MS2501-24b-Ink-v2-ep2-adpt / trainer_state.json
Fizzarolli's picture
Upload folder using huggingface_hub
aa219b5 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 544,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003676470588235294,
"grad_norm": 0.20527108013629913,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.7943,
"step": 1
},
{
"epoch": 0.007352941176470588,
"grad_norm": 0.24291086196899414,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.8345,
"step": 2
},
{
"epoch": 0.011029411764705883,
"grad_norm": 0.20741452276706696,
"learning_rate": 6.000000000000001e-07,
"loss": 1.998,
"step": 3
},
{
"epoch": 0.014705882352941176,
"grad_norm": 0.242351695895195,
"learning_rate": 8.000000000000001e-07,
"loss": 1.814,
"step": 4
},
{
"epoch": 0.01838235294117647,
"grad_norm": 0.20806454122066498,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.8541,
"step": 5
},
{
"epoch": 0.022058823529411766,
"grad_norm": 0.2232266068458557,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.8491,
"step": 6
},
{
"epoch": 0.025735294117647058,
"grad_norm": 0.24171140789985657,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.7849,
"step": 7
},
{
"epoch": 0.029411764705882353,
"grad_norm": 0.2743687331676483,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.7188,
"step": 8
},
{
"epoch": 0.03308823529411765,
"grad_norm": 0.17957501113414764,
"learning_rate": 1.8000000000000001e-06,
"loss": 2.0571,
"step": 9
},
{
"epoch": 0.03676470588235294,
"grad_norm": 0.201459139585495,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.9404,
"step": 10
},
{
"epoch": 0.04044117647058824,
"grad_norm": 0.2665609121322632,
"learning_rate": 2.2e-06,
"loss": 1.9073,
"step": 11
},
{
"epoch": 0.04411764705882353,
"grad_norm": 0.2173280417919159,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.8472,
"step": 12
},
{
"epoch": 0.04779411764705882,
"grad_norm": 1.175387978553772,
"learning_rate": 2.6e-06,
"loss": 1.9748,
"step": 13
},
{
"epoch": 0.051470588235294115,
"grad_norm": 0.22836479544639587,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.799,
"step": 14
},
{
"epoch": 0.05514705882352941,
"grad_norm": 0.22088803350925446,
"learning_rate": 3e-06,
"loss": 1.735,
"step": 15
},
{
"epoch": 0.058823529411764705,
"grad_norm": 0.20766077935695648,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.8871,
"step": 16
},
{
"epoch": 0.0625,
"grad_norm": 0.2599094808101654,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.6319,
"step": 17
},
{
"epoch": 0.0661764705882353,
"grad_norm": 0.21609334647655487,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.845,
"step": 18
},
{
"epoch": 0.06985294117647059,
"grad_norm": 0.21392510831356049,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.925,
"step": 19
},
{
"epoch": 0.07352941176470588,
"grad_norm": 0.22204363346099854,
"learning_rate": 4.000000000000001e-06,
"loss": 1.83,
"step": 20
},
{
"epoch": 0.07720588235294118,
"grad_norm": 0.24028566479682922,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.7375,
"step": 21
},
{
"epoch": 0.08088235294117647,
"grad_norm": 0.26085978746414185,
"learning_rate": 4.4e-06,
"loss": 1.7598,
"step": 22
},
{
"epoch": 0.08455882352941177,
"grad_norm": 0.29647570848464966,
"learning_rate": 4.600000000000001e-06,
"loss": 1.7059,
"step": 23
},
{
"epoch": 0.08823529411764706,
"grad_norm": 0.2606264054775238,
"learning_rate": 4.800000000000001e-06,
"loss": 1.7886,
"step": 24
},
{
"epoch": 0.09191176470588236,
"grad_norm": 0.20121997594833374,
"learning_rate": 5e-06,
"loss": 2.069,
"step": 25
},
{
"epoch": 0.09558823529411764,
"grad_norm": 0.23160211741924286,
"learning_rate": 4.9999541991311605e-06,
"loss": 1.9422,
"step": 26
},
{
"epoch": 0.09926470588235294,
"grad_norm": 0.2353476583957672,
"learning_rate": 4.999816798202817e-06,
"loss": 1.8665,
"step": 27
},
{
"epoch": 0.10294117647058823,
"grad_norm": 0.24767348170280457,
"learning_rate": 4.999587802249433e-06,
"loss": 1.7446,
"step": 28
},
{
"epoch": 0.10661764705882353,
"grad_norm": 0.24555310606956482,
"learning_rate": 4.999267219661583e-06,
"loss": 1.8733,
"step": 29
},
{
"epoch": 0.11029411764705882,
"grad_norm": 2.8036930561065674,
"learning_rate": 4.9988550621856336e-06,
"loss": 1.829,
"step": 30
},
{
"epoch": 0.11397058823529412,
"grad_norm": 0.1886998862028122,
"learning_rate": 4.998351344923323e-06,
"loss": 1.8259,
"step": 31
},
{
"epoch": 0.11764705882352941,
"grad_norm": 0.23365086317062378,
"learning_rate": 4.997756086331198e-06,
"loss": 1.9775,
"step": 32
},
{
"epoch": 0.1213235294117647,
"grad_norm": 0.18128617107868195,
"learning_rate": 4.997069308219952e-06,
"loss": 1.8398,
"step": 33
},
{
"epoch": 0.125,
"grad_norm": 0.21758407354354858,
"learning_rate": 4.996291035753608e-06,
"loss": 1.7429,
"step": 34
},
{
"epoch": 0.12867647058823528,
"grad_norm": 0.17675141990184784,
"learning_rate": 4.995421297448614e-06,
"loss": 1.9317,
"step": 35
},
{
"epoch": 0.1323529411764706,
"grad_norm": 0.18282735347747803,
"learning_rate": 4.994460125172783e-06,
"loss": 1.8007,
"step": 36
},
{
"epoch": 0.13602941176470587,
"grad_norm": 0.17422187328338623,
"learning_rate": 4.993407554144137e-06,
"loss": 1.9293,
"step": 37
},
{
"epoch": 0.13970588235294118,
"grad_norm": 0.20074324309825897,
"learning_rate": 4.992263622929609e-06,
"loss": 1.7154,
"step": 38
},
{
"epoch": 0.14338235294117646,
"grad_norm": 0.16826124489307404,
"learning_rate": 4.991028373443635e-06,
"loss": 1.8976,
"step": 39
},
{
"epoch": 0.14705882352941177,
"grad_norm": 0.1737484484910965,
"learning_rate": 4.989701850946613e-06,
"loss": 1.9842,
"step": 40
},
{
"epoch": 0.15073529411764705,
"grad_norm": 0.17321883141994476,
"learning_rate": 4.988284104043251e-06,
"loss": 1.691,
"step": 41
},
{
"epoch": 0.15441176470588236,
"grad_norm": 0.16045594215393066,
"learning_rate": 4.986775184680782e-06,
"loss": 1.9388,
"step": 42
},
{
"epoch": 0.15808823529411764,
"grad_norm": 0.1485867202281952,
"learning_rate": 4.985175148147057e-06,
"loss": 1.9412,
"step": 43
},
{
"epoch": 0.16176470588235295,
"grad_norm": 0.15609779953956604,
"learning_rate": 4.983484053068529e-06,
"loss": 1.8341,
"step": 44
},
{
"epoch": 0.16544117647058823,
"grad_norm": 0.14435409009456635,
"learning_rate": 4.981701961408096e-06,
"loss": 1.8187,
"step": 45
},
{
"epoch": 0.16911764705882354,
"grad_norm": 0.14100129902362823,
"learning_rate": 4.979828938462836e-06,
"loss": 1.8016,
"step": 46
},
{
"epoch": 0.17279411764705882,
"grad_norm": 0.13618572056293488,
"learning_rate": 4.977865052861611e-06,
"loss": 1.7432,
"step": 47
},
{
"epoch": 0.17647058823529413,
"grad_norm": 0.6496532559394836,
"learning_rate": 4.975810376562555e-06,
"loss": 1.7203,
"step": 48
},
{
"epoch": 0.1801470588235294,
"grad_norm": 0.1398126780986786,
"learning_rate": 4.973664984850435e-06,
"loss": 1.6294,
"step": 49
},
{
"epoch": 0.18382352941176472,
"grad_norm": 0.13840018212795258,
"learning_rate": 4.971428956333896e-06,
"loss": 1.7861,
"step": 50
},
{
"epoch": 0.1875,
"grad_norm": 0.11684554070234299,
"learning_rate": 4.969102372942575e-06,
"loss": 1.9307,
"step": 51
},
{
"epoch": 0.19117647058823528,
"grad_norm": 0.12185637652873993,
"learning_rate": 4.966685319924105e-06,
"loss": 1.8909,
"step": 52
},
{
"epoch": 0.1948529411764706,
"grad_norm": 0.11658155173063278,
"learning_rate": 4.96417788584099e-06,
"loss": 1.8678,
"step": 53
},
{
"epoch": 0.19852941176470587,
"grad_norm": 0.1272619068622589,
"learning_rate": 4.961580162567358e-06,
"loss": 1.8719,
"step": 54
},
{
"epoch": 0.20220588235294118,
"grad_norm": 0.12430202960968018,
"learning_rate": 4.958892245285594e-06,
"loss": 1.849,
"step": 55
},
{
"epoch": 0.20588235294117646,
"grad_norm": 0.11343392729759216,
"learning_rate": 4.956114232482854e-06,
"loss": 2.0185,
"step": 56
},
{
"epoch": 0.20955882352941177,
"grad_norm": 0.10528125613927841,
"learning_rate": 4.953246225947461e-06,
"loss": 1.8122,
"step": 57
},
{
"epoch": 0.21323529411764705,
"grad_norm": 0.11682136356830597,
"learning_rate": 4.950288330765167e-06,
"loss": 1.819,
"step": 58
},
{
"epoch": 0.21691176470588236,
"grad_norm": 0.10487648099660873,
"learning_rate": 4.947240655315306e-06,
"loss": 1.8969,
"step": 59
},
{
"epoch": 0.22058823529411764,
"grad_norm": 0.10324753075838089,
"learning_rate": 4.944103311266827e-06,
"loss": 1.9131,
"step": 60
},
{
"epoch": 0.22426470588235295,
"grad_norm": 0.11773849278688431,
"learning_rate": 4.9408764135741955e-06,
"loss": 1.8771,
"step": 61
},
{
"epoch": 0.22794117647058823,
"grad_norm": 0.10764829814434052,
"learning_rate": 4.937560080473186e-06,
"loss": 1.8007,
"step": 62
},
{
"epoch": 0.23161764705882354,
"grad_norm": 0.09463761746883392,
"learning_rate": 4.934154433476548e-06,
"loss": 1.7991,
"step": 63
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.11386000365018845,
"learning_rate": 4.9306595973695545e-06,
"loss": 1.8523,
"step": 64
},
{
"epoch": 0.23897058823529413,
"grad_norm": 0.10437195003032684,
"learning_rate": 4.927075700205431e-06,
"loss": 1.7621,
"step": 65
},
{
"epoch": 0.2426470588235294,
"grad_norm": 0.11349602788686752,
"learning_rate": 4.923402873300659e-06,
"loss": 1.7418,
"step": 66
},
{
"epoch": 0.24632352941176472,
"grad_norm": 0.16744942963123322,
"learning_rate": 4.91964125123017e-06,
"loss": 1.8985,
"step": 67
},
{
"epoch": 0.25,
"grad_norm": 0.11906889081001282,
"learning_rate": 4.915790971822412e-06,
"loss": 1.7156,
"step": 68
},
{
"epoch": 0.2536764705882353,
"grad_norm": 0.10837192088365555,
"learning_rate": 4.911852176154298e-06,
"loss": 1.9407,
"step": 69
},
{
"epoch": 0.25735294117647056,
"grad_norm": 0.12406501919031143,
"learning_rate": 4.907825008546039e-06,
"loss": 1.8413,
"step": 70
},
{
"epoch": 0.2610294117647059,
"grad_norm": 0.10750163346529007,
"learning_rate": 4.903709616555854e-06,
"loss": 1.9044,
"step": 71
},
{
"epoch": 0.2647058823529412,
"grad_norm": 0.11223389953374863,
"learning_rate": 4.899506150974568e-06,
"loss": 1.7632,
"step": 72
},
{
"epoch": 0.26838235294117646,
"grad_norm": 0.1262880116701126,
"learning_rate": 4.8952147658200815e-06,
"loss": 1.6656,
"step": 73
},
{
"epoch": 0.27205882352941174,
"grad_norm": 0.1263853758573532,
"learning_rate": 4.890835618331729e-06,
"loss": 1.7326,
"step": 74
},
{
"epoch": 0.2757352941176471,
"grad_norm": 0.10728048533201218,
"learning_rate": 4.886368868964517e-06,
"loss": 1.6571,
"step": 75
},
{
"epoch": 0.27941176470588236,
"grad_norm": 0.10924043506383896,
"learning_rate": 4.8818146813832475e-06,
"loss": 1.8581,
"step": 76
},
{
"epoch": 0.28308823529411764,
"grad_norm": 0.12963686883449554,
"learning_rate": 4.877173222456521e-06,
"loss": 1.6903,
"step": 77
},
{
"epoch": 0.2867647058823529,
"grad_norm": 0.11058393120765686,
"learning_rate": 4.872444662250617e-06,
"loss": 1.9864,
"step": 78
},
{
"epoch": 0.29044117647058826,
"grad_norm": 0.10086097568273544,
"learning_rate": 4.867629174023269e-06,
"loss": 1.7289,
"step": 79
},
{
"epoch": 0.29411764705882354,
"grad_norm": 0.19624628126621246,
"learning_rate": 4.862726934217311e-06,
"loss": 1.8319,
"step": 80
},
{
"epoch": 0.2977941176470588,
"grad_norm": 0.10972262173891068,
"learning_rate": 4.857738122454219e-06,
"loss": 1.9062,
"step": 81
},
{
"epoch": 0.3014705882352941,
"grad_norm": 0.1163967102766037,
"learning_rate": 4.852662921527523e-06,
"loss": 1.7979,
"step": 82
},
{
"epoch": 0.30514705882352944,
"grad_norm": 0.26644641160964966,
"learning_rate": 4.847501517396111e-06,
"loss": 1.8551,
"step": 83
},
{
"epoch": 0.3088235294117647,
"grad_norm": 0.097193643450737,
"learning_rate": 4.84225409917742e-06,
"loss": 1.8247,
"step": 84
},
{
"epoch": 0.3125,
"grad_norm": 0.09594879299402237,
"learning_rate": 4.8369208591405e-06,
"loss": 1.6824,
"step": 85
},
{
"epoch": 0.3161764705882353,
"grad_norm": 0.10404081642627716,
"learning_rate": 4.831501992698972e-06,
"loss": 1.8342,
"step": 86
},
{
"epoch": 0.31985294117647056,
"grad_norm": 0.10298274457454681,
"learning_rate": 4.825997698403871e-06,
"loss": 1.9121,
"step": 87
},
{
"epoch": 0.3235294117647059,
"grad_norm": 0.12172818183898926,
"learning_rate": 4.820408177936365e-06,
"loss": 1.8046,
"step": 88
},
{
"epoch": 0.3272058823529412,
"grad_norm": 0.10750740766525269,
"learning_rate": 4.814733636100369e-06,
"loss": 1.9354,
"step": 89
},
{
"epoch": 0.33088235294117646,
"grad_norm": 0.10367678105831146,
"learning_rate": 4.808974280815039e-06,
"loss": 2.0473,
"step": 90
},
{
"epoch": 0.33455882352941174,
"grad_norm": 0.10471010208129883,
"learning_rate": 4.803130323107157e-06,
"loss": 1.8647,
"step": 91
},
{
"epoch": 0.3382352941176471,
"grad_norm": 0.11597222834825516,
"learning_rate": 4.797201977103395e-06,
"loss": 1.6487,
"step": 92
},
{
"epoch": 0.34191176470588236,
"grad_norm": 0.10744346678256989,
"learning_rate": 4.791189460022472e-06,
"loss": 1.8301,
"step": 93
},
{
"epoch": 0.34558823529411764,
"grad_norm": 0.1046157106757164,
"learning_rate": 4.785092992167192e-06,
"loss": 1.8117,
"step": 94
},
{
"epoch": 0.3492647058823529,
"grad_norm": 0.10467652231454849,
"learning_rate": 4.778912796916374e-06,
"loss": 1.932,
"step": 95
},
{
"epoch": 0.35294117647058826,
"grad_norm": 0.10906960070133209,
"learning_rate": 4.77264910071667e-06,
"loss": 1.8958,
"step": 96
},
{
"epoch": 0.35661764705882354,
"grad_norm": 0.1271820366382599,
"learning_rate": 4.766302133074261e-06,
"loss": 1.5758,
"step": 97
},
{
"epoch": 0.3602941176470588,
"grad_norm": 0.10758531838655472,
"learning_rate": 4.759872126546452e-06,
"loss": 1.7336,
"step": 98
},
{
"epoch": 0.3639705882352941,
"grad_norm": 0.11692432314157486,
"learning_rate": 4.753359316733154e-06,
"loss": 1.7036,
"step": 99
},
{
"epoch": 0.36764705882352944,
"grad_norm": 0.11493757367134094,
"learning_rate": 4.746763942268243e-06,
"loss": 1.648,
"step": 100
},
{
"epoch": 0.3713235294117647,
"grad_norm": 0.11132095009088516,
"learning_rate": 4.740086244810825e-06,
"loss": 1.8441,
"step": 101
},
{
"epoch": 0.375,
"grad_norm": 0.11245792359113693,
"learning_rate": 4.733326469036377e-06,
"loss": 1.8768,
"step": 102
},
{
"epoch": 0.3786764705882353,
"grad_norm": 0.11370263248682022,
"learning_rate": 4.726484862627779e-06,
"loss": 1.8407,
"step": 103
},
{
"epoch": 0.38235294117647056,
"grad_norm": 0.1197124496102333,
"learning_rate": 4.719561676266249e-06,
"loss": 1.6798,
"step": 104
},
{
"epoch": 0.3860294117647059,
"grad_norm": 0.114149309694767,
"learning_rate": 4.712557163622145e-06,
"loss": 1.7261,
"step": 105
},
{
"epoch": 0.3897058823529412,
"grad_norm": 0.1167868822813034,
"learning_rate": 4.7054715813456795e-06,
"loss": 1.7402,
"step": 106
},
{
"epoch": 0.39338235294117646,
"grad_norm": 0.10888572782278061,
"learning_rate": 4.698305189057512e-06,
"loss": 1.7388,
"step": 107
},
{
"epoch": 0.39705882352941174,
"grad_norm": 0.11005932092666626,
"learning_rate": 4.691058249339238e-06,
"loss": 1.8437,
"step": 108
},
{
"epoch": 0.4007352941176471,
"grad_norm": 0.11730121821165085,
"learning_rate": 4.683731027723764e-06,
"loss": 1.7581,
"step": 109
},
{
"epoch": 0.40441176470588236,
"grad_norm": 0.10690360516309738,
"learning_rate": 4.676323792685585e-06,
"loss": 1.8829,
"step": 110
},
{
"epoch": 0.40808823529411764,
"grad_norm": 0.11015469580888748,
"learning_rate": 4.668836815630939e-06,
"loss": 1.856,
"step": 111
},
{
"epoch": 0.4117647058823529,
"grad_norm": 0.12613315880298615,
"learning_rate": 4.661270370887872e-06,
"loss": 1.737,
"step": 112
},
{
"epoch": 0.41544117647058826,
"grad_norm": 0.12080664932727814,
"learning_rate": 4.6536247356961775e-06,
"loss": 1.7363,
"step": 113
},
{
"epoch": 0.41911764705882354,
"grad_norm": 0.11145513504743576,
"learning_rate": 4.645900190197242e-06,
"loss": 1.6671,
"step": 114
},
{
"epoch": 0.4227941176470588,
"grad_norm": 0.11746977269649506,
"learning_rate": 4.638097017423783e-06,
"loss": 1.6717,
"step": 115
},
{
"epoch": 0.4264705882352941,
"grad_norm": 0.10946469008922577,
"learning_rate": 4.6302155032894745e-06,
"loss": 1.7082,
"step": 116
},
{
"epoch": 0.43014705882352944,
"grad_norm": 0.11189309507608414,
"learning_rate": 4.622255936578473e-06,
"loss": 1.7705,
"step": 117
},
{
"epoch": 0.4338235294117647,
"grad_norm": 0.12764301896095276,
"learning_rate": 4.614218608934834e-06,
"loss": 1.6192,
"step": 118
},
{
"epoch": 0.4375,
"grad_norm": 0.11146359890699387,
"learning_rate": 4.606103814851829e-06,
"loss": 1.8188,
"step": 119
},
{
"epoch": 0.4411764705882353,
"grad_norm": 0.15550808608531952,
"learning_rate": 4.597911851661155e-06,
"loss": 1.9713,
"step": 120
},
{
"epoch": 0.44485294117647056,
"grad_norm": 0.12268511950969696,
"learning_rate": 4.589643019522036e-06,
"loss": 1.575,
"step": 121
},
{
"epoch": 0.4485294117647059,
"grad_norm": 0.11507008224725723,
"learning_rate": 4.581297621410231e-06,
"loss": 1.8125,
"step": 122
},
{
"epoch": 0.4522058823529412,
"grad_norm": 0.13275226950645447,
"learning_rate": 4.572875963106924e-06,
"loss": 1.6451,
"step": 123
},
{
"epoch": 0.45588235294117646,
"grad_norm": 0.1096140444278717,
"learning_rate": 4.564378353187533e-06,
"loss": 1.879,
"step": 124
},
{
"epoch": 0.45955882352941174,
"grad_norm": 0.11247947067022324,
"learning_rate": 4.555805103010388e-06,
"loss": 1.5851,
"step": 125
},
{
"epoch": 0.4632352941176471,
"grad_norm": 0.12080512940883636,
"learning_rate": 4.5471565267053365e-06,
"loss": 1.6433,
"step": 126
},
{
"epoch": 0.46691176470588236,
"grad_norm": 0.1230054646730423,
"learning_rate": 4.538432941162227e-06,
"loss": 1.9295,
"step": 127
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.11661480367183685,
"learning_rate": 4.529634666019294e-06,
"loss": 1.8518,
"step": 128
},
{
"epoch": 0.4742647058823529,
"grad_norm": 0.10992003977298737,
"learning_rate": 4.520762023651456e-06,
"loss": 1.8141,
"step": 129
},
{
"epoch": 0.47794117647058826,
"grad_norm": 0.11171197146177292,
"learning_rate": 4.511815339158497e-06,
"loss": 1.8021,
"step": 130
},
{
"epoch": 0.48161764705882354,
"grad_norm": 0.1056443303823471,
"learning_rate": 4.502794940353155e-06,
"loss": 1.9079,
"step": 131
},
{
"epoch": 0.4852941176470588,
"grad_norm": 0.10654587298631668,
"learning_rate": 4.493701157749112e-06,
"loss": 1.7223,
"step": 132
},
{
"epoch": 0.4889705882352941,
"grad_norm": 0.11630342155694962,
"learning_rate": 4.484534324548883e-06,
"loss": 1.7638,
"step": 133
},
{
"epoch": 0.49264705882352944,
"grad_norm": 0.10967176407575607,
"learning_rate": 4.4752947766316094e-06,
"loss": 1.7722,
"step": 134
},
{
"epoch": 0.4963235294117647,
"grad_norm": 0.10630583018064499,
"learning_rate": 4.465982852540747e-06,
"loss": 1.7088,
"step": 135
},
{
"epoch": 0.5,
"grad_norm": 0.12106428295373917,
"learning_rate": 4.456598893471668e-06,
"loss": 1.7594,
"step": 136
},
{
"epoch": 0.5036764705882353,
"grad_norm": 0.11997083574533463,
"learning_rate": 4.447143243259155e-06,
"loss": 1.9027,
"step": 137
},
{
"epoch": 0.5073529411764706,
"grad_norm": 0.12016075849533081,
"learning_rate": 4.437616248364805e-06,
"loss": 1.8426,
"step": 138
},
{
"epoch": 0.5110294117647058,
"grad_norm": 0.1092967689037323,
"learning_rate": 4.428018257864333e-06,
"loss": 1.7154,
"step": 139
},
{
"epoch": 0.5147058823529411,
"grad_norm": 0.12067247182130814,
"learning_rate": 4.41834962343478e-06,
"loss": 1.7537,
"step": 140
},
{
"epoch": 0.5183823529411765,
"grad_norm": 0.10798091441392899,
"learning_rate": 4.408610699341634e-06,
"loss": 1.8887,
"step": 141
},
{
"epoch": 0.5220588235294118,
"grad_norm": 0.12265045195817947,
"learning_rate": 4.398801842425842e-06,
"loss": 1.5765,
"step": 142
},
{
"epoch": 0.5257352941176471,
"grad_norm": 0.112462118268013,
"learning_rate": 4.38892341209074e-06,
"loss": 1.924,
"step": 143
},
{
"epoch": 0.5294117647058824,
"grad_norm": 0.11943833529949188,
"learning_rate": 4.378975770288881e-06,
"loss": 1.793,
"step": 144
},
{
"epoch": 0.5330882352941176,
"grad_norm": 0.13217097520828247,
"learning_rate": 4.368959281508776e-06,
"loss": 1.7578,
"step": 145
},
{
"epoch": 0.5367647058823529,
"grad_norm": 0.13434284925460815,
"learning_rate": 4.358874312761535e-06,
"loss": 1.8399,
"step": 146
},
{
"epoch": 0.5404411764705882,
"grad_norm": 0.11146039515733719,
"learning_rate": 4.348721233567424e-06,
"loss": 1.9686,
"step": 147
},
{
"epoch": 0.5441176470588235,
"grad_norm": 0.11943879723548889,
"learning_rate": 4.3385004159423195e-06,
"loss": 1.7348,
"step": 148
},
{
"epoch": 0.5477941176470589,
"grad_norm": 0.1157916858792305,
"learning_rate": 4.328212234384085e-06,
"loss": 1.7748,
"step": 149
},
{
"epoch": 0.5514705882352942,
"grad_norm": 0.1146322712302208,
"learning_rate": 4.317857065858843e-06,
"loss": 1.607,
"step": 150
},
{
"epoch": 0.5551470588235294,
"grad_norm": 0.11403496563434601,
"learning_rate": 4.307435289787169e-06,
"loss": 1.8049,
"step": 151
},
{
"epoch": 0.5588235294117647,
"grad_norm": 0.11926942318677902,
"learning_rate": 4.296947288030178e-06,
"loss": 1.8483,
"step": 152
},
{
"epoch": 0.5625,
"grad_norm": 0.11152566224336624,
"learning_rate": 4.286393444875546e-06,
"loss": 1.8334,
"step": 153
},
{
"epoch": 0.5661764705882353,
"grad_norm": 0.11430760473012924,
"learning_rate": 4.2757741470234214e-06,
"loss": 1.9025,
"step": 154
},
{
"epoch": 0.5698529411764706,
"grad_norm": 0.11427745968103409,
"learning_rate": 4.26508978357226e-06,
"loss": 1.787,
"step": 155
},
{
"epoch": 0.5735294117647058,
"grad_norm": 0.11434771120548248,
"learning_rate": 4.254340746004564e-06,
"loss": 1.868,
"step": 156
},
{
"epoch": 0.5772058823529411,
"grad_norm": 0.11024492233991623,
"learning_rate": 4.243527428172541e-06,
"loss": 1.7228,
"step": 157
},
{
"epoch": 0.5808823529411765,
"grad_norm": 0.11791540682315826,
"learning_rate": 4.232650226283672e-06,
"loss": 1.9425,
"step": 158
},
{
"epoch": 0.5845588235294118,
"grad_norm": 0.1227816492319107,
"learning_rate": 4.221709538886197e-06,
"loss": 1.6622,
"step": 159
},
{
"epoch": 0.5882352941176471,
"grad_norm": 0.12510444223880768,
"learning_rate": 4.210705766854505e-06,
"loss": 1.7713,
"step": 160
},
{
"epoch": 0.5919117647058824,
"grad_norm": 0.1140972450375557,
"learning_rate": 4.199639313374451e-06,
"loss": 1.936,
"step": 161
},
{
"epoch": 0.5955882352941176,
"grad_norm": 0.12299997359514236,
"learning_rate": 4.188510583928583e-06,
"loss": 1.7037,
"step": 162
},
{
"epoch": 0.5992647058823529,
"grad_norm": 0.13803604245185852,
"learning_rate": 4.177319986281285e-06,
"loss": 1.9056,
"step": 163
},
{
"epoch": 0.6029411764705882,
"grad_norm": 0.13489902019500732,
"learning_rate": 4.166067930463831e-06,
"loss": 1.6053,
"step": 164
},
{
"epoch": 0.6066176470588235,
"grad_norm": 0.11227516829967499,
"learning_rate": 4.154754828759368e-06,
"loss": 1.7738,
"step": 165
},
{
"epoch": 0.6102941176470589,
"grad_norm": 0.1096302792429924,
"learning_rate": 4.143381095687805e-06,
"loss": 1.8948,
"step": 166
},
{
"epoch": 0.6139705882352942,
"grad_norm": 0.11712092161178589,
"learning_rate": 4.131947147990629e-06,
"loss": 1.8913,
"step": 167
},
{
"epoch": 0.6176470588235294,
"grad_norm": 0.12525494396686554,
"learning_rate": 4.120453404615628e-06,
"loss": 1.7482,
"step": 168
},
{
"epoch": 0.6213235294117647,
"grad_norm": 0.11608923971652985,
"learning_rate": 4.108900286701553e-06,
"loss": 1.6711,
"step": 169
},
{
"epoch": 0.625,
"grad_norm": 0.12490076571702957,
"learning_rate": 4.097288217562669e-06,
"loss": 1.797,
"step": 170
},
{
"epoch": 0.6286764705882353,
"grad_norm": 0.11467793583869934,
"learning_rate": 4.085617622673265e-06,
"loss": 1.7588,
"step": 171
},
{
"epoch": 0.6323529411764706,
"grad_norm": 0.11370235681533813,
"learning_rate": 4.073888929652048e-06,
"loss": 1.8153,
"step": 172
},
{
"epoch": 0.6360294117647058,
"grad_norm": 0.11949747055768967,
"learning_rate": 4.062102568246482e-06,
"loss": 1.6628,
"step": 173
},
{
"epoch": 0.6397058823529411,
"grad_norm": 0.11381959915161133,
"learning_rate": 4.050258970317042e-06,
"loss": 1.7939,
"step": 174
},
{
"epoch": 0.6433823529411765,
"grad_norm": 0.12485770136117935,
"learning_rate": 4.0383585698213874e-06,
"loss": 1.9112,
"step": 175
},
{
"epoch": 0.6470588235294118,
"grad_norm": 0.12126533687114716,
"learning_rate": 4.0264018027984654e-06,
"loss": 1.7908,
"step": 176
},
{
"epoch": 0.6507352941176471,
"grad_norm": 0.10532236844301224,
"learning_rate": 4.01438910735253e-06,
"loss": 1.857,
"step": 177
},
{
"epoch": 0.6544117647058824,
"grad_norm": 0.11462371796369553,
"learning_rate": 4.002320923637091e-06,
"loss": 1.961,
"step": 178
},
{
"epoch": 0.6580882352941176,
"grad_norm": 0.11390509456396103,
"learning_rate": 3.99019769383879e-06,
"loss": 1.7004,
"step": 179
},
{
"epoch": 0.6617647058823529,
"grad_norm": 0.11384452134370804,
"learning_rate": 3.978019862161191e-06,
"loss": 1.719,
"step": 180
},
{
"epoch": 0.6654411764705882,
"grad_norm": 0.12574177980422974,
"learning_rate": 3.965787874808513e-06,
"loss": 1.7728,
"step": 181
},
{
"epoch": 0.6691176470588235,
"grad_norm": 0.11700152605772018,
"learning_rate": 3.953502179969274e-06,
"loss": 1.7115,
"step": 182
},
{
"epoch": 0.6727941176470589,
"grad_norm": 0.15770043432712555,
"learning_rate": 3.941163227799872e-06,
"loss": 1.711,
"step": 183
},
{
"epoch": 0.6764705882352942,
"grad_norm": 0.11906450986862183,
"learning_rate": 3.928771470408092e-06,
"loss": 1.7453,
"step": 184
},
{
"epoch": 0.6801470588235294,
"grad_norm": 0.1260623037815094,
"learning_rate": 3.916327361836536e-06,
"loss": 1.8312,
"step": 185
},
{
"epoch": 0.6838235294117647,
"grad_norm": 0.11971792578697205,
"learning_rate": 3.903831358045994e-06,
"loss": 1.6005,
"step": 186
},
{
"epoch": 0.6875,
"grad_norm": 0.12143218517303467,
"learning_rate": 3.891283916898729e-06,
"loss": 1.7532,
"step": 187
},
{
"epoch": 0.6911764705882353,
"grad_norm": 0.13371998071670532,
"learning_rate": 3.8786854981417064e-06,
"loss": 1.7866,
"step": 188
},
{
"epoch": 0.6948529411764706,
"grad_norm": 0.13985855877399445,
"learning_rate": 3.866036563389747e-06,
"loss": 1.7175,
"step": 189
},
{
"epoch": 0.6985294117647058,
"grad_norm": 0.11457841098308563,
"learning_rate": 3.85333757610861e-06,
"loss": 2.0277,
"step": 190
},
{
"epoch": 0.7022058823529411,
"grad_norm": 0.1197565421462059,
"learning_rate": 3.840589001598018e-06,
"loss": 1.939,
"step": 191
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.117172010242939,
"learning_rate": 3.827791306974602e-06,
"loss": 1.8403,
"step": 192
},
{
"epoch": 0.7095588235294118,
"grad_norm": 0.11386429518461227,
"learning_rate": 3.814944961154788e-06,
"loss": 1.8311,
"step": 193
},
{
"epoch": 0.7132352941176471,
"grad_norm": 0.11297334730625153,
"learning_rate": 3.802050434837615e-06,
"loss": 1.8393,
"step": 194
},
{
"epoch": 0.7169117647058824,
"grad_norm": 0.11836741119623184,
"learning_rate": 3.789108200487493e-06,
"loss": 1.7931,
"step": 195
},
{
"epoch": 0.7205882352941176,
"grad_norm": 0.1100229024887085,
"learning_rate": 3.77611873231688e-06,
"loss": 1.8247,
"step": 196
},
{
"epoch": 0.7242647058823529,
"grad_norm": 0.13370762765407562,
"learning_rate": 3.763082506268922e-06,
"loss": 1.5297,
"step": 197
},
{
"epoch": 0.7279411764705882,
"grad_norm": 0.11646929383277893,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.6622,
"step": 198
},
{
"epoch": 0.7316176470588235,
"grad_norm": 0.17706073820590973,
"learning_rate": 3.736871692862239e-06,
"loss": 1.7438,
"step": 199
},
{
"epoch": 0.7352941176470589,
"grad_norm": 0.1213546171784401,
"learning_rate": 3.723698065885936e-06,
"loss": 1.5792,
"step": 200
},
{
"epoch": 0.7389705882352942,
"grad_norm": 0.11523102223873138,
"learning_rate": 3.7104796017619416e-06,
"loss": 1.8361,
"step": 201
},
{
"epoch": 0.7426470588235294,
"grad_norm": 0.11374305188655853,
"learning_rate": 3.6972167848239677e-06,
"loss": 1.7189,
"step": 202
},
{
"epoch": 0.7463235294117647,
"grad_norm": 0.2544911503791809,
"learning_rate": 3.683910101030846e-06,
"loss": 1.7848,
"step": 203
},
{
"epoch": 0.75,
"grad_norm": 0.11955133825540543,
"learning_rate": 3.6705600379487204e-06,
"loss": 1.8594,
"step": 204
},
{
"epoch": 0.7536764705882353,
"grad_norm": 0.11935199052095413,
"learning_rate": 3.6571670847331802e-06,
"loss": 1.9072,
"step": 205
},
{
"epoch": 0.7573529411764706,
"grad_norm": 0.12341475486755371,
"learning_rate": 3.6437317321113415e-06,
"loss": 1.6725,
"step": 206
},
{
"epoch": 0.7610294117647058,
"grad_norm": 0.1144358366727829,
"learning_rate": 3.6302544723638623e-06,
"loss": 1.7683,
"step": 207
},
{
"epoch": 0.7647058823529411,
"grad_norm": 0.11911847442388535,
"learning_rate": 3.6167357993069075e-06,
"loss": 1.7537,
"step": 208
},
{
"epoch": 0.7683823529411765,
"grad_norm": 0.11903777718544006,
"learning_rate": 3.603176208274054e-06,
"loss": 1.8182,
"step": 209
},
{
"epoch": 0.7720588235294118,
"grad_norm": 0.1177648976445198,
"learning_rate": 3.5895761960981423e-06,
"loss": 1.8037,
"step": 210
},
{
"epoch": 0.7757352941176471,
"grad_norm": 0.11667867004871368,
"learning_rate": 3.5759362610930733e-06,
"loss": 1.7944,
"step": 211
},
{
"epoch": 0.7794117647058824,
"grad_norm": 0.11360645294189453,
"learning_rate": 3.5622569030355434e-06,
"loss": 1.6972,
"step": 212
},
{
"epoch": 0.7830882352941176,
"grad_norm": 0.12743256986141205,
"learning_rate": 3.5485386231467417e-06,
"loss": 1.6698,
"step": 213
},
{
"epoch": 0.7867647058823529,
"grad_norm": 0.1186446100473404,
"learning_rate": 3.5347819240739783e-06,
"loss": 1.7351,
"step": 214
},
{
"epoch": 0.7904411764705882,
"grad_norm": 0.13953393697738647,
"learning_rate": 3.5209873098722693e-06,
"loss": 1.8201,
"step": 215
},
{
"epoch": 0.7941176470588235,
"grad_norm": 0.11326795816421509,
"learning_rate": 3.507155285985866e-06,
"loss": 1.7309,
"step": 216
},
{
"epoch": 0.7977941176470589,
"grad_norm": 0.11922699958086014,
"learning_rate": 3.4932863592297393e-06,
"loss": 1.7146,
"step": 217
},
{
"epoch": 0.8014705882352942,
"grad_norm": 0.13740628957748413,
"learning_rate": 3.4793810377710048e-06,
"loss": 1.7193,
"step": 218
},
{
"epoch": 0.8051470588235294,
"grad_norm": 0.12394211441278458,
"learning_rate": 3.465439831110306e-06,
"loss": 1.8768,
"step": 219
},
{
"epoch": 0.8088235294117647,
"grad_norm": 0.12386978417634964,
"learning_rate": 3.451463250063146e-06,
"loss": 1.5901,
"step": 220
},
{
"epoch": 0.8125,
"grad_norm": 0.12373729050159454,
"learning_rate": 3.4374518067411674e-06,
"loss": 1.712,
"step": 221
},
{
"epoch": 0.8161764705882353,
"grad_norm": 0.12053066492080688,
"learning_rate": 3.4234060145333937e-06,
"loss": 1.8149,
"step": 222
},
{
"epoch": 0.8198529411764706,
"grad_norm": 0.1302081197500229,
"learning_rate": 3.409326388087414e-06,
"loss": 1.8146,
"step": 223
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.12400317192077637,
"learning_rate": 3.3952134432905275e-06,
"loss": 1.6355,
"step": 224
},
{
"epoch": 0.8272058823529411,
"grad_norm": 0.1284218281507492,
"learning_rate": 3.3810676972508405e-06,
"loss": 1.8496,
"step": 225
},
{
"epoch": 0.8308823529411765,
"grad_norm": 0.12382645905017853,
"learning_rate": 3.3668896682783216e-06,
"loss": 1.7146,
"step": 226
},
{
"epoch": 0.8345588235294118,
"grad_norm": 0.12787872552871704,
"learning_rate": 3.3526798758658062e-06,
"loss": 1.8304,
"step": 227
},
{
"epoch": 0.8382352941176471,
"grad_norm": 0.16194690763950348,
"learning_rate": 3.338438840669964e-06,
"loss": 1.6469,
"step": 228
},
{
"epoch": 0.8419117647058824,
"grad_norm": 0.1237102746963501,
"learning_rate": 3.324167084492226e-06,
"loss": 1.7982,
"step": 229
},
{
"epoch": 0.8455882352941176,
"grad_norm": 0.12435570359230042,
"learning_rate": 3.3098651302596565e-06,
"loss": 1.7429,
"step": 230
},
{
"epoch": 0.8492647058823529,
"grad_norm": 0.12152646481990814,
"learning_rate": 3.2955335020057994e-06,
"loss": 1.8001,
"step": 231
},
{
"epoch": 0.8529411764705882,
"grad_norm": 0.11374185979366302,
"learning_rate": 3.281172724851476e-06,
"loss": 1.8071,
"step": 232
},
{
"epoch": 0.8566176470588235,
"grad_norm": 0.12285105139017105,
"learning_rate": 3.266783324985543e-06,
"loss": 1.6797,
"step": 233
},
{
"epoch": 0.8602941176470589,
"grad_norm": 0.12639090418815613,
"learning_rate": 3.252365829645612e-06,
"loss": 1.7243,
"step": 234
},
{
"epoch": 0.8639705882352942,
"grad_norm": 0.11098191887140274,
"learning_rate": 3.2379207670987352e-06,
"loss": 1.8761,
"step": 235
},
{
"epoch": 0.8676470588235294,
"grad_norm": 0.11968422681093216,
"learning_rate": 3.2234486666220437e-06,
"loss": 1.8813,
"step": 236
},
{
"epoch": 0.8713235294117647,
"grad_norm": 0.12833014130592346,
"learning_rate": 3.2089500584833577e-06,
"loss": 1.6301,
"step": 237
},
{
"epoch": 0.875,
"grad_norm": 0.1229129508137703,
"learning_rate": 3.1944254739217584e-06,
"loss": 1.8822,
"step": 238
},
{
"epoch": 0.8786764705882353,
"grad_norm": 0.12925758957862854,
"learning_rate": 3.179875445128119e-06,
"loss": 1.9632,
"step": 239
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.12624742090702057,
"learning_rate": 3.165300505225608e-06,
"loss": 1.7239,
"step": 240
},
{
"epoch": 0.8860294117647058,
"grad_norm": 0.1159316673874855,
"learning_rate": 3.150701188250153e-06,
"loss": 1.7914,
"step": 241
},
{
"epoch": 0.8897058823529411,
"grad_norm": 0.12504705786705017,
"learning_rate": 3.136078029130877e-06,
"loss": 1.9467,
"step": 242
},
{
"epoch": 0.8933823529411765,
"grad_norm": 0.12300896644592285,
"learning_rate": 3.1214315636704928e-06,
"loss": 1.8546,
"step": 243
},
{
"epoch": 0.8970588235294118,
"grad_norm": 0.15699361264705658,
"learning_rate": 3.106762328525677e-06,
"loss": 1.7644,
"step": 244
},
{
"epoch": 0.9007352941176471,
"grad_norm": 1.1493163108825684,
"learning_rate": 3.0920708611874006e-06,
"loss": 1.8477,
"step": 245
},
{
"epoch": 0.9044117647058824,
"grad_norm": 0.1137952208518982,
"learning_rate": 3.0773576999612375e-06,
"loss": 1.7101,
"step": 246
},
{
"epoch": 0.9080882352941176,
"grad_norm": 0.1225443109869957,
"learning_rate": 3.0626233839476434e-06,
"loss": 1.8754,
"step": 247
},
{
"epoch": 0.9117647058823529,
"grad_norm": 0.12916752696037292,
"learning_rate": 3.0478684530221977e-06,
"loss": 1.8084,
"step": 248
},
{
"epoch": 0.9154411764705882,
"grad_norm": 0.12458333373069763,
"learning_rate": 3.033093447815825e-06,
"loss": 1.9428,
"step": 249
},
{
"epoch": 0.9191176470588235,
"grad_norm": 0.1416483372449875,
"learning_rate": 3.018298909694986e-06,
"loss": 1.8178,
"step": 250
},
{
"epoch": 0.9227941176470589,
"grad_norm": 0.1401221603155136,
"learning_rate": 3.0034853807418412e-06,
"loss": 1.7829,
"step": 251
},
{
"epoch": 0.9264705882352942,
"grad_norm": 0.11938274651765823,
"learning_rate": 2.9886534037343872e-06,
"loss": 1.6625,
"step": 252
},
{
"epoch": 0.9301470588235294,
"grad_norm": 0.1268637776374817,
"learning_rate": 2.973803522126571e-06,
"loss": 1.7527,
"step": 253
},
{
"epoch": 0.9338235294117647,
"grad_norm": 0.13126443326473236,
"learning_rate": 2.9589362800283774e-06,
"loss": 1.6907,
"step": 254
},
{
"epoch": 0.9375,
"grad_norm": 0.13425901532173157,
"learning_rate": 2.9440522221858886e-06,
"loss": 1.7868,
"step": 255
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.12785860896110535,
"learning_rate": 2.9291518939613317e-06,
"loss": 1.8627,
"step": 256
},
{
"epoch": 0.9448529411764706,
"grad_norm": 0.12060689181089401,
"learning_rate": 2.914235841313088e-06,
"loss": 1.7925,
"step": 257
},
{
"epoch": 0.9485294117647058,
"grad_norm": 0.12087797373533249,
"learning_rate": 2.899304610775695e-06,
"loss": 1.741,
"step": 258
},
{
"epoch": 0.9522058823529411,
"grad_norm": 0.1257392317056656,
"learning_rate": 2.8843587494398177e-06,
"loss": 1.9245,
"step": 259
},
{
"epoch": 0.9558823529411765,
"grad_norm": 0.13595956563949585,
"learning_rate": 2.869398804932204e-06,
"loss": 1.6015,
"step": 260
},
{
"epoch": 0.9595588235294118,
"grad_norm": 0.11528779566287994,
"learning_rate": 2.854425325395619e-06,
"loss": 1.8843,
"step": 261
},
{
"epoch": 0.9632352941176471,
"grad_norm": 0.17899318039417267,
"learning_rate": 2.83943885946876e-06,
"loss": 1.7506,
"step": 262
},
{
"epoch": 0.9669117647058824,
"grad_norm": 0.12404326349496841,
"learning_rate": 2.824439956266156e-06,
"loss": 1.8862,
"step": 263
},
{
"epoch": 0.9705882352941176,
"grad_norm": 0.14045919477939606,
"learning_rate": 2.809429165358045e-06,
"loss": 1.6858,
"step": 264
},
{
"epoch": 0.9742647058823529,
"grad_norm": 0.1323767602443695,
"learning_rate": 2.7944070367502404e-06,
"loss": 1.7981,
"step": 265
},
{
"epoch": 0.9779411764705882,
"grad_norm": 0.1255924552679062,
"learning_rate": 2.7793741208639746e-06,
"loss": 1.8193,
"step": 266
},
{
"epoch": 0.9816176470588235,
"grad_norm": 0.13566090166568756,
"learning_rate": 2.7643309685157355e-06,
"loss": 1.8234,
"step": 267
},
{
"epoch": 0.9852941176470589,
"grad_norm": 0.14218254387378693,
"learning_rate": 2.7492781308970805e-06,
"loss": 1.6932,
"step": 268
},
{
"epoch": 0.9889705882352942,
"grad_norm": 0.14237117767333984,
"learning_rate": 2.7342161595544443e-06,
"loss": 1.9034,
"step": 269
},
{
"epoch": 0.9926470588235294,
"grad_norm": 0.1268174946308136,
"learning_rate": 2.7191456063689235e-06,
"loss": 1.7338,
"step": 270
},
{
"epoch": 0.9963235294117647,
"grad_norm": 0.11195939034223557,
"learning_rate": 2.7040670235360643e-06,
"loss": 1.7169,
"step": 271
},
{
"epoch": 1.0,
"grad_norm": 0.12390360236167908,
"learning_rate": 2.688980963545621e-06,
"loss": 1.5995,
"step": 272
},
{
"epoch": 1.0036764705882353,
"grad_norm": 0.13594205677509308,
"learning_rate": 2.6738879791613183e-06,
"loss": 1.881,
"step": 273
},
{
"epoch": 1.0073529411764706,
"grad_norm": 0.12911252677440643,
"learning_rate": 2.658788623400595e-06,
"loss": 1.96,
"step": 274
},
{
"epoch": 1.0110294117647058,
"grad_norm": 0.12003444135189056,
"learning_rate": 2.6436834495143398e-06,
"loss": 1.8635,
"step": 275
},
{
"epoch": 1.0147058823529411,
"grad_norm": 0.15043824911117554,
"learning_rate": 2.6285730109666245e-06,
"loss": 1.6302,
"step": 276
},
{
"epoch": 1.0183823529411764,
"grad_norm": 0.12194662541151047,
"learning_rate": 2.61345786141442e-06,
"loss": 1.8137,
"step": 277
},
{
"epoch": 1.0220588235294117,
"grad_norm": 0.13448664546012878,
"learning_rate": 2.598338554687312e-06,
"loss": 1.6858,
"step": 278
},
{
"epoch": 1.025735294117647,
"grad_norm": 0.12283740192651749,
"learning_rate": 2.5832156447672074e-06,
"loss": 1.6681,
"step": 279
},
{
"epoch": 1.0294117647058822,
"grad_norm": 0.1438128501176834,
"learning_rate": 2.568089685768038e-06,
"loss": 1.8845,
"step": 280
},
{
"epoch": 1.0330882352941178,
"grad_norm": 0.1411317139863968,
"learning_rate": 2.552961231915455e-06,
"loss": 1.5879,
"step": 281
},
{
"epoch": 1.036764705882353,
"grad_norm": 0.12831564247608185,
"learning_rate": 2.537830837526522e-06,
"loss": 1.9616,
"step": 282
},
{
"epoch": 1.0404411764705883,
"grad_norm": 0.12855687737464905,
"learning_rate": 2.522699056989408e-06,
"loss": 1.6792,
"step": 283
},
{
"epoch": 1.0441176470588236,
"grad_norm": 0.1262204647064209,
"learning_rate": 2.5075664447430675e-06,
"loss": 1.8003,
"step": 284
},
{
"epoch": 1.0477941176470589,
"grad_norm": 0.1396966278553009,
"learning_rate": 2.492433555256933e-06,
"loss": 1.7243,
"step": 285
},
{
"epoch": 1.0514705882352942,
"grad_norm": 0.12766163051128387,
"learning_rate": 2.4773009430105923e-06,
"loss": 1.8063,
"step": 286
},
{
"epoch": 1.0551470588235294,
"grad_norm": 0.118883416056633,
"learning_rate": 2.462169162473478e-06,
"loss": 1.9198,
"step": 287
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.1223691925406456,
"learning_rate": 2.4470387680845462e-06,
"loss": 1.865,
"step": 288
},
{
"epoch": 1.0625,
"grad_norm": 0.12109634280204773,
"learning_rate": 2.4319103142319624e-06,
"loss": 1.6201,
"step": 289
},
{
"epoch": 1.0661764705882353,
"grad_norm": 0.1238977313041687,
"learning_rate": 2.4167843552327934e-06,
"loss": 1.84,
"step": 290
},
{
"epoch": 1.0698529411764706,
"grad_norm": 0.11403720825910568,
"learning_rate": 2.4016614453126886e-06,
"loss": 1.9161,
"step": 291
},
{
"epoch": 1.0735294117647058,
"grad_norm": 0.12475405633449554,
"learning_rate": 2.3865421385855807e-06,
"loss": 1.8774,
"step": 292
},
{
"epoch": 1.0772058823529411,
"grad_norm": 0.11545270681381226,
"learning_rate": 2.3714269890333755e-06,
"loss": 1.5962,
"step": 293
},
{
"epoch": 1.0808823529411764,
"grad_norm": 0.13439597189426422,
"learning_rate": 2.3563165504856615e-06,
"loss": 1.7551,
"step": 294
},
{
"epoch": 1.0845588235294117,
"grad_norm": 0.17008645832538605,
"learning_rate": 2.341211376599406e-06,
"loss": 1.5653,
"step": 295
},
{
"epoch": 1.088235294117647,
"grad_norm": 0.13042360544204712,
"learning_rate": 2.3261120208386825e-06,
"loss": 1.6963,
"step": 296
},
{
"epoch": 1.0919117647058822,
"grad_norm": 0.11813171952962875,
"learning_rate": 2.3110190364543794e-06,
"loss": 1.6211,
"step": 297
},
{
"epoch": 1.0955882352941178,
"grad_norm": 0.13147880136966705,
"learning_rate": 2.2959329764639366e-06,
"loss": 1.7167,
"step": 298
},
{
"epoch": 1.099264705882353,
"grad_norm": 0.12413813173770905,
"learning_rate": 2.280854393631077e-06,
"loss": 1.8632,
"step": 299
},
{
"epoch": 1.1029411764705883,
"grad_norm": 0.11693166196346283,
"learning_rate": 2.265783840445557e-06,
"loss": 1.7328,
"step": 300
},
{
"epoch": 1.1066176470588236,
"grad_norm": 0.12898504734039307,
"learning_rate": 2.2507218691029204e-06,
"loss": 1.9324,
"step": 301
},
{
"epoch": 1.1102941176470589,
"grad_norm": 0.12191050499677658,
"learning_rate": 2.2356690314842657e-06,
"loss": 1.6725,
"step": 302
},
{
"epoch": 1.1139705882352942,
"grad_norm": 0.1292850822210312,
"learning_rate": 2.2206258791360262e-06,
"loss": 1.7853,
"step": 303
},
{
"epoch": 1.1176470588235294,
"grad_norm": 0.11706730723381042,
"learning_rate": 2.20559296324976e-06,
"loss": 1.7502,
"step": 304
},
{
"epoch": 1.1213235294117647,
"grad_norm": 0.12003550678491592,
"learning_rate": 2.1905708346419554e-06,
"loss": 1.8697,
"step": 305
},
{
"epoch": 1.125,
"grad_norm": 0.13008306920528412,
"learning_rate": 2.1755600437338443e-06,
"loss": 1.8624,
"step": 306
},
{
"epoch": 1.1286764705882353,
"grad_norm": 0.12180894613265991,
"learning_rate": 2.1605611405312406e-06,
"loss": 1.9183,
"step": 307
},
{
"epoch": 1.1323529411764706,
"grad_norm": 0.13236035406589508,
"learning_rate": 2.1455746746043822e-06,
"loss": 1.6329,
"step": 308
},
{
"epoch": 1.1360294117647058,
"grad_norm": 0.13113169372081757,
"learning_rate": 2.1306011950677967e-06,
"loss": 1.6412,
"step": 309
},
{
"epoch": 1.1397058823529411,
"grad_norm": 0.12876993417739868,
"learning_rate": 2.115641250560183e-06,
"loss": 1.8172,
"step": 310
},
{
"epoch": 1.1433823529411764,
"grad_norm": 0.11529678106307983,
"learning_rate": 2.1006953892243055e-06,
"loss": 1.9124,
"step": 311
},
{
"epoch": 1.1470588235294117,
"grad_norm": 0.13113842904567719,
"learning_rate": 2.0857641586869126e-06,
"loss": 1.6567,
"step": 312
},
{
"epoch": 1.150735294117647,
"grad_norm": 0.1215846836566925,
"learning_rate": 2.07084810603867e-06,
"loss": 1.689,
"step": 313
},
{
"epoch": 1.1544117647058822,
"grad_norm": 0.1332322210073471,
"learning_rate": 2.055947777814112e-06,
"loss": 1.7132,
"step": 314
},
{
"epoch": 1.1580882352941178,
"grad_norm": 0.13042818009853363,
"learning_rate": 2.041063719971624e-06,
"loss": 1.6855,
"step": 315
},
{
"epoch": 1.161764705882353,
"grad_norm": 0.13144513964653015,
"learning_rate": 2.02619647787343e-06,
"loss": 1.7519,
"step": 316
},
{
"epoch": 1.1654411764705883,
"grad_norm": 0.12236303836107254,
"learning_rate": 2.011346596265613e-06,
"loss": 1.7687,
"step": 317
},
{
"epoch": 1.1691176470588236,
"grad_norm": 0.13190139830112457,
"learning_rate": 1.9965146192581596e-06,
"loss": 1.6745,
"step": 318
},
{
"epoch": 1.1727941176470589,
"grad_norm": 0.1366875171661377,
"learning_rate": 1.981701090305014e-06,
"loss": 1.7524,
"step": 319
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.1305026262998581,
"learning_rate": 1.9669065521841757e-06,
"loss": 1.7546,
"step": 320
},
{
"epoch": 1.1801470588235294,
"grad_norm": 0.11931124329566956,
"learning_rate": 1.9521315469778036e-06,
"loss": 1.7536,
"step": 321
},
{
"epoch": 1.1838235294117647,
"grad_norm": 0.1269647479057312,
"learning_rate": 1.937376616052357e-06,
"loss": 1.7513,
"step": 322
},
{
"epoch": 1.1875,
"grad_norm": 0.13300736248493195,
"learning_rate": 1.922642300038763e-06,
"loss": 1.7341,
"step": 323
},
{
"epoch": 1.1911764705882353,
"grad_norm": 0.1518763303756714,
"learning_rate": 1.9079291388126002e-06,
"loss": 1.7928,
"step": 324
},
{
"epoch": 1.1948529411764706,
"grad_norm": 0.13257811963558197,
"learning_rate": 1.8932376714743237e-06,
"loss": 1.6585,
"step": 325
},
{
"epoch": 1.1985294117647058,
"grad_norm": 0.12374303489923477,
"learning_rate": 1.878568436329508e-06,
"loss": 1.8799,
"step": 326
},
{
"epoch": 1.2022058823529411,
"grad_norm": 0.12358583509922028,
"learning_rate": 1.8639219708691241e-06,
"loss": 1.7606,
"step": 327
},
{
"epoch": 1.2058823529411764,
"grad_norm": 0.12911354005336761,
"learning_rate": 1.8492988117498478e-06,
"loss": 1.5737,
"step": 328
},
{
"epoch": 1.2095588235294117,
"grad_norm": 0.14360639452934265,
"learning_rate": 1.834699494774393e-06,
"loss": 1.7317,
"step": 329
},
{
"epoch": 1.213235294117647,
"grad_norm": 0.13579204678535461,
"learning_rate": 1.8201245548718816e-06,
"loss": 1.6728,
"step": 330
},
{
"epoch": 1.2169117647058822,
"grad_norm": 0.12950024008750916,
"learning_rate": 1.8055745260782415e-06,
"loss": 1.7669,
"step": 331
},
{
"epoch": 1.2205882352941178,
"grad_norm": 0.12464011460542679,
"learning_rate": 1.7910499415166428e-06,
"loss": 1.8882,
"step": 332
},
{
"epoch": 1.224264705882353,
"grad_norm": 0.12047456204891205,
"learning_rate": 1.7765513333779574e-06,
"loss": 1.714,
"step": 333
},
{
"epoch": 1.2279411764705883,
"grad_norm": 0.11501781642436981,
"learning_rate": 1.7620792329012656e-06,
"loss": 1.9265,
"step": 334
},
{
"epoch": 1.2316176470588236,
"grad_norm": 0.12257359176874161,
"learning_rate": 1.7476341703543887e-06,
"loss": 1.7632,
"step": 335
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.13445287942886353,
"learning_rate": 1.733216675014458e-06,
"loss": 1.834,
"step": 336
},
{
"epoch": 1.2389705882352942,
"grad_norm": 0.13747021555900574,
"learning_rate": 1.7188272751485246e-06,
"loss": 1.7433,
"step": 337
},
{
"epoch": 1.2426470588235294,
"grad_norm": 0.138136625289917,
"learning_rate": 1.7044664979942014e-06,
"loss": 1.8748,
"step": 338
},
{
"epoch": 1.2463235294117647,
"grad_norm": 0.13729244470596313,
"learning_rate": 1.690134869740345e-06,
"loss": 1.7655,
"step": 339
},
{
"epoch": 1.25,
"grad_norm": 0.12297158688306808,
"learning_rate": 1.6758329155077746e-06,
"loss": 1.8723,
"step": 340
},
{
"epoch": 1.2536764705882353,
"grad_norm": 0.13168908655643463,
"learning_rate": 1.6615611593300361e-06,
"loss": 1.7511,
"step": 341
},
{
"epoch": 1.2573529411764706,
"grad_norm": 0.13656580448150635,
"learning_rate": 1.6473201241341946e-06,
"loss": 1.8123,
"step": 342
},
{
"epoch": 1.2610294117647058,
"grad_norm": 0.12358947098255157,
"learning_rate": 1.633110331721679e-06,
"loss": 1.8874,
"step": 343
},
{
"epoch": 1.2647058823529411,
"grad_norm": 0.13836705684661865,
"learning_rate": 1.6189323027491593e-06,
"loss": 1.828,
"step": 344
},
{
"epoch": 1.2683823529411764,
"grad_norm": 0.12668243050575256,
"learning_rate": 1.6047865567094735e-06,
"loss": 1.8937,
"step": 345
},
{
"epoch": 1.2720588235294117,
"grad_norm": 0.21876026690006256,
"learning_rate": 1.5906736119125871e-06,
"loss": 1.8524,
"step": 346
},
{
"epoch": 1.2757352941176472,
"grad_norm": 0.12802840769290924,
"learning_rate": 1.576593985466607e-06,
"loss": 1.7938,
"step": 347
},
{
"epoch": 1.2794117647058822,
"grad_norm": 0.1320679783821106,
"learning_rate": 1.5625481932588332e-06,
"loss": 1.7729,
"step": 348
},
{
"epoch": 1.2830882352941178,
"grad_norm": 0.12416329979896545,
"learning_rate": 1.5485367499368547e-06,
"loss": 1.8229,
"step": 349
},
{
"epoch": 1.2867647058823528,
"grad_norm": 0.13838981091976166,
"learning_rate": 1.5345601688896944e-06,
"loss": 1.8826,
"step": 350
},
{
"epoch": 1.2904411764705883,
"grad_norm": 0.12297661602497101,
"learning_rate": 1.5206189622289963e-06,
"loss": 1.847,
"step": 351
},
{
"epoch": 1.2941176470588236,
"grad_norm": 0.13412605226039886,
"learning_rate": 1.5067136407702615e-06,
"loss": 1.8287,
"step": 352
},
{
"epoch": 1.2977941176470589,
"grad_norm": 0.14892882108688354,
"learning_rate": 1.492844714014135e-06,
"loss": 1.881,
"step": 353
},
{
"epoch": 1.3014705882352942,
"grad_norm": 0.13254675269126892,
"learning_rate": 1.479012690127732e-06,
"loss": 2.0435,
"step": 354
},
{
"epoch": 1.3051470588235294,
"grad_norm": 0.13538019359111786,
"learning_rate": 1.4652180759260224e-06,
"loss": 1.7622,
"step": 355
},
{
"epoch": 1.3088235294117647,
"grad_norm": 0.1283712536096573,
"learning_rate": 1.451461376853258e-06,
"loss": 1.7107,
"step": 356
},
{
"epoch": 1.3125,
"grad_norm": 0.13089057803153992,
"learning_rate": 1.4377430969644568e-06,
"loss": 1.7766,
"step": 357
},
{
"epoch": 1.3161764705882353,
"grad_norm": 0.13077013194561005,
"learning_rate": 1.4240637389069284e-06,
"loss": 1.7357,
"step": 358
},
{
"epoch": 1.3198529411764706,
"grad_norm": 0.13065537810325623,
"learning_rate": 1.4104238039018575e-06,
"loss": 1.7515,
"step": 359
},
{
"epoch": 1.3235294117647058,
"grad_norm": 0.1312171220779419,
"learning_rate": 1.3968237917259463e-06,
"loss": 1.9154,
"step": 360
},
{
"epoch": 1.3272058823529411,
"grad_norm": 0.13449817895889282,
"learning_rate": 1.3832642006930935e-06,
"loss": 1.5684,
"step": 361
},
{
"epoch": 1.3308823529411764,
"grad_norm": 0.13083653151988983,
"learning_rate": 1.3697455276361388e-06,
"loss": 1.7913,
"step": 362
},
{
"epoch": 1.3345588235294117,
"grad_norm": 0.134398952126503,
"learning_rate": 1.3562682678886585e-06,
"loss": 1.8831,
"step": 363
},
{
"epoch": 1.3382352941176472,
"grad_norm": 0.12796249985694885,
"learning_rate": 1.342832915266821e-06,
"loss": 1.7452,
"step": 364
},
{
"epoch": 1.3419117647058822,
"grad_norm": 0.1314622461795807,
"learning_rate": 1.329439962051281e-06,
"loss": 1.8515,
"step": 365
},
{
"epoch": 1.3455882352941178,
"grad_norm": 0.13672280311584473,
"learning_rate": 1.3160898989691545e-06,
"loss": 1.6131,
"step": 366
},
{
"epoch": 1.3492647058823528,
"grad_norm": 0.12695570290088654,
"learning_rate": 1.3027832151760328e-06,
"loss": 1.7702,
"step": 367
},
{
"epoch": 1.3529411764705883,
"grad_norm": 0.12619709968566895,
"learning_rate": 1.2895203982380592e-06,
"loss": 1.7551,
"step": 368
},
{
"epoch": 1.3566176470588236,
"grad_norm": 0.1321066915988922,
"learning_rate": 1.2763019341140642e-06,
"loss": 1.7449,
"step": 369
},
{
"epoch": 1.3602941176470589,
"grad_norm": 0.1324441283941269,
"learning_rate": 1.263128307137762e-06,
"loss": 1.8789,
"step": 370
},
{
"epoch": 1.3639705882352942,
"grad_norm": 0.24257655441761017,
"learning_rate": 1.2500000000000007e-06,
"loss": 1.843,
"step": 371
},
{
"epoch": 1.3676470588235294,
"grad_norm": 0.13487295806407928,
"learning_rate": 1.2369174937310794e-06,
"loss": 1.7277,
"step": 372
},
{
"epoch": 1.3713235294117647,
"grad_norm": 0.1332412213087082,
"learning_rate": 1.22388126768312e-06,
"loss": 1.7645,
"step": 373
},
{
"epoch": 1.375,
"grad_norm": 0.15057015419006348,
"learning_rate": 1.2108917995125078e-06,
"loss": 1.7781,
"step": 374
},
{
"epoch": 1.3786764705882353,
"grad_norm": 0.12643615901470184,
"learning_rate": 1.1979495651623846e-06,
"loss": 1.7512,
"step": 375
},
{
"epoch": 1.3823529411764706,
"grad_norm": 0.1606963723897934,
"learning_rate": 1.1850550388452125e-06,
"loss": 1.7735,
"step": 376
},
{
"epoch": 1.3860294117647058,
"grad_norm": 0.15707363188266754,
"learning_rate": 1.1722086930253983e-06,
"loss": 1.7344,
"step": 377
},
{
"epoch": 1.3897058823529411,
"grad_norm": 0.12574374675750732,
"learning_rate": 1.159410998401982e-06,
"loss": 1.7748,
"step": 378
},
{
"epoch": 1.3933823529411764,
"grad_norm": 0.13599219918251038,
"learning_rate": 1.1466624238913907e-06,
"loss": 1.6268,
"step": 379
},
{
"epoch": 1.3970588235294117,
"grad_norm": 0.13079968094825745,
"learning_rate": 1.1339634366102544e-06,
"loss": 1.6599,
"step": 380
},
{
"epoch": 1.4007352941176472,
"grad_norm": 0.1395445168018341,
"learning_rate": 1.1213145018582938e-06,
"loss": 1.753,
"step": 381
},
{
"epoch": 1.4044117647058822,
"grad_norm": 0.13108734786510468,
"learning_rate": 1.1087160831012713e-06,
"loss": 1.7661,
"step": 382
},
{
"epoch": 1.4080882352941178,
"grad_norm": 0.13053135573863983,
"learning_rate": 1.0961686419540073e-06,
"loss": 1.9817,
"step": 383
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.12607184052467346,
"learning_rate": 1.0836726381634644e-06,
"loss": 1.8506,
"step": 384
},
{
"epoch": 1.4154411764705883,
"grad_norm": 0.12961024045944214,
"learning_rate": 1.0712285295919092e-06,
"loss": 1.8133,
"step": 385
},
{
"epoch": 1.4191176470588236,
"grad_norm": 0.13077639043331146,
"learning_rate": 1.0588367722001288e-06,
"loss": 1.9577,
"step": 386
},
{
"epoch": 1.4227941176470589,
"grad_norm": 0.12397018074989319,
"learning_rate": 1.0464978200307272e-06,
"loss": 1.7779,
"step": 387
},
{
"epoch": 1.4264705882352942,
"grad_norm": 0.1258416622877121,
"learning_rate": 1.034212125191487e-06,
"loss": 1.8292,
"step": 388
},
{
"epoch": 1.4301470588235294,
"grad_norm": 0.13193950057029724,
"learning_rate": 1.02198013783881e-06,
"loss": 1.7834,
"step": 389
},
{
"epoch": 1.4338235294117647,
"grad_norm": 0.12783780694007874,
"learning_rate": 1.0098023061612113e-06,
"loss": 1.7964,
"step": 390
},
{
"epoch": 1.4375,
"grad_norm": 0.12796999514102936,
"learning_rate": 9.97679076362909e-07,
"loss": 1.7532,
"step": 391
},
{
"epoch": 1.4411764705882353,
"grad_norm": 0.13268254697322845,
"learning_rate": 9.856108926474706e-07,
"loss": 1.8421,
"step": 392
},
{
"epoch": 1.4448529411764706,
"grad_norm": 0.1279154270887375,
"learning_rate": 9.735981972015352e-07,
"loss": 1.9176,
"step": 393
},
{
"epoch": 1.4485294117647058,
"grad_norm": 0.1238161027431488,
"learning_rate": 9.616414301786128e-07,
"loss": 1.815,
"step": 394
},
{
"epoch": 1.4522058823529411,
"grad_norm": 0.1436426043510437,
"learning_rate": 9.497410296829582e-07,
"loss": 1.8147,
"step": 395
},
{
"epoch": 1.4558823529411764,
"grad_norm": 0.12739156186580658,
"learning_rate": 9.378974317535189e-07,
"loss": 1.6875,
"step": 396
},
{
"epoch": 1.4595588235294117,
"grad_norm": 0.13925060629844666,
"learning_rate": 9.261110703479531e-07,
"loss": 1.7835,
"step": 397
},
{
"epoch": 1.4632352941176472,
"grad_norm": 0.1358778327703476,
"learning_rate": 9.143823773267349e-07,
"loss": 1.7969,
"step": 398
},
{
"epoch": 1.4669117647058822,
"grad_norm": 0.1351114958524704,
"learning_rate": 9.027117824373308e-07,
"loss": 1.736,
"step": 399
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.12710829079151154,
"learning_rate": 8.910997132984481e-07,
"loss": 1.7682,
"step": 400
},
{
"epoch": 1.4742647058823528,
"grad_norm": 0.19130057096481323,
"learning_rate": 8.795465953843715e-07,
"loss": 1.7461,
"step": 401
},
{
"epoch": 1.4779411764705883,
"grad_norm": 0.13635779917240143,
"learning_rate": 8.68052852009372e-07,
"loss": 1.6813,
"step": 402
},
{
"epoch": 1.4816176470588236,
"grad_norm": 0.1377837061882019,
"learning_rate": 8.566189043121953e-07,
"loss": 1.7816,
"step": 403
},
{
"epoch": 1.4852941176470589,
"grad_norm": 0.12702207267284393,
"learning_rate": 8.452451712406329e-07,
"loss": 1.7292,
"step": 404
},
{
"epoch": 1.4889705882352942,
"grad_norm": 0.13705359399318695,
"learning_rate": 8.339320695361697e-07,
"loss": 1.7473,
"step": 405
},
{
"epoch": 1.4926470588235294,
"grad_norm": 0.1425228714942932,
"learning_rate": 8.22680013718715e-07,
"loss": 1.6972,
"step": 406
},
{
"epoch": 1.4963235294117647,
"grad_norm": 0.12557172775268555,
"learning_rate": 8.114894160714165e-07,
"loss": 1.8611,
"step": 407
},
{
"epoch": 1.5,
"grad_norm": 0.13769450783729553,
"learning_rate": 8.003606866255504e-07,
"loss": 1.5658,
"step": 408
},
{
"epoch": 1.5036764705882353,
"grad_norm": 0.12127039581537247,
"learning_rate": 7.892942331454959e-07,
"loss": 1.7032,
"step": 409
},
{
"epoch": 1.5073529411764706,
"grad_norm": 0.13212333619594574,
"learning_rate": 7.782904611138037e-07,
"loss": 1.6767,
"step": 410
},
{
"epoch": 1.5110294117647058,
"grad_norm": 0.18305175006389618,
"learning_rate": 7.67349773716328e-07,
"loss": 1.8421,
"step": 411
},
{
"epoch": 1.5147058823529411,
"grad_norm": 0.13587015867233276,
"learning_rate": 7.564725718274601e-07,
"loss": 1.6409,
"step": 412
},
{
"epoch": 1.5183823529411766,
"grad_norm": 0.14161740243434906,
"learning_rate": 7.456592539954366e-07,
"loss": 1.767,
"step": 413
},
{
"epoch": 1.5220588235294117,
"grad_norm": 0.7190842628479004,
"learning_rate": 7.349102164277403e-07,
"loss": 1.6086,
"step": 414
},
{
"epoch": 1.5257352941176472,
"grad_norm": 0.13130687177181244,
"learning_rate": 7.242258529765794e-07,
"loss": 1.7691,
"step": 415
},
{
"epoch": 1.5294117647058822,
"grad_norm": 0.13257664442062378,
"learning_rate": 7.136065551244544e-07,
"loss": 1.7076,
"step": 416
},
{
"epoch": 1.5330882352941178,
"grad_norm": 0.1287647783756256,
"learning_rate": 7.030527119698227e-07,
"loss": 1.8224,
"step": 417
},
{
"epoch": 1.5367647058823528,
"grad_norm": 0.12957987189292908,
"learning_rate": 6.925647102128319e-07,
"loss": 1.8724,
"step": 418
},
{
"epoch": 1.5404411764705883,
"grad_norm": 0.1263745278120041,
"learning_rate": 6.821429341411564e-07,
"loss": 1.7504,
"step": 419
},
{
"epoch": 1.5441176470588234,
"grad_norm": 0.14651522040367126,
"learning_rate": 6.717877656159147e-07,
"loss": 1.6761,
"step": 420
},
{
"epoch": 1.5477941176470589,
"grad_norm": 0.14042457938194275,
"learning_rate": 6.614995840576816e-07,
"loss": 1.788,
"step": 421
},
{
"epoch": 1.5514705882352942,
"grad_norm": 0.1395975798368454,
"learning_rate": 6.512787664325776e-07,
"loss": 1.8696,
"step": 422
},
{
"epoch": 1.5551470588235294,
"grad_norm": 0.13296611607074738,
"learning_rate": 6.411256872384658e-07,
"loss": 1.9549,
"step": 423
},
{
"epoch": 1.5588235294117647,
"grad_norm": 0.14705660939216614,
"learning_rate": 6.31040718491224e-07,
"loss": 1.7868,
"step": 424
},
{
"epoch": 1.5625,
"grad_norm": 0.12362435460090637,
"learning_rate": 6.210242297111189e-07,
"loss": 1.7684,
"step": 425
},
{
"epoch": 1.5661764705882353,
"grad_norm": 0.1328449845314026,
"learning_rate": 6.110765879092604e-07,
"loss": 1.9487,
"step": 426
},
{
"epoch": 1.5698529411764706,
"grad_norm": 0.14349178969860077,
"learning_rate": 6.011981575741583e-07,
"loss": 1.7547,
"step": 427
},
{
"epoch": 1.5735294117647058,
"grad_norm": 0.132049098610878,
"learning_rate": 5.913893006583668e-07,
"loss": 1.7595,
"step": 428
},
{
"epoch": 1.5772058823529411,
"grad_norm": 0.14611732959747314,
"learning_rate": 5.816503765652209e-07,
"loss": 1.6253,
"step": 429
},
{
"epoch": 1.5808823529411766,
"grad_norm": 0.5706349015235901,
"learning_rate": 5.719817421356686e-07,
"loss": 1.861,
"step": 430
},
{
"epoch": 1.5845588235294117,
"grad_norm": 0.12961260974407196,
"learning_rate": 5.623837516351949e-07,
"loss": 1.8815,
"step": 431
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.13493973016738892,
"learning_rate": 5.52856756740845e-07,
"loss": 1.8026,
"step": 432
},
{
"epoch": 1.5919117647058822,
"grad_norm": 0.12814201414585114,
"learning_rate": 5.434011065283326e-07,
"loss": 1.7631,
"step": 433
},
{
"epoch": 1.5955882352941178,
"grad_norm": 0.13228128850460052,
"learning_rate": 5.340171474592537e-07,
"loss": 1.6501,
"step": 434
},
{
"epoch": 1.5992647058823528,
"grad_norm": 0.12718702852725983,
"learning_rate": 5.247052233683916e-07,
"loss": 1.7465,
"step": 435
},
{
"epoch": 1.6029411764705883,
"grad_norm": 0.12795035541057587,
"learning_rate": 5.154656754511175e-07,
"loss": 1.769,
"step": 436
},
{
"epoch": 1.6066176470588234,
"grad_norm": 0.138128861784935,
"learning_rate": 5.062988422508886e-07,
"loss": 1.8483,
"step": 437
},
{
"epoch": 1.6102941176470589,
"grad_norm": 0.13711762428283691,
"learning_rate": 4.972050596468448e-07,
"loss": 1.4195,
"step": 438
},
{
"epoch": 1.6139705882352942,
"grad_norm": 0.13988971710205078,
"learning_rate": 4.881846608415028e-07,
"loss": 1.887,
"step": 439
},
{
"epoch": 1.6176470588235294,
"grad_norm": 0.1385488212108612,
"learning_rate": 4.792379763485446e-07,
"loss": 1.8329,
"step": 440
},
{
"epoch": 1.6213235294117647,
"grad_norm": 0.142958402633667,
"learning_rate": 4.703653339807074e-07,
"loss": 1.736,
"step": 441
},
{
"epoch": 1.625,
"grad_norm": 0.14326037466526031,
"learning_rate": 4.6156705883777434e-07,
"loss": 1.9025,
"step": 442
},
{
"epoch": 1.6286764705882353,
"grad_norm": 0.14004214107990265,
"learning_rate": 4.528434732946638e-07,
"loss": 1.7955,
"step": 443
},
{
"epoch": 1.6323529411764706,
"grad_norm": 0.14174719154834747,
"learning_rate": 4.441948969896126e-07,
"loss": 1.919,
"step": 444
},
{
"epoch": 1.6360294117647058,
"grad_norm": 0.13766086101531982,
"learning_rate": 4.3562164681246825e-07,
"loss": 1.6951,
"step": 445
},
{
"epoch": 1.6397058823529411,
"grad_norm": 0.14875784516334534,
"learning_rate": 4.2712403689307623e-07,
"loss": 1.8235,
"step": 446
},
{
"epoch": 1.6433823529411766,
"grad_norm": 0.12639568746089935,
"learning_rate": 4.187023785897701e-07,
"loss": 1.9602,
"step": 447
},
{
"epoch": 1.6470588235294117,
"grad_norm": 0.12960933148860931,
"learning_rate": 4.103569804779642e-07,
"loss": 1.7242,
"step": 448
},
{
"epoch": 1.6507352941176472,
"grad_norm": 0.1281520575284958,
"learning_rate": 4.02088148338845e-07,
"loss": 1.8703,
"step": 449
},
{
"epoch": 1.6544117647058822,
"grad_norm": 0.15090522170066833,
"learning_rate": 3.9389618514817124e-07,
"loss": 1.8158,
"step": 450
},
{
"epoch": 1.6580882352941178,
"grad_norm": 0.15420696139335632,
"learning_rate": 3.85781391065167e-07,
"loss": 1.6869,
"step": 451
},
{
"epoch": 1.6617647058823528,
"grad_norm": 0.13783778250217438,
"learning_rate": 3.777440634215282e-07,
"loss": 1.8575,
"step": 452
},
{
"epoch": 1.6654411764705883,
"grad_norm": 0.1347050964832306,
"learning_rate": 3.697844967105263e-07,
"loss": 1.7492,
"step": 453
},
{
"epoch": 1.6691176470588234,
"grad_norm": 0.14325948059558868,
"learning_rate": 3.619029825762177e-07,
"loss": 1.5755,
"step": 454
},
{
"epoch": 1.6727941176470589,
"grad_norm": 0.12705065310001373,
"learning_rate": 3.54099809802759e-07,
"loss": 1.8027,
"step": 455
},
{
"epoch": 1.6764705882352942,
"grad_norm": 0.13613003492355347,
"learning_rate": 3.463752643038232e-07,
"loss": 1.8703,
"step": 456
},
{
"epoch": 1.6801470588235294,
"grad_norm": 0.13046707212924957,
"learning_rate": 3.3872962911212836e-07,
"loss": 1.7997,
"step": 457
},
{
"epoch": 1.6838235294117647,
"grad_norm": 0.13141557574272156,
"learning_rate": 3.31163184369061e-07,
"loss": 1.885,
"step": 458
},
{
"epoch": 1.6875,
"grad_norm": 0.13569918274879456,
"learning_rate": 3.236762073144159e-07,
"loss": 1.6728,
"step": 459
},
{
"epoch": 1.6911764705882353,
"grad_norm": 0.1508040875196457,
"learning_rate": 3.162689722762366e-07,
"loss": 1.7277,
"step": 460
},
{
"epoch": 1.6948529411764706,
"grad_norm": 0.1174527034163475,
"learning_rate": 3.089417506607631e-07,
"loss": 1.8911,
"step": 461
},
{
"epoch": 1.6985294117647058,
"grad_norm": 0.1410108208656311,
"learning_rate": 3.016948109424883e-07,
"loss": 1.8093,
"step": 462
},
{
"epoch": 1.7022058823529411,
"grad_norm": 0.13780896365642548,
"learning_rate": 2.9452841865432126e-07,
"loss": 1.5834,
"step": 463
},
{
"epoch": 1.7058823529411766,
"grad_norm": 0.13869351148605347,
"learning_rate": 2.874428363778553e-07,
"loss": 1.6792,
"step": 464
},
{
"epoch": 1.7095588235294117,
"grad_norm": 0.14043569564819336,
"learning_rate": 2.8043832373375187e-07,
"loss": 1.6761,
"step": 465
},
{
"epoch": 1.7132352941176472,
"grad_norm": 0.1396993100643158,
"learning_rate": 2.735151373722214e-07,
"loss": 1.8474,
"step": 466
},
{
"epoch": 1.7169117647058822,
"grad_norm": 0.14382068812847137,
"learning_rate": 2.6667353096362395e-07,
"loss": 1.7655,
"step": 467
},
{
"epoch": 1.7205882352941178,
"grad_norm": 0.13429930806159973,
"learning_rate": 2.5991375518917535e-07,
"loss": 1.881,
"step": 468
},
{
"epoch": 1.7242647058823528,
"grad_norm": 0.15133066475391388,
"learning_rate": 2.5323605773175775e-07,
"loss": 1.6794,
"step": 469
},
{
"epoch": 1.7279411764705883,
"grad_norm": 0.18690359592437744,
"learning_rate": 2.4664068326684723e-07,
"loss": 1.9156,
"step": 470
},
{
"epoch": 1.7316176470588234,
"grad_norm": 0.1307322084903717,
"learning_rate": 2.401278734535484e-07,
"loss": 1.8899,
"step": 471
},
{
"epoch": 1.7352941176470589,
"grad_norm": 0.12253100425004959,
"learning_rate": 2.336978669257406e-07,
"loss": 1.7231,
"step": 472
},
{
"epoch": 1.7389705882352942,
"grad_norm": 0.1309325098991394,
"learning_rate": 2.273508992833312e-07,
"loss": 1.9309,
"step": 473
},
{
"epoch": 1.7426470588235294,
"grad_norm": 0.1405581384897232,
"learning_rate": 2.2108720308362613e-07,
"loss": 1.5802,
"step": 474
},
{
"epoch": 1.7463235294117647,
"grad_norm": 0.12829530239105225,
"learning_rate": 2.149070078328089e-07,
"loss": 1.8891,
"step": 475
},
{
"epoch": 1.75,
"grad_norm": 0.1326986700296402,
"learning_rate": 2.0881053997752893e-07,
"loss": 1.7289,
"step": 476
},
{
"epoch": 1.7536764705882353,
"grad_norm": 0.13683846592903137,
"learning_rate": 2.0279802289660533e-07,
"loss": 1.7529,
"step": 477
},
{
"epoch": 1.7573529411764706,
"grad_norm": 0.1288047432899475,
"learning_rate": 1.9686967689284354e-07,
"loss": 1.8638,
"step": 478
},
{
"epoch": 1.7610294117647058,
"grad_norm": 0.13414016366004944,
"learning_rate": 1.910257191849618e-07,
"loss": 1.8131,
"step": 479
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.13683004677295685,
"learning_rate": 1.8526636389963216e-07,
"loss": 1.8441,
"step": 480
},
{
"epoch": 1.7683823529411766,
"grad_norm": 0.13091401755809784,
"learning_rate": 1.7959182206363502e-07,
"loss": 1.7694,
"step": 481
},
{
"epoch": 1.7720588235294117,
"grad_norm": 0.13786040246486664,
"learning_rate": 1.740023015961287e-07,
"loss": 1.6129,
"step": 482
},
{
"epoch": 1.7757352941176472,
"grad_norm": 0.1324450820684433,
"learning_rate": 1.6849800730102755e-07,
"loss": 1.6062,
"step": 483
},
{
"epoch": 1.7794117647058822,
"grad_norm": 0.13514846563339233,
"learning_rate": 1.6307914085950117e-07,
"loss": 1.7428,
"step": 484
},
{
"epoch": 1.7830882352941178,
"grad_norm": 0.13554343581199646,
"learning_rate": 1.577459008225807e-07,
"loss": 1.6424,
"step": 485
},
{
"epoch": 1.7867647058823528,
"grad_norm": 0.13436546921730042,
"learning_rate": 1.5249848260388943e-07,
"loss": 1.8134,
"step": 486
},
{
"epoch": 1.7904411764705883,
"grad_norm": 0.14063028991222382,
"learning_rate": 1.4733707847247814e-07,
"loss": 1.8225,
"step": 487
},
{
"epoch": 1.7941176470588234,
"grad_norm": 0.13966462016105652,
"learning_rate": 1.4226187754578153e-07,
"loss": 1.8506,
"step": 488
},
{
"epoch": 1.7977941176470589,
"grad_norm": 0.13221842050552368,
"learning_rate": 1.37273065782689e-07,
"loss": 1.8214,
"step": 489
},
{
"epoch": 1.8014705882352942,
"grad_norm": 0.13220606744289398,
"learning_rate": 1.3237082597673173e-07,
"loss": 1.6753,
"step": 490
},
{
"epoch": 1.8051470588235294,
"grad_norm": 0.1535516381263733,
"learning_rate": 1.275553377493835e-07,
"loss": 1.7165,
"step": 491
},
{
"epoch": 1.8088235294117647,
"grad_norm": 0.1355370581150055,
"learning_rate": 1.228267775434791e-07,
"loss": 1.899,
"step": 492
},
{
"epoch": 1.8125,
"grad_norm": 0.1324004977941513,
"learning_rate": 1.1818531861675226e-07,
"loss": 1.8999,
"step": 493
},
{
"epoch": 1.8161764705882353,
"grad_norm": 0.13550646603107452,
"learning_rate": 1.1363113103548374e-07,
"loss": 1.7658,
"step": 494
},
{
"epoch": 1.8198529411764706,
"grad_norm": 0.12017887830734253,
"learning_rate": 1.091643816682722e-07,
"loss": 1.7258,
"step": 495
},
{
"epoch": 1.8235294117647058,
"grad_norm": 0.12774993479251862,
"learning_rate": 1.0478523417991882e-07,
"loss": 1.9321,
"step": 496
},
{
"epoch": 1.8272058823529411,
"grad_norm": 0.12875042855739594,
"learning_rate": 1.004938490254323e-07,
"loss": 1.8203,
"step": 497
},
{
"epoch": 1.8308823529411766,
"grad_norm": 0.14257602393627167,
"learning_rate": 9.629038344414643e-08,
"loss": 1.5868,
"step": 498
},
{
"epoch": 1.8345588235294117,
"grad_norm": 0.13076844811439514,
"learning_rate": 9.217499145396187e-08,
"loss": 1.7004,
"step": 499
},
{
"epoch": 1.8382352941176472,
"grad_norm": 0.13167351484298706,
"learning_rate": 8.814782384570259e-08,
"loss": 1.8342,
"step": 500
},
{
"epoch": 1.8419117647058822,
"grad_norm": 0.1307523101568222,
"learning_rate": 8.420902817758814e-08,
"loss": 1.8168,
"step": 501
},
{
"epoch": 1.8455882352941178,
"grad_norm": 0.13685598969459534,
"learning_rate": 8.035874876982957e-08,
"loss": 1.7374,
"step": 502
},
{
"epoch": 1.8492647058823528,
"grad_norm": 0.13376078009605408,
"learning_rate": 7.659712669934121e-08,
"loss": 1.7374,
"step": 503
},
{
"epoch": 1.8529411764705883,
"grad_norm": 0.13515940308570862,
"learning_rate": 7.29242997945695e-08,
"loss": 1.8987,
"step": 504
},
{
"epoch": 1.8566176470588234,
"grad_norm": 0.1335776448249817,
"learning_rate": 6.93404026304459e-08,
"loss": 1.9652,
"step": 505
},
{
"epoch": 1.8602941176470589,
"grad_norm": 0.13335004448890686,
"learning_rate": 6.58455665234528e-08,
"loss": 1.6804,
"step": 506
},
{
"epoch": 1.8639705882352942,
"grad_norm": 0.13274776935577393,
"learning_rate": 6.243991952681466e-08,
"loss": 1.6935,
"step": 507
},
{
"epoch": 1.8676470588235294,
"grad_norm": 0.13108554482460022,
"learning_rate": 5.912358642580496e-08,
"loss": 1.8228,
"step": 508
},
{
"epoch": 1.8713235294117647,
"grad_norm": 0.13036131858825684,
"learning_rate": 5.589668873317328e-08,
"loss": 1.7531,
"step": 509
},
{
"epoch": 1.875,
"grad_norm": 0.13800695538520813,
"learning_rate": 5.275934468469385e-08,
"loss": 1.7133,
"step": 510
},
{
"epoch": 1.8786764705882353,
"grad_norm": 0.13230036199092865,
"learning_rate": 4.971166923483372e-08,
"loss": 1.8917,
"step": 511
},
{
"epoch": 1.8823529411764706,
"grad_norm": 0.16875067353248596,
"learning_rate": 4.675377405253923e-08,
"loss": 1.6803,
"step": 512
},
{
"epoch": 1.8860294117647058,
"grad_norm": 0.13764630258083344,
"learning_rate": 4.388576751714613e-08,
"loss": 1.6458,
"step": 513
},
{
"epoch": 1.8897058823529411,
"grad_norm": 0.15355347096920013,
"learning_rate": 4.110775471440703e-08,
"loss": 1.7683,
"step": 514
},
{
"epoch": 1.8933823529411766,
"grad_norm": 0.12625183165073395,
"learning_rate": 3.841983743264249e-08,
"loss": 1.935,
"step": 515
},
{
"epoch": 1.8970588235294117,
"grad_norm": 0.13688094913959503,
"learning_rate": 3.582211415900983e-08,
"loss": 1.8904,
"step": 516
},
{
"epoch": 1.9007352941176472,
"grad_norm": 0.1413528323173523,
"learning_rate": 3.331468007589489e-08,
"loss": 1.7597,
"step": 517
},
{
"epoch": 1.9044117647058822,
"grad_norm": 0.12093937397003174,
"learning_rate": 3.089762705742572e-08,
"loss": 1.777,
"step": 518
},
{
"epoch": 1.9080882352941178,
"grad_norm": 0.13151197135448456,
"learning_rate": 2.8571043666105193e-08,
"loss": 1.767,
"step": 519
},
{
"epoch": 1.9117647058823528,
"grad_norm": 0.13570436835289001,
"learning_rate": 2.6335015149565324e-08,
"loss": 1.8249,
"step": 520
},
{
"epoch": 1.9154411764705883,
"grad_norm": 0.12817277014255524,
"learning_rate": 2.4189623437445576e-08,
"loss": 1.6606,
"step": 521
},
{
"epoch": 1.9191176470588234,
"grad_norm": 0.13553419709205627,
"learning_rate": 2.2134947138389708e-08,
"loss": 1.8117,
"step": 522
},
{
"epoch": 1.9227941176470589,
"grad_norm": 0.13775591552257538,
"learning_rate": 2.017106153716475e-08,
"loss": 1.683,
"step": 523
},
{
"epoch": 1.9264705882352942,
"grad_norm": 0.1265813112258911,
"learning_rate": 1.82980385919046e-08,
"loss": 1.7364,
"step": 524
},
{
"epoch": 1.9301470588235294,
"grad_norm": 0.13504433631896973,
"learning_rate": 1.651594693147185e-08,
"loss": 1.8376,
"step": 525
},
{
"epoch": 1.9338235294117647,
"grad_norm": 0.1287858486175537,
"learning_rate": 1.482485185294369e-08,
"loss": 1.7019,
"step": 526
},
{
"epoch": 1.9375,
"grad_norm": 0.1338948756456375,
"learning_rate": 1.3224815319219098e-08,
"loss": 1.7236,
"step": 527
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.13983018696308136,
"learning_rate": 1.1715895956748713e-08,
"loss": 1.9374,
"step": 528
},
{
"epoch": 1.9448529411764706,
"grad_norm": 0.1403280794620514,
"learning_rate": 1.0298149053387107e-08,
"loss": 1.6141,
"step": 529
},
{
"epoch": 1.9485294117647058,
"grad_norm": 0.1476178765296936,
"learning_rate": 8.971626556365798e-09,
"loss": 1.7692,
"step": 530
},
{
"epoch": 1.9522058823529411,
"grad_norm": 0.13044756650924683,
"learning_rate": 7.736377070391432e-09,
"loss": 1.7169,
"step": 531
},
{
"epoch": 1.9558823529411766,
"grad_norm": 0.13228951394557953,
"learning_rate": 6.592445855863883e-09,
"loss": 1.904,
"step": 532
},
{
"epoch": 1.9595588235294117,
"grad_norm": 0.12543240189552307,
"learning_rate": 5.5398748272175725e-09,
"loss": 1.8149,
"step": 533
},
{
"epoch": 1.9632352941176472,
"grad_norm": 0.1290730983018875,
"learning_rate": 4.578702551386871e-09,
"loss": 1.7823,
"step": 534
},
{
"epoch": 1.9669117647058822,
"grad_norm": 0.12492910772562027,
"learning_rate": 3.7089642463922217e-09,
"loss": 1.844,
"step": 535
},
{
"epoch": 1.9705882352941178,
"grad_norm": 0.1257825791835785,
"learning_rate": 2.930691780049233e-09,
"loss": 1.7038,
"step": 536
},
{
"epoch": 1.9742647058823528,
"grad_norm": 0.13755561411380768,
"learning_rate": 2.2439136688021113e-09,
"loss": 1.7303,
"step": 537
},
{
"epoch": 1.9779411764705883,
"grad_norm": 0.13982956111431122,
"learning_rate": 1.6486550766783849e-09,
"loss": 1.8592,
"step": 538
},
{
"epoch": 1.9816176470588234,
"grad_norm": 0.1264951378107071,
"learning_rate": 1.1449378143665867e-09,
"loss": 1.8278,
"step": 539
},
{
"epoch": 1.9852941176470589,
"grad_norm": 0.13675741851329803,
"learning_rate": 7.327803384171717e-10,
"loss": 1.7962,
"step": 540
},
{
"epoch": 1.9889705882352942,
"grad_norm": 0.14268295466899872,
"learning_rate": 4.1219775056666746e-10,
"loss": 1.7077,
"step": 541
},
{
"epoch": 1.9926470588235294,
"grad_norm": 0.13337546586990356,
"learning_rate": 1.832017971839517e-10,
"loss": 1.7953,
"step": 542
},
{
"epoch": 1.9963235294117647,
"grad_norm": 0.13678434491157532,
"learning_rate": 4.5800868840040204e-11,
"loss": 1.8995,
"step": 543
},
{
"epoch": 2.0,
"grad_norm": 0.13058674335479736,
"learning_rate": 0.0,
"loss": 1.6561,
"step": 544
}
],
"logging_steps": 1,
"max_steps": 544,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 28,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 9.837150617531843e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}