,
"learning_rate": 1.0732160807889211e-05,
"loss": 2.8567,
"step": 3620
},
{
"epoch": 0.726,
"grad_norm": 1.416352391242981,
"learning_rate": 1.058919209575517e-05,
"loss": 1.3896,
"step": 3630
},
{
"epoch": 0.728,
"grad_norm": 3.275563955307007,
"learning_rate": 1.0446925746067768e-05,
"loss": 1.186,
"step": 3640
},
{
"epoch": 0.73,
"grad_norm": 0.0,
"learning_rate": 1.0305368692688174e-05,
"loss": 1.9734,
"step": 3650
},
{
"epoch": 0.732,
"grad_norm": 5.804834365844727,
"learning_rate": 1.0164527834907467e-05,
"loss": 1.3322,
"step": 3660
},
{
"epoch": 0.734,
"grad_norm": 0.6750529408454895,
"learning_rate": 1.0024410037110357e-05,
"loss": 1.4241,
"step": 3670
},
{
"epoch": 0.736,
"grad_norm": 0.5938382148742676,
"learning_rate": 9.88502212844063e-06,
"loss": 1.1161,
"step": 3680
},
{
"epoch": 0.738,
"grad_norm": 8.857674598693848,
"learning_rate": 9.746370902468311e-06,
"loss": 1.4002,
"step": 3690
},
{
"epoch": 0.74,
"grad_norm": 0.0,
"learning_rate": 9.608463116858542e-06,
"loss": 0.6686,
"step": 3700
},
{
"epoch": 0.742,
"grad_norm": 6.373303413391113,
"learning_rate": 9.471305493042243e-06,
"loss": 1.1473,
"step": 3710
},
{
"epoch": 0.744,
"grad_norm": 1.3617981672286987,
"learning_rate": 9.334904715888495e-06,
"loss": 1.3623,
"step": 3720
},
{
"epoch": 0.746,
"grad_norm": 9.023974418640137,
"learning_rate": 9.199267433378727e-06,
"loss": 2.5043,
"step": 3730
},
{
"epoch": 0.748,
"grad_norm": 2.228018045425415,
"learning_rate": 9.064400256282757e-06,
"loss": 0.807,
"step": 3740
},
{
"epoch": 0.75,
"grad_norm": 4.2638630867004395,
"learning_rate": 8.930309757836517e-06,
"loss": 1.078,
"step": 3750
},
{
"epoch": 0.752,
"grad_norm": 3.281846284866333,
"learning_rate": 8.797002473421728e-06,
"loss": 0.8329,
"step": 3760
},
{
"epoch": 0.754,
"grad_norm": 5.123702049255371,
"learning_rate": 8.664484900247363e-06,
"loss": 1.9266,
"step": 3770
},
{
"epoch": 0.756,
"grad_norm": 10.131644248962402,
"learning_rate": 8.532763497032987e-06,
"loss": 1.7022,
"step": 3780
},
{
"epoch": 0.758,
"grad_norm": 1.5231475830078125,
"learning_rate": 8.40184468369396e-06,
"loss": 0.8765,
"step": 3790
},
{
"epoch": 0.76,
"grad_norm": 9.655771255493164,
"learning_rate": 8.271734841028553e-06,
"loss": 1.8413,
"step": 3800
},
{
"epoch": 0.762,
"grad_norm": 5.206886291503906,
"learning_rate": 8.142440310406924e-06,
"loss": 1.2464,
"step": 3810
},
{
"epoch": 0.764,
"grad_norm": 0.4430970549583435,
"learning_rate": 8.013967393462094e-06,
"loss": 1.1392,
"step": 3820
},
{
"epoch": 0.766,
"grad_norm": 0.3004146218299866,
"learning_rate": 7.886322351782783e-06,
"loss": 1.8922,
"step": 3830
},
{
"epoch": 0.768,
"grad_norm": 0.0,
"learning_rate": 7.759511406608255e-06,
"loss": 1.3631,
"step": 3840
},
{
"epoch": 0.77,
"grad_norm": 2.2798960208892822,
"learning_rate": 7.633540738525066e-06,
"loss": 2.1561,
"step": 3850
},
{
"epoch": 0.772,
"grad_norm": 6.353856563568115,
"learning_rate": 7.508416487165862e-06,
"loss": 1.3865,
"step": 3860
},
{
"epoch": 0.774,
"grad_norm": 23.976675033569336,
"learning_rate": 7.384144750910133e-06,
"loss": 1.3,
"step": 3870
},
{
"epoch": 0.776,
"grad_norm": 5.609853744506836,
"learning_rate": 7.260731586586983e-06,
"loss": 1.4537,
"step": 3880
},
{
"epoch": 0.778,
"grad_norm": 0.0,
"learning_rate": 7.138183009179922e-06,
"loss": 1.0312,
"step": 3890
},
{
"epoch": 0.78,
"grad_norm": 5.4580230712890625,
"learning_rate": 7.016504991533726e-06,
"loss": 2.1389,
"step": 3900
},
{
"epoch": 0.782,
"grad_norm": 2.1210005283355713,
"learning_rate": 6.895703464063319e-06,
"loss": 0.5286,
"step": 3910
},
{
"epoch": 0.784,
"grad_norm": 4.246682643890381,
"learning_rate": 6.775784314464717e-06,
"loss": 1.1388,
"step": 3920
},
{
"epoch": 0.786,
"grad_norm": 0.31312233209609985,
"learning_rate": 6.656753387428089e-06,
"loss": 1.7807,
"step": 3930
},
{
"epoch": 0.788,
"grad_norm": 5.389893054962158,
"learning_rate": 6.538616484352902e-06,
"loss": 0.8653,
"step": 3940
},
{
"epoch": 0.79,
"grad_norm": 0.0,
"learning_rate": 6.421379363065142e-06,
"loss": 2.2613,
"step": 3950
},
{
"epoch": 0.792,
"grad_norm": 2.296661376953125,
"learning_rate": 6.305047737536707e-06,
"loss": 0.7061,
"step": 3960
},
{
"epoch": 0.794,
"grad_norm": 4.505575656890869,
"learning_rate": 6.189627277606894e-06,
"loss": 1.1075,
"step": 3970
},
{
"epoch": 0.796,
"grad_norm": 1.5126315355300903,
"learning_rate": 6.075123608706093e-06,
"loss": 0.9158,
"step": 3980
},
{
"epoch": 0.798,
"grad_norm": 9.576681137084961,
"learning_rate": 5.961542311581586e-06,
"loss": 1.5019,
"step": 3990
},
{
"epoch": 0.8,
"grad_norm": 5.753732204437256,
"learning_rate": 5.848888922025553e-06,
"loss": 0.8174,
"step": 4000
},
{
"epoch": 0.802,
"grad_norm": 1.5566200017929077,
"learning_rate": 5.737168930605272e-06,
"loss": 0.6246,
"step": 4010
},
{
"epoch": 0.804,
"grad_norm": 0.31750229001045227,
"learning_rate": 5.626387782395512e-06,
"loss": 1.1496,
"step": 4020
},
{
"epoch": 0.806,
"grad_norm": 21.146928787231445,
"learning_rate": 5.5165508767131415e-06,
"loss": 1.4641,
"step": 4030
},
{
"epoch": 0.808,
"grad_norm": 20.838895797729492,
"learning_rate": 5.4076635668540075e-06,
"loss": 2.6414,
"step": 4040
},
{
"epoch": 0.81,
"grad_norm": 0.0,
"learning_rate": 5.299731159831953e-06,
"loss": 1.5888,
"step": 4050
},
{
"epoch": 0.812,
"grad_norm": 3.5689587593078613,
"learning_rate": 5.192758916120236e-06,
"loss": 0.755,
"step": 4060
},
{
"epoch": 0.814,
"grad_norm": 0.0,
"learning_rate": 5.086752049395094e-06,
"loss": 1.1083,
"step": 4070
},
{
"epoch": 0.816,
"grad_norm": 2.862039566040039,
"learning_rate": 4.981715726281666e-06,
"loss": 0.5798,
"step": 4080
},
{
"epoch": 0.818,
"grad_norm": 21.366016387939453,
"learning_rate": 4.877655066102149e-06,
"loss": 1.061,
"step": 4090
},
{
"epoch": 0.82,
"grad_norm": 1.1818464994430542,
"learning_rate": 4.7745751406263165e-06,
"loss": 0.4177,
"step": 4100
},
{
"epoch": 0.822,
"grad_norm": 17.175460815429688,
"learning_rate": 4.672480973824311e-06,
"loss": 1.3882,
"step": 4110
},
{
"epoch": 0.824,
"grad_norm": 7.882750511169434,
"learning_rate": 4.571377541621788e-06,
"loss": 1.16,
"step": 4120
},
{
"epoch": 0.826,
"grad_norm": 0.0,
"learning_rate": 4.4712697716574e-06,
"loss": 1.0392,
"step": 4130
},
{
"epoch": 0.828,
"grad_norm": 3.8057103157043457,
"learning_rate": 4.372162543042624e-06,
"loss": 1.1697,
"step": 4140
},
{
"epoch": 0.83,
"grad_norm": 3.2752530574798584,
"learning_rate": 4.274060686123959e-06,
"loss": 1.3669,
"step": 4150
},
{
"epoch": 0.832,
"grad_norm": 2.768428325653076,
"learning_rate": 4.176968982247514e-06,
"loss": 0.6651,
"step": 4160
},
{
"epoch": 0.834,
"grad_norm": 37.53615188598633,
"learning_rate": 4.08089216352596e-06,
"loss": 1.4382,
"step": 4170
},
{
"epoch": 0.836,
"grad_norm": 20.771350860595703,
"learning_rate": 3.985834912607894e-06,
"loss": 1.1561,
"step": 4180
},
{
"epoch": 0.838,
"grad_norm": 9.721131324768066,
"learning_rate": 3.891801862449629e-06,
"loss": 0.8151,
"step": 4190
},
{
"epoch": 0.84,
"grad_norm": 70.80477142333984,
"learning_rate": 3.798797596089351e-06,
"loss": 2.6738,
"step": 4200
},
{
"epoch": 0.842,
"grad_norm": 5.896388053894043,
"learning_rate": 3.7068266464238084e-06,
"loss": 1.4829,
"step": 4210
},
{
"epoch": 0.844,
"grad_norm": 0.883860170841217,
"learning_rate": 3.6158934959873353e-06,
"loss": 0.7147,
"step": 4220
},
{
"epoch": 0.846,
"grad_norm": 0.0,
"learning_rate": 3.5260025767333893e-06,
"loss": 1.6543,
"step": 4230
},
{
"epoch": 0.848,
"grad_norm": 3.300741672515869,
"learning_rate": 3.4371582698185633e-06,
"loss": 0.3801,
"step": 4240
},
{
"epoch": 0.85,
"grad_norm": 0.0,
"learning_rate": 3.3493649053890326e-06,
"loss": 0.6843,
"step": 4250
},
{
"epoch": 0.852,
"grad_norm": 5.147795677185059,
"learning_rate": 3.262626762369525e-06,
"loss": 2.1468,
"step": 4260
},
{
"epoch": 0.854,
"grad_norm": 1.0076605081558228,
"learning_rate": 3.176948068254762e-06,
"loss": 0.6189,
"step": 4270
},
{
"epoch": 0.856,
"grad_norm": 10.264873504638672,
"learning_rate": 3.092332998903416e-06,
"loss": 1.1855,
"step": 4280
},
{
"epoch": 0.858,
"grad_norm": 0.3832724392414093,
"learning_rate": 3.0087856783345914e-06,
"loss": 0.5795,
"step": 4290
},
{
"epoch": 0.86,
"grad_norm": 2.3567116260528564,
"learning_rate": 2.9263101785268254e-06,
"loss": 1.6466,
"step": 4300
},
{
"epoch": 0.862,
"grad_norm": 22.869237899780273,
"learning_rate": 2.8449105192196316e-06,
"loss": 1.3837,
"step": 4310
},
{
"epoch": 0.864,
"grad_norm": 0.0,
"learning_rate": 2.764590667717562e-06,
"loss": 0.9491,
"step": 4320
},
{
"epoch": 0.866,
"grad_norm": 5.939896106719971,
"learning_rate": 2.6853545386968606e-06,
"loss": 1.1473,
"step": 4330
},
{
"epoch": 0.868,
"grad_norm": 0.0,
"learning_rate": 2.6072059940146775e-06,
"loss": 1.9689,
"step": 4340
},
{
"epoch": 0.87,
"grad_norm": 0.42899826169013977,
"learning_rate": 2.5301488425208296e-06,
"loss": 0.8104,
"step": 4350
},
{
"epoch": 0.872,
"grad_norm": 3.8288755416870117,
"learning_rate": 2.454186839872158e-06,
"loss": 1.2391,
"step": 4360
},
{
"epoch": 0.874,
"grad_norm": 14.07198715209961,
"learning_rate": 2.379323688349516e-06,
"loss": 2.2037,
"step": 4370
},
{
"epoch": 0.876,
"grad_norm": 10.503342628479004,
"learning_rate": 2.3055630366772856e-06,
"loss": 1.78,
"step": 4380
},
{
"epoch": 0.878,
"grad_norm": 0.36506208777427673,
"learning_rate": 2.2329084798455746e-06,
"loss": 1.4213,
"step": 4390
},
{
"epoch": 0.88,
"grad_norm": 4.561337947845459,
"learning_rate": 2.1613635589349756e-06,
"loss": 1.4604,
"step": 4400
},
{
"epoch": 0.882,
"grad_norm": 3.658083915710449,
"learning_rate": 2.0909317609440095e-06,
"loss": 1.631,
"step": 4410
},
{
"epoch": 0.884,
"grad_norm": 4.377072811126709,
"learning_rate": 2.0216165186191407e-06,
"loss": 1.7125,
"step": 4420
},
{
"epoch": 0.886,
"grad_norm": 9.698302268981934,
"learning_rate": 1.95342121028749e-06,
"loss": 2.2037,
"step": 4430
},
{
"epoch": 0.888,
"grad_norm": 4.783137798309326,
"learning_rate": 1.8863491596921745e-06,
"loss": 1.4982,
"step": 4440
},
{
"epoch": 0.89,
"grad_norm": 5.755715370178223,
"learning_rate": 1.8204036358303173e-06,
"loss": 1.1612,
"step": 4450
},
{
"epoch": 0.892,
"grad_norm": 12.483977317810059,
"learning_rate": 1.7555878527937164e-06,
"loss": 1.2837,
"step": 4460
},
{
"epoch": 0.894,
"grad_norm": 0.0,
"learning_rate": 1.6919049696121958e-06,
"loss": 1.6072,
"step": 4470
},
{
"epoch": 0.896,
"grad_norm": 6.6550774574279785,
"learning_rate": 1.629358090099639e-06,
"loss": 0.8136,
"step": 4480
},
{
"epoch": 0.898,
"grad_norm": 3.452754020690918,
"learning_rate": 1.5679502627027136e-06,
"loss": 0.9143,
"step": 4490
},
{
"epoch": 0.9,
"grad_norm": 12.866599082946777,
"learning_rate": 1.5076844803522922e-06,
"loss": 0.568,
"step": 4500
},
{
"epoch": 0.902,
"grad_norm": 4.006110668182373,
"learning_rate": 1.4485636803175829e-06,
"loss": 1.1768,
"step": 4510
},
{
"epoch": 0.904,
"grad_norm": 1.1515779495239258,
"learning_rate": 1.3905907440629752e-06,
"loss": 0.9012,
"step": 4520
},
{
"epoch": 0.906,
"grad_norm": 11.245774269104004,
"learning_rate": 1.333768497107593e-06,
"loss": 1.1081,
"step": 4530
},
{
"epoch": 0.908,
"grad_norm": 6.475898742675781,
"learning_rate": 1.2780997088875869e-06,
"loss": 0.9983,
"step": 4540
},
{
"epoch": 0.91,
"grad_norm": 2.4155514240264893,
"learning_rate": 1.2235870926211619e-06,
"loss": 0.9166,
"step": 4550
},
{
"epoch": 0.912,
"grad_norm": 8.179412841796875,
"learning_rate": 1.170233305176327e-06,
"loss": 1.4171,
"step": 4560
},
{
"epoch": 0.914,
"grad_norm": 4.91797399520874,
"learning_rate": 1.1180409469414094e-06,
"loss": 2.5297,
"step": 4570
},
{
"epoch": 0.916,
"grad_norm": 0.5031004548072815,
"learning_rate": 1.067012561698319e-06,
"loss": 1.2909,
"step": 4580
},
{
"epoch": 0.918,
"grad_norm": 20.431604385375977,
"learning_rate": 1.0171506364985622e-06,
"loss": 3.3333,
"step": 4590
},
{
"epoch": 0.92,
"grad_norm": 18.30546760559082,
"learning_rate": 9.684576015420278e-07,
"loss": 1.3298,
"step": 4600
},
{
"epoch": 0.922,
"grad_norm": 1.9837167263031006,
"learning_rate": 9.209358300585474e-07,
"loss": 1.202,
"step": 4610
},
{
"epoch": 0.924,
"grad_norm": 0.7103928327560425,
"learning_rate": 8.745876381922147e-07,
"loss": 2.3325,
"step": 4620
},
{
"epoch": 0.926,
"grad_norm": 4.550110816955566,
"learning_rate": 8.294152848885157e-07,
"loss": 1.1146,
"step": 4630
},
{
"epoch": 0.928,
"grad_norm": 0.0,
"learning_rate": 7.854209717842231e-07,
"loss": 0.5371,
"step": 4640
},
{
"epoch": 0.93,
"grad_norm": 1.3056379556655884,
"learning_rate": 7.426068431000882e-07,
"loss": 1.3357,
"step": 4650
},
{
"epoch": 0.932,
"grad_norm": 46.47900390625,
"learning_rate": 7.009749855363456e-07,
"loss": 1.3991,
"step": 4660
},
{
"epoch": 0.934,
"grad_norm": 0.2128896415233612,
"learning_rate": 6.605274281709928e-07,
"loss": 0.6493,
"step": 4670
},
{
"epoch": 0.936,
"grad_norm": 0.0,
"learning_rate": 6.212661423609184e-07,
"loss": 1.5364,
"step": 4680
},
{
"epoch": 0.938,
"grad_norm": 1.3817780017852783,
"learning_rate": 5.83193041645802e-07,
"loss": 0.7224,
"step": 4690
},
{
"epoch": 0.94,
"grad_norm": 0.0,
"learning_rate": 5.463099816548579e-07,
"loss": 0.6421,
"step": 4700
},
{
"epoch": 0.942,
"grad_norm": 2.6664462089538574,
"learning_rate": 5.106187600163987e-07,
"loss": 1.108,
"step": 4710
},
{
"epoch": 0.944,
"grad_norm": 4.678840160369873,
"learning_rate": 4.7612111627021175e-07,
"loss": 1.6731,
"step": 4720
},
{
"epoch": 0.946,
"grad_norm": 1.585053563117981,
"learning_rate": 4.4281873178278475e-07,
"loss": 0.9536,
"step": 4730
},
{
"epoch": 0.948,
"grad_norm": 0.0,
"learning_rate": 4.107132296653549e-07,
"loss": 0.811,
"step": 4740
},
{
"epoch": 0.95,
"grad_norm": 3.8119912147521973,
"learning_rate": 3.7980617469479953e-07,
"loss": 1.1257,
"step": 4750
},
{
"epoch": 0.952,
"grad_norm": 0.0,
"learning_rate": 3.5009907323737825e-07,
"loss": 1.5451,
"step": 4760
},
{
"epoch": 0.954,
"grad_norm": 13.099092483520508,
"learning_rate": 3.215933731753024e-07,
"loss": 1.2267,
"step": 4770
},
{
"epoch": 0.956,
"grad_norm": 0.0,
"learning_rate": 2.942904638361804e-07,
"loss": 0.7739,
"step": 4780
},
{
"epoch": 0.958,
"grad_norm": 4.4148335456848145,
"learning_rate": 2.681916759252917e-07,
"loss": 1.1737,
"step": 4790
},
{
"epoch": 0.96,
"grad_norm": 0.0,
"learning_rate": 2.4329828146074095e-07,
"loss": 2.5418,
"step": 4800
},
{
"epoch": 0.962,
"grad_norm": 1.5763373374938965,
"learning_rate": 2.1961149371145795e-07,
"loss": 1.3623,
"step": 4810
},
{
"epoch": 0.964,
"grad_norm": 0.0,
"learning_rate": 1.9713246713805588e-07,
"loss": 1.2778,
"step": 4820
},
{
"epoch": 0.966,
"grad_norm": 5.108182907104492,
"learning_rate": 1.7586229733657644e-07,
"loss": 1.3646,
"step": 4830
},
{
"epoch": 0.968,
"grad_norm": 21.954763412475586,
"learning_rate": 1.5580202098509077e-07,
"loss": 2.2615,
"step": 4840
},
{
"epoch": 0.97,
"grad_norm": 1.4642102718353271,
"learning_rate": 1.3695261579316777e-07,
"loss": 1.3617,
"step": 4850
},
{
"epoch": 0.972,
"grad_norm": 1.9794245958328247,
"learning_rate": 1.193150004542204e-07,
"loss": 1.1849,
"step": 4860
},
{
"epoch": 0.974,
"grad_norm": 3.215317487716675,
"learning_rate": 1.0289003460074165e-07,
"loss": 0.8638,
"step": 4870
},
{
"epoch": 0.976,
"grad_norm": 2.1437785625457764,
"learning_rate": 8.767851876239074e-08,
"loss": 1.9723,
"step": 4880
},
{
"epoch": 0.978,
"grad_norm": 0.0,
"learning_rate": 7.368119432699383e-08,
"loss": 0.862,
"step": 4890
},
{
"epoch": 0.98,
"grad_norm": 2.7813355922698975,
"learning_rate": 6.089874350439506e-08,
"loss": 1.0444,
"step": 4900
},
{
"epoch": 0.982,
"grad_norm": 2.3938515186309814,
"learning_rate": 4.9331789293211026e-08,
"loss": 1.3459,
"step": 4910
},
{
"epoch": 0.984,
"grad_norm": 1.8486565351486206,
"learning_rate": 3.8980895450474455e-08,
"loss": 0.8181,
"step": 4920
},
{
"epoch": 0.986,
"grad_norm": 5.9725494384765625,
"learning_rate": 2.9846566464150626e-08,
"loss": 0.8222,
"step": 4930
},
{
"epoch": 0.988,
"grad_norm": 0.0,
"learning_rate": 2.192924752854042e-08,
"loss": 1.6087,
"step": 4940
},
{
"epoch": 0.99,
"grad_norm": 1.0317845344543457,
"learning_rate": 1.522932452260595e-08,
"loss": 0.8004,
"step": 4950
},
{
"epoch": 0.992,
"grad_norm": 14.324555397033691,
"learning_rate": 9.747123991141194e-09,
"loss": 0.9372,
"step": 4960
},
{
"epoch": 0.994,
"grad_norm": 2.1189212799072266,
"learning_rate": 5.48291312886251e-09,
"loss": 1.4132,
"step": 4970
},
{
"epoch": 0.996,
"grad_norm": 2.8049211502075195,
"learning_rate": 2.4368997673940297e-09,
"loss": 1.172,
"step": 4980
},
{
"epoch": 0.998,
"grad_norm": 1.8685941696166992,
"learning_rate": 6.092323651313292e-10,
"loss": 1.6231,
"step": 4990
},
{
"epoch": 1.0,
"grad_norm": 0.8898677229881287,
"learning_rate": 0.0,
"loss": 0.94,
"step": 5000
},
{
"epoch": 1.0,
"step": 5000,
"total_flos": 6806517818228736.0,
"train_loss": 1.3851989124298096,
"train_runtime": 652.6593,
"train_samples_per_second": 7.661,
"train_steps_per_second": 7.661
}
],
"logging_steps": 10,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 4000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6806517818228736.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}