| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0294117647058822, | |
| "eval_steps": 500, | |
| "global_step": 280, | |
| "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 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.063239288435507e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |