model_c2b62d8d / checkpoint-504 /trainer_state.json
ugaoo's picture
Upload folder using huggingface_hub
63def14 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.9736147757255935,
"eval_steps": 500,
"global_step": 504,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0079155672823219,
"grad_norm": 34.582191467285156,
"learning_rate": 5.0000000000000004e-08,
"loss": 2.1743,
"step": 1
},
{
"epoch": 0.0158311345646438,
"grad_norm": 35.2754020690918,
"learning_rate": 1.0000000000000001e-07,
"loss": 2.195,
"step": 2
},
{
"epoch": 0.023746701846965697,
"grad_norm": 35.12391662597656,
"learning_rate": 1.5000000000000002e-07,
"loss": 2.1971,
"step": 3
},
{
"epoch": 0.0316622691292876,
"grad_norm": 34.45759201049805,
"learning_rate": 2.0000000000000002e-07,
"loss": 2.1701,
"step": 4
},
{
"epoch": 0.0395778364116095,
"grad_norm": 35.35009002685547,
"learning_rate": 2.5000000000000004e-07,
"loss": 2.2099,
"step": 5
},
{
"epoch": 0.047493403693931395,
"grad_norm": 35.039520263671875,
"learning_rate": 3.0000000000000004e-07,
"loss": 2.2169,
"step": 6
},
{
"epoch": 0.055408970976253295,
"grad_norm": 35.103248596191406,
"learning_rate": 3.5000000000000004e-07,
"loss": 2.2121,
"step": 7
},
{
"epoch": 0.0633245382585752,
"grad_norm": 34.65024185180664,
"learning_rate": 4.0000000000000003e-07,
"loss": 2.1715,
"step": 8
},
{
"epoch": 0.0712401055408971,
"grad_norm": 35.055023193359375,
"learning_rate": 4.5000000000000003e-07,
"loss": 2.1571,
"step": 9
},
{
"epoch": 0.079155672823219,
"grad_norm": 34.82114028930664,
"learning_rate": 5.000000000000001e-07,
"loss": 2.1428,
"step": 10
},
{
"epoch": 0.0870712401055409,
"grad_norm": 33.663883209228516,
"learning_rate": 5.5e-07,
"loss": 2.0656,
"step": 11
},
{
"epoch": 0.09498680738786279,
"grad_norm": 33.614967346191406,
"learning_rate": 6.000000000000001e-07,
"loss": 2.05,
"step": 12
},
{
"epoch": 0.10290237467018469,
"grad_norm": 32.66334915161133,
"learning_rate": 6.5e-07,
"loss": 1.9905,
"step": 13
},
{
"epoch": 0.11081794195250659,
"grad_norm": 31.83934211730957,
"learning_rate": 7.000000000000001e-07,
"loss": 1.8981,
"step": 14
},
{
"epoch": 0.11873350923482849,
"grad_norm": 32.067840576171875,
"learning_rate": 7.5e-07,
"loss": 1.8803,
"step": 15
},
{
"epoch": 0.1266490765171504,
"grad_norm": 31.775592803955078,
"learning_rate": 8.000000000000001e-07,
"loss": 1.8215,
"step": 16
},
{
"epoch": 0.1345646437994723,
"grad_norm": 30.205060958862305,
"learning_rate": 8.500000000000001e-07,
"loss": 1.6654,
"step": 17
},
{
"epoch": 0.1424802110817942,
"grad_norm": 30.699304580688477,
"learning_rate": 9.000000000000001e-07,
"loss": 1.6149,
"step": 18
},
{
"epoch": 0.1503957783641161,
"grad_norm": 30.16368293762207,
"learning_rate": 9.500000000000001e-07,
"loss": 1.4905,
"step": 19
},
{
"epoch": 0.158311345646438,
"grad_norm": 30.227079391479492,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.3845,
"step": 20
},
{
"epoch": 0.1662269129287599,
"grad_norm": 30.42757797241211,
"learning_rate": 1.0500000000000001e-06,
"loss": 1.2462,
"step": 21
},
{
"epoch": 0.1741424802110818,
"grad_norm": 30.44344711303711,
"learning_rate": 1.1e-06,
"loss": 1.0948,
"step": 22
},
{
"epoch": 0.1820580474934037,
"grad_norm": 30.735376358032227,
"learning_rate": 1.1500000000000002e-06,
"loss": 0.9618,
"step": 23
},
{
"epoch": 0.18997361477572558,
"grad_norm": 30.048675537109375,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.7738,
"step": 24
},
{
"epoch": 0.19788918205804748,
"grad_norm": 30.161460876464844,
"learning_rate": 1.25e-06,
"loss": 0.6424,
"step": 25
},
{
"epoch": 0.20580474934036938,
"grad_norm": 28.648540496826172,
"learning_rate": 1.3e-06,
"loss": 0.4602,
"step": 26
},
{
"epoch": 0.21372031662269128,
"grad_norm": 25.193084716796875,
"learning_rate": 1.3500000000000002e-06,
"loss": 0.3309,
"step": 27
},
{
"epoch": 0.22163588390501318,
"grad_norm": 18.437116622924805,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.2206,
"step": 28
},
{
"epoch": 0.22955145118733508,
"grad_norm": 12.642715454101562,
"learning_rate": 1.45e-06,
"loss": 0.139,
"step": 29
},
{
"epoch": 0.23746701846965698,
"grad_norm": 6.1605305671691895,
"learning_rate": 1.5e-06,
"loss": 0.0982,
"step": 30
},
{
"epoch": 0.24538258575197888,
"grad_norm": 2.986103057861328,
"learning_rate": 1.5500000000000002e-06,
"loss": 0.0707,
"step": 31
},
{
"epoch": 0.2532981530343008,
"grad_norm": 2.030493974685669,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.0637,
"step": 32
},
{
"epoch": 0.2612137203166227,
"grad_norm": 1.5127642154693604,
"learning_rate": 1.6500000000000003e-06,
"loss": 0.0616,
"step": 33
},
{
"epoch": 0.2691292875989446,
"grad_norm": 1.8087722063064575,
"learning_rate": 1.7000000000000002e-06,
"loss": 0.0555,
"step": 34
},
{
"epoch": 0.2770448548812665,
"grad_norm": 1.3533340692520142,
"learning_rate": 1.75e-06,
"loss": 0.0517,
"step": 35
},
{
"epoch": 0.2849604221635884,
"grad_norm": 1.3595997095108032,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.0537,
"step": 36
},
{
"epoch": 0.2928759894459103,
"grad_norm": 1.2778421640396118,
"learning_rate": 1.85e-06,
"loss": 0.0483,
"step": 37
},
{
"epoch": 0.3007915567282322,
"grad_norm": 1.0834269523620605,
"learning_rate": 1.9000000000000002e-06,
"loss": 0.0471,
"step": 38
},
{
"epoch": 0.3087071240105541,
"grad_norm": 1.175179362297058,
"learning_rate": 1.9500000000000004e-06,
"loss": 0.0528,
"step": 39
},
{
"epoch": 0.316622691292876,
"grad_norm": 0.9915143847465515,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.0525,
"step": 40
},
{
"epoch": 0.3245382585751979,
"grad_norm": 0.7996736168861389,
"learning_rate": 2.05e-06,
"loss": 0.0479,
"step": 41
},
{
"epoch": 0.3324538258575198,
"grad_norm": 0.7375659942626953,
"learning_rate": 2.1000000000000002e-06,
"loss": 0.0509,
"step": 42
},
{
"epoch": 0.3403693931398417,
"grad_norm": 0.7712483406066895,
"learning_rate": 2.15e-06,
"loss": 0.0484,
"step": 43
},
{
"epoch": 0.3482849604221636,
"grad_norm": 0.9321597218513489,
"learning_rate": 2.2e-06,
"loss": 0.0448,
"step": 44
},
{
"epoch": 0.3562005277044855,
"grad_norm": 1.02247953414917,
"learning_rate": 2.25e-06,
"loss": 0.0484,
"step": 45
},
{
"epoch": 0.3641160949868074,
"grad_norm": 0.857718825340271,
"learning_rate": 2.3000000000000004e-06,
"loss": 0.0425,
"step": 46
},
{
"epoch": 0.3720316622691293,
"grad_norm": 1.0721040964126587,
"learning_rate": 2.35e-06,
"loss": 0.0472,
"step": 47
},
{
"epoch": 0.37994722955145116,
"grad_norm": 0.811705470085144,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.039,
"step": 48
},
{
"epoch": 0.38786279683377306,
"grad_norm": 0.9680790901184082,
"learning_rate": 2.4500000000000003e-06,
"loss": 0.0423,
"step": 49
},
{
"epoch": 0.39577836411609496,
"grad_norm": 0.9470660090446472,
"learning_rate": 2.5e-06,
"loss": 0.0411,
"step": 50
},
{
"epoch": 0.40369393139841686,
"grad_norm": 0.8540447950363159,
"learning_rate": 2.55e-06,
"loss": 0.049,
"step": 51
},
{
"epoch": 0.41160949868073876,
"grad_norm": 1.39769446849823,
"learning_rate": 2.6e-06,
"loss": 0.0478,
"step": 52
},
{
"epoch": 0.41952506596306066,
"grad_norm": 0.8914775848388672,
"learning_rate": 2.6500000000000005e-06,
"loss": 0.0415,
"step": 53
},
{
"epoch": 0.42744063324538256,
"grad_norm": 0.759353518486023,
"learning_rate": 2.7000000000000004e-06,
"loss": 0.046,
"step": 54
},
{
"epoch": 0.43535620052770446,
"grad_norm": 0.8632487058639526,
"learning_rate": 2.7500000000000004e-06,
"loss": 0.0435,
"step": 55
},
{
"epoch": 0.44327176781002636,
"grad_norm": 0.6662724018096924,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.0496,
"step": 56
},
{
"epoch": 0.45118733509234826,
"grad_norm": 0.8853585124015808,
"learning_rate": 2.85e-06,
"loss": 0.0384,
"step": 57
},
{
"epoch": 0.45910290237467016,
"grad_norm": 0.7003194093704224,
"learning_rate": 2.9e-06,
"loss": 0.0446,
"step": 58
},
{
"epoch": 0.46701846965699206,
"grad_norm": 0.8177786469459534,
"learning_rate": 2.95e-06,
"loss": 0.0417,
"step": 59
},
{
"epoch": 0.47493403693931396,
"grad_norm": 0.9565925598144531,
"learning_rate": 3e-06,
"loss": 0.044,
"step": 60
},
{
"epoch": 0.48284960422163586,
"grad_norm": 0.7401254177093506,
"learning_rate": 3.05e-06,
"loss": 0.0439,
"step": 61
},
{
"epoch": 0.49076517150395776,
"grad_norm": 0.9952861666679382,
"learning_rate": 3.1000000000000004e-06,
"loss": 0.0371,
"step": 62
},
{
"epoch": 0.49868073878627966,
"grad_norm": 0.9622153639793396,
"learning_rate": 3.1500000000000003e-06,
"loss": 0.0457,
"step": 63
},
{
"epoch": 0.5065963060686016,
"grad_norm": 0.7078511118888855,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.0396,
"step": 64
},
{
"epoch": 0.5145118733509235,
"grad_norm": 0.9370516538619995,
"learning_rate": 3.2500000000000002e-06,
"loss": 0.0421,
"step": 65
},
{
"epoch": 0.5224274406332454,
"grad_norm": 0.6465123891830444,
"learning_rate": 3.3000000000000006e-06,
"loss": 0.0341,
"step": 66
},
{
"epoch": 0.5303430079155673,
"grad_norm": 0.7659086585044861,
"learning_rate": 3.3500000000000005e-06,
"loss": 0.0429,
"step": 67
},
{
"epoch": 0.5382585751978892,
"grad_norm": 0.7772212028503418,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.036,
"step": 68
},
{
"epoch": 0.5461741424802111,
"grad_norm": 0.7115844488143921,
"learning_rate": 3.45e-06,
"loss": 0.0327,
"step": 69
},
{
"epoch": 0.554089709762533,
"grad_norm": 0.6542766690254211,
"learning_rate": 3.5e-06,
"loss": 0.0362,
"step": 70
},
{
"epoch": 0.5620052770448549,
"grad_norm": 0.7394366264343262,
"learning_rate": 3.5500000000000003e-06,
"loss": 0.0426,
"step": 71
},
{
"epoch": 0.5699208443271768,
"grad_norm": 0.8163363337516785,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.0344,
"step": 72
},
{
"epoch": 0.5778364116094987,
"grad_norm": 0.7115806937217712,
"learning_rate": 3.65e-06,
"loss": 0.0379,
"step": 73
},
{
"epoch": 0.5857519788918206,
"grad_norm": 0.6192121505737305,
"learning_rate": 3.7e-06,
"loss": 0.0276,
"step": 74
},
{
"epoch": 0.5936675461741425,
"grad_norm": 0.822684645652771,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.0308,
"step": 75
},
{
"epoch": 0.6015831134564644,
"grad_norm": 0.8095663189888,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.0303,
"step": 76
},
{
"epoch": 0.6094986807387863,
"grad_norm": 0.7700952887535095,
"learning_rate": 3.85e-06,
"loss": 0.0328,
"step": 77
},
{
"epoch": 0.6174142480211082,
"grad_norm": 0.7777529358863831,
"learning_rate": 3.900000000000001e-06,
"loss": 0.0344,
"step": 78
},
{
"epoch": 0.6253298153034301,
"grad_norm": 0.826386034488678,
"learning_rate": 3.95e-06,
"loss": 0.033,
"step": 79
},
{
"epoch": 0.633245382585752,
"grad_norm": 0.788942277431488,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0331,
"step": 80
},
{
"epoch": 0.6411609498680739,
"grad_norm": 0.7351260185241699,
"learning_rate": 4.05e-06,
"loss": 0.0346,
"step": 81
},
{
"epoch": 0.6490765171503958,
"grad_norm": 0.5737812519073486,
"learning_rate": 4.1e-06,
"loss": 0.027,
"step": 82
},
{
"epoch": 0.6569920844327177,
"grad_norm": 0.7053394913673401,
"learning_rate": 4.15e-06,
"loss": 0.034,
"step": 83
},
{
"epoch": 0.6649076517150396,
"grad_norm": 0.7508617639541626,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.0323,
"step": 84
},
{
"epoch": 0.6728232189973615,
"grad_norm": 1.0235975980758667,
"learning_rate": 4.25e-06,
"loss": 0.0395,
"step": 85
},
{
"epoch": 0.6807387862796834,
"grad_norm": 0.6522803902626038,
"learning_rate": 4.3e-06,
"loss": 0.0262,
"step": 86
},
{
"epoch": 0.6886543535620053,
"grad_norm": 1.182160496711731,
"learning_rate": 4.350000000000001e-06,
"loss": 0.0387,
"step": 87
},
{
"epoch": 0.6965699208443272,
"grad_norm": 0.6571248769760132,
"learning_rate": 4.4e-06,
"loss": 0.0262,
"step": 88
},
{
"epoch": 0.7044854881266491,
"grad_norm": 0.6641517877578735,
"learning_rate": 4.450000000000001e-06,
"loss": 0.0265,
"step": 89
},
{
"epoch": 0.712401055408971,
"grad_norm": 0.8751306533813477,
"learning_rate": 4.5e-06,
"loss": 0.0307,
"step": 90
},
{
"epoch": 0.7203166226912929,
"grad_norm": 0.7627044320106506,
"learning_rate": 4.5500000000000005e-06,
"loss": 0.0332,
"step": 91
},
{
"epoch": 0.7282321899736148,
"grad_norm": 0.7124688625335693,
"learning_rate": 4.600000000000001e-06,
"loss": 0.0299,
"step": 92
},
{
"epoch": 0.7361477572559367,
"grad_norm": 0.7688237428665161,
"learning_rate": 4.65e-06,
"loss": 0.0258,
"step": 93
},
{
"epoch": 0.7440633245382586,
"grad_norm": 0.8663469552993774,
"learning_rate": 4.7e-06,
"loss": 0.0342,
"step": 94
},
{
"epoch": 0.7519788918205804,
"grad_norm": 0.6536213755607605,
"learning_rate": 4.75e-06,
"loss": 0.0226,
"step": 95
},
{
"epoch": 0.7598944591029023,
"grad_norm": 0.5616201758384705,
"learning_rate": 4.800000000000001e-06,
"loss": 0.022,
"step": 96
},
{
"epoch": 0.7678100263852242,
"grad_norm": 0.7985079884529114,
"learning_rate": 4.85e-06,
"loss": 0.0305,
"step": 97
},
{
"epoch": 0.7757255936675461,
"grad_norm": 0.726844847202301,
"learning_rate": 4.9000000000000005e-06,
"loss": 0.0243,
"step": 98
},
{
"epoch": 0.783641160949868,
"grad_norm": 0.656874418258667,
"learning_rate": 4.95e-06,
"loss": 0.0243,
"step": 99
},
{
"epoch": 0.7915567282321899,
"grad_norm": 0.8744286894798279,
"learning_rate": 5e-06,
"loss": 0.0261,
"step": 100
},
{
"epoch": 0.7994722955145118,
"grad_norm": 1.0138187408447266,
"learning_rate": 4.9999713317456065e-06,
"loss": 0.0254,
"step": 101
},
{
"epoch": 0.8073878627968337,
"grad_norm": 0.6964263319969177,
"learning_rate": 4.9998853276399215e-06,
"loss": 0.0263,
"step": 102
},
{
"epoch": 0.8153034300791556,
"grad_norm": 0.6374551057815552,
"learning_rate": 4.999741989655415e-06,
"loss": 0.0228,
"step": 103
},
{
"epoch": 0.8232189973614775,
"grad_norm": 0.6274930834770203,
"learning_rate": 4.999541321079486e-06,
"loss": 0.0286,
"step": 104
},
{
"epoch": 0.8311345646437994,
"grad_norm": 0.6344143748283386,
"learning_rate": 4.99928332651439e-06,
"loss": 0.0258,
"step": 105
},
{
"epoch": 0.8390501319261213,
"grad_norm": 0.8304360508918762,
"learning_rate": 4.998968011877129e-06,
"loss": 0.022,
"step": 106
},
{
"epoch": 0.8469656992084432,
"grad_norm": 0.6463897228240967,
"learning_rate": 4.998595384399319e-06,
"loss": 0.0184,
"step": 107
},
{
"epoch": 0.8548812664907651,
"grad_norm": 0.5647016763687134,
"learning_rate": 4.998165452627025e-06,
"loss": 0.0221,
"step": 108
},
{
"epoch": 0.862796833773087,
"grad_norm": 0.5621638298034668,
"learning_rate": 4.997678226420561e-06,
"loss": 0.022,
"step": 109
},
{
"epoch": 0.8707124010554089,
"grad_norm": 0.8295672535896301,
"learning_rate": 4.997133716954266e-06,
"loss": 0.0206,
"step": 110
},
{
"epoch": 0.8786279683377308,
"grad_norm": 0.8057272434234619,
"learning_rate": 4.99653193671625e-06,
"loss": 0.0204,
"step": 111
},
{
"epoch": 0.8865435356200527,
"grad_norm": 0.6956275105476379,
"learning_rate": 4.995872899508103e-06,
"loss": 0.0186,
"step": 112
},
{
"epoch": 0.8944591029023746,
"grad_norm": 0.7873161435127258,
"learning_rate": 4.995156620444584e-06,
"loss": 0.0149,
"step": 113
},
{
"epoch": 0.9023746701846965,
"grad_norm": 0.9187519550323486,
"learning_rate": 4.994383115953266e-06,
"loss": 0.0183,
"step": 114
},
{
"epoch": 0.9102902374670184,
"grad_norm": 0.5987403392791748,
"learning_rate": 4.9935524037741705e-06,
"loss": 0.0158,
"step": 115
},
{
"epoch": 0.9182058047493403,
"grad_norm": 0.6011790037155151,
"learning_rate": 4.992664502959351e-06,
"loss": 0.016,
"step": 116
},
{
"epoch": 0.9261213720316622,
"grad_norm": 0.8429304957389832,
"learning_rate": 4.991719433872461e-06,
"loss": 0.0107,
"step": 117
},
{
"epoch": 0.9340369393139841,
"grad_norm": 0.8618797659873962,
"learning_rate": 4.990717218188286e-06,
"loss": 0.0162,
"step": 118
},
{
"epoch": 0.941952506596306,
"grad_norm": 1.1907824277877808,
"learning_rate": 4.989657878892245e-06,
"loss": 0.0173,
"step": 119
},
{
"epoch": 0.9498680738786279,
"grad_norm": 0.7317540645599365,
"learning_rate": 4.988541440279862e-06,
"loss": 0.0183,
"step": 120
},
{
"epoch": 0.9577836411609498,
"grad_norm": 0.6827302575111389,
"learning_rate": 4.987367927956218e-06,
"loss": 0.0192,
"step": 121
},
{
"epoch": 0.9656992084432717,
"grad_norm": 0.6512569189071655,
"learning_rate": 4.986137368835351e-06,
"loss": 0.0169,
"step": 122
},
{
"epoch": 0.9736147757255936,
"grad_norm": 0.5439648032188416,
"learning_rate": 4.984849791139647e-06,
"loss": 0.0136,
"step": 123
},
{
"epoch": 0.9815303430079155,
"grad_norm": 0.5652760863304138,
"learning_rate": 4.983505224399188e-06,
"loss": 0.0171,
"step": 124
},
{
"epoch": 0.9894459102902374,
"grad_norm": 0.6217207908630371,
"learning_rate": 4.9821036994510816e-06,
"loss": 0.0175,
"step": 125
},
{
"epoch": 0.9973614775725593,
"grad_norm": 0.5073563456535339,
"learning_rate": 4.980645248438746e-06,
"loss": 0.0166,
"step": 126
},
{
"epoch": 1.0,
"grad_norm": 0.5073563456535339,
"learning_rate": 4.979129904811177e-06,
"loss": 0.0103,
"step": 127
},
{
"epoch": 1.007915567282322,
"grad_norm": 0.9930127263069153,
"learning_rate": 4.977557703322178e-06,
"loss": 0.0116,
"step": 128
},
{
"epoch": 1.0158311345646438,
"grad_norm": 0.4781687259674072,
"learning_rate": 4.975928680029571e-06,
"loss": 0.0097,
"step": 129
},
{
"epoch": 1.0237467018469657,
"grad_norm": 0.5260384678840637,
"learning_rate": 4.974242872294354e-06,
"loss": 0.0131,
"step": 130
},
{
"epoch": 1.0316622691292876,
"grad_norm": 0.6748254895210266,
"learning_rate": 4.972500318779864e-06,
"loss": 0.0093,
"step": 131
},
{
"epoch": 1.0395778364116095,
"grad_norm": 0.42848217487335205,
"learning_rate": 4.970701059450872e-06,
"loss": 0.0086,
"step": 132
},
{
"epoch": 1.0474934036939314,
"grad_norm": 0.8815901279449463,
"learning_rate": 4.968845135572678e-06,
"loss": 0.0103,
"step": 133
},
{
"epoch": 1.0554089709762533,
"grad_norm": 0.5253162980079651,
"learning_rate": 4.96693258971016e-06,
"loss": 0.0088,
"step": 134
},
{
"epoch": 1.0633245382585752,
"grad_norm": 0.6259243488311768,
"learning_rate": 4.9649634657268e-06,
"loss": 0.0074,
"step": 135
},
{
"epoch": 1.071240105540897,
"grad_norm": 1.0937525033950806,
"learning_rate": 4.962937808783675e-06,
"loss": 0.0115,
"step": 136
},
{
"epoch": 1.079155672823219,
"grad_norm": 1.0612727403640747,
"learning_rate": 4.960855665338424e-06,
"loss": 0.009,
"step": 137
},
{
"epoch": 1.087071240105541,
"grad_norm": 0.9188615679740906,
"learning_rate": 4.958717083144182e-06,
"loss": 0.0127,
"step": 138
},
{
"epoch": 1.0949868073878628,
"grad_norm": 0.6099897623062134,
"learning_rate": 4.956522111248483e-06,
"loss": 0.0088,
"step": 139
},
{
"epoch": 1.1029023746701847,
"grad_norm": 0.6967251896858215,
"learning_rate": 4.954270799992138e-06,
"loss": 0.0088,
"step": 140
},
{
"epoch": 1.1108179419525066,
"grad_norm": 0.6113755702972412,
"learning_rate": 4.9519632010080765e-06,
"loss": 0.0079,
"step": 141
},
{
"epoch": 1.1187335092348285,
"grad_norm": 0.6708074808120728,
"learning_rate": 4.9495993672201675e-06,
"loss": 0.0093,
"step": 142
},
{
"epoch": 1.1266490765171504,
"grad_norm": 0.425584614276886,
"learning_rate": 4.947179352842001e-06,
"loss": 0.0093,
"step": 143
},
{
"epoch": 1.1345646437994723,
"grad_norm": 0.3952641785144806,
"learning_rate": 4.944703213375648e-06,
"loss": 0.0057,
"step": 144
},
{
"epoch": 1.1424802110817942,
"grad_norm": 0.39029237627983093,
"learning_rate": 4.942171005610385e-06,
"loss": 0.0106,
"step": 145
},
{
"epoch": 1.150395778364116,
"grad_norm": 0.5346894264221191,
"learning_rate": 4.939582787621394e-06,
"loss": 0.0062,
"step": 146
},
{
"epoch": 1.158311345646438,
"grad_norm": 0.48500752449035645,
"learning_rate": 4.936938618768427e-06,
"loss": 0.0077,
"step": 147
},
{
"epoch": 1.16622691292876,
"grad_norm": 0.5172581076622009,
"learning_rate": 4.934238559694448e-06,
"loss": 0.0122,
"step": 148
},
{
"epoch": 1.1741424802110818,
"grad_norm": 0.571660578250885,
"learning_rate": 4.9314826723242425e-06,
"loss": 0.0107,
"step": 149
},
{
"epoch": 1.1820580474934037,
"grad_norm": 0.558224081993103,
"learning_rate": 4.928671019862995e-06,
"loss": 0.0062,
"step": 150
},
{
"epoch": 1.1899736147757256,
"grad_norm": 0.46116045117378235,
"learning_rate": 4.925803666794839e-06,
"loss": 0.0089,
"step": 151
},
{
"epoch": 1.1978891820580475,
"grad_norm": 0.5193414092063904,
"learning_rate": 4.92288067888138e-06,
"loss": 0.0084,
"step": 152
},
{
"epoch": 1.2058047493403694,
"grad_norm": 0.6195410490036011,
"learning_rate": 4.919902123160187e-06,
"loss": 0.0115,
"step": 153
},
{
"epoch": 1.2137203166226913,
"grad_norm": 0.5878890752792358,
"learning_rate": 4.9168680679432565e-06,
"loss": 0.0075,
"step": 154
},
{
"epoch": 1.2216358839050132,
"grad_norm": 0.3049341142177582,
"learning_rate": 4.913778582815439e-06,
"loss": 0.0071,
"step": 155
},
{
"epoch": 1.229551451187335,
"grad_norm": 0.9924205541610718,
"learning_rate": 4.9106337386328524e-06,
"loss": 0.0121,
"step": 156
},
{
"epoch": 1.237467018469657,
"grad_norm": 0.44824329018592834,
"learning_rate": 4.907433607521252e-06,
"loss": 0.0066,
"step": 157
},
{
"epoch": 1.245382585751979,
"grad_norm": 0.7608261108398438,
"learning_rate": 4.904178262874374e-06,
"loss": 0.0046,
"step": 158
},
{
"epoch": 1.2532981530343008,
"grad_norm": 0.4237998425960541,
"learning_rate": 4.900867779352258e-06,
"loss": 0.0032,
"step": 159
},
{
"epoch": 1.2612137203166227,
"grad_norm": 0.9337185025215149,
"learning_rate": 4.897502232879533e-06,
"loss": 0.0061,
"step": 160
},
{
"epoch": 1.2691292875989446,
"grad_norm": 0.6228697299957275,
"learning_rate": 4.89408170064367e-06,
"loss": 0.0057,
"step": 161
},
{
"epoch": 1.2770448548812665,
"grad_norm": 0.41712895035743713,
"learning_rate": 4.890606261093221e-06,
"loss": 0.0046,
"step": 162
},
{
"epoch": 1.2849604221635884,
"grad_norm": 0.6179379820823669,
"learning_rate": 4.887075993936014e-06,
"loss": 0.0045,
"step": 163
},
{
"epoch": 1.2928759894459103,
"grad_norm": 0.3739337921142578,
"learning_rate": 4.883490980137327e-06,
"loss": 0.0061,
"step": 164
},
{
"epoch": 1.3007915567282322,
"grad_norm": 0.5171535015106201,
"learning_rate": 4.8798513019180295e-06,
"loss": 0.0073,
"step": 165
},
{
"epoch": 1.3087071240105541,
"grad_norm": 0.640513002872467,
"learning_rate": 4.876157042752698e-06,
"loss": 0.0043,
"step": 166
},
{
"epoch": 1.316622691292876,
"grad_norm": 0.4904448390007019,
"learning_rate": 4.872408287367702e-06,
"loss": 0.0044,
"step": 167
},
{
"epoch": 1.324538258575198,
"grad_norm": 0.4022831618785858,
"learning_rate": 4.868605121739261e-06,
"loss": 0.0019,
"step": 168
},
{
"epoch": 1.3324538258575198,
"grad_norm": 0.35701292753219604,
"learning_rate": 4.86474763309147e-06,
"loss": 0.0048,
"step": 169
},
{
"epoch": 1.3403693931398417,
"grad_norm": 1.2709388732910156,
"learning_rate": 4.8608359098943014e-06,
"loss": 0.0073,
"step": 170
},
{
"epoch": 1.3482849604221636,
"grad_norm": 1.1490532159805298,
"learning_rate": 4.856870041861576e-06,
"loss": 0.0082,
"step": 171
},
{
"epoch": 1.3562005277044855,
"grad_norm": 1.9450149536132812,
"learning_rate": 4.8528501199489045e-06,
"loss": 0.0047,
"step": 172
},
{
"epoch": 1.3641160949868074,
"grad_norm": 0.2517525553703308,
"learning_rate": 4.848776236351602e-06,
"loss": 0.0101,
"step": 173
},
{
"epoch": 1.3720316622691293,
"grad_norm": 0.612678050994873,
"learning_rate": 4.8446484845025754e-06,
"loss": 0.0064,
"step": 174
},
{
"epoch": 1.3799472295514512,
"grad_norm": 0.8020515441894531,
"learning_rate": 4.840466959070174e-06,
"loss": 0.0077,
"step": 175
},
{
"epoch": 1.3878627968337731,
"grad_norm": 0.46455827355384827,
"learning_rate": 4.836231755956028e-06,
"loss": 0.0034,
"step": 176
},
{
"epoch": 1.395778364116095,
"grad_norm": 0.4173589050769806,
"learning_rate": 4.83194297229284e-06,
"loss": 0.0065,
"step": 177
},
{
"epoch": 1.403693931398417,
"grad_norm": 0.37849071621894836,
"learning_rate": 4.827600706442164e-06,
"loss": 0.0062,
"step": 178
},
{
"epoch": 1.4116094986807388,
"grad_norm": 0.6507642269134521,
"learning_rate": 4.823205057992145e-06,
"loss": 0.0054,
"step": 179
},
{
"epoch": 1.4195250659630607,
"grad_norm": 0.41411933302879333,
"learning_rate": 4.8187561277552376e-06,
"loss": 0.0056,
"step": 180
},
{
"epoch": 1.4274406332453826,
"grad_norm": 0.2742733359336853,
"learning_rate": 4.8142540177658925e-06,
"loss": 0.0022,
"step": 181
},
{
"epoch": 1.4353562005277045,
"grad_norm": 0.4650178849697113,
"learning_rate": 4.809698831278217e-06,
"loss": 0.0054,
"step": 182
},
{
"epoch": 1.4432717678100264,
"grad_norm": 0.2831101715564728,
"learning_rate": 4.805090672763609e-06,
"loss": 0.0041,
"step": 183
},
{
"epoch": 1.4511873350923483,
"grad_norm": 0.4332253634929657,
"learning_rate": 4.800429647908354e-06,
"loss": 0.0024,
"step": 184
},
{
"epoch": 1.4591029023746702,
"grad_norm": 0.13565762341022491,
"learning_rate": 4.795715863611212e-06,
"loss": 0.0025,
"step": 185
},
{
"epoch": 1.4670184696569921,
"grad_norm": 0.36338773369789124,
"learning_rate": 4.790949427980956e-06,
"loss": 0.0006,
"step": 186
},
{
"epoch": 1.474934036939314,
"grad_norm": 0.6935271620750427,
"learning_rate": 4.786130450333897e-06,
"loss": 0.0034,
"step": 187
},
{
"epoch": 1.482849604221636,
"grad_norm": 0.7220961451530457,
"learning_rate": 4.7812590411913755e-06,
"loss": 0.0041,
"step": 188
},
{
"epoch": 1.4907651715039578,
"grad_norm": 0.26949548721313477,
"learning_rate": 4.77633531227723e-06,
"loss": 0.0025,
"step": 189
},
{
"epoch": 1.4986807387862797,
"grad_norm": 0.6303775906562805,
"learning_rate": 4.771359376515231e-06,
"loss": 0.005,
"step": 190
},
{
"epoch": 1.5065963060686016,
"grad_norm": 0.7352619171142578,
"learning_rate": 4.766331348026493e-06,
"loss": 0.0061,
"step": 191
},
{
"epoch": 1.5145118733509235,
"grad_norm": 0.6918866038322449,
"learning_rate": 4.7612513421268546e-06,
"loss": 0.0056,
"step": 192
},
{
"epoch": 1.5224274406332454,
"grad_norm": 1.1295803785324097,
"learning_rate": 4.756119475324237e-06,
"loss": 0.0042,
"step": 193
},
{
"epoch": 1.5303430079155673,
"grad_norm": 0.4663296043872833,
"learning_rate": 4.750935865315972e-06,
"loss": 0.0026,
"step": 194
},
{
"epoch": 1.5382585751978892,
"grad_norm": 0.19712182879447937,
"learning_rate": 4.745700630986097e-06,
"loss": 0.0012,
"step": 195
},
{
"epoch": 1.5461741424802111,
"grad_norm": 0.322085440158844,
"learning_rate": 4.740413892402639e-06,
"loss": 0.0053,
"step": 196
},
{
"epoch": 1.554089709762533,
"grad_norm": 0.4675672650337219,
"learning_rate": 4.73507577081485e-06,
"loss": 0.0043,
"step": 197
},
{
"epoch": 1.562005277044855,
"grad_norm": 0.5347087979316711,
"learning_rate": 4.7296863886504315e-06,
"loss": 0.0063,
"step": 198
},
{
"epoch": 1.5699208443271768,
"grad_norm": 0.48838165402412415,
"learning_rate": 4.7242458695127275e-06,
"loss": 0.0029,
"step": 199
},
{
"epoch": 1.5778364116094987,
"grad_norm": 0.42908889055252075,
"learning_rate": 4.718754338177887e-06,
"loss": 0.0049,
"step": 200
},
{
"epoch": 1.5857519788918206,
"grad_norm": 0.2672640085220337,
"learning_rate": 4.713211920592003e-06,
"loss": 0.0025,
"step": 201
},
{
"epoch": 1.5936675461741425,
"grad_norm": 0.7370240092277527,
"learning_rate": 4.707618743868226e-06,
"loss": 0.0033,
"step": 202
},
{
"epoch": 1.6015831134564644,
"grad_norm": 0.3037024736404419,
"learning_rate": 4.701974936283848e-06,
"loss": 0.0028,
"step": 203
},
{
"epoch": 1.6094986807387863,
"grad_norm": 0.3099062442779541,
"learning_rate": 4.696280627277356e-06,
"loss": 0.0026,
"step": 204
},
{
"epoch": 1.6174142480211082,
"grad_norm": 0.16460958123207092,
"learning_rate": 4.690535947445471e-06,
"loss": 0.006,
"step": 205
},
{
"epoch": 1.6253298153034301,
"grad_norm": 0.43918171525001526,
"learning_rate": 4.6847410285401465e-06,
"loss": 0.0037,
"step": 206
},
{
"epoch": 1.633245382585752,
"grad_norm": 0.2510988414287567,
"learning_rate": 4.67889600346555e-06,
"loss": 0.0014,
"step": 207
},
{
"epoch": 1.641160949868074,
"grad_norm": 0.27360427379608154,
"learning_rate": 4.673001006275013e-06,
"loss": 0.0019,
"step": 208
},
{
"epoch": 1.6490765171503958,
"grad_norm": 0.16950200498104095,
"learning_rate": 4.667056172167962e-06,
"loss": 0.001,
"step": 209
},
{
"epoch": 1.6569920844327177,
"grad_norm": 0.27843964099884033,
"learning_rate": 4.6610616374868066e-06,
"loss": 0.0037,
"step": 210
},
{
"epoch": 1.6649076517150396,
"grad_norm": 0.5085106492042542,
"learning_rate": 4.655017539713826e-06,
"loss": 0.0028,
"step": 211
},
{
"epoch": 1.6728232189973615,
"grad_norm": 0.28183189034461975,
"learning_rate": 4.648924017468003e-06,
"loss": 0.0003,
"step": 212
},
{
"epoch": 1.6807387862796834,
"grad_norm": 0.38587337732315063,
"learning_rate": 4.642781210501858e-06,
"loss": 0.0052,
"step": 213
},
{
"epoch": 1.6886543535620053,
"grad_norm": 0.4533829987049103,
"learning_rate": 4.6365892596982295e-06,
"loss": 0.0036,
"step": 214
},
{
"epoch": 1.6965699208443272,
"grad_norm": 0.9035217761993408,
"learning_rate": 4.6303483070670574e-06,
"loss": 0.0022,
"step": 215
},
{
"epoch": 1.7044854881266491,
"grad_norm": 0.545513391494751,
"learning_rate": 4.624058495742115e-06,
"loss": 0.0019,
"step": 216
},
{
"epoch": 1.712401055408971,
"grad_norm": 0.09330940991640091,
"learning_rate": 4.617719969977729e-06,
"loss": 0.0027,
"step": 217
},
{
"epoch": 1.720316622691293,
"grad_norm": 0.4759463667869568,
"learning_rate": 4.611332875145476e-06,
"loss": 0.002,
"step": 218
},
{
"epoch": 1.7282321899736148,
"grad_norm": 0.2812918722629547,
"learning_rate": 4.604897357730846e-06,
"loss": 0.0033,
"step": 219
},
{
"epoch": 1.7361477572559367,
"grad_norm": 0.5131822228431702,
"learning_rate": 4.598413565329876e-06,
"loss": 0.0019,
"step": 220
},
{
"epoch": 1.7440633245382586,
"grad_norm": 0.17534379661083221,
"learning_rate": 4.591881646645775e-06,
"loss": 0.0014,
"step": 221
},
{
"epoch": 1.7519788918205803,
"grad_norm": 0.2808787524700165,
"learning_rate": 4.585301751485508e-06,
"loss": 0.0011,
"step": 222
},
{
"epoch": 1.7598944591029024,
"grad_norm": 0.1664104461669922,
"learning_rate": 4.578674030756364e-06,
"loss": 0.0013,
"step": 223
},
{
"epoch": 1.767810026385224,
"grad_norm": 0.11474815756082535,
"learning_rate": 4.571998636462487e-06,
"loss": 0.0003,
"step": 224
},
{
"epoch": 1.7757255936675462,
"grad_norm": 0.2994535267353058,
"learning_rate": 4.5652757217014e-06,
"loss": 0.0043,
"step": 225
},
{
"epoch": 1.783641160949868,
"grad_norm": 0.2962826192378998,
"learning_rate": 4.5585054406604865e-06,
"loss": 0.0012,
"step": 226
},
{
"epoch": 1.79155672823219,
"grad_norm": 0.10045085847377777,
"learning_rate": 4.551687948613459e-06,
"loss": 0.0016,
"step": 227
},
{
"epoch": 1.7994722955145117,
"grad_norm": 0.8167795538902283,
"learning_rate": 4.544823401916794e-06,
"loss": 0.0024,
"step": 228
},
{
"epoch": 1.8073878627968338,
"grad_norm": 0.562492847442627,
"learning_rate": 4.537911958006149e-06,
"loss": 0.0029,
"step": 229
},
{
"epoch": 1.8153034300791555,
"grad_norm": 0.12345612794160843,
"learning_rate": 4.530953775392749e-06,
"loss": 0.0006,
"step": 230
},
{
"epoch": 1.8232189973614776,
"grad_norm": 0.1261640191078186,
"learning_rate": 4.523949013659754e-06,
"loss": 0.0006,
"step": 231
},
{
"epoch": 1.8311345646437993,
"grad_norm": 0.2126503586769104,
"learning_rate": 4.5168978334585955e-06,
"loss": 0.0005,
"step": 232
},
{
"epoch": 1.8390501319261214,
"grad_norm": 0.23717330396175385,
"learning_rate": 4.509800396505298e-06,
"loss": 0.0019,
"step": 233
},
{
"epoch": 1.8469656992084431,
"grad_norm": 0.2814564108848572,
"learning_rate": 4.502656865576762e-06,
"loss": 0.0019,
"step": 234
},
{
"epoch": 1.8548812664907652,
"grad_norm": 0.12869513034820557,
"learning_rate": 4.495467404507039e-06,
"loss": 0.0007,
"step": 235
},
{
"epoch": 1.862796833773087,
"grad_norm": 0.45971599221229553,
"learning_rate": 4.4882321781835666e-06,
"loss": 0.0007,
"step": 236
},
{
"epoch": 1.870712401055409,
"grad_norm": 0.15887171030044556,
"learning_rate": 4.4809513525433925e-06,
"loss": 0.0009,
"step": 237
},
{
"epoch": 1.8786279683377307,
"grad_norm": 0.26040664315223694,
"learning_rate": 4.473625094569366e-06,
"loss": 0.0005,
"step": 238
},
{
"epoch": 1.8865435356200528,
"grad_norm": 0.6354907155036926,
"learning_rate": 4.466253572286308e-06,
"loss": 0.0012,
"step": 239
},
{
"epoch": 1.8944591029023745,
"grad_norm": 0.5247849225997925,
"learning_rate": 4.458836954757161e-06,
"loss": 0.0018,
"step": 240
},
{
"epoch": 1.9023746701846966,
"grad_norm": 0.14983093738555908,
"learning_rate": 4.4513754120791065e-06,
"loss": 0.0004,
"step": 241
},
{
"epoch": 1.9102902374670183,
"grad_norm": 0.1560702919960022,
"learning_rate": 4.443869115379667e-06,
"loss": 0.0011,
"step": 242
},
{
"epoch": 1.9182058047493404,
"grad_norm": 0.5521268248558044,
"learning_rate": 4.436318236812782e-06,
"loss": 0.0002,
"step": 243
},
{
"epoch": 1.9261213720316621,
"grad_norm": 0.34709402918815613,
"learning_rate": 4.428722949554858e-06,
"loss": 0.0009,
"step": 244
},
{
"epoch": 1.9340369393139842,
"grad_norm": 0.4721614122390747,
"learning_rate": 4.421083427800795e-06,
"loss": 0.0014,
"step": 245
},
{
"epoch": 1.941952506596306,
"grad_norm": 0.12338114529848099,
"learning_rate": 4.413399846759998e-06,
"loss": 0.0005,
"step": 246
},
{
"epoch": 1.949868073878628,
"grad_norm": 0.2648129463195801,
"learning_rate": 4.405672382652349e-06,
"loss": 0.0035,
"step": 247
},
{
"epoch": 1.9577836411609497,
"grad_norm": 0.41379258036613464,
"learning_rate": 4.397901212704176e-06,
"loss": 0.0001,
"step": 248
},
{
"epoch": 1.9656992084432718,
"grad_norm": 0.04220602661371231,
"learning_rate": 4.390086515144179e-06,
"loss": 0.0002,
"step": 249
},
{
"epoch": 1.9736147757255935,
"grad_norm": 0.04110397771000862,
"learning_rate": 4.38222846919935e-06,
"loss": 0.0001,
"step": 250
},
{
"epoch": 1.9815303430079156,
"grad_norm": 0.027100196108222008,
"learning_rate": 4.3743272550908545e-06,
"loss": 0.0001,
"step": 251
},
{
"epoch": 1.9894459102902373,
"grad_norm": 0.0841871052980423,
"learning_rate": 4.366383054029907e-06,
"loss": 0.0002,
"step": 252
},
{
"epoch": 1.9973614775725594,
"grad_norm": 0.038113586604595184,
"learning_rate": 4.358396048213609e-06,
"loss": 0.0002,
"step": 253
},
{
"epoch": 2.0,
"grad_norm": 0.038113586604595184,
"learning_rate": 4.350366420820771e-06,
"loss": 0.0001,
"step": 254
},
{
"epoch": 2.007915567282322,
"grad_norm": 0.1468001902103424,
"learning_rate": 4.342294356007715e-06,
"loss": 0.0003,
"step": 255
},
{
"epoch": 2.015831134564644,
"grad_norm": 0.1259719580411911,
"learning_rate": 4.3341800389040465e-06,
"loss": 0.0003,
"step": 256
},
{
"epoch": 2.0237467018469655,
"grad_norm": 0.042028915137052536,
"learning_rate": 4.326023655608412e-06,
"loss": 0.0,
"step": 257
},
{
"epoch": 2.0316622691292876,
"grad_norm": 0.007664548233151436,
"learning_rate": 4.317825393184226e-06,
"loss": 0.0,
"step": 258
},
{
"epoch": 2.0395778364116097,
"grad_norm": 0.007942886091768742,
"learning_rate": 4.30958543965539e-06,
"loss": 0.0,
"step": 259
},
{
"epoch": 2.0474934036939314,
"grad_norm": 0.010280617512762547,
"learning_rate": 4.3013039840019675e-06,
"loss": 0.0,
"step": 260
},
{
"epoch": 2.055408970976253,
"grad_norm": 0.007112486753612757,
"learning_rate": 4.292981216155864e-06,
"loss": 0.0,
"step": 261
},
{
"epoch": 2.063324538258575,
"grad_norm": 0.013413694687187672,
"learning_rate": 4.284617326996458e-06,
"loss": 0.0,
"step": 262
},
{
"epoch": 2.0712401055408973,
"grad_norm": 0.0064396364614367485,
"learning_rate": 4.276212508346232e-06,
"loss": 0.0,
"step": 263
},
{
"epoch": 2.079155672823219,
"grad_norm": 0.008679047226905823,
"learning_rate": 4.267766952966369e-06,
"loss": 0.0002,
"step": 264
},
{
"epoch": 2.0870712401055407,
"grad_norm": 0.06370187550783157,
"learning_rate": 4.2592808545523335e-06,
"loss": 0.0007,
"step": 265
},
{
"epoch": 2.094986807387863,
"grad_norm": 0.3114399313926697,
"learning_rate": 4.250754407729428e-06,
"loss": 0.0009,
"step": 266
},
{
"epoch": 2.1029023746701845,
"grad_norm": 0.38784605264663696,
"learning_rate": 4.242187808048329e-06,
"loss": 0.0,
"step": 267
},
{
"epoch": 2.1108179419525066,
"grad_norm": 0.013528178445994854,
"learning_rate": 4.233581251980604e-06,
"loss": 0.0009,
"step": 268
},
{
"epoch": 2.1187335092348283,
"grad_norm": 0.7630422711372375,
"learning_rate": 4.224934936914206e-06,
"loss": 0.0002,
"step": 269
},
{
"epoch": 2.1266490765171504,
"grad_norm": 0.06802177429199219,
"learning_rate": 4.216249061148939e-06,
"loss": 0.0,
"step": 270
},
{
"epoch": 2.1345646437994725,
"grad_norm": 0.015802081674337387,
"learning_rate": 4.207523823891924e-06,
"loss": 0.0008,
"step": 271
},
{
"epoch": 2.142480211081794,
"grad_norm": 0.6553952693939209,
"learning_rate": 4.198759425253015e-06,
"loss": 0.0,
"step": 272
},
{
"epoch": 2.150395778364116,
"grad_norm": 0.007900790311396122,
"learning_rate": 4.1899560662402204e-06,
"loss": 0.0001,
"step": 273
},
{
"epoch": 2.158311345646438,
"grad_norm": 0.058594342321157455,
"learning_rate": 4.18111394875509e-06,
"loss": 0.0,
"step": 274
},
{
"epoch": 2.16622691292876,
"grad_norm": 0.008023036643862724,
"learning_rate": 4.172233275588082e-06,
"loss": 0.0001,
"step": 275
},
{
"epoch": 2.174142480211082,
"grad_norm": 0.06547389924526215,
"learning_rate": 4.163314250413913e-06,
"loss": 0.0,
"step": 276
},
{
"epoch": 2.1820580474934035,
"grad_norm": 0.007858374156057835,
"learning_rate": 4.154357077786892e-06,
"loss": 0.0006,
"step": 277
},
{
"epoch": 2.1899736147757256,
"grad_norm": 0.2166021764278412,
"learning_rate": 4.145361963136219e-06,
"loss": 0.0,
"step": 278
},
{
"epoch": 2.1978891820580473,
"grad_norm": 0.36737874150276184,
"learning_rate": 4.136329112761285e-06,
"loss": 0.0004,
"step": 279
},
{
"epoch": 2.2058047493403694,
"grad_norm": 0.07245466858148575,
"learning_rate": 4.127258733826929e-06,
"loss": 0.0,
"step": 280
},
{
"epoch": 2.213720316622691,
"grad_norm": 0.029037440195679665,
"learning_rate": 4.118151034358696e-06,
"loss": 0.0,
"step": 281
},
{
"epoch": 2.221635883905013,
"grad_norm": 0.06319958716630936,
"learning_rate": 4.109006223238064e-06,
"loss": 0.0002,
"step": 282
},
{
"epoch": 2.229551451187335,
"grad_norm": 0.04793912172317505,
"learning_rate": 4.0998245101976495e-06,
"loss": 0.0,
"step": 283
},
{
"epoch": 2.237467018469657,
"grad_norm": 0.006495122332125902,
"learning_rate": 4.0906061058164e-06,
"loss": 0.0012,
"step": 284
},
{
"epoch": 2.2453825857519787,
"grad_norm": 0.14944440126419067,
"learning_rate": 4.081351221514765e-06,
"loss": 0.0001,
"step": 285
},
{
"epoch": 2.253298153034301,
"grad_norm": 0.05634067952632904,
"learning_rate": 4.072060069549848e-06,
"loss": 0.0,
"step": 286
},
{
"epoch": 2.261213720316623,
"grad_norm": 0.003317984053865075,
"learning_rate": 4.062732863010534e-06,
"loss": 0.0008,
"step": 287
},
{
"epoch": 2.2691292875989446,
"grad_norm": 0.10713385790586472,
"learning_rate": 4.053369815812608e-06,
"loss": 0.0001,
"step": 288
},
{
"epoch": 2.2770448548812663,
"grad_norm": 0.026916418224573135,
"learning_rate": 4.043971142693845e-06,
"loss": 0.0,
"step": 289
},
{
"epoch": 2.2849604221635884,
"grad_norm": 0.16971777379512787,
"learning_rate": 4.034537059209085e-06,
"loss": 0.0012,
"step": 290
},
{
"epoch": 2.2928759894459105,
"grad_norm": 0.2465554028749466,
"learning_rate": 4.025067781725294e-06,
"loss": 0.0009,
"step": 291
},
{
"epoch": 2.300791556728232,
"grad_norm": 0.017931131646037102,
"learning_rate": 4.015563527416596e-06,
"loss": 0.0019,
"step": 292
},
{
"epoch": 2.308707124010554,
"grad_norm": 0.5008033514022827,
"learning_rate": 4.006024514259295e-06,
"loss": 0.0016,
"step": 293
},
{
"epoch": 2.316622691292876,
"grad_norm": 0.2697279155254364,
"learning_rate": 3.996450961026876e-06,
"loss": 0.0003,
"step": 294
},
{
"epoch": 2.324538258575198,
"grad_norm": 0.04123712331056595,
"learning_rate": 3.986843087284986e-06,
"loss": 0.0,
"step": 295
},
{
"epoch": 2.33245382585752,
"grad_norm": 0.018080947920680046,
"learning_rate": 3.977201113386402e-06,
"loss": 0.0001,
"step": 296
},
{
"epoch": 2.3403693931398415,
"grad_norm": 0.1296033263206482,
"learning_rate": 3.96752526046597e-06,
"loss": 0.0022,
"step": 297
},
{
"epoch": 2.3482849604221636,
"grad_norm": 0.053319524973630905,
"learning_rate": 3.957815750435542e-06,
"loss": 0.0008,
"step": 298
},
{
"epoch": 2.3562005277044857,
"grad_norm": 0.48205140233039856,
"learning_rate": 3.948072805978879e-06,
"loss": 0.0001,
"step": 299
},
{
"epoch": 2.3641160949868074,
"grad_norm": 0.031876154243946075,
"learning_rate": 3.938296650546552e-06,
"loss": 0.0001,
"step": 300
},
{
"epoch": 2.372031662269129,
"grad_norm": 0.03231018781661987,
"learning_rate": 3.928487508350808e-06,
"loss": 0.0002,
"step": 301
},
{
"epoch": 2.379947229551451,
"grad_norm": 0.04064059630036354,
"learning_rate": 3.918645604360434e-06,
"loss": 0.0001,
"step": 302
},
{
"epoch": 2.387862796833773,
"grad_norm": 0.04606090486049652,
"learning_rate": 3.908771164295595e-06,
"loss": 0.0002,
"step": 303
},
{
"epoch": 2.395778364116095,
"grad_norm": 0.031620122492313385,
"learning_rate": 3.898864414622661e-06,
"loss": 0.0001,
"step": 304
},
{
"epoch": 2.4036939313984167,
"grad_norm": 0.03372185304760933,
"learning_rate": 3.888925582549006e-06,
"loss": 0.0001,
"step": 305
},
{
"epoch": 2.411609498680739,
"grad_norm": 0.025314636528491974,
"learning_rate": 3.878954896017804e-06,
"loss": 0.0003,
"step": 306
},
{
"epoch": 2.4195250659630605,
"grad_norm": 0.07136713713407516,
"learning_rate": 3.868952583702798e-06,
"loss": 0.0017,
"step": 307
},
{
"epoch": 2.4274406332453826,
"grad_norm": 0.1380709558725357,
"learning_rate": 3.858918875003053e-06,
"loss": 0.0001,
"step": 308
},
{
"epoch": 2.4353562005277043,
"grad_norm": 0.020328925922513008,
"learning_rate": 3.848854000037702e-06,
"loss": 0.0001,
"step": 309
},
{
"epoch": 2.4432717678100264,
"grad_norm": 0.08837846666574478,
"learning_rate": 3.83875818964066e-06,
"loss": 0.0001,
"step": 310
},
{
"epoch": 2.451187335092348,
"grad_norm": 0.03620443120598793,
"learning_rate": 3.828631675355338e-06,
"loss": 0.0001,
"step": 311
},
{
"epoch": 2.45910290237467,
"grad_norm": 0.01960093528032303,
"learning_rate": 3.818474689429324e-06,
"loss": 0.0001,
"step": 312
},
{
"epoch": 2.467018469656992,
"grad_norm": 0.1628216803073883,
"learning_rate": 3.808287464809063e-06,
"loss": 0.0025,
"step": 313
},
{
"epoch": 2.474934036939314,
"grad_norm": 0.19068323075771332,
"learning_rate": 3.7980702351345146e-06,
"loss": 0.0002,
"step": 314
},
{
"epoch": 2.4828496042216357,
"grad_norm": 0.06465128064155579,
"learning_rate": 3.787823234733788e-06,
"loss": 0.0001,
"step": 315
},
{
"epoch": 2.490765171503958,
"grad_norm": 0.016947569325566292,
"learning_rate": 3.7775466986177763e-06,
"loss": 0.0001,
"step": 316
},
{
"epoch": 2.4986807387862795,
"grad_norm": 0.010782090947031975,
"learning_rate": 3.7672408624747598e-06,
"loss": 0.0,
"step": 317
},
{
"epoch": 2.5065963060686016,
"grad_norm": 0.02703935280442238,
"learning_rate": 3.756905962665005e-06,
"loss": 0.0001,
"step": 318
},
{
"epoch": 2.5145118733509237,
"grad_norm": 0.010577079840004444,
"learning_rate": 3.7465422362153416e-06,
"loss": 0.0,
"step": 319
},
{
"epoch": 2.5224274406332454,
"grad_norm": 0.006201338488608599,
"learning_rate": 3.736149920813726e-06,
"loss": 0.0,
"step": 320
},
{
"epoch": 2.530343007915567,
"grad_norm": 0.12876874208450317,
"learning_rate": 3.7257292548037917e-06,
"loss": 0.0009,
"step": 321
},
{
"epoch": 2.538258575197889,
"grad_norm": 0.012635966762900352,
"learning_rate": 3.715280477179382e-06,
"loss": 0.0,
"step": 322
},
{
"epoch": 2.5461741424802113,
"grad_norm": 0.0092966603115201,
"learning_rate": 3.7048038275790695e-06,
"loss": 0.0,
"step": 323
},
{
"epoch": 2.554089709762533,
"grad_norm": 0.013077512383460999,
"learning_rate": 3.6942995462806574e-06,
"loss": 0.0,
"step": 324
},
{
"epoch": 2.5620052770448547,
"grad_norm": 0.01774498261511326,
"learning_rate": 3.6837678741956747e-06,
"loss": 0.0001,
"step": 325
},
{
"epoch": 2.569920844327177,
"grad_norm": 0.021662306040525436,
"learning_rate": 3.6732090528638432e-06,
"loss": 0.0,
"step": 326
},
{
"epoch": 2.577836411609499,
"grad_norm": 0.005042460281401873,
"learning_rate": 3.6626233244475445e-06,
"loss": 0.0001,
"step": 327
},
{
"epoch": 2.5857519788918206,
"grad_norm": 0.07442116737365723,
"learning_rate": 3.6520109317262624e-06,
"loss": 0.0,
"step": 328
},
{
"epoch": 2.5936675461741423,
"grad_norm": 0.005822429433465004,
"learning_rate": 3.6413721180910165e-06,
"loss": 0.0,
"step": 329
},
{
"epoch": 2.6015831134564644,
"grad_norm": 0.0028016779106110334,
"learning_rate": 3.6307071275387807e-06,
"loss": 0.0,
"step": 330
},
{
"epoch": 2.6094986807387865,
"grad_norm": 0.005941660143435001,
"learning_rate": 3.6200162046668826e-06,
"loss": 0.0,
"step": 331
},
{
"epoch": 2.6174142480211082,
"grad_norm": 0.007512587122619152,
"learning_rate": 3.6092995946673996e-06,
"loss": 0.0013,
"step": 332
},
{
"epoch": 2.62532981530343,
"grad_norm": 0.1878112256526947,
"learning_rate": 3.5985575433215345e-06,
"loss": 0.0,
"step": 333
},
{
"epoch": 2.633245382585752,
"grad_norm": 0.24462495744228363,
"learning_rate": 3.587790296993976e-06,
"loss": 0.0024,
"step": 334
},
{
"epoch": 2.641160949868074,
"grad_norm": 0.010779382660984993,
"learning_rate": 3.5769981026272477e-06,
"loss": 0.0001,
"step": 335
},
{
"epoch": 2.649076517150396,
"grad_norm": 0.024886395782232285,
"learning_rate": 3.5661812077360496e-06,
"loss": 0.0,
"step": 336
},
{
"epoch": 2.6569920844327175,
"grad_norm": 0.27707141637802124,
"learning_rate": 3.5553398604015777e-06,
"loss": 0.0014,
"step": 337
},
{
"epoch": 2.6649076517150396,
"grad_norm": 0.006285437382757664,
"learning_rate": 3.544474309265834e-06,
"loss": 0.0,
"step": 338
},
{
"epoch": 2.6728232189973617,
"grad_norm": 0.006901748012751341,
"learning_rate": 3.5335848035259257e-06,
"loss": 0.0,
"step": 339
},
{
"epoch": 2.6807387862796834,
"grad_norm": 0.005699906963855028,
"learning_rate": 3.5226715929283507e-06,
"loss": 0.0,
"step": 340
},
{
"epoch": 2.688654353562005,
"grad_norm": 0.010477579198777676,
"learning_rate": 3.511734927763265e-06,
"loss": 0.0,
"step": 341
},
{
"epoch": 2.6965699208443272,
"grad_norm": 0.0179861169308424,
"learning_rate": 3.5007750588587495e-06,
"loss": 0.0007,
"step": 342
},
{
"epoch": 2.7044854881266494,
"grad_norm": 0.1684534102678299,
"learning_rate": 3.4897922375750517e-06,
"loss": 0.0001,
"step": 343
},
{
"epoch": 2.712401055408971,
"grad_norm": 0.05259430781006813,
"learning_rate": 3.478786715798823e-06,
"loss": 0.0001,
"step": 344
},
{
"epoch": 2.7203166226912927,
"grad_norm": 0.6708176732063293,
"learning_rate": 3.4677587459373417e-06,
"loss": 0.0003,
"step": 345
},
{
"epoch": 2.728232189973615,
"grad_norm": 0.02362961135804653,
"learning_rate": 3.4567085809127247e-06,
"loss": 0.0,
"step": 346
},
{
"epoch": 2.736147757255937,
"grad_norm": 0.008437946438789368,
"learning_rate": 3.4456364741561256e-06,
"loss": 0.0,
"step": 347
},
{
"epoch": 2.7440633245382586,
"grad_norm": 0.011274375952780247,
"learning_rate": 3.434542679601922e-06,
"loss": 0.0023,
"step": 348
},
{
"epoch": 2.7519788918205803,
"grad_norm": 0.2040291279554367,
"learning_rate": 3.423427451681895e-06,
"loss": 0.0001,
"step": 349
},
{
"epoch": 2.7598944591029024,
"grad_norm": 0.01595059037208557,
"learning_rate": 3.4122910453193885e-06,
"loss": 0.0001,
"step": 350
},
{
"epoch": 2.767810026385224,
"grad_norm": 0.02352329157292843,
"learning_rate": 3.4011337159234674e-06,
"loss": 0.0012,
"step": 351
},
{
"epoch": 2.7757255936675462,
"grad_norm": 0.10779980570077896,
"learning_rate": 3.3899557193830585e-06,
"loss": 0.0003,
"step": 352
},
{
"epoch": 2.783641160949868,
"grad_norm": 0.08968038111925125,
"learning_rate": 3.3787573120610794e-06,
"loss": 0.0001,
"step": 353
},
{
"epoch": 2.79155672823219,
"grad_norm": 0.009699780493974686,
"learning_rate": 3.367538750788563e-06,
"loss": 0.0,
"step": 354
},
{
"epoch": 2.7994722955145117,
"grad_norm": 0.01223874930292368,
"learning_rate": 3.356300292858763e-06,
"loss": 0.0001,
"step": 355
},
{
"epoch": 2.807387862796834,
"grad_norm": 0.015586338005959988,
"learning_rate": 3.345042196021257e-06,
"loss": 0.0001,
"step": 356
},
{
"epoch": 2.8153034300791555,
"grad_norm": 0.10576637834310532,
"learning_rate": 3.333764718476032e-06,
"loss": 0.0011,
"step": 357
},
{
"epoch": 2.8232189973614776,
"grad_norm": 0.010167757980525494,
"learning_rate": 3.3224681188675643e-06,
"loss": 0.0,
"step": 358
},
{
"epoch": 2.8311345646437993,
"grad_norm": 0.09357836842536926,
"learning_rate": 3.3111526562788864e-06,
"loss": 0.0003,
"step": 359
},
{
"epoch": 2.8390501319261214,
"grad_norm": 0.013279908336699009,
"learning_rate": 3.2998185902256475e-06,
"loss": 0.0001,
"step": 360
},
{
"epoch": 2.846965699208443,
"grad_norm": 0.030286984518170357,
"learning_rate": 3.2884661806501576e-06,
"loss": 0.0001,
"step": 361
},
{
"epoch": 2.8548812664907652,
"grad_norm": 0.01646578684449196,
"learning_rate": 3.2770956879154305e-06,
"loss": 0.0001,
"step": 362
},
{
"epoch": 2.862796833773087,
"grad_norm": 0.018039187416434288,
"learning_rate": 3.2657073727992078e-06,
"loss": 0.0,
"step": 363
},
{
"epoch": 2.870712401055409,
"grad_norm": 0.009995294734835625,
"learning_rate": 3.2543014964879814e-06,
"loss": 0.0,
"step": 364
},
{
"epoch": 2.8786279683377307,
"grad_norm": 0.01269359327852726,
"learning_rate": 3.2428783205710023e-06,
"loss": 0.0001,
"step": 365
},
{
"epoch": 2.886543535620053,
"grad_norm": 0.01195498462766409,
"learning_rate": 3.2314381070342815e-06,
"loss": 0.0,
"step": 366
},
{
"epoch": 2.8944591029023745,
"grad_norm": 0.009492097422480583,
"learning_rate": 3.21998111825458e-06,
"loss": 0.0001,
"step": 367
},
{
"epoch": 2.9023746701846966,
"grad_norm": 0.012942278757691383,
"learning_rate": 3.208507616993393e-06,
"loss": 0.0001,
"step": 368
},
{
"epoch": 2.9102902374670183,
"grad_norm": 0.014576618559658527,
"learning_rate": 3.1970178663909233e-06,
"loss": 0.0,
"step": 369
},
{
"epoch": 2.9182058047493404,
"grad_norm": 0.014812547713518143,
"learning_rate": 3.1855121299600454e-06,
"loss": 0.0,
"step": 370
},
{
"epoch": 2.926121372031662,
"grad_norm": 0.005972759798169136,
"learning_rate": 3.173990671580263e-06,
"loss": 0.0008,
"step": 371
},
{
"epoch": 2.9340369393139842,
"grad_norm": 0.16172465682029724,
"learning_rate": 3.162453755491655e-06,
"loss": 0.0,
"step": 372
},
{
"epoch": 2.941952506596306,
"grad_norm": 0.005183232482522726,
"learning_rate": 3.1509016462888175e-06,
"loss": 0.0,
"step": 373
},
{
"epoch": 2.949868073878628,
"grad_norm": 0.013237228617072105,
"learning_rate": 3.139334608914795e-06,
"loss": 0.0001,
"step": 374
},
{
"epoch": 2.9577836411609497,
"grad_norm": 0.003913429565727711,
"learning_rate": 3.1277529086550044e-06,
"loss": 0.0,
"step": 375
},
{
"epoch": 2.965699208443272,
"grad_norm": 0.123182512819767,
"learning_rate": 3.1161568111311487e-06,
"loss": 0.0002,
"step": 376
},
{
"epoch": 2.9736147757255935,
"grad_norm": 0.007754236459732056,
"learning_rate": 3.1045465822951265e-06,
"loss": 0.0,
"step": 377
},
{
"epoch": 2.9815303430079156,
"grad_norm": 0.003976548556238413,
"learning_rate": 3.092922488422933e-06,
"loss": 0.0,
"step": 378
},
{
"epoch": 2.9894459102902373,
"grad_norm": 0.0028897474985569715,
"learning_rate": 3.0812847961085527e-06,
"loss": 0.0,
"step": 379
},
{
"epoch": 2.9973614775725594,
"grad_norm": 0.0054747299291193485,
"learning_rate": 3.0696337722578444e-06,
"loss": 0.0,
"step": 380
},
{
"epoch": 3.0,
"grad_norm": 0.005969229619950056,
"learning_rate": 3.057969684082421e-06,
"loss": 0.0,
"step": 381
},
{
"epoch": 3.007915567282322,
"grad_norm": 0.004056141711771488,
"learning_rate": 3.04629279909352e-06,
"loss": 0.0,
"step": 382
},
{
"epoch": 3.015831134564644,
"grad_norm": 0.006289948709309101,
"learning_rate": 3.0346033850958685e-06,
"loss": 0.0,
"step": 383
},
{
"epoch": 3.0237467018469655,
"grad_norm": 0.06978850066661835,
"learning_rate": 3.0229017101815424e-06,
"loss": 0.0001,
"step": 384
},
{
"epoch": 3.0316622691292876,
"grad_norm": 0.003850232344120741,
"learning_rate": 3.011188042723816e-06,
"loss": 0.0,
"step": 385
},
{
"epoch": 3.0395778364116097,
"grad_norm": 0.006014230661094189,
"learning_rate": 2.9994626513710085e-06,
"loss": 0.0,
"step": 386
},
{
"epoch": 3.0474934036939314,
"grad_norm": 0.008263601921498775,
"learning_rate": 2.9877258050403214e-06,
"loss": 0.0,
"step": 387
},
{
"epoch": 3.055408970976253,
"grad_norm": 0.0018157489830628037,
"learning_rate": 2.975977772911671e-06,
"loss": 0.0,
"step": 388
},
{
"epoch": 3.063324538258575,
"grad_norm": 0.0012247132835909724,
"learning_rate": 2.964218824421518e-06,
"loss": 0.0,
"step": 389
},
{
"epoch": 3.0712401055408973,
"grad_norm": 0.005896579474210739,
"learning_rate": 2.9524492292566824e-06,
"loss": 0.0,
"step": 390
},
{
"epoch": 3.079155672823219,
"grad_norm": 0.0017256075516343117,
"learning_rate": 2.9406692573481634e-06,
"loss": 0.0,
"step": 391
},
{
"epoch": 3.0870712401055407,
"grad_norm": 0.001976009923964739,
"learning_rate": 2.928879178864946e-06,
"loss": 0.0,
"step": 392
},
{
"epoch": 3.094986807387863,
"grad_norm": 0.31570467352867126,
"learning_rate": 2.9170792642078057e-06,
"loss": 0.0026,
"step": 393
},
{
"epoch": 3.1029023746701845,
"grad_norm": 0.004713984671980143,
"learning_rate": 2.9052697840031065e-06,
"loss": 0.0,
"step": 394
},
{
"epoch": 3.1108179419525066,
"grad_norm": 0.135233074426651,
"learning_rate": 2.8934510090965943e-06,
"loss": 0.0005,
"step": 395
},
{
"epoch": 3.1187335092348283,
"grad_norm": 0.0030627846717834473,
"learning_rate": 2.8816232105471864e-06,
"loss": 0.0,
"step": 396
},
{
"epoch": 3.1266490765171504,
"grad_norm": 0.005015626084059477,
"learning_rate": 2.8697866596207524e-06,
"loss": 0.0,
"step": 397
},
{
"epoch": 3.1345646437994725,
"grad_norm": 0.0035850070416927338,
"learning_rate": 2.8579416277838952e-06,
"loss": 0.0,
"step": 398
},
{
"epoch": 3.142480211081794,
"grad_norm": 0.007022330071777105,
"learning_rate": 2.846088386697723e-06,
"loss": 0.0,
"step": 399
},
{
"epoch": 3.150395778364116,
"grad_norm": 0.003523892490193248,
"learning_rate": 2.8342272082116214e-06,
"loss": 0.0,
"step": 400
},
{
"epoch": 3.158311345646438,
"grad_norm": 0.04955633729696274,
"learning_rate": 2.822358364357015e-06,
"loss": 0.0001,
"step": 401
},
{
"epoch": 3.16622691292876,
"grad_norm": 0.006145668216049671,
"learning_rate": 2.8104821273411333e-06,
"loss": 0.0,
"step": 402
},
{
"epoch": 3.174142480211082,
"grad_norm": 0.00394571665674448,
"learning_rate": 2.7985987695407618e-06,
"loss": 0.0,
"step": 403
},
{
"epoch": 3.1820580474934035,
"grad_norm": 0.003726326860487461,
"learning_rate": 2.786708563496002e-06,
"loss": 0.0,
"step": 404
},
{
"epoch": 3.1899736147757256,
"grad_norm": 0.003756965510547161,
"learning_rate": 2.774811781904013e-06,
"loss": 0.0,
"step": 405
},
{
"epoch": 3.1978891820580473,
"grad_norm": 0.008439401164650917,
"learning_rate": 2.762908697612765e-06,
"loss": 0.0,
"step": 406
},
{
"epoch": 3.2058047493403694,
"grad_norm": 0.0023035435006022453,
"learning_rate": 2.750999583614777e-06,
"loss": 0.0,
"step": 407
},
{
"epoch": 3.213720316622691,
"grad_norm": 0.003911241423338652,
"learning_rate": 2.739084713040856e-06,
"loss": 0.0,
"step": 408
},
{
"epoch": 3.221635883905013,
"grad_norm": 0.030907543376088142,
"learning_rate": 2.7271643591538355e-06,
"loss": 0.0001,
"step": 409
},
{
"epoch": 3.229551451187335,
"grad_norm": 0.004778213333338499,
"learning_rate": 2.7152387953423047e-06,
"loss": 0.0,
"step": 410
},
{
"epoch": 3.237467018469657,
"grad_norm": 0.007156203966587782,
"learning_rate": 2.703308295114342e-06,
"loss": 0.0,
"step": 411
},
{
"epoch": 3.2453825857519787,
"grad_norm": 0.003489116206765175,
"learning_rate": 2.69137313209124e-06,
"loss": 0.0,
"step": 412
},
{
"epoch": 3.253298153034301,
"grad_norm": 0.004960685037076473,
"learning_rate": 2.6794335800012294e-06,
"loss": 0.0,
"step": 413
},
{
"epoch": 3.261213720316623,
"grad_norm": 0.004319795873016119,
"learning_rate": 2.6674899126732045e-06,
"loss": 0.0,
"step": 414
},
{
"epoch": 3.2691292875989446,
"grad_norm": 0.007416910026222467,
"learning_rate": 2.65554240403044e-06,
"loss": 0.0,
"step": 415
},
{
"epoch": 3.2770448548812663,
"grad_norm": 0.00757510494440794,
"learning_rate": 2.643591328084309e-06,
"loss": 0.0,
"step": 416
},
{
"epoch": 3.2849604221635884,
"grad_norm": 0.006031471770256758,
"learning_rate": 2.631636958928e-06,
"loss": 0.0,
"step": 417
},
{
"epoch": 3.2928759894459105,
"grad_norm": 0.025160983204841614,
"learning_rate": 2.6196795707302304e-06,
"loss": 0.0001,
"step": 418
},
{
"epoch": 3.300791556728232,
"grad_norm": 0.002678148215636611,
"learning_rate": 2.607719437728957e-06,
"loss": 0.0,
"step": 419
},
{
"epoch": 3.308707124010554,
"grad_norm": 0.0023166383616626263,
"learning_rate": 2.595756834225089e-06,
"loss": 0.0,
"step": 420
},
{
"epoch": 3.316622691292876,
"grad_norm": 0.004792348947376013,
"learning_rate": 2.583792034576194e-06,
"loss": 0.0,
"step": 421
},
{
"epoch": 3.324538258575198,
"grad_norm": 0.002316155703738332,
"learning_rate": 2.5718253131902084e-06,
"loss": 0.0,
"step": 422
},
{
"epoch": 3.33245382585752,
"grad_norm": 0.002854425460100174,
"learning_rate": 2.5598569445191418e-06,
"loss": 0.0,
"step": 423
},
{
"epoch": 3.3403693931398415,
"grad_norm": 0.0025122410152107477,
"learning_rate": 2.547887203052786e-06,
"loss": 0.0,
"step": 424
},
{
"epoch": 3.3482849604221636,
"grad_norm": 0.007665908429771662,
"learning_rate": 2.535916363312414e-06,
"loss": 0.0,
"step": 425
},
{
"epoch": 3.3562005277044857,
"grad_norm": 0.0029636449180543423,
"learning_rate": 2.52394469984449e-06,
"loss": 0.0,
"step": 426
},
{
"epoch": 3.3641160949868074,
"grad_norm": 0.003854286391288042,
"learning_rate": 2.5119724872143693e-06,
"loss": 0.0,
"step": 427
},
{
"epoch": 3.372031662269129,
"grad_norm": 0.01237494871020317,
"learning_rate": 2.5e-06,
"loss": 0.0,
"step": 428
},
{
"epoch": 3.379947229551451,
"grad_norm": 0.005211927928030491,
"learning_rate": 2.4880275127856324e-06,
"loss": 0.0,
"step": 429
},
{
"epoch": 3.387862796833773,
"grad_norm": 0.008481196127831936,
"learning_rate": 2.4760553001555105e-06,
"loss": 0.0,
"step": 430
},
{
"epoch": 3.395778364116095,
"grad_norm": 0.0037402070593088865,
"learning_rate": 2.4640836366875872e-06,
"loss": 0.0,
"step": 431
},
{
"epoch": 3.4036939313984167,
"grad_norm": 0.002274413825944066,
"learning_rate": 2.452112796947215e-06,
"loss": 0.0,
"step": 432
},
{
"epoch": 3.411609498680739,
"grad_norm": 0.003822160651907325,
"learning_rate": 2.440143055480859e-06,
"loss": 0.0,
"step": 433
},
{
"epoch": 3.4195250659630605,
"grad_norm": 0.0016057207249104977,
"learning_rate": 2.428174686809793e-06,
"loss": 0.0,
"step": 434
},
{
"epoch": 3.4274406332453826,
"grad_norm": 0.0016935165040194988,
"learning_rate": 2.4162079654238073e-06,
"loss": 0.0,
"step": 435
},
{
"epoch": 3.4353562005277043,
"grad_norm": 0.0018719176296144724,
"learning_rate": 2.404243165774912e-06,
"loss": 0.0,
"step": 436
},
{
"epoch": 3.4432717678100264,
"grad_norm": 0.0013446552911773324,
"learning_rate": 2.392280562271044e-06,
"loss": 0.0,
"step": 437
},
{
"epoch": 3.451187335092348,
"grad_norm": 0.0062268865294754505,
"learning_rate": 2.3803204292697705e-06,
"loss": 0.0,
"step": 438
},
{
"epoch": 3.45910290237467,
"grad_norm": 0.0036832697223871946,
"learning_rate": 2.3683630410720013e-06,
"loss": 0.0,
"step": 439
},
{
"epoch": 3.467018469656992,
"grad_norm": 0.0031680618412792683,
"learning_rate": 2.356408671915692e-06,
"loss": 0.0,
"step": 440
},
{
"epoch": 3.474934036939314,
"grad_norm": 0.00204239459708333,
"learning_rate": 2.3444575959695615e-06,
"loss": 0.0,
"step": 441
},
{
"epoch": 3.4828496042216357,
"grad_norm": 0.002014332450926304,
"learning_rate": 2.3325100873267963e-06,
"loss": 0.0,
"step": 442
},
{
"epoch": 3.490765171503958,
"grad_norm": 0.004848967771977186,
"learning_rate": 2.320566419998772e-06,
"loss": 0.0,
"step": 443
},
{
"epoch": 3.4986807387862795,
"grad_norm": 0.0017829296411946416,
"learning_rate": 2.308626867908761e-06,
"loss": 0.0,
"step": 444
},
{
"epoch": 3.5065963060686016,
"grad_norm": 0.046102311462163925,
"learning_rate": 2.296691704885659e-06,
"loss": 0.0001,
"step": 445
},
{
"epoch": 3.5145118733509237,
"grad_norm": 0.0010413214331492782,
"learning_rate": 2.284761204657696e-06,
"loss": 0.0,
"step": 446
},
{
"epoch": 3.5224274406332454,
"grad_norm": 0.0014018567744642496,
"learning_rate": 2.2728356408461653e-06,
"loss": 0.0,
"step": 447
},
{
"epoch": 3.530343007915567,
"grad_norm": 0.001453349948860705,
"learning_rate": 2.2609152869591445e-06,
"loss": 0.0,
"step": 448
},
{
"epoch": 3.538258575197889,
"grad_norm": 0.0017676043789833784,
"learning_rate": 2.249000416385224e-06,
"loss": 0.0,
"step": 449
},
{
"epoch": 3.5461741424802113,
"grad_norm": 0.002675161464139819,
"learning_rate": 2.2370913023872357e-06,
"loss": 0.0,
"step": 450
},
{
"epoch": 3.554089709762533,
"grad_norm": 0.0015697075286880136,
"learning_rate": 2.2251882180959876e-06,
"loss": 0.0,
"step": 451
},
{
"epoch": 3.5620052770448547,
"grad_norm": 0.0011259898310527205,
"learning_rate": 2.2132914365039993e-06,
"loss": 0.0,
"step": 452
},
{
"epoch": 3.569920844327177,
"grad_norm": 0.002902733162045479,
"learning_rate": 2.201401230459239e-06,
"loss": 0.0,
"step": 453
},
{
"epoch": 3.577836411609499,
"grad_norm": 0.0017832021694630384,
"learning_rate": 2.189517872658867e-06,
"loss": 0.0,
"step": 454
},
{
"epoch": 3.5857519788918206,
"grad_norm": 0.0015155959408730268,
"learning_rate": 2.1776416356429857e-06,
"loss": 0.0,
"step": 455
},
{
"epoch": 3.5936675461741423,
"grad_norm": 0.000976615003310144,
"learning_rate": 2.165772791788379e-06,
"loss": 0.0,
"step": 456
},
{
"epoch": 3.6015831134564644,
"grad_norm": 0.020495692268013954,
"learning_rate": 2.1539116133022776e-06,
"loss": 0.0,
"step": 457
},
{
"epoch": 3.6094986807387865,
"grad_norm": 0.0031040646135807037,
"learning_rate": 2.142058372216105e-06,
"loss": 0.0,
"step": 458
},
{
"epoch": 3.6174142480211082,
"grad_norm": 0.003270247019827366,
"learning_rate": 2.130213340379248e-06,
"loss": 0.0,
"step": 459
},
{
"epoch": 3.62532981530343,
"grad_norm": 0.003336927155032754,
"learning_rate": 2.1183767894528135e-06,
"loss": 0.0,
"step": 460
},
{
"epoch": 3.633245382585752,
"grad_norm": 0.0010945361573249102,
"learning_rate": 2.1065489909034065e-06,
"loss": 0.0,
"step": 461
},
{
"epoch": 3.641160949868074,
"grad_norm": 0.0009257032070308924,
"learning_rate": 2.094730215996894e-06,
"loss": 0.0,
"step": 462
},
{
"epoch": 3.649076517150396,
"grad_norm": 0.001292677130550146,
"learning_rate": 2.082920735792195e-06,
"loss": 0.0,
"step": 463
},
{
"epoch": 3.6569920844327175,
"grad_norm": 0.02042875811457634,
"learning_rate": 2.0711208211350543e-06,
"loss": 0.0,
"step": 464
},
{
"epoch": 3.6649076517150396,
"grad_norm": 0.14115594327449799,
"learning_rate": 2.059330742651837e-06,
"loss": 0.0003,
"step": 465
},
{
"epoch": 3.6728232189973617,
"grad_norm": 0.0009795207297429442,
"learning_rate": 2.047550770743318e-06,
"loss": 0.0,
"step": 466
},
{
"epoch": 3.6807387862796834,
"grad_norm": 0.005499033723026514,
"learning_rate": 2.035781175578483e-06,
"loss": 0.0,
"step": 467
},
{
"epoch": 3.688654353562005,
"grad_norm": 0.0013082363875582814,
"learning_rate": 2.024022227088329e-06,
"loss": 0.0,
"step": 468
},
{
"epoch": 3.6965699208443272,
"grad_norm": 0.0016859848983585835,
"learning_rate": 2.01227419495968e-06,
"loss": 0.0,
"step": 469
},
{
"epoch": 3.7044854881266494,
"grad_norm": 0.0014348351396620274,
"learning_rate": 2.0005373486289932e-06,
"loss": 0.0,
"step": 470
},
{
"epoch": 3.712401055408971,
"grad_norm": 0.0014519428368657827,
"learning_rate": 1.988811957276185e-06,
"loss": 0.0,
"step": 471
},
{
"epoch": 3.7203166226912927,
"grad_norm": 0.0020222277380526066,
"learning_rate": 1.977098289818459e-06,
"loss": 0.0,
"step": 472
},
{
"epoch": 3.728232189973615,
"grad_norm": 0.0014894002815708518,
"learning_rate": 1.9653966149041323e-06,
"loss": 0.0,
"step": 473
},
{
"epoch": 3.736147757255937,
"grad_norm": 0.0010474161244928837,
"learning_rate": 1.9537072009064814e-06,
"loss": 0.0,
"step": 474
},
{
"epoch": 3.7440633245382586,
"grad_norm": 0.00139061629306525,
"learning_rate": 1.9420303159175795e-06,
"loss": 0.0,
"step": 475
},
{
"epoch": 3.7519788918205803,
"grad_norm": 0.003557840595021844,
"learning_rate": 1.930366227742157e-06,
"loss": 0.0,
"step": 476
},
{
"epoch": 3.7598944591029024,
"grad_norm": 0.0006267677526921034,
"learning_rate": 1.918715203891448e-06,
"loss": 0.0,
"step": 477
},
{
"epoch": 3.767810026385224,
"grad_norm": 0.009051766246557236,
"learning_rate": 1.907077511577068e-06,
"loss": 0.0,
"step": 478
},
{
"epoch": 3.7757255936675462,
"grad_norm": 0.0014555881498381495,
"learning_rate": 1.8954534177048744e-06,
"loss": 0.0,
"step": 479
},
{
"epoch": 3.783641160949868,
"grad_norm": 0.002180263167247176,
"learning_rate": 1.8838431888688528e-06,
"loss": 0.0,
"step": 480
},
{
"epoch": 3.79155672823219,
"grad_norm": 0.0008298380998894572,
"learning_rate": 1.8722470913449962e-06,
"loss": 0.0,
"step": 481
},
{
"epoch": 3.7994722955145117,
"grad_norm": 0.08523594588041306,
"learning_rate": 1.8606653910852058e-06,
"loss": 0.0002,
"step": 482
},
{
"epoch": 3.807387862796834,
"grad_norm": 0.0013316926779225469,
"learning_rate": 1.8490983537111831e-06,
"loss": 0.0,
"step": 483
},
{
"epoch": 3.8153034300791555,
"grad_norm": 0.0014392542652785778,
"learning_rate": 1.8375462445083464e-06,
"loss": 0.0,
"step": 484
},
{
"epoch": 3.8232189973614776,
"grad_norm": 0.0012960598105564713,
"learning_rate": 1.8260093284197371e-06,
"loss": 0.0,
"step": 485
},
{
"epoch": 3.8311345646437993,
"grad_norm": 0.0018995273858308792,
"learning_rate": 1.814487870039955e-06,
"loss": 0.0,
"step": 486
},
{
"epoch": 3.8390501319261214,
"grad_norm": 0.0011370782740414143,
"learning_rate": 1.8029821336090769e-06,
"loss": 0.0,
"step": 487
},
{
"epoch": 3.846965699208443,
"grad_norm": 0.001328559941612184,
"learning_rate": 1.7914923830066074e-06,
"loss": 0.0,
"step": 488
},
{
"epoch": 3.8548812664907652,
"grad_norm": 0.006753549445420504,
"learning_rate": 1.7800188817454209e-06,
"loss": 0.0,
"step": 489
},
{
"epoch": 3.862796833773087,
"grad_norm": 0.0009229669230990112,
"learning_rate": 1.7685618929657193e-06,
"loss": 0.0,
"step": 490
},
{
"epoch": 3.870712401055409,
"grad_norm": 0.0008038815576583147,
"learning_rate": 1.7571216794289985e-06,
"loss": 0.0,
"step": 491
},
{
"epoch": 3.8786279683377307,
"grad_norm": 0.0008389101712964475,
"learning_rate": 1.7456985035120194e-06,
"loss": 0.0,
"step": 492
},
{
"epoch": 3.886543535620053,
"grad_norm": 0.0009506504866294563,
"learning_rate": 1.734292627200793e-06,
"loss": 0.0,
"step": 493
},
{
"epoch": 3.8944591029023745,
"grad_norm": 0.0009613066795282066,
"learning_rate": 1.7229043120845705e-06,
"loss": 0.0,
"step": 494
},
{
"epoch": 3.9023746701846966,
"grad_norm": 0.050488874316215515,
"learning_rate": 1.711533819349842e-06,
"loss": 0.0001,
"step": 495
},
{
"epoch": 3.9102902374670183,
"grad_norm": 0.0020255844574421644,
"learning_rate": 1.700181409774353e-06,
"loss": 0.0,
"step": 496
},
{
"epoch": 3.9182058047493404,
"grad_norm": 0.0408753827214241,
"learning_rate": 1.6888473437211134e-06,
"loss": 0.0001,
"step": 497
},
{
"epoch": 3.926121372031662,
"grad_norm": 0.0009621501667425036,
"learning_rate": 1.6775318811324365e-06,
"loss": 0.0,
"step": 498
},
{
"epoch": 3.9340369393139842,
"grad_norm": 0.00070030870847404,
"learning_rate": 1.666235281523968e-06,
"loss": 0.0,
"step": 499
},
{
"epoch": 3.941952506596306,
"grad_norm": 0.0018594212597236037,
"learning_rate": 1.6549578039787436e-06,
"loss": 0.0,
"step": 500
},
{
"epoch": 3.949868073878628,
"grad_norm": 0.0013388278894126415,
"learning_rate": 1.643699707141237e-06,
"loss": 0.0,
"step": 501
},
{
"epoch": 3.9577836411609497,
"grad_norm": 0.0010811896063387394,
"learning_rate": 1.6324612492114378e-06,
"loss": 0.0,
"step": 502
},
{
"epoch": 3.965699208443272,
"grad_norm": 0.001241902238689363,
"learning_rate": 1.6212426879389205e-06,
"loss": 0.0,
"step": 503
},
{
"epoch": 3.9736147757255935,
"grad_norm": 0.0013246425660327077,
"learning_rate": 1.6100442806169423e-06,
"loss": 0.0,
"step": 504
}
],
"logging_steps": 1,
"max_steps": 756,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 126,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.2680704217736806e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}