| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 471, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006369426751592357, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1e-05, | |
| "loss": 2.5992, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012738853503184714, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 9.987261146496817e-06, | |
| "loss": 2.332, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01910828025477707, | |
| "grad_norm": 4.625, | |
| "learning_rate": 9.97452229299363e-06, | |
| "loss": 2.1657, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025477707006369428, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.961783439490446e-06, | |
| "loss": 2.0276, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03184713375796178, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 9.949044585987262e-06, | |
| "loss": 2.1257, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03821656050955414, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 9.936305732484078e-06, | |
| "loss": 2.0139, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.044585987261146494, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 9.923566878980892e-06, | |
| "loss": 1.9839, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.050955414012738856, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 9.910828025477707e-06, | |
| "loss": 1.9895, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05732484076433121, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 9.898089171974523e-06, | |
| "loss": 2.0393, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06369426751592357, | |
| "grad_norm": 4.25, | |
| "learning_rate": 9.885350318471339e-06, | |
| "loss": 1.976, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07006369426751592, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 9.872611464968153e-06, | |
| "loss": 1.9623, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 9.859872611464969e-06, | |
| "loss": 2.0266, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08280254777070063, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 9.847133757961784e-06, | |
| "loss": 1.921, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08917197452229299, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 9.8343949044586e-06, | |
| "loss": 2.0458, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09554140127388536, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 9.821656050955414e-06, | |
| "loss": 1.8837, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10191082802547771, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 9.80891719745223e-06, | |
| "loss": 1.909, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.10828025477707007, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 9.796178343949046e-06, | |
| "loss": 1.808, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11464968152866242, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 9.783439490445861e-06, | |
| "loss": 1.8784, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12101910828025478, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 9.770700636942675e-06, | |
| "loss": 1.8299, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 9.757961783439491e-06, | |
| "loss": 1.8889, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1337579617834395, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 9.745222929936307e-06, | |
| "loss": 1.9387, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14012738853503184, | |
| "grad_norm": 3.25, | |
| "learning_rate": 9.732484076433122e-06, | |
| "loss": 1.867, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1464968152866242, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 9.719745222929937e-06, | |
| "loss": 1.8889, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 9.707006369426752e-06, | |
| "loss": 1.8178, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 9.694267515923568e-06, | |
| "loss": 1.851, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16560509554140126, | |
| "grad_norm": 2.875, | |
| "learning_rate": 9.681528662420384e-06, | |
| "loss": 1.8757, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17197452229299362, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 9.668789808917198e-06, | |
| "loss": 1.7925, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.17834394904458598, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 9.656050955414013e-06, | |
| "loss": 1.7666, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18471337579617833, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 9.64331210191083e-06, | |
| "loss": 1.7757, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 9.630573248407645e-06, | |
| "loss": 1.744, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19745222929936307, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 9.617834394904459e-06, | |
| "loss": 1.7344, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.20382165605095542, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 9.605095541401275e-06, | |
| "loss": 1.8763, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.21019108280254778, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 9.59235668789809e-06, | |
| "loss": 1.6856, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.21656050955414013, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 9.579617834394906e-06, | |
| "loss": 1.6951, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2229299363057325, | |
| "grad_norm": 4.25, | |
| "learning_rate": 9.56687898089172e-06, | |
| "loss": 1.9347, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 3.375, | |
| "learning_rate": 9.554140127388536e-06, | |
| "loss": 1.7508, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2356687898089172, | |
| "grad_norm": 3.125, | |
| "learning_rate": 9.541401273885352e-06, | |
| "loss": 1.6907, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.24203821656050956, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 9.528662420382167e-06, | |
| "loss": 1.5077, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2484076433121019, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 9.515923566878981e-06, | |
| "loss": 1.9329, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 9.503184713375797e-06, | |
| "loss": 1.8119, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2611464968152866, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 9.490445859872613e-06, | |
| "loss": 1.6772, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.267515923566879, | |
| "grad_norm": 4.0, | |
| "learning_rate": 9.477707006369428e-06, | |
| "loss": 1.7283, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.27388535031847133, | |
| "grad_norm": 4.0, | |
| "learning_rate": 9.464968152866242e-06, | |
| "loss": 1.71, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2802547770700637, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 9.452229299363058e-06, | |
| "loss": 1.8664, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.28662420382165604, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 9.439490445859874e-06, | |
| "loss": 1.5812, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2929936305732484, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 9.426751592356688e-06, | |
| "loss": 1.7476, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.29936305732484075, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 9.414012738853504e-06, | |
| "loss": 1.6807, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 9.40127388535032e-06, | |
| "loss": 1.6162, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.31210191082802546, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 9.388535031847135e-06, | |
| "loss": 1.7376, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 9.375796178343949e-06, | |
| "loss": 1.7971, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3248407643312102, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 9.363057324840765e-06, | |
| "loss": 1.611, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.33121019108280253, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 9.35031847133758e-06, | |
| "loss": 1.6623, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3375796178343949, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 9.337579617834396e-06, | |
| "loss": 1.9183, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.34394904458598724, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 9.32484076433121e-06, | |
| "loss": 1.5339, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3503184713375796, | |
| "grad_norm": 3.375, | |
| "learning_rate": 9.312101910828026e-06, | |
| "loss": 1.6644, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.35668789808917195, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 9.299363057324842e-06, | |
| "loss": 1.5716, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3630573248407643, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 9.286624203821658e-06, | |
| "loss": 1.5645, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.36942675159235666, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 9.273885350318472e-06, | |
| "loss": 1.6377, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.37579617834394907, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 9.261146496815287e-06, | |
| "loss": 1.6325, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 3.875, | |
| "learning_rate": 9.248407643312103e-06, | |
| "loss": 1.5807, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3885350318471338, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 9.235668789808919e-06, | |
| "loss": 1.4947, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.39490445859872614, | |
| "grad_norm": 3.390625, | |
| "learning_rate": 9.222929936305733e-06, | |
| "loss": 1.5759, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4012738853503185, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 9.210191082802548e-06, | |
| "loss": 1.5119, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.40764331210191085, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 9.197452229299364e-06, | |
| "loss": 1.5607, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4140127388535032, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 9.18471337579618e-06, | |
| "loss": 1.7135, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42038216560509556, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 9.171974522292994e-06, | |
| "loss": 1.3702, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4267515923566879, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 9.15923566878981e-06, | |
| "loss": 1.664, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.43312101910828027, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 9.146496815286625e-06, | |
| "loss": 1.5558, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4394904458598726, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 9.133757961783441e-06, | |
| "loss": 1.439, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.445859872611465, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 9.121019108280255e-06, | |
| "loss": 1.518, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.45222929936305734, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 9.10828025477707e-06, | |
| "loss": 1.4205, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 9.095541401273887e-06, | |
| "loss": 1.3655, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.46496815286624205, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 9.082802547770702e-06, | |
| "loss": 1.5735, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4713375796178344, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 9.070063694267516e-06, | |
| "loss": 1.4343, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 3.0, | |
| "learning_rate": 9.057324840764332e-06, | |
| "loss": 1.4682, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4840764331210191, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 9.044585987261148e-06, | |
| "loss": 1.3758, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.49044585987261147, | |
| "grad_norm": 5.375, | |
| "learning_rate": 9.031847133757963e-06, | |
| "loss": 1.5826, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4968152866242038, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 9.019108280254777e-06, | |
| "loss": 1.5303, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5031847133757962, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 9.006369426751593e-06, | |
| "loss": 1.2987, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5095541401273885, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 8.993630573248409e-06, | |
| "loss": 1.3336, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5159235668789809, | |
| "grad_norm": 3.25, | |
| "learning_rate": 8.980891719745225e-06, | |
| "loss": 1.4831, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5222929936305732, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 8.968152866242039e-06, | |
| "loss": 1.5189, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5286624203821656, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 8.955414012738854e-06, | |
| "loss": 1.3341, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 3.0, | |
| "learning_rate": 8.94267515923567e-06, | |
| "loss": 1.307, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5414012738853503, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 8.929936305732486e-06, | |
| "loss": 1.4342, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5477707006369427, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 8.9171974522293e-06, | |
| "loss": 1.3204, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.554140127388535, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 8.904458598726116e-06, | |
| "loss": 1.4072, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5605095541401274, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 8.891719745222931e-06, | |
| "loss": 1.4137, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5668789808917197, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 8.878980891719747e-06, | |
| "loss": 1.4142, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 8.866242038216561e-06, | |
| "loss": 1.476, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5796178343949044, | |
| "grad_norm": 3.625, | |
| "learning_rate": 8.853503184713377e-06, | |
| "loss": 1.2305, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5859872611464968, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 8.840764331210193e-06, | |
| "loss": 1.2832, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5923566878980892, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 8.828025477707007e-06, | |
| "loss": 1.501, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5987261146496815, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 8.815286624203822e-06, | |
| "loss": 1.3384, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6050955414012739, | |
| "grad_norm": 3.125, | |
| "learning_rate": 8.802547770700638e-06, | |
| "loss": 1.1737, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 8.789808917197454e-06, | |
| "loss": 1.271, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6178343949044586, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 8.777070063694268e-06, | |
| "loss": 1.4223, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6242038216560509, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 8.764331210191083e-06, | |
| "loss": 1.5493, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6305732484076433, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 8.7515923566879e-06, | |
| "loss": 1.3415, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 8.738853503184715e-06, | |
| "loss": 1.5326, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.643312101910828, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 8.726114649681529e-06, | |
| "loss": 1.2059, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6496815286624203, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.713375796178345e-06, | |
| "loss": 1.3436, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6560509554140127, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 8.70063694267516e-06, | |
| "loss": 1.4409, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6624203821656051, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 8.687898089171976e-06, | |
| "loss": 1.1457, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6687898089171974, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.67515923566879e-06, | |
| "loss": 1.5, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6751592356687898, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 8.662420382165606e-06, | |
| "loss": 1.578, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6815286624203821, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 8.649681528662422e-06, | |
| "loss": 1.4769, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 3.125, | |
| "learning_rate": 8.636942675159237e-06, | |
| "loss": 1.4304, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6942675159235668, | |
| "grad_norm": 2.75, | |
| "learning_rate": 8.624203821656051e-06, | |
| "loss": 1.195, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7006369426751592, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 8.611464968152867e-06, | |
| "loss": 1.6497, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7070063694267515, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 8.598726114649683e-06, | |
| "loss": 1.452, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7133757961783439, | |
| "grad_norm": 3.375, | |
| "learning_rate": 8.585987261146498e-06, | |
| "loss": 1.258, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7197452229299363, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 8.573248407643312e-06, | |
| "loss": 1.3933, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7261146496815286, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 8.560509554140128e-06, | |
| "loss": 1.4396, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.732484076433121, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 8.547770700636944e-06, | |
| "loss": 1.418, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7388535031847133, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 8.53503184713376e-06, | |
| "loss": 1.4755, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7452229299363057, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 8.522292993630574e-06, | |
| "loss": 1.3728, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7515923566878981, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 8.50955414012739e-06, | |
| "loss": 1.2451, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7579617834394905, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 8.496815286624205e-06, | |
| "loss": 1.3311, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 8.484076433121021e-06, | |
| "loss": 1.3561, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7707006369426752, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 8.471337579617835e-06, | |
| "loss": 1.0827, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7770700636942676, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 8.45859872611465e-06, | |
| "loss": 1.3941, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7834394904458599, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 8.445859872611466e-06, | |
| "loss": 1.3919, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7898089171974523, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 8.433121019108282e-06, | |
| "loss": 1.2135, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 8.420382165605096e-06, | |
| "loss": 1.3053, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.802547770700637, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 8.407643312101912e-06, | |
| "loss": 1.586, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8089171974522293, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 8.394904458598728e-06, | |
| "loss": 1.3054, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8152866242038217, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 8.382165605095543e-06, | |
| "loss": 1.0584, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.821656050955414, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 8.369426751592357e-06, | |
| "loss": 1.2344, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8280254777070064, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 8.356687898089173e-06, | |
| "loss": 1.3238, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8343949044585988, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 8.343949044585989e-06, | |
| "loss": 1.2827, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 8.331210191082804e-06, | |
| "loss": 1.2038, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8471337579617835, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 8.318471337579618e-06, | |
| "loss": 1.2496, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8535031847133758, | |
| "grad_norm": 3.390625, | |
| "learning_rate": 8.305732484076434e-06, | |
| "loss": 1.5994, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8598726114649682, | |
| "grad_norm": 3.125, | |
| "learning_rate": 8.29299363057325e-06, | |
| "loss": 1.2164, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8662420382165605, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 8.280254777070064e-06, | |
| "loss": 1.3794, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8726114649681529, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 8.26751592356688e-06, | |
| "loss": 1.2825, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8789808917197452, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 8.254777070063695e-06, | |
| "loss": 1.156, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8853503184713376, | |
| "grad_norm": 3.0, | |
| "learning_rate": 8.242038216560511e-06, | |
| "loss": 1.3448, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.89171974522293, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.229299363057325e-06, | |
| "loss": 1.2518, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8980891719745223, | |
| "grad_norm": 3.25, | |
| "learning_rate": 8.21656050955414e-06, | |
| "loss": 1.3553, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9044585987261147, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 8.203821656050957e-06, | |
| "loss": 1.5213, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.910828025477707, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 8.191082802547772e-06, | |
| "loss": 0.989, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 8.178343949044586e-06, | |
| "loss": 1.5073, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9235668789808917, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 8.165605095541402e-06, | |
| "loss": 1.2304, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9299363057324841, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 8.152866242038218e-06, | |
| "loss": 1.2064, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9363057324840764, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.140127388535033e-06, | |
| "loss": 1.2503, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9426751592356688, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 8.127388535031848e-06, | |
| "loss": 1.1985, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9490445859872612, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 8.114649681528663e-06, | |
| "loss": 1.3619, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 8.101910828025479e-06, | |
| "loss": 1.4085, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9617834394904459, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 8.089171974522295e-06, | |
| "loss": 1.21, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9681528662420382, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 8.076433121019109e-06, | |
| "loss": 1.3612, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9745222929936306, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 8.063694267515924e-06, | |
| "loss": 1.2914, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9808917197452229, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 8.05095541401274e-06, | |
| "loss": 1.2966, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.9872611464968153, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.038216560509556e-06, | |
| "loss": 1.2187, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9936305732484076, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 8.02547770700637e-06, | |
| "loss": 1.3362, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 8.012738853503186e-06, | |
| "loss": 1.2948, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.1024558544158936, | |
| "eval_runtime": 10.1101, | |
| "eval_samples_per_second": 98.911, | |
| "eval_steps_per_second": 3.165, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0063694267515924, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.1057, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0127388535031847, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 7.987261146496817e-06, | |
| "loss": 1.3991, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.019108280254777, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 7.974522292993631e-06, | |
| "loss": 1.3091, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0254777070063694, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 7.961783439490447e-06, | |
| "loss": 1.215, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0318471337579618, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 7.949044585987263e-06, | |
| "loss": 1.3057, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0382165605095541, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 7.936305732484078e-06, | |
| "loss": 1.0569, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0445859872611465, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 7.923566878980892e-06, | |
| "loss": 1.4477, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0509554140127388, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 7.910828025477708e-06, | |
| "loss": 1.2675, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0573248407643312, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 7.898089171974524e-06, | |
| "loss": 1.1781, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0636942675159236, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 7.88535031847134e-06, | |
| "loss": 1.3547, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.070063694267516, | |
| "grad_norm": 3.875, | |
| "learning_rate": 7.872611464968153e-06, | |
| "loss": 1.4928, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0764331210191083, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 7.85987261146497e-06, | |
| "loss": 1.2927, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0828025477707006, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 7.847133757961785e-06, | |
| "loss": 1.1105, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.089171974522293, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 7.8343949044586e-06, | |
| "loss": 1.1904, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.0955414012738853, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 7.821656050955415e-06, | |
| "loss": 1.2626, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1019108280254777, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 7.80891719745223e-06, | |
| "loss": 1.0206, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.10828025477707, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 7.796178343949046e-06, | |
| "loss": 1.3053, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 4.75, | |
| "learning_rate": 7.783439490445862e-06, | |
| "loss": 1.1811, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1210191082802548, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 7.770700636942676e-06, | |
| "loss": 1.2443, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.127388535031847, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 7.757961783439492e-06, | |
| "loss": 1.3596, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1337579617834395, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 7.745222929936307e-06, | |
| "loss": 1.1795, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1401273885350318, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 7.732484076433121e-06, | |
| "loss": 1.1409, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 7.719745222929937e-06, | |
| "loss": 1.1503, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1528662420382165, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 7.707006369426753e-06, | |
| "loss": 1.1541, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1592356687898089, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 7.694267515923568e-06, | |
| "loss": 1.1671, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1656050955414012, | |
| "grad_norm": 3.5, | |
| "learning_rate": 7.681528662420383e-06, | |
| "loss": 1.1447, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1719745222929936, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 7.668789808917198e-06, | |
| "loss": 1.157, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.178343949044586, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 7.656050955414014e-06, | |
| "loss": 1.2289, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1847133757961783, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 7.64331210191083e-06, | |
| "loss": 0.7996, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.1910828025477707, | |
| "grad_norm": 4.0, | |
| "learning_rate": 7.630573248407644e-06, | |
| "loss": 1.4251, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.197452229299363, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 7.6178343949044594e-06, | |
| "loss": 1.048, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2038216560509554, | |
| "grad_norm": 3.5, | |
| "learning_rate": 7.605095541401275e-06, | |
| "loss": 1.0727, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2101910828025477, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 7.59235668789809e-06, | |
| "loss": 1.2227, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.21656050955414, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 7.579617834394906e-06, | |
| "loss": 1.2099, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2229299363057324, | |
| "grad_norm": 3.25, | |
| "learning_rate": 7.566878980891721e-06, | |
| "loss": 0.9908, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2292993630573248, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 7.554140127388536e-06, | |
| "loss": 1.3231, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2356687898089171, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 7.541401273885351e-06, | |
| "loss": 1.2752, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2420382165605095, | |
| "grad_norm": 3.875, | |
| "learning_rate": 7.528662420382167e-06, | |
| "loss": 1.3112, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2484076433121019, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 7.515923566878982e-06, | |
| "loss": 1.4386, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.2547770700636942, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 7.5031847133757975e-06, | |
| "loss": 1.2486, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2611464968152866, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 7.4904458598726116e-06, | |
| "loss": 1.3977, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.267515923566879, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 7.4777070063694264e-06, | |
| "loss": 0.8955, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 7.464968152866242e-06, | |
| "loss": 1.1061, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2802547770700636, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 7.452229299363057e-06, | |
| "loss": 1.1828, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.286624203821656, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 7.439490445859873e-06, | |
| "loss": 1.1345, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.2929936305732483, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 7.426751592356688e-06, | |
| "loss": 1.081, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.2993630573248407, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 7.414012738853503e-06, | |
| "loss": 1.1087, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.305732484076433, | |
| "grad_norm": 3.625, | |
| "learning_rate": 7.401273885350318e-06, | |
| "loss": 1.01, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3121019108280254, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 7.388535031847134e-06, | |
| "loss": 1.0275, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.3184713375796178, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 7.375796178343949e-06, | |
| "loss": 1.1017, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3248407643312101, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 7.3630573248407645e-06, | |
| "loss": 0.8965, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3312101910828025, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 7.350318471337579e-06, | |
| "loss": 0.8589, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3375796178343948, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 7.337579617834395e-06, | |
| "loss": 1.1455, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3439490445859872, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 7.32484076433121e-06, | |
| "loss": 1.0583, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.3503184713375795, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 7.312101910828026e-06, | |
| "loss": 1.0971, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.356687898089172, | |
| "grad_norm": 3.875, | |
| "learning_rate": 7.299363057324841e-06, | |
| "loss": 1.3848, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3630573248407643, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 7.286624203821656e-06, | |
| "loss": 1.3677, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.3694267515923566, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 7.273885350318471e-06, | |
| "loss": 0.9285, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3757961783439492, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 7.261146496815287e-06, | |
| "loss": 1.1623, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.3821656050955413, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 7.248407643312102e-06, | |
| "loss": 1.0579, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.388535031847134, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 7.2356687898089175e-06, | |
| "loss": 1.0718, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.394904458598726, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 7.222929936305732e-06, | |
| "loss": 1.0097, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.4012738853503186, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 7.210191082802548e-06, | |
| "loss": 1.1821, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4076433121019107, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 7.197452229299363e-06, | |
| "loss": 1.1137, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.4140127388535033, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 7.184713375796179e-06, | |
| "loss": 1.0751, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4203821656050954, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 7.171974522292994e-06, | |
| "loss": 1.2457, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.426751592356688, | |
| "grad_norm": 4.5, | |
| "learning_rate": 7.159235668789809e-06, | |
| "loss": 1.003, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 7.146496815286624e-06, | |
| "loss": 1.1019, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4394904458598727, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 7.13375796178344e-06, | |
| "loss": 1.1613, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.4458598726114649, | |
| "grad_norm": 4.125, | |
| "learning_rate": 7.121019108280255e-06, | |
| "loss": 1.1851, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4522292993630574, | |
| "grad_norm": 4.375, | |
| "learning_rate": 7.1082802547770705e-06, | |
| "loss": 1.3668, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4585987261146496, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 7.095541401273885e-06, | |
| "loss": 1.0416, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4649681528662422, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 7.082802547770701e-06, | |
| "loss": 1.0807, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4713375796178343, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 7.070063694267516e-06, | |
| "loss": 1.1577, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.4777070063694269, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 7.057324840764332e-06, | |
| "loss": 1.1354, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.484076433121019, | |
| "grad_norm": 4.75, | |
| "learning_rate": 7.0445859872611466e-06, | |
| "loss": 1.0867, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4904458598726116, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 7.031847133757962e-06, | |
| "loss": 1.0491, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.4968152866242037, | |
| "grad_norm": 4.5, | |
| "learning_rate": 7.019108280254777e-06, | |
| "loss": 1.1881, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5031847133757963, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.006369426751593e-06, | |
| "loss": 1.0543, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.5095541401273884, | |
| "grad_norm": 4.125, | |
| "learning_rate": 6.993630573248408e-06, | |
| "loss": 1.1899, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.515923566878981, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 6.9808917197452235e-06, | |
| "loss": 1.0804, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5222929936305731, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 6.968152866242038e-06, | |
| "loss": 1.0973, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5286624203821657, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 6.955414012738853e-06, | |
| "loss": 1.0888, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5350318471337578, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.942675159235669e-06, | |
| "loss": 1.0908, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.5414012738853504, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 6.929936305732484e-06, | |
| "loss": 1.0698, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5477707006369426, | |
| "grad_norm": 4.0, | |
| "learning_rate": 6.9171974522292995e-06, | |
| "loss": 1.1819, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5541401273885351, | |
| "grad_norm": 5.625, | |
| "learning_rate": 6.904458598726114e-06, | |
| "loss": 1.3576, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5605095541401273, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 6.89171974522293e-06, | |
| "loss": 1.0149, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5668789808917198, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 6.878980891719745e-06, | |
| "loss": 0.7369, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.573248407643312, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 6.866242038216561e-06, | |
| "loss": 1.2456, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.5796178343949046, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 6.853503184713376e-06, | |
| "loss": 1.0428, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.5859872611464967, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 6.840764331210191e-06, | |
| "loss": 1.0207, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 6.828025477707006e-06, | |
| "loss": 1.145, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.5987261146496814, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 6.815286624203822e-06, | |
| "loss": 1.0502, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.605095541401274, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 6.802547770700637e-06, | |
| "loss": 0.9335, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.611464968152866, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 6.7898089171974525e-06, | |
| "loss": 1.0329, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6178343949044587, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 6.777070063694267e-06, | |
| "loss": 1.2377, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6242038216560508, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 6.764331210191083e-06, | |
| "loss": 0.9866, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6305732484076434, | |
| "grad_norm": 4.125, | |
| "learning_rate": 6.751592356687898e-06, | |
| "loss": 1.1617, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6369426751592355, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 6.738853503184714e-06, | |
| "loss": 1.0133, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.643312101910828, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 6.726114649681529e-06, | |
| "loss": 1.1279, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6496815286624202, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 6.713375796178344e-06, | |
| "loss": 1.063, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6560509554140128, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 6.700636942675159e-06, | |
| "loss": 0.6675, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.662420382165605, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 6.687898089171975e-06, | |
| "loss": 1.2324, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6687898089171975, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 6.67515923566879e-06, | |
| "loss": 0.9511, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.6751592356687897, | |
| "grad_norm": 3.875, | |
| "learning_rate": 6.6624203821656055e-06, | |
| "loss": 1.015, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.6815286624203822, | |
| "grad_norm": 4.375, | |
| "learning_rate": 6.64968152866242e-06, | |
| "loss": 1.242, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.6878980891719744, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 6.636942675159236e-06, | |
| "loss": 1.2436, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.694267515923567, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 6.624203821656051e-06, | |
| "loss": 0.862, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.700636942675159, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 6.611464968152867e-06, | |
| "loss": 0.9759, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.7070063694267517, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 6.5987261146496816e-06, | |
| "loss": 1.208, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7133757961783438, | |
| "grad_norm": 3.875, | |
| "learning_rate": 6.585987261146497e-06, | |
| "loss": 0.978, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.7197452229299364, | |
| "grad_norm": 4.625, | |
| "learning_rate": 6.573248407643312e-06, | |
| "loss": 1.1733, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7261146496815285, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 6.560509554140128e-06, | |
| "loss": 1.1611, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.732484076433121, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 6.547770700636943e-06, | |
| "loss": 1.1775, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.7388535031847132, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 6.5350318471337585e-06, | |
| "loss": 1.0329, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7452229299363058, | |
| "grad_norm": 4.375, | |
| "learning_rate": 6.522292993630573e-06, | |
| "loss": 1.0495, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 6.509554140127389e-06, | |
| "loss": 1.0536, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7579617834394905, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 6.496815286624204e-06, | |
| "loss": 0.8779, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.7643312101910829, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 6.48407643312102e-06, | |
| "loss": 0.8187, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7707006369426752, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 6.4713375796178345e-06, | |
| "loss": 1.0167, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.7770700636942676, | |
| "grad_norm": 4.625, | |
| "learning_rate": 6.45859872611465e-06, | |
| "loss": 1.3205, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.78343949044586, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 6.445859872611465e-06, | |
| "loss": 0.986, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7898089171974523, | |
| "grad_norm": 3.375, | |
| "learning_rate": 6.433121019108281e-06, | |
| "loss": 0.7035, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.7961783439490446, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 6.420382165605096e-06, | |
| "loss": 1.2722, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.802547770700637, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 6.407643312101911e-06, | |
| "loss": 0.9495, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.8089171974522293, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 6.394904458598726e-06, | |
| "loss": 1.1044, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8152866242038217, | |
| "grad_norm": 4.875, | |
| "learning_rate": 6.382165605095541e-06, | |
| "loss": 1.3434, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.821656050955414, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 6.369426751592357e-06, | |
| "loss": 0.9644, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8280254777070064, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 6.356687898089172e-06, | |
| "loss": 1.1358, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8343949044585988, | |
| "grad_norm": 4.5, | |
| "learning_rate": 6.3439490445859875e-06, | |
| "loss": 1.1577, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.8407643312101911, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 6.331210191082802e-06, | |
| "loss": 0.9626, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8471337579617835, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 6.318471337579618e-06, | |
| "loss": 1.2491, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8535031847133758, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.305732484076433e-06, | |
| "loss": 0.9339, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.8598726114649682, | |
| "grad_norm": 5.875, | |
| "learning_rate": 6.292993630573249e-06, | |
| "loss": 1.2729, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.8662420382165605, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 6.280254777070064e-06, | |
| "loss": 1.1192, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.872611464968153, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 6.267515923566879e-06, | |
| "loss": 0.9164, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.8789808917197452, | |
| "grad_norm": 4.375, | |
| "learning_rate": 6.254777070063694e-06, | |
| "loss": 0.9535, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.8853503184713376, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.24203821656051e-06, | |
| "loss": 0.89, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.89171974522293, | |
| "grad_norm": 4.5, | |
| "learning_rate": 6.229299363057325e-06, | |
| "loss": 1.0462, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.8980891719745223, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 6.2165605095541405e-06, | |
| "loss": 1.0853, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9044585987261147, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 6.203821656050955e-06, | |
| "loss": 0.9663, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 4.5, | |
| "learning_rate": 6.191082802547771e-06, | |
| "loss": 1.053, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9171974522292994, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 6.178343949044586e-06, | |
| "loss": 0.953, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.9235668789808917, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 6.165605095541402e-06, | |
| "loss": 1.0281, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.929936305732484, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 6.1528662420382166e-06, | |
| "loss": 0.8134, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9363057324840764, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 6.140127388535032e-06, | |
| "loss": 0.9999, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.9426751592356688, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 6.127388535031847e-06, | |
| "loss": 1.2659, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9490445859872612, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.114649681528663e-06, | |
| "loss": 1.2972, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.9554140127388535, | |
| "grad_norm": 4.625, | |
| "learning_rate": 6.101910828025478e-06, | |
| "loss": 0.972, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.9617834394904459, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 6.0891719745222935e-06, | |
| "loss": 1.1367, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.9681528662420382, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 6.076433121019108e-06, | |
| "loss": 1.1563, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.9745222929936306, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.063694267515924e-06, | |
| "loss": 1.174, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.980891719745223, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 6.050955414012739e-06, | |
| "loss": 0.9944, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.9872611464968153, | |
| "grad_norm": 4.125, | |
| "learning_rate": 6.038216560509555e-06, | |
| "loss": 0.9938, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.9936305732484076, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 6.0254777070063696e-06, | |
| "loss": 1.0508, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.0, | |
| "learning_rate": 6.012738853503185e-06, | |
| "loss": 0.8859, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.256985902786255, | |
| "eval_runtime": 9.4852, | |
| "eval_samples_per_second": 105.427, | |
| "eval_steps_per_second": 3.374, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.0063694267515926, | |
| "grad_norm": 5.5, | |
| "learning_rate": 6e-06, | |
| "loss": 1.1515, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.0127388535031847, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 5.987261146496816e-06, | |
| "loss": 1.1307, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.0191082802547773, | |
| "grad_norm": 4.25, | |
| "learning_rate": 5.974522292993631e-06, | |
| "loss": 0.7498, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.0254777070063694, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.9617834394904465e-06, | |
| "loss": 0.9773, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.031847133757962, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 5.949044585987261e-06, | |
| "loss": 1.0051, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.038216560509554, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 5.936305732484077e-06, | |
| "loss": 1.0551, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0445859872611467, | |
| "grad_norm": 5.75, | |
| "learning_rate": 5.923566878980892e-06, | |
| "loss": 0.8955, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.050955414012739, | |
| "grad_norm": 4.375, | |
| "learning_rate": 5.910828025477708e-06, | |
| "loss": 0.889, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.0573248407643314, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 5.8980891719745225e-06, | |
| "loss": 1.2016, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.0636942675159236, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 5.885350318471338e-06, | |
| "loss": 0.9772, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.872611464968153e-06, | |
| "loss": 0.8408, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.0764331210191083, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 5.859872611464969e-06, | |
| "loss": 0.967, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.082802547770701, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 5.847133757961784e-06, | |
| "loss": 0.7869, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.089171974522293, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 5.834394904458599e-06, | |
| "loss": 1.0154, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.0955414012738856, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 5.821656050955414e-06, | |
| "loss": 0.9255, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.1019108280254777, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 5.808917197452229e-06, | |
| "loss": 0.8462, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.1082802547770703, | |
| "grad_norm": 5.375, | |
| "learning_rate": 5.796178343949045e-06, | |
| "loss": 0.9195, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.1146496815286624, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.78343949044586e-06, | |
| "loss": 0.9329, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.121019108280255, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 5.7707006369426755e-06, | |
| "loss": 1.0013, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.127388535031847, | |
| "grad_norm": 5.375, | |
| "learning_rate": 5.75796178343949e-06, | |
| "loss": 0.9989, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.1337579617834397, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 5.745222929936306e-06, | |
| "loss": 0.9828, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.140127388535032, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.732484076433121e-06, | |
| "loss": 1.0511, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.1464968152866244, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 5.719745222929937e-06, | |
| "loss": 0.8785, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.1528662420382165, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 5.707006369426752e-06, | |
| "loss": 0.8857, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.159235668789809, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 5.694267515923567e-06, | |
| "loss": 0.8445, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.1656050955414012, | |
| "grad_norm": 5.875, | |
| "learning_rate": 5.681528662420382e-06, | |
| "loss": 0.8713, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.171974522292994, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.668789808917198e-06, | |
| "loss": 0.7253, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.178343949044586, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 5.656050955414013e-06, | |
| "loss": 0.8478, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.1847133757961785, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 5.6433121019108285e-06, | |
| "loss": 1.0531, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.1910828025477707, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 5.630573248407643e-06, | |
| "loss": 0.9684, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.1974522292993632, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 5.617834394904459e-06, | |
| "loss": 0.8787, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.2038216560509554, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 5.605095541401274e-06, | |
| "loss": 0.9049, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.210191082802548, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 5.59235668789809e-06, | |
| "loss": 0.8971, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.21656050955414, | |
| "grad_norm": 5.5, | |
| "learning_rate": 5.5796178343949046e-06, | |
| "loss": 0.8962, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.2229299363057327, | |
| "grad_norm": 5.625, | |
| "learning_rate": 5.56687898089172e-06, | |
| "loss": 0.962, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 6.0, | |
| "learning_rate": 5.554140127388535e-06, | |
| "loss": 0.9135, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.2356687898089174, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 5.541401273885351e-06, | |
| "loss": 0.6419, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.2420382165605095, | |
| "grad_norm": 5.875, | |
| "learning_rate": 5.528662420382166e-06, | |
| "loss": 0.9344, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.248407643312102, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 5.5159235668789815e-06, | |
| "loss": 0.9457, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.254777070063694, | |
| "grad_norm": 5.25, | |
| "learning_rate": 5.503184713375796e-06, | |
| "loss": 0.9527, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.261146496815287, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 5.490445859872612e-06, | |
| "loss": 0.9957, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.267515923566879, | |
| "grad_norm": 5.5, | |
| "learning_rate": 5.477707006369427e-06, | |
| "loss": 0.9117, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.2738853503184715, | |
| "grad_norm": 5.625, | |
| "learning_rate": 5.464968152866243e-06, | |
| "loss": 0.9524, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.2802547770700636, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.4522292993630575e-06, | |
| "loss": 0.9644, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.286624203821656, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 5.439490445859873e-06, | |
| "loss": 0.8749, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.2929936305732483, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 5.426751592356688e-06, | |
| "loss": 0.8673, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.299363057324841, | |
| "grad_norm": 5.0, | |
| "learning_rate": 5.414012738853504e-06, | |
| "loss": 0.7435, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.305732484076433, | |
| "grad_norm": 6.125, | |
| "learning_rate": 5.401273885350319e-06, | |
| "loss": 0.9566, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.3121019108280256, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 5.3885350318471345e-06, | |
| "loss": 0.9705, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.3184713375796178, | |
| "grad_norm": 5.125, | |
| "learning_rate": 5.375796178343949e-06, | |
| "loss": 0.882, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.3248407643312103, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 5.363057324840765e-06, | |
| "loss": 0.8602, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.3312101910828025, | |
| "grad_norm": 5.375, | |
| "learning_rate": 5.35031847133758e-06, | |
| "loss": 0.8275, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.337579617834395, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 5.337579617834396e-06, | |
| "loss": 0.7082, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.343949044585987, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.3248407643312105e-06, | |
| "loss": 1.077, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.3503184713375798, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 5.312101910828026e-06, | |
| "loss": 0.9805, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.356687898089172, | |
| "grad_norm": 5.75, | |
| "learning_rate": 5.299363057324841e-06, | |
| "loss": 0.9766, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.3630573248407645, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.286624203821657e-06, | |
| "loss": 0.7754, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.3694267515923566, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 5.273885350318472e-06, | |
| "loss": 0.8902, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.375796178343949, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.261146496815287e-06, | |
| "loss": 0.9486, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.3821656050955413, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 5.248407643312102e-06, | |
| "loss": 0.8804, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 5.235668789808917e-06, | |
| "loss": 0.9215, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.394904458598726, | |
| "grad_norm": 8.0, | |
| "learning_rate": 5.222929936305733e-06, | |
| "loss": 0.912, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.4012738853503186, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 5.210191082802548e-06, | |
| "loss": 0.8471, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.4076433121019107, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 5.1974522292993635e-06, | |
| "loss": 0.9813, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.4140127388535033, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 5.184713375796178e-06, | |
| "loss": 1.0401, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.4203821656050954, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 5.171974522292994e-06, | |
| "loss": 0.9632, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.426751592356688, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 5.159235668789809e-06, | |
| "loss": 0.7636, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.43312101910828, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 5.146496815286625e-06, | |
| "loss": 0.8122, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.4394904458598727, | |
| "grad_norm": 5.875, | |
| "learning_rate": 5.1337579617834396e-06, | |
| "loss": 0.9331, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.445859872611465, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 5.121019108280255e-06, | |
| "loss": 0.7487, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.4522292993630574, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 5.10828025477707e-06, | |
| "loss": 0.8742, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.4585987261146496, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 5.095541401273886e-06, | |
| "loss": 1.0264, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.464968152866242, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.082802547770701e-06, | |
| "loss": 0.7233, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.4713375796178343, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 5.0700636942675165e-06, | |
| "loss": 0.8406, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.477707006369427, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 5.057324840764331e-06, | |
| "loss": 0.9097, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.484076433121019, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 5.044585987261147e-06, | |
| "loss": 1.0201, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.4904458598726116, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.031847133757962e-06, | |
| "loss": 0.8389, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.4968152866242037, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 5.019108280254778e-06, | |
| "loss": 0.8662, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.5031847133757963, | |
| "grad_norm": 5.75, | |
| "learning_rate": 5.0063694267515925e-06, | |
| "loss": 0.962, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.5095541401273884, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 4.993630573248408e-06, | |
| "loss": 0.8663, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.515923566878981, | |
| "grad_norm": 5.5, | |
| "learning_rate": 4.980891719745223e-06, | |
| "loss": 0.7373, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.522292993630573, | |
| "grad_norm": 5.375, | |
| "learning_rate": 4.968152866242039e-06, | |
| "loss": 0.7439, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.5286624203821657, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 4.955414012738854e-06, | |
| "loss": 0.907, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.535031847133758, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.9426751592356695e-06, | |
| "loss": 0.913, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.5414012738853504, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 4.929936305732484e-06, | |
| "loss": 0.91, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.9171974522293e-06, | |
| "loss": 0.7951, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.554140127388535, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.904458598726115e-06, | |
| "loss": 0.9046, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.5605095541401273, | |
| "grad_norm": 8.875, | |
| "learning_rate": 4.891719745222931e-06, | |
| "loss": 0.9787, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.56687898089172, | |
| "grad_norm": 6.0, | |
| "learning_rate": 4.8789808917197455e-06, | |
| "loss": 0.9956, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.573248407643312, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 4.866242038216561e-06, | |
| "loss": 0.7218, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.5796178343949046, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 4.853503184713376e-06, | |
| "loss": 0.7912, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.5859872611464967, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.840764331210192e-06, | |
| "loss": 0.9452, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.5923566878980893, | |
| "grad_norm": 5.875, | |
| "learning_rate": 4.828025477707007e-06, | |
| "loss": 0.8666, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.5987261146496814, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 4.8152866242038224e-06, | |
| "loss": 0.7909, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.605095541401274, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.802547770700637e-06, | |
| "loss": 1.0329, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.611464968152866, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.789808917197453e-06, | |
| "loss": 0.8773, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.6178343949044587, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.777070063694268e-06, | |
| "loss": 1.0021, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.624203821656051, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.764331210191084e-06, | |
| "loss": 0.9092, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.6305732484076434, | |
| "grad_norm": 5.875, | |
| "learning_rate": 4.7515923566878985e-06, | |
| "loss": 0.8987, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.6369426751592355, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.738853503184714e-06, | |
| "loss": 0.8914, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.643312101910828, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.726114649681529e-06, | |
| "loss": 0.8273, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.6496815286624202, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.713375796178344e-06, | |
| "loss": 0.858, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.656050955414013, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 4.70063694267516e-06, | |
| "loss": 0.8168, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.662420382165605, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.6878980891719746e-06, | |
| "loss": 0.9304, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.6687898089171975, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.67515923566879e-06, | |
| "loss": 0.7596, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.6751592356687897, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 4.662420382165605e-06, | |
| "loss": 0.8506, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.6815286624203822, | |
| "grad_norm": 7.625, | |
| "learning_rate": 4.649681528662421e-06, | |
| "loss": 1.0233, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.6878980891719744, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 4.636942675159236e-06, | |
| "loss": 0.7036, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.694267515923567, | |
| "grad_norm": 7.5, | |
| "learning_rate": 4.6242038216560515e-06, | |
| "loss": 1.0356, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.700636942675159, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 4.611464968152866e-06, | |
| "loss": 0.8359, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.7070063694267517, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.598726114649682e-06, | |
| "loss": 0.8872, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.713375796178344, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 4.585987261146497e-06, | |
| "loss": 0.6886, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.7197452229299364, | |
| "grad_norm": 10.0, | |
| "learning_rate": 4.573248407643313e-06, | |
| "loss": 0.9165, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.7261146496815285, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.5605095541401276e-06, | |
| "loss": 0.9986, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.732484076433121, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.547770700636943e-06, | |
| "loss": 0.8863, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.738853503184713, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.535031847133758e-06, | |
| "loss": 0.735, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.745222929936306, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.522292993630574e-06, | |
| "loss": 0.8812, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.7515923566878984, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.509554140127389e-06, | |
| "loss": 0.6897, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.7579617834394905, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.4968152866242045e-06, | |
| "loss": 1.0259, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.7643312101910826, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.484076433121019e-06, | |
| "loss": 0.824, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.770700636942675, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.471337579617835e-06, | |
| "loss": 0.8058, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.777070063694268, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 4.45859872611465e-06, | |
| "loss": 0.7136, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.78343949044586, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.445859872611466e-06, | |
| "loss": 1.0948, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.789808917197452, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 4.4331210191082805e-06, | |
| "loss": 0.7173, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.7961783439490446, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 4.420382165605096e-06, | |
| "loss": 0.9048, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.802547770700637, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 4.407643312101911e-06, | |
| "loss": 0.7007, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.8089171974522293, | |
| "grad_norm": 5.625, | |
| "learning_rate": 4.394904458598727e-06, | |
| "loss": 0.7283, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.8152866242038215, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.382165605095542e-06, | |
| "loss": 0.8633, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.821656050955414, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.3694267515923574e-06, | |
| "loss": 0.7952, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.8280254777070066, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 4.356687898089172e-06, | |
| "loss": 0.9529, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.8343949044585988, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.343949044585988e-06, | |
| "loss": 0.7239, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.840764331210191, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.331210191082803e-06, | |
| "loss": 0.928, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.8471337579617835, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.318471337579619e-06, | |
| "loss": 0.9538, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.853503184713376, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.3057324840764335e-06, | |
| "loss": 0.826, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.859872611464968, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.292993630573249e-06, | |
| "loss": 0.8187, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 7.625, | |
| "learning_rate": 4.280254777070064e-06, | |
| "loss": 1.0319, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.872611464968153, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 4.26751592356688e-06, | |
| "loss": 0.7012, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.8789808917197455, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 4.254777070063695e-06, | |
| "loss": 0.7482, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.8853503184713376, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 4.2420382165605104e-06, | |
| "loss": 0.8301, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.8917197452229297, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 4.229299363057325e-06, | |
| "loss": 0.8581, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.8980891719745223, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.216560509554141e-06, | |
| "loss": 0.8197, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.904458598726115, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.203821656050956e-06, | |
| "loss": 0.8834, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.910828025477707, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.191082802547772e-06, | |
| "loss": 0.7329, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.917197452229299, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 4.1783439490445865e-06, | |
| "loss": 1.0238, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.9235668789808917, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 4.165605095541402e-06, | |
| "loss": 0.8598, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.9299363057324843, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.152866242038217e-06, | |
| "loss": 0.9251, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.9363057324840764, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.140127388535032e-06, | |
| "loss": 0.9192, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.9426751592356686, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.127388535031848e-06, | |
| "loss": 0.8875, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.949044585987261, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 4.1146496815286626e-06, | |
| "loss": 0.6841, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.9554140127388537, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.101910828025478e-06, | |
| "loss": 0.7161, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.961783439490446, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.089171974522293e-06, | |
| "loss": 0.9488, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.968152866242038, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 4.076433121019109e-06, | |
| "loss": 1.0557, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.9745222929936306, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 4.063694267515924e-06, | |
| "loss": 0.808, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.980891719745223, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.0509554140127395e-06, | |
| "loss": 0.8483, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.9872611464968153, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.038216560509554e-06, | |
| "loss": 0.738, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.9936305732484074, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 4.02547770700637e-06, | |
| "loss": 0.7864, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 12.875, | |
| "learning_rate": 4.012738853503185e-06, | |
| "loss": 0.7464, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.4188008308410645, | |
| "eval_runtime": 9.5026, | |
| "eval_samples_per_second": 105.234, | |
| "eval_steps_per_second": 3.367, | |
| "step": 471 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 785, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.940873567971008e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |