| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9725490196078432, | |
| "eval_steps": 500, | |
| "global_step": 126, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.023529411764705882, | |
| "grad_norm": 6.901778221130371, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.1119, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.047058823529411764, | |
| "grad_norm": 7.4228010177612305, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.1537, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.07058823529411765, | |
| "grad_norm": 6.921288967132568, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.1207, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.09411764705882353, | |
| "grad_norm": 6.287109851837158, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0477, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 5.439420700073242, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.0753, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1411764705882353, | |
| "grad_norm": 3.6856906414031982, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9761, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.16470588235294117, | |
| "grad_norm": 2.859872341156006, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.958, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.18823529411764706, | |
| "grad_norm": 4.026933670043945, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.0194, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.21176470588235294, | |
| "grad_norm": 4.821041107177734, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9508, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 4.604616641998291, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.0138, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.25882352941176473, | |
| "grad_norm": 3.8580994606018066, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.9089, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2823529411764706, | |
| "grad_norm": 3.7659692764282227, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0376, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3058823529411765, | |
| "grad_norm": 2.5426435470581055, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9651, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.32941176470588235, | |
| "grad_norm": 2.092756986618042, | |
| "learning_rate": 9.998067787472772e-06, | |
| "loss": 0.9098, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 2.1536059379577637, | |
| "learning_rate": 9.992272643269181e-06, | |
| "loss": 0.8308, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3764705882352941, | |
| "grad_norm": 1.9728121757507324, | |
| "learning_rate": 9.982619046369321e-06, | |
| "loss": 0.9148, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.6017658710479736, | |
| "learning_rate": 9.96911445789354e-06, | |
| "loss": 0.8949, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.4235294117647059, | |
| "grad_norm": 1.4864757061004639, | |
| "learning_rate": 9.951769315335843e-06, | |
| "loss": 0.8593, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.4470588235294118, | |
| "grad_norm": 1.447627305984497, | |
| "learning_rate": 9.930597024496933e-06, | |
| "loss": 0.8316, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 1.1549137830734253, | |
| "learning_rate": 9.905613949123036e-06, | |
| "loss": 0.8079, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.49411764705882355, | |
| "grad_norm": 1.2319858074188232, | |
| "learning_rate": 9.87683939825864e-06, | |
| "loss": 0.8833, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.5176470588235295, | |
| "grad_norm": 1.2558043003082275, | |
| "learning_rate": 9.844295611322804e-06, | |
| "loss": 0.8729, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.5411764705882353, | |
| "grad_norm": 1.0202245712280273, | |
| "learning_rate": 9.808007740920647e-06, | |
| "loss": 0.7801, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5647058823529412, | |
| "grad_norm": 0.9334889054298401, | |
| "learning_rate": 9.768003833403278e-06, | |
| "loss": 0.8134, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 1.1106406450271606, | |
| "learning_rate": 9.724314807191197e-06, | |
| "loss": 0.8359, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.611764705882353, | |
| "grad_norm": 0.9234170317649841, | |
| "learning_rate": 9.6769744288779e-06, | |
| "loss": 0.8229, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.6352941176470588, | |
| "grad_norm": 0.7843503355979919, | |
| "learning_rate": 9.626019287132202e-06, | |
| "loss": 0.7927, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.6588235294117647, | |
| "grad_norm": 0.8040333390235901, | |
| "learning_rate": 9.571488764419381e-06, | |
| "loss": 0.8129, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6823529411764706, | |
| "grad_norm": 0.7696279287338257, | |
| "learning_rate": 9.51342500656308e-06, | |
| "loss": 0.8572, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.6732362508773804, | |
| "learning_rate": 9.451872890171419e-06, | |
| "loss": 0.8103, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7294117647058823, | |
| "grad_norm": 0.6914750337600708, | |
| "learning_rate": 9.386879987952549e-06, | |
| "loss": 0.8969, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.7529411764705882, | |
| "grad_norm": 0.6257173418998718, | |
| "learning_rate": 9.318496531946411e-06, | |
| "loss": 0.818, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7764705882352941, | |
| "grad_norm": 0.7246304750442505, | |
| "learning_rate": 9.246775374701139e-06, | |
| "loss": 0.8332, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.7830196022987366, | |
| "learning_rate": 9.171771948424138e-06, | |
| "loss": 0.8585, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 0.653729259967804, | |
| "learning_rate": 9.093544222139338e-06, | |
| "loss": 0.8725, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.8470588235294118, | |
| "grad_norm": 0.742987871170044, | |
| "learning_rate": 9.012152656883824e-06, | |
| "loss": 0.785, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.8705882352941177, | |
| "grad_norm": 0.8727246522903442, | |
| "learning_rate": 8.927660158978392e-06, | |
| "loss": 0.8348, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8941176470588236, | |
| "grad_norm": 0.6196137070655823, | |
| "learning_rate": 8.84013203140821e-06, | |
| "loss": 0.8418, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.9176470588235294, | |
| "grad_norm": 0.610687255859375, | |
| "learning_rate": 8.749635923351108e-06, | |
| "loss": 0.776, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.6278626322746277, | |
| "learning_rate": 8.656241777892544e-06, | |
| "loss": 0.7207, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.9647058823529412, | |
| "grad_norm": 0.5417529344558716, | |
| "learning_rate": 8.56002177796765e-06, | |
| "loss": 0.7694, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.9882352941176471, | |
| "grad_norm": 0.5784769654273987, | |
| "learning_rate": 8.461050290572114e-06, | |
| "loss": 0.7733, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.0156862745098039, | |
| "grad_norm": 1.1008639335632324, | |
| "learning_rate": 8.359403809285054e-06, | |
| "loss": 1.312, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.0392156862745099, | |
| "grad_norm": 0.5411049723625183, | |
| "learning_rate": 8.255160895148263e-06, | |
| "loss": 0.7666, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.0627450980392157, | |
| "grad_norm": 0.602947473526001, | |
| "learning_rate": 8.14840211594757e-06, | |
| "loss": 0.7368, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0862745098039215, | |
| "grad_norm": 0.6235800385475159, | |
| "learning_rate": 8.039209983943201e-06, | |
| "loss": 0.7976, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.1098039215686275, | |
| "grad_norm": 0.569098174571991, | |
| "learning_rate": 7.927668892097288e-06, | |
| "loss": 0.7109, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.1333333333333333, | |
| "grad_norm": 0.5677081942558289, | |
| "learning_rate": 7.81386504884782e-06, | |
| "loss": 0.738, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.156862745098039, | |
| "grad_norm": 0.7025531530380249, | |
| "learning_rate": 7.697886411479422e-06, | |
| "loss": 0.8267, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.1803921568627451, | |
| "grad_norm": 0.5158477425575256, | |
| "learning_rate": 7.579822618142505e-06, | |
| "loss": 0.7993, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.203921568627451, | |
| "grad_norm": 0.7206972241401672, | |
| "learning_rate": 7.459764918573264e-06, | |
| "loss": 0.8324, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.227450980392157, | |
| "grad_norm": 0.5990767478942871, | |
| "learning_rate": 7.3378061035681415e-06, | |
| "loss": 0.73, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.2509803921568627, | |
| "grad_norm": 0.5361316204071045, | |
| "learning_rate": 7.2140404332671986e-06, | |
| "loss": 0.7399, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.2745098039215685, | |
| "grad_norm": 0.6394158601760864, | |
| "learning_rate": 7.088563564301874e-06, | |
| "loss": 0.8289, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.2980392156862746, | |
| "grad_norm": 0.5778906345367432, | |
| "learning_rate": 6.961472475863406e-06, | |
| "loss": 0.7455, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.3215686274509804, | |
| "grad_norm": 0.453545480966568, | |
| "learning_rate": 6.832865394749065e-06, | |
| "loss": 0.6694, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.3450980392156864, | |
| "grad_norm": 0.5790985226631165, | |
| "learning_rate": 6.702841719444141e-06, | |
| "loss": 0.8093, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.3686274509803922, | |
| "grad_norm": 0.48918935656547546, | |
| "learning_rate": 6.571501943298335e-06, | |
| "loss": 0.7096, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.392156862745098, | |
| "grad_norm": 0.6267134547233582, | |
| "learning_rate": 6.4389475768559675e-06, | |
| "loss": 0.814, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.415686274509804, | |
| "grad_norm": 0.4953418970108032, | |
| "learning_rate": 6.305281069399989e-06, | |
| "loss": 0.6618, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.4392156862745098, | |
| "grad_norm": 0.5042392611503601, | |
| "learning_rate": 6.17060572977047e-06, | |
| "loss": 0.6929, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.4627450980392158, | |
| "grad_norm": 0.5538609027862549, | |
| "learning_rate": 6.035025646518747e-06, | |
| "loss": 0.7561, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.4862745098039216, | |
| "grad_norm": 0.6071920394897461, | |
| "learning_rate": 5.898645607458941e-06, | |
| "loss": 0.7797, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.5098039215686274, | |
| "grad_norm": 0.5658022165298462, | |
| "learning_rate": 5.761571018679025e-06, | |
| "loss": 0.7374, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.5333333333333332, | |
| "grad_norm": 0.5699365139007568, | |
| "learning_rate": 5.623907823074044e-06, | |
| "loss": 0.8136, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.5568627450980392, | |
| "grad_norm": 0.47883421182632446, | |
| "learning_rate": 5.48576241846443e-06, | |
| "loss": 0.7933, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.5803921568627453, | |
| "grad_norm": 0.476575642824173, | |
| "learning_rate": 5.347241575362729e-06, | |
| "loss": 0.7209, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.603921568627451, | |
| "grad_norm": 0.5210950970649719, | |
| "learning_rate": 5.208452354452275e-06, | |
| "loss": 0.7747, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.6274509803921569, | |
| "grad_norm": 0.47601795196533203, | |
| "learning_rate": 5.069502023841576e-06, | |
| "loss": 0.7635, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.6509803921568627, | |
| "grad_norm": 0.521950364112854, | |
| "learning_rate": 4.9304979761584256e-06, | |
| "loss": 0.7707, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.6745098039215687, | |
| "grad_norm": 0.4452243745326996, | |
| "learning_rate": 4.791547645547727e-06, | |
| "loss": 0.6827, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.6980392156862745, | |
| "grad_norm": 0.5476846098899841, | |
| "learning_rate": 4.652758424637271e-06, | |
| "loss": 0.7939, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.7215686274509805, | |
| "grad_norm": 0.480186402797699, | |
| "learning_rate": 4.514237581535571e-06, | |
| "loss": 0.7367, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.7450980392156863, | |
| "grad_norm": 0.5076435804367065, | |
| "learning_rate": 4.3760921769259585e-06, | |
| "loss": 0.6935, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.768627450980392, | |
| "grad_norm": 0.5008230805397034, | |
| "learning_rate": 4.2384289813209754e-06, | |
| "loss": 0.7474, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.792156862745098, | |
| "grad_norm": 0.44636473059654236, | |
| "learning_rate": 4.101354392541061e-06, | |
| "loss": 0.7357, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.815686274509804, | |
| "grad_norm": 0.4537220597267151, | |
| "learning_rate": 3.964974353481254e-06, | |
| "loss": 0.7329, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.83921568627451, | |
| "grad_norm": 0.4459396302700043, | |
| "learning_rate": 3.829394270229531e-06, | |
| "loss": 0.7294, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.8627450980392157, | |
| "grad_norm": 0.3966139256954193, | |
| "learning_rate": 3.694718930600012e-06, | |
| "loss": 0.6419, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.8862745098039215, | |
| "grad_norm": 0.4331508278846741, | |
| "learning_rate": 3.5610524231440324e-06, | |
| "loss": 0.789, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.9098039215686273, | |
| "grad_norm": 0.44077518582344055, | |
| "learning_rate": 3.428498056701665e-06, | |
| "loss": 0.7499, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.9333333333333333, | |
| "grad_norm": 0.428218275308609, | |
| "learning_rate": 3.2971582805558622e-06, | |
| "loss": 0.7664, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.9568627450980394, | |
| "grad_norm": 0.45457276701927185, | |
| "learning_rate": 3.167134605250938e-06, | |
| "loss": 0.7651, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.9803921568627452, | |
| "grad_norm": 0.4581441283226013, | |
| "learning_rate": 3.0385275241365965e-06, | |
| "loss": 0.771, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.007843137254902, | |
| "grad_norm": 0.8593717813491821, | |
| "learning_rate": 2.9114364356981274e-06, | |
| "loss": 1.2373, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.0313725490196077, | |
| "grad_norm": 0.38274502754211426, | |
| "learning_rate": 2.7859595667328027e-06, | |
| "loss": 0.7255, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.0549019607843135, | |
| "grad_norm": 0.4255603551864624, | |
| "learning_rate": 2.6621938964318593e-06, | |
| "loss": 0.6407, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.0784313725490198, | |
| "grad_norm": 0.4372337758541107, | |
| "learning_rate": 2.5402350814267364e-06, | |
| "loss": 0.6874, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.1019607843137256, | |
| "grad_norm": 0.4747924506664276, | |
| "learning_rate": 2.4201773818574956e-06, | |
| "loss": 0.654, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.1254901960784314, | |
| "grad_norm": 0.45802468061447144, | |
| "learning_rate": 2.302113588520578e-06, | |
| "loss": 0.6809, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.149019607843137, | |
| "grad_norm": 0.39464399218559265, | |
| "learning_rate": 2.1861349511521817e-06, | |
| "loss": 0.6086, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.172549019607843, | |
| "grad_norm": 0.4720733165740967, | |
| "learning_rate": 2.072331107902713e-06, | |
| "loss": 0.9134, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.196078431372549, | |
| "grad_norm": 0.4298717677593231, | |
| "learning_rate": 1.960790016056801e-06, | |
| "loss": 0.6735, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.219607843137255, | |
| "grad_norm": 0.4402537941932678, | |
| "learning_rate": 1.8515978840524302e-06, | |
| "loss": 0.6972, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.243137254901961, | |
| "grad_norm": 0.45702147483825684, | |
| "learning_rate": 1.7448391048517378e-06, | |
| "loss": 0.7225, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.2666666666666666, | |
| "grad_norm": 0.4096600413322449, | |
| "learning_rate": 1.640596190714947e-06, | |
| "loss": 0.7225, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.2901960784313724, | |
| "grad_norm": 0.43652820587158203, | |
| "learning_rate": 1.5389497094278861e-06, | |
| "loss": 0.7208, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.313725490196078, | |
| "grad_norm": 0.44025924801826477, | |
| "learning_rate": 1.4399782220323515e-06, | |
| "loss": 0.6707, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.3372549019607844, | |
| "grad_norm": 0.5045623779296875, | |
| "learning_rate": 1.3437582221074574e-06, | |
| "loss": 0.7924, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.3607843137254902, | |
| "grad_norm": 0.48436567187309265, | |
| "learning_rate": 1.250364076648894e-06, | |
| "loss": 0.7384, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.384313725490196, | |
| "grad_norm": 0.3896447718143463, | |
| "learning_rate": 1.1598679685917901e-06, | |
| "loss": 0.6666, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.407843137254902, | |
| "grad_norm": 0.40689510107040405, | |
| "learning_rate": 1.0723398410216085e-06, | |
| "loss": 0.8292, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.431372549019608, | |
| "grad_norm": 0.40202853083610535, | |
| "learning_rate": 9.878473431161767e-07, | |
| "loss": 0.6669, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.454901960784314, | |
| "grad_norm": 0.3605956733226776, | |
| "learning_rate": 9.064557778606631e-07, | |
| "loss": 0.6017, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.4784313725490197, | |
| "grad_norm": 0.4316107928752899, | |
| "learning_rate": 8.282280515758639e-07, | |
| "loss": 0.7825, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.5019607843137255, | |
| "grad_norm": 0.46594148874282837, | |
| "learning_rate": 7.532246252988617e-07, | |
| "loss": 0.7446, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.5254901960784313, | |
| "grad_norm": 0.42160096764564514, | |
| "learning_rate": 6.815034680535915e-07, | |
| "loss": 0.713, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.549019607843137, | |
| "grad_norm": 0.4095713794231415, | |
| "learning_rate": 6.131200120474512e-07, | |
| "loss": 0.7409, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.572549019607843, | |
| "grad_norm": 0.40359240770339966, | |
| "learning_rate": 5.481271098285818e-07, | |
| "loss": 0.7501, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.596078431372549, | |
| "grad_norm": 0.3566288352012634, | |
| "learning_rate": 4.865749934369224e-07, | |
| "loss": 0.6082, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.619607843137255, | |
| "grad_norm": 0.417287141084671, | |
| "learning_rate": 4.2851123558061927e-07, | |
| "loss": 0.7517, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.6431372549019607, | |
| "grad_norm": 0.36074298620224, | |
| "learning_rate": 3.739807128677986e-07, | |
| "loss": 0.6589, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.37484005093574524, | |
| "learning_rate": 3.230255711220992e-07, | |
| "loss": 0.7009, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.6901960784313728, | |
| "grad_norm": 0.3959641754627228, | |
| "learning_rate": 2.756851928088056e-07, | |
| "loss": 0.7579, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.7137254901960786, | |
| "grad_norm": 0.3696475625038147, | |
| "learning_rate": 2.3199616659672352e-07, | |
| "loss": 0.8004, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.7372549019607844, | |
| "grad_norm": 0.3429110646247864, | |
| "learning_rate": 1.9199225907935492e-07, | |
| "loss": 0.6913, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.76078431372549, | |
| "grad_norm": 0.36563417315483093, | |
| "learning_rate": 1.5570438867719695e-07, | |
| "loss": 0.6749, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.784313725490196, | |
| "grad_norm": 0.41839203238487244, | |
| "learning_rate": 1.2316060174136e-07, | |
| "loss": 0.93, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.8078431372549018, | |
| "grad_norm": 0.329913854598999, | |
| "learning_rate": 9.43860508769645e-08, | |
| "loss": 0.5853, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.831372549019608, | |
| "grad_norm": 0.44181394577026367, | |
| "learning_rate": 6.940297550306895e-08, | |
| "loss": 0.7548, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.854901960784314, | |
| "grad_norm": 0.3787192702293396, | |
| "learning_rate": 4.823068466415615e-08, | |
| "loss": 0.7454, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.8784313725490196, | |
| "grad_norm": 0.4256850481033325, | |
| "learning_rate": 3.088554210646133e-08, | |
| "loss": 0.8, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.9019607843137254, | |
| "grad_norm": 0.3509824573993683, | |
| "learning_rate": 1.7380953630678488e-08, | |
| "loss": 0.7289, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.9254901960784316, | |
| "grad_norm": 0.4128064513206482, | |
| "learning_rate": 7.727356730820035e-09, | |
| "loss": 0.6974, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.9490196078431374, | |
| "grad_norm": 0.36814892292022705, | |
| "learning_rate": 1.9322125272297488e-09, | |
| "loss": 0.765, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.9725490196078432, | |
| "grad_norm": 0.3678930401802063, | |
| "learning_rate": 0.0, | |
| "loss": 0.6713, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.9725490196078432, | |
| "step": 126, | |
| "total_flos": 121055548211200.0, | |
| "train_loss": 0.7961941848671625, | |
| "train_runtime": 4090.1751, | |
| "train_samples_per_second": 2.985, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 126, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 121055548211200.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |