| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 628, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006369426751592357, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1e-05, | |
| "loss": 2.464, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012738853503184714, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 9.987261146496817e-06, | |
| "loss": 2.2779, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01910828025477707, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 9.97452229299363e-06, | |
| "loss": 2.3239, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025477707006369428, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 9.961783439490446e-06, | |
| "loss": 2.2466, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03184713375796178, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 9.949044585987262e-06, | |
| "loss": 2.2339, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03821656050955414, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 9.936305732484078e-06, | |
| "loss": 2.1748, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.044585987261146494, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 9.923566878980892e-06, | |
| "loss": 2.1675, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.050955414012738856, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 9.910828025477707e-06, | |
| "loss": 2.0934, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05732484076433121, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 9.898089171974523e-06, | |
| "loss": 2.0622, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06369426751592357, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 9.885350318471339e-06, | |
| "loss": 2.125, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07006369426751592, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 9.872611464968153e-06, | |
| "loss": 2.0822, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 9.859872611464969e-06, | |
| "loss": 2.0995, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08280254777070063, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 9.847133757961784e-06, | |
| "loss": 1.9927, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08917197452229299, | |
| "grad_norm": 3.25, | |
| "learning_rate": 9.8343949044586e-06, | |
| "loss": 1.9817, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09554140127388536, | |
| "grad_norm": 4.125, | |
| "learning_rate": 9.821656050955414e-06, | |
| "loss": 2.0455, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10191082802547771, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 9.80891719745223e-06, | |
| "loss": 1.9892, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.10828025477707007, | |
| "grad_norm": 3.875, | |
| "learning_rate": 9.796178343949046e-06, | |
| "loss": 1.9728, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11464968152866242, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 9.783439490445861e-06, | |
| "loss": 2.0389, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12101910828025478, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 9.770700636942675e-06, | |
| "loss": 1.9921, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 2.875, | |
| "learning_rate": 9.757961783439491e-06, | |
| "loss": 1.9869, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1337579617834395, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 9.745222929936307e-06, | |
| "loss": 2.0115, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14012738853503184, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 9.732484076433122e-06, | |
| "loss": 2.0177, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1464968152866242, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 9.719745222929937e-06, | |
| "loss": 1.9454, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 9.707006369426752e-06, | |
| "loss": 1.8294, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 9.694267515923568e-06, | |
| "loss": 2.0004, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16560509554140126, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 9.681528662420384e-06, | |
| "loss": 1.8173, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17197452229299362, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 9.668789808917198e-06, | |
| "loss": 1.8761, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.17834394904458598, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 9.656050955414013e-06, | |
| "loss": 2.0223, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18471337579617833, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 9.64331210191083e-06, | |
| "loss": 2.0232, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 9.630573248407645e-06, | |
| "loss": 2.0382, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19745222929936307, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 9.617834394904459e-06, | |
| "loss": 1.9112, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.20382165605095542, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 9.605095541401275e-06, | |
| "loss": 1.7688, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.21019108280254778, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 9.59235668789809e-06, | |
| "loss": 1.8924, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.21656050955414013, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 9.579617834394906e-06, | |
| "loss": 1.7671, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2229299363057325, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 9.56687898089172e-06, | |
| "loss": 1.7464, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 9.554140127388536e-06, | |
| "loss": 1.7568, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2356687898089172, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 9.541401273885352e-06, | |
| "loss": 1.7649, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.24203821656050956, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 9.528662420382167e-06, | |
| "loss": 1.9104, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2484076433121019, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 9.515923566878981e-06, | |
| "loss": 1.7278, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 9.503184713375797e-06, | |
| "loss": 1.7522, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2611464968152866, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 9.490445859872613e-06, | |
| "loss": 1.8213, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.267515923566879, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9.477707006369428e-06, | |
| "loss": 1.7487, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.27388535031847133, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 9.464968152866242e-06, | |
| "loss": 1.8848, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2802547770700637, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 9.452229299363058e-06, | |
| "loss": 1.7972, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.28662420382165604, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 9.439490445859874e-06, | |
| "loss": 1.6838, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2929936305732484, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 9.426751592356688e-06, | |
| "loss": 1.6363, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.29936305732484075, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 9.414012738853504e-06, | |
| "loss": 1.817, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 9.40127388535032e-06, | |
| "loss": 1.4969, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.31210191082802546, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 9.388535031847135e-06, | |
| "loss": 1.6618, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 9.375796178343949e-06, | |
| "loss": 1.9315, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3248407643312102, | |
| "grad_norm": 4.5, | |
| "learning_rate": 9.363057324840765e-06, | |
| "loss": 1.7195, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.33121019108280253, | |
| "grad_norm": 4.375, | |
| "learning_rate": 9.35031847133758e-06, | |
| "loss": 1.8091, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3375796178343949, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 9.337579617834396e-06, | |
| "loss": 1.6198, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.34394904458598724, | |
| "grad_norm": 4.875, | |
| "learning_rate": 9.32484076433121e-06, | |
| "loss": 1.9276, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3503184713375796, | |
| "grad_norm": 7.375, | |
| "learning_rate": 9.312101910828026e-06, | |
| "loss": 1.5707, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.35668789808917195, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 9.299363057324842e-06, | |
| "loss": 1.8087, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3630573248407643, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 9.286624203821658e-06, | |
| "loss": 1.7809, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.36942675159235666, | |
| "grad_norm": 3.5, | |
| "learning_rate": 9.273885350318472e-06, | |
| "loss": 1.7255, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.37579617834394907, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 9.261146496815287e-06, | |
| "loss": 1.694, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 6.0, | |
| "learning_rate": 9.248407643312103e-06, | |
| "loss": 1.761, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3885350318471338, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 9.235668789808919e-06, | |
| "loss": 1.556, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.39490445859872614, | |
| "grad_norm": 3.25, | |
| "learning_rate": 9.222929936305733e-06, | |
| "loss": 1.8402, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4012738853503185, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 9.210191082802548e-06, | |
| "loss": 1.7088, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.40764331210191085, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 9.197452229299364e-06, | |
| "loss": 1.5687, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4140127388535032, | |
| "grad_norm": 3.875, | |
| "learning_rate": 9.18471337579618e-06, | |
| "loss": 1.648, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42038216560509556, | |
| "grad_norm": 4.0, | |
| "learning_rate": 9.171974522292994e-06, | |
| "loss": 1.5228, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4267515923566879, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 9.15923566878981e-06, | |
| "loss": 1.758, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.43312101910828027, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 9.146496815286625e-06, | |
| "loss": 1.4379, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4394904458598726, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 9.133757961783441e-06, | |
| "loss": 1.591, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.445859872611465, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 9.121019108280255e-06, | |
| "loss": 1.5161, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.45222929936305734, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 9.10828025477707e-06, | |
| "loss": 1.6205, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 4.375, | |
| "learning_rate": 9.095541401273887e-06, | |
| "loss": 1.5561, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.46496815286624205, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 9.082802547770702e-06, | |
| "loss": 1.7969, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4713375796178344, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 9.070063694267516e-06, | |
| "loss": 1.5636, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 9.057324840764332e-06, | |
| "loss": 1.5697, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4840764331210191, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 9.044585987261148e-06, | |
| "loss": 1.6277, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.49044585987261147, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 9.031847133757963e-06, | |
| "loss": 1.5438, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4968152866242038, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 9.019108280254777e-06, | |
| "loss": 1.5346, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5031847133757962, | |
| "grad_norm": 3.5, | |
| "learning_rate": 9.006369426751593e-06, | |
| "loss": 1.8924, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5095541401273885, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 8.993630573248409e-06, | |
| "loss": 1.5297, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5159235668789809, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 8.980891719745225e-06, | |
| "loss": 1.5757, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5222929936305732, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 8.968152866242039e-06, | |
| "loss": 1.4139, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5286624203821656, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 8.955414012738854e-06, | |
| "loss": 1.8459, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 8.94267515923567e-06, | |
| "loss": 1.5654, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5414012738853503, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 8.929936305732486e-06, | |
| "loss": 1.5617, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5477707006369427, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 8.9171974522293e-06, | |
| "loss": 1.6943, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.554140127388535, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 8.904458598726116e-06, | |
| "loss": 1.4783, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5605095541401274, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 8.891719745222931e-06, | |
| "loss": 1.9107, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5668789808917197, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 8.878980891719747e-06, | |
| "loss": 1.7035, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 8.866242038216561e-06, | |
| "loss": 1.6064, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5796178343949044, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 8.853503184713377e-06, | |
| "loss": 1.5809, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5859872611464968, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 8.840764331210193e-06, | |
| "loss": 1.4688, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5923566878980892, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 8.828025477707007e-06, | |
| "loss": 1.6023, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5987261146496815, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 8.815286624203822e-06, | |
| "loss": 1.2422, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6050955414012739, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 8.802547770700638e-06, | |
| "loss": 1.3482, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 3.125, | |
| "learning_rate": 8.789808917197454e-06, | |
| "loss": 1.4544, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6178343949044586, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 8.777070063694268e-06, | |
| "loss": 1.4195, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6242038216560509, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 8.764331210191083e-06, | |
| "loss": 1.5878, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6305732484076433, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 8.7515923566879e-06, | |
| "loss": 1.6565, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 8.738853503184715e-06, | |
| "loss": 1.4203, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.643312101910828, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 8.726114649681529e-06, | |
| "loss": 1.3304, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6496815286624203, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 8.713375796178345e-06, | |
| "loss": 1.3758, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6560509554140127, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 8.70063694267516e-06, | |
| "loss": 1.3662, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6624203821656051, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 8.687898089171976e-06, | |
| "loss": 1.3483, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6687898089171974, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.67515923566879e-06, | |
| "loss": 1.3097, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6751592356687898, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 8.662420382165606e-06, | |
| "loss": 1.3437, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6815286624203821, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.649681528662422e-06, | |
| "loss": 1.4274, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 8.636942675159237e-06, | |
| "loss": 1.3018, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6942675159235668, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.624203821656051e-06, | |
| "loss": 1.6072, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7006369426751592, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.611464968152867e-06, | |
| "loss": 1.5044, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7070063694267515, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 8.598726114649683e-06, | |
| "loss": 1.4844, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7133757961783439, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 8.585987261146498e-06, | |
| "loss": 1.365, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7197452229299363, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 8.573248407643312e-06, | |
| "loss": 1.4717, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7261146496815286, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.560509554140128e-06, | |
| "loss": 1.5203, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.732484076433121, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 8.547770700636944e-06, | |
| "loss": 1.5183, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7388535031847133, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 8.53503184713376e-06, | |
| "loss": 1.4564, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7452229299363057, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 8.522292993630574e-06, | |
| "loss": 1.2536, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7515923566878981, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 8.50955414012739e-06, | |
| "loss": 1.1554, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7579617834394905, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 8.496815286624205e-06, | |
| "loss": 1.5978, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 8.484076433121021e-06, | |
| "loss": 1.6433, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7707006369426752, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 8.471337579617835e-06, | |
| "loss": 1.4163, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7770700636942676, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 8.45859872611465e-06, | |
| "loss": 1.468, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7834394904458599, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 8.445859872611466e-06, | |
| "loss": 1.2846, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7898089171974523, | |
| "grad_norm": 3.25, | |
| "learning_rate": 8.433121019108282e-06, | |
| "loss": 1.378, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 8.420382165605096e-06, | |
| "loss": 1.1916, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.802547770700637, | |
| "grad_norm": 4.375, | |
| "learning_rate": 8.407643312101912e-06, | |
| "loss": 1.1367, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8089171974522293, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 8.394904458598728e-06, | |
| "loss": 1.4716, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8152866242038217, | |
| "grad_norm": 3.75, | |
| "learning_rate": 8.382165605095543e-06, | |
| "loss": 1.694, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.821656050955414, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 8.369426751592357e-06, | |
| "loss": 1.4589, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8280254777070064, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 8.356687898089173e-06, | |
| "loss": 1.2544, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8343949044585988, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 8.343949044585989e-06, | |
| "loss": 1.6617, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 8.331210191082804e-06, | |
| "loss": 1.6309, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8471337579617835, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 8.318471337579618e-06, | |
| "loss": 1.6673, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8535031847133758, | |
| "grad_norm": 3.625, | |
| "learning_rate": 8.305732484076434e-06, | |
| "loss": 1.5529, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8598726114649682, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 8.29299363057325e-06, | |
| "loss": 1.1977, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8662420382165605, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 8.280254777070064e-06, | |
| "loss": 1.5706, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8726114649681529, | |
| "grad_norm": 3.75, | |
| "learning_rate": 8.26751592356688e-06, | |
| "loss": 1.5236, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8789808917197452, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 8.254777070063695e-06, | |
| "loss": 1.4243, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8853503184713376, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 8.242038216560511e-06, | |
| "loss": 1.2357, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.89171974522293, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 8.229299363057325e-06, | |
| "loss": 1.2332, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8980891719745223, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 8.21656050955414e-06, | |
| "loss": 1.4376, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9044585987261147, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.203821656050957e-06, | |
| "loss": 1.1844, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.910828025477707, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 8.191082802547772e-06, | |
| "loss": 1.6206, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.178343949044586e-06, | |
| "loss": 1.4769, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9235668789808917, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 8.165605095541402e-06, | |
| "loss": 1.6485, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9299363057324841, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 8.152866242038218e-06, | |
| "loss": 1.4478, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9363057324840764, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 8.140127388535033e-06, | |
| "loss": 1.341, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9426751592356688, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 8.127388535031848e-06, | |
| "loss": 1.5807, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9490445859872612, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 8.114649681528663e-06, | |
| "loss": 1.4992, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 8.101910828025479e-06, | |
| "loss": 1.2082, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9617834394904459, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 8.089171974522295e-06, | |
| "loss": 1.459, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9681528662420382, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 8.076433121019109e-06, | |
| "loss": 1.3577, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9745222929936306, | |
| "grad_norm": 2.875, | |
| "learning_rate": 8.063694267515924e-06, | |
| "loss": 1.1222, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9808917197452229, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 8.05095541401274e-06, | |
| "loss": 1.3321, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.9872611464968153, | |
| "grad_norm": 3.625, | |
| "learning_rate": 8.038216560509556e-06, | |
| "loss": 1.5208, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9936305732484076, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 8.02547770700637e-06, | |
| "loss": 1.2333, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 8.012738853503186e-06, | |
| "loss": 1.7693, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.246551752090454, | |
| "eval_runtime": 9.531, | |
| "eval_samples_per_second": 104.921, | |
| "eval_steps_per_second": 3.357, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0063694267515924, | |
| "grad_norm": 4.25, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.1428, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0127388535031847, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 7.987261146496817e-06, | |
| "loss": 1.1755, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.019108280254777, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 7.974522292993631e-06, | |
| "loss": 1.2935, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0254777070063694, | |
| "grad_norm": 3.5, | |
| "learning_rate": 7.961783439490447e-06, | |
| "loss": 1.4293, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0318471337579618, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 7.949044585987263e-06, | |
| "loss": 1.2692, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0382165605095541, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 7.936305732484078e-06, | |
| "loss": 1.2946, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0445859872611465, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 7.923566878980892e-06, | |
| "loss": 1.0998, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0509554140127388, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 7.910828025477708e-06, | |
| "loss": 1.4483, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0573248407643312, | |
| "grad_norm": 4.75, | |
| "learning_rate": 7.898089171974524e-06, | |
| "loss": 1.4334, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0636942675159236, | |
| "grad_norm": 3.5, | |
| "learning_rate": 7.88535031847134e-06, | |
| "loss": 1.2831, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.070063694267516, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 7.872611464968153e-06, | |
| "loss": 1.1684, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0764331210191083, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 7.85987261146497e-06, | |
| "loss": 1.4103, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0828025477707006, | |
| "grad_norm": 14.5, | |
| "learning_rate": 7.847133757961785e-06, | |
| "loss": 1.4969, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.089171974522293, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 7.8343949044586e-06, | |
| "loss": 1.3231, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.0955414012738853, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 7.821656050955415e-06, | |
| "loss": 1.1973, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1019108280254777, | |
| "grad_norm": 3.75, | |
| "learning_rate": 7.80891719745223e-06, | |
| "loss": 1.2171, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.10828025477707, | |
| "grad_norm": 4.125, | |
| "learning_rate": 7.796178343949046e-06, | |
| "loss": 1.2995, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 3.75, | |
| "learning_rate": 7.783439490445862e-06, | |
| "loss": 1.3068, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1210191082802548, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 7.770700636942676e-06, | |
| "loss": 1.0732, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.127388535031847, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 7.757961783439492e-06, | |
| "loss": 1.3082, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1337579617834395, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 7.745222929936307e-06, | |
| "loss": 1.2516, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1401273885350318, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 7.732484076433121e-06, | |
| "loss": 1.2596, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 7.719745222929937e-06, | |
| "loss": 1.3712, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1528662420382165, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 7.707006369426753e-06, | |
| "loss": 1.4099, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1592356687898089, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 7.694267515923568e-06, | |
| "loss": 1.0945, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1656050955414012, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 7.681528662420383e-06, | |
| "loss": 1.3058, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1719745222929936, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 7.668789808917198e-06, | |
| "loss": 1.3675, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.178343949044586, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 7.656050955414014e-06, | |
| "loss": 1.237, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1847133757961783, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 7.64331210191083e-06, | |
| "loss": 1.4323, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.1910828025477707, | |
| "grad_norm": 4.625, | |
| "learning_rate": 7.630573248407644e-06, | |
| "loss": 1.2617, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.197452229299363, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 7.6178343949044594e-06, | |
| "loss": 1.3992, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2038216560509554, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 7.605095541401275e-06, | |
| "loss": 1.4794, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2101910828025477, | |
| "grad_norm": 3.5, | |
| "learning_rate": 7.59235668789809e-06, | |
| "loss": 0.8253, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.21656050955414, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 7.579617834394906e-06, | |
| "loss": 1.1067, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2229299363057324, | |
| "grad_norm": 4.25, | |
| "learning_rate": 7.566878980891721e-06, | |
| "loss": 1.3815, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2292993630573248, | |
| "grad_norm": 3.5, | |
| "learning_rate": 7.554140127388536e-06, | |
| "loss": 0.8663, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2356687898089171, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 7.541401273885351e-06, | |
| "loss": 1.0237, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2420382165605095, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.528662420382167e-06, | |
| "loss": 1.3546, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2484076433121019, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 7.515923566878982e-06, | |
| "loss": 1.2349, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.2547770700636942, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.5031847133757975e-06, | |
| "loss": 1.322, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2611464968152866, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 7.4904458598726116e-06, | |
| "loss": 1.191, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.267515923566879, | |
| "grad_norm": 4.375, | |
| "learning_rate": 7.4777070063694264e-06, | |
| "loss": 1.2896, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 7.464968152866242e-06, | |
| "loss": 1.2666, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2802547770700636, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.452229299363057e-06, | |
| "loss": 1.3656, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.286624203821656, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 7.439490445859873e-06, | |
| "loss": 1.1488, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.2929936305732483, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 7.426751592356688e-06, | |
| "loss": 1.3889, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.2993630573248407, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 7.414012738853503e-06, | |
| "loss": 1.2333, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.305732484076433, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 7.401273885350318e-06, | |
| "loss": 1.4232, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3121019108280254, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 7.388535031847134e-06, | |
| "loss": 1.3532, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.3184713375796178, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 7.375796178343949e-06, | |
| "loss": 1.058, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3248407643312101, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 7.3630573248407645e-06, | |
| "loss": 1.0788, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3312101910828025, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 7.350318471337579e-06, | |
| "loss": 1.231, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3375796178343948, | |
| "grad_norm": 4.625, | |
| "learning_rate": 7.337579617834395e-06, | |
| "loss": 1.5649, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3439490445859872, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 7.32484076433121e-06, | |
| "loss": 1.2633, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.3503184713375795, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 7.312101910828026e-06, | |
| "loss": 1.1841, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.356687898089172, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 7.299363057324841e-06, | |
| "loss": 1.3064, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3630573248407643, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 7.286624203821656e-06, | |
| "loss": 1.1253, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.3694267515923566, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 7.273885350318471e-06, | |
| "loss": 1.0368, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3757961783439492, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 7.261146496815287e-06, | |
| "loss": 1.0507, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.3821656050955413, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 7.248407643312102e-06, | |
| "loss": 1.1262, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.388535031847134, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 7.2356687898089175e-06, | |
| "loss": 1.1362, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.394904458598726, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 7.222929936305732e-06, | |
| "loss": 1.2026, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.4012738853503186, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 7.210191082802548e-06, | |
| "loss": 1.3407, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4076433121019107, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 7.197452229299363e-06, | |
| "loss": 1.2616, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.4140127388535033, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.184713375796179e-06, | |
| "loss": 1.1387, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4203821656050954, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 7.171974522292994e-06, | |
| "loss": 1.3138, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.426751592356688, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 7.159235668789809e-06, | |
| "loss": 1.2038, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 4.125, | |
| "learning_rate": 7.146496815286624e-06, | |
| "loss": 1.1395, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4394904458598727, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 7.13375796178344e-06, | |
| "loss": 0.9929, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.4458598726114649, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 7.121019108280255e-06, | |
| "loss": 1.2248, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4522292993630574, | |
| "grad_norm": 4.75, | |
| "learning_rate": 7.1082802547770705e-06, | |
| "loss": 1.2465, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4585987261146496, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 7.095541401273885e-06, | |
| "loss": 1.2351, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4649681528662422, | |
| "grad_norm": 4.125, | |
| "learning_rate": 7.082802547770701e-06, | |
| "loss": 1.0333, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4713375796178343, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 7.070063694267516e-06, | |
| "loss": 1.1574, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.4777070063694269, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.057324840764332e-06, | |
| "loss": 0.9559, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.484076433121019, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 7.0445859872611466e-06, | |
| "loss": 1.3477, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4904458598726116, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 7.031847133757962e-06, | |
| "loss": 1.1675, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.4968152866242037, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 7.019108280254777e-06, | |
| "loss": 1.3436, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5031847133757963, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 7.006369426751593e-06, | |
| "loss": 1.1373, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.5095541401273884, | |
| "grad_norm": 4.25, | |
| "learning_rate": 6.993630573248408e-06, | |
| "loss": 1.1741, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.515923566878981, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 6.9808917197452235e-06, | |
| "loss": 1.1214, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5222929936305731, | |
| "grad_norm": 4.5, | |
| "learning_rate": 6.968152866242038e-06, | |
| "loss": 1.1291, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5286624203821657, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 6.955414012738853e-06, | |
| "loss": 1.0282, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5350318471337578, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 6.942675159235669e-06, | |
| "loss": 1.1718, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.5414012738853504, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 6.929936305732484e-06, | |
| "loss": 1.189, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5477707006369426, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 6.9171974522292995e-06, | |
| "loss": 1.1356, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5541401273885351, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.904458598726114e-06, | |
| "loss": 1.145, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5605095541401273, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 6.89171974522293e-06, | |
| "loss": 1.1753, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5668789808917198, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.878980891719745e-06, | |
| "loss": 1.2176, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.573248407643312, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 6.866242038216561e-06, | |
| "loss": 1.4333, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.5796178343949046, | |
| "grad_norm": 4.25, | |
| "learning_rate": 6.853503184713376e-06, | |
| "loss": 1.2131, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.5859872611464967, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 6.840764331210191e-06, | |
| "loss": 1.2344, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 5.0, | |
| "learning_rate": 6.828025477707006e-06, | |
| "loss": 1.2126, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.5987261146496814, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 6.815286624203822e-06, | |
| "loss": 1.2191, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.605095541401274, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.802547770700637e-06, | |
| "loss": 1.0808, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.611464968152866, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 6.7898089171974525e-06, | |
| "loss": 1.0224, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6178343949044587, | |
| "grad_norm": 4.75, | |
| "learning_rate": 6.777070063694267e-06, | |
| "loss": 1.3284, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6242038216560508, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 6.764331210191083e-06, | |
| "loss": 1.1836, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6305732484076434, | |
| "grad_norm": 4.75, | |
| "learning_rate": 6.751592356687898e-06, | |
| "loss": 1.3268, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6369426751592355, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 6.738853503184714e-06, | |
| "loss": 0.9918, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.643312101910828, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 6.726114649681529e-06, | |
| "loss": 1.315, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6496815286624202, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.713375796178344e-06, | |
| "loss": 1.1504, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6560509554140128, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 6.700636942675159e-06, | |
| "loss": 1.3433, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.662420382165605, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 6.687898089171975e-06, | |
| "loss": 0.8864, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6687898089171975, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 6.67515923566879e-06, | |
| "loss": 1.2297, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.6751592356687897, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 6.6624203821656055e-06, | |
| "loss": 1.0687, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.6815286624203822, | |
| "grad_norm": 4.125, | |
| "learning_rate": 6.64968152866242e-06, | |
| "loss": 0.896, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.6878980891719744, | |
| "grad_norm": 4.375, | |
| "learning_rate": 6.636942675159236e-06, | |
| "loss": 0.999, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.694267515923567, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 6.624203821656051e-06, | |
| "loss": 1.2021, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.700636942675159, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 6.611464968152867e-06, | |
| "loss": 1.2678, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.7070063694267517, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 6.5987261146496816e-06, | |
| "loss": 1.035, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7133757961783438, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 6.585987261146497e-06, | |
| "loss": 1.2687, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.7197452229299364, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.573248407643312e-06, | |
| "loss": 1.0557, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7261146496815285, | |
| "grad_norm": 5.5, | |
| "learning_rate": 6.560509554140128e-06, | |
| "loss": 1.2142, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.732484076433121, | |
| "grad_norm": 4.75, | |
| "learning_rate": 6.547770700636943e-06, | |
| "loss": 1.0477, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.7388535031847132, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 6.5350318471337585e-06, | |
| "loss": 1.338, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7452229299363058, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.522292993630573e-06, | |
| "loss": 0.9228, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 6.509554140127389e-06, | |
| "loss": 1.1064, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7579617834394905, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 6.496815286624204e-06, | |
| "loss": 1.1275, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.7643312101910829, | |
| "grad_norm": 12.25, | |
| "learning_rate": 6.48407643312102e-06, | |
| "loss": 1.1311, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7707006369426752, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 6.4713375796178345e-06, | |
| "loss": 1.1379, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.7770700636942676, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 6.45859872611465e-06, | |
| "loss": 1.1072, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.78343949044586, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 6.445859872611465e-06, | |
| "loss": 0.9658, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7898089171974523, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 6.433121019108281e-06, | |
| "loss": 1.1306, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.7961783439490446, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 6.420382165605096e-06, | |
| "loss": 0.9347, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.802547770700637, | |
| "grad_norm": 6.0, | |
| "learning_rate": 6.407643312101911e-06, | |
| "loss": 1.2534, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.8089171974522293, | |
| "grad_norm": 4.5, | |
| "learning_rate": 6.394904458598726e-06, | |
| "loss": 1.0046, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8152866242038217, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 6.382165605095541e-06, | |
| "loss": 1.1839, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.821656050955414, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 6.369426751592357e-06, | |
| "loss": 0.9906, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8280254777070064, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 6.356687898089172e-06, | |
| "loss": 1.0906, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8343949044585988, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 6.3439490445859875e-06, | |
| "loss": 1.4091, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.8407643312101911, | |
| "grad_norm": 5.0, | |
| "learning_rate": 6.331210191082802e-06, | |
| "loss": 1.0953, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8471337579617835, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 6.318471337579618e-06, | |
| "loss": 1.0599, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8535031847133758, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 6.305732484076433e-06, | |
| "loss": 1.1559, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.8598726114649682, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 6.292993630573249e-06, | |
| "loss": 0.9877, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.8662420382165605, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 6.280254777070064e-06, | |
| "loss": 1.0641, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.872611464968153, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.267515923566879e-06, | |
| "loss": 1.1279, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.8789808917197452, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 6.254777070063694e-06, | |
| "loss": 0.8615, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.8853503184713376, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 6.24203821656051e-06, | |
| "loss": 1.0757, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.89171974522293, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.229299363057325e-06, | |
| "loss": 1.1164, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.8980891719745223, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 6.2165605095541405e-06, | |
| "loss": 0.8994, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9044585987261147, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 6.203821656050955e-06, | |
| "loss": 1.1392, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 6.191082802547771e-06, | |
| "loss": 1.0445, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9171974522292994, | |
| "grad_norm": 4.625, | |
| "learning_rate": 6.178343949044586e-06, | |
| "loss": 0.8624, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.9235668789808917, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.165605095541402e-06, | |
| "loss": 1.2205, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.929936305732484, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 6.1528662420382166e-06, | |
| "loss": 1.2422, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9363057324840764, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 6.140127388535032e-06, | |
| "loss": 0.9838, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.9426751592356688, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 6.127388535031847e-06, | |
| "loss": 1.2007, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9490445859872612, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 6.114649681528663e-06, | |
| "loss": 1.0595, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.9554140127388535, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 6.101910828025478e-06, | |
| "loss": 1.1795, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.9617834394904459, | |
| "grad_norm": 5.375, | |
| "learning_rate": 6.0891719745222935e-06, | |
| "loss": 1.0842, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.9681528662420382, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.076433121019108e-06, | |
| "loss": 0.9748, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.9745222929936306, | |
| "grad_norm": 5.375, | |
| "learning_rate": 6.063694267515924e-06, | |
| "loss": 1.112, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.980891719745223, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.050955414012739e-06, | |
| "loss": 0.9138, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.9872611464968153, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 6.038216560509555e-06, | |
| "loss": 0.9162, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.9936305732484076, | |
| "grad_norm": 6.125, | |
| "learning_rate": 6.0254777070063696e-06, | |
| "loss": 1.1772, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 6.012738853503185e-06, | |
| "loss": 0.974, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.41422963142395, | |
| "eval_runtime": 9.0706, | |
| "eval_samples_per_second": 110.247, | |
| "eval_steps_per_second": 3.528, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.0063694267515926, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 6e-06, | |
| "loss": 1.1089, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.0127388535031847, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 5.987261146496816e-06, | |
| "loss": 1.1255, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.0191082802547773, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 5.974522292993631e-06, | |
| "loss": 0.9402, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.0254777070063694, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 5.9617834394904465e-06, | |
| "loss": 1.1566, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.031847133757962, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 5.949044585987261e-06, | |
| "loss": 0.9663, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.038216560509554, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.936305732484077e-06, | |
| "loss": 0.9331, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0445859872611467, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.923566878980892e-06, | |
| "loss": 0.953, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.050955414012739, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 5.910828025477708e-06, | |
| "loss": 1.0141, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.0573248407643314, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 5.8980891719745225e-06, | |
| "loss": 0.8474, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.0636942675159236, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.885350318471338e-06, | |
| "loss": 1.0246, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 5.872611464968153e-06, | |
| "loss": 1.0792, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.0764331210191083, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.859872611464969e-06, | |
| "loss": 0.8486, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.082802547770701, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 5.847133757961784e-06, | |
| "loss": 0.9662, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.089171974522293, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 5.834394904458599e-06, | |
| "loss": 0.9775, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.0955414012738856, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 5.821656050955414e-06, | |
| "loss": 1.0779, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.1019108280254777, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 5.808917197452229e-06, | |
| "loss": 0.7921, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.1082802547770703, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 5.796178343949045e-06, | |
| "loss": 0.8923, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.1146496815286624, | |
| "grad_norm": 7.125, | |
| "learning_rate": 5.78343949044586e-06, | |
| "loss": 1.0133, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.121019108280255, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 5.7707006369426755e-06, | |
| "loss": 1.1287, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.127388535031847, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 5.75796178343949e-06, | |
| "loss": 0.9475, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.1337579617834397, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 5.745222929936306e-06, | |
| "loss": 1.1251, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.140127388535032, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 5.732484076433121e-06, | |
| "loss": 1.1615, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.1464968152866244, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 5.719745222929937e-06, | |
| "loss": 1.1937, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.1528662420382165, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 5.707006369426752e-06, | |
| "loss": 0.8947, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.159235668789809, | |
| "grad_norm": 6.375, | |
| "learning_rate": 5.694267515923567e-06, | |
| "loss": 1.0208, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.1656050955414012, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 5.681528662420382e-06, | |
| "loss": 0.8956, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.171974522292994, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 5.668789808917198e-06, | |
| "loss": 0.8284, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.178343949044586, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 5.656050955414013e-06, | |
| "loss": 1.0183, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.1847133757961785, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 5.6433121019108285e-06, | |
| "loss": 0.8664, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.1910828025477707, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 5.630573248407643e-06, | |
| "loss": 0.9622, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.1974522292993632, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 5.617834394904459e-06, | |
| "loss": 0.94, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.2038216560509554, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 5.605095541401274e-06, | |
| "loss": 1.1599, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.210191082802548, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 5.59235668789809e-06, | |
| "loss": 1.0154, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.21656050955414, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 5.5796178343949046e-06, | |
| "loss": 0.9312, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.2229299363057327, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 5.56687898089172e-06, | |
| "loss": 0.9159, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 5.554140127388535e-06, | |
| "loss": 0.9921, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.2356687898089174, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.541401273885351e-06, | |
| "loss": 0.8543, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.2420382165605095, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 5.528662420382166e-06, | |
| "loss": 0.9698, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.248407643312102, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 5.5159235668789815e-06, | |
| "loss": 0.8562, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.254777070063694, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 5.503184713375796e-06, | |
| "loss": 0.9015, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.261146496815287, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 5.490445859872612e-06, | |
| "loss": 1.0079, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.267515923566879, | |
| "grad_norm": 6.875, | |
| "learning_rate": 5.477707006369427e-06, | |
| "loss": 0.948, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.2738853503184715, | |
| "grad_norm": 6.25, | |
| "learning_rate": 5.464968152866243e-06, | |
| "loss": 1.0708, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.2802547770700636, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 5.4522292993630575e-06, | |
| "loss": 1.0075, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.286624203821656, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 5.439490445859873e-06, | |
| "loss": 0.9092, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.2929936305732483, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 5.426751592356688e-06, | |
| "loss": 0.7904, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.299363057324841, | |
| "grad_norm": 6.875, | |
| "learning_rate": 5.414012738853504e-06, | |
| "loss": 1.0069, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.305732484076433, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 5.401273885350319e-06, | |
| "loss": 1.1094, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.3121019108280256, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 5.3885350318471345e-06, | |
| "loss": 1.0723, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.3184713375796178, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 5.375796178343949e-06, | |
| "loss": 0.7936, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.3248407643312103, | |
| "grad_norm": 7.125, | |
| "learning_rate": 5.363057324840765e-06, | |
| "loss": 0.9827, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.3312101910828025, | |
| "grad_norm": 6.125, | |
| "learning_rate": 5.35031847133758e-06, | |
| "loss": 1.0221, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.337579617834395, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 5.337579617834396e-06, | |
| "loss": 0.9385, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.343949044585987, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 5.3248407643312105e-06, | |
| "loss": 0.8619, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.3503184713375798, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 5.312101910828026e-06, | |
| "loss": 0.87, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.356687898089172, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 5.299363057324841e-06, | |
| "loss": 0.7255, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.3630573248407645, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 5.286624203821657e-06, | |
| "loss": 0.9238, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.3694267515923566, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 5.273885350318472e-06, | |
| "loss": 0.8998, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.375796178343949, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 5.261146496815287e-06, | |
| "loss": 0.8195, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.3821656050955413, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 5.248407643312102e-06, | |
| "loss": 0.9705, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 5.235668789808917e-06, | |
| "loss": 0.7799, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.394904458598726, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 5.222929936305733e-06, | |
| "loss": 0.9264, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.4012738853503186, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 5.210191082802548e-06, | |
| "loss": 0.8549, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.4076433121019107, | |
| "grad_norm": 8.625, | |
| "learning_rate": 5.1974522292993635e-06, | |
| "loss": 0.9288, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.4140127388535033, | |
| "grad_norm": 7.25, | |
| "learning_rate": 5.184713375796178e-06, | |
| "loss": 0.8305, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.4203821656050954, | |
| "grad_norm": 6.625, | |
| "learning_rate": 5.171974522292994e-06, | |
| "loss": 0.8398, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.426751592356688, | |
| "grad_norm": 7.625, | |
| "learning_rate": 5.159235668789809e-06, | |
| "loss": 1.1755, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.43312101910828, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 5.146496815286625e-06, | |
| "loss": 0.9441, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.4394904458598727, | |
| "grad_norm": 8.625, | |
| "learning_rate": 5.1337579617834396e-06, | |
| "loss": 0.9724, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.445859872611465, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 5.121019108280255e-06, | |
| "loss": 0.8549, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.4522292993630574, | |
| "grad_norm": 8.0, | |
| "learning_rate": 5.10828025477707e-06, | |
| "loss": 0.9343, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.4585987261146496, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.095541401273886e-06, | |
| "loss": 0.6092, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.464968152866242, | |
| "grad_norm": 6.875, | |
| "learning_rate": 5.082802547770701e-06, | |
| "loss": 0.9078, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.4713375796178343, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 5.0700636942675165e-06, | |
| "loss": 0.7282, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.477707006369427, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 5.057324840764331e-06, | |
| "loss": 0.8971, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.484076433121019, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 5.044585987261147e-06, | |
| "loss": 0.9202, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.4904458598726116, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.031847133757962e-06, | |
| "loss": 0.8603, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.4968152866242037, | |
| "grad_norm": 7.875, | |
| "learning_rate": 5.019108280254778e-06, | |
| "loss": 1.0743, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.5031847133757963, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 5.0063694267515925e-06, | |
| "loss": 0.9304, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.5095541401273884, | |
| "grad_norm": 6.75, | |
| "learning_rate": 4.993630573248408e-06, | |
| "loss": 1.0682, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.515923566878981, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 4.980891719745223e-06, | |
| "loss": 0.9261, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.522292993630573, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.968152866242039e-06, | |
| "loss": 0.9446, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.5286624203821657, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.955414012738854e-06, | |
| "loss": 0.9587, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.535031847133758, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.9426751592356695e-06, | |
| "loss": 0.8462, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.5414012738853504, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 4.929936305732484e-06, | |
| "loss": 0.9314, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 4.9171974522293e-06, | |
| "loss": 0.9174, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.554140127388535, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.904458598726115e-06, | |
| "loss": 0.9356, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.5605095541401273, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.891719745222931e-06, | |
| "loss": 0.89, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.56687898089172, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 4.8789808917197455e-06, | |
| "loss": 1.0187, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.573248407643312, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.866242038216561e-06, | |
| "loss": 0.9083, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.5796178343949046, | |
| "grad_norm": 7.25, | |
| "learning_rate": 4.853503184713376e-06, | |
| "loss": 0.9705, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.5859872611464967, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 4.840764331210192e-06, | |
| "loss": 1.069, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.5923566878980893, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 4.828025477707007e-06, | |
| "loss": 0.8275, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.5987261146496814, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 4.8152866242038224e-06, | |
| "loss": 0.8257, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.605095541401274, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 4.802547770700637e-06, | |
| "loss": 1.0051, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.611464968152866, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.789808917197453e-06, | |
| "loss": 1.1191, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.6178343949044587, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 4.777070063694268e-06, | |
| "loss": 0.9982, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.624203821656051, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.764331210191084e-06, | |
| "loss": 0.7671, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.6305732484076434, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.7515923566878985e-06, | |
| "loss": 0.8938, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.6369426751592355, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.738853503184714e-06, | |
| "loss": 1.0574, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.643312101910828, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 4.726114649681529e-06, | |
| "loss": 0.8727, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.6496815286624202, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.713375796178344e-06, | |
| "loss": 0.9062, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.656050955414013, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.70063694267516e-06, | |
| "loss": 0.9495, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.662420382165605, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 4.6878980891719746e-06, | |
| "loss": 1.0316, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.6687898089171975, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.67515923566879e-06, | |
| "loss": 0.9483, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.6751592356687897, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 4.662420382165605e-06, | |
| "loss": 0.8035, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.6815286624203822, | |
| "grad_norm": 6.625, | |
| "learning_rate": 4.649681528662421e-06, | |
| "loss": 0.7403, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.6878980891719744, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.636942675159236e-06, | |
| "loss": 0.8955, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.694267515923567, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 4.6242038216560515e-06, | |
| "loss": 0.9146, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.700636942675159, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 4.611464968152866e-06, | |
| "loss": 0.9436, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.7070063694267517, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 4.598726114649682e-06, | |
| "loss": 1.0588, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.713375796178344, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 4.585987261146497e-06, | |
| "loss": 0.9712, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.7197452229299364, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.573248407643313e-06, | |
| "loss": 0.7337, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.7261146496815285, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 4.5605095541401276e-06, | |
| "loss": 0.8345, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.732484076433121, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 4.547770700636943e-06, | |
| "loss": 0.8751, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.738853503184713, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 4.535031847133758e-06, | |
| "loss": 0.9105, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.745222929936306, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.522292993630574e-06, | |
| "loss": 0.9287, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.7515923566878984, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.509554140127389e-06, | |
| "loss": 0.9341, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.7579617834394905, | |
| "grad_norm": 6.75, | |
| "learning_rate": 4.4968152866242045e-06, | |
| "loss": 0.8777, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.7643312101910826, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 4.484076433121019e-06, | |
| "loss": 0.744, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.770700636942675, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 4.471337579617835e-06, | |
| "loss": 0.9441, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.777070063694268, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.45859872611465e-06, | |
| "loss": 0.8909, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.78343949044586, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 4.445859872611466e-06, | |
| "loss": 0.9307, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.789808917197452, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 4.4331210191082805e-06, | |
| "loss": 1.0462, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.7961783439490446, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.420382165605096e-06, | |
| "loss": 0.8442, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.802547770700637, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.407643312101911e-06, | |
| "loss": 0.7325, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.8089171974522293, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.394904458598727e-06, | |
| "loss": 0.8699, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.8152866242038215, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 4.382165605095542e-06, | |
| "loss": 0.9305, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.821656050955414, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 4.3694267515923574e-06, | |
| "loss": 0.8791, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.8280254777070066, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 4.356687898089172e-06, | |
| "loss": 1.0769, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.8343949044585988, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 4.343949044585988e-06, | |
| "loss": 0.9787, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.840764331210191, | |
| "grad_norm": 6.625, | |
| "learning_rate": 4.331210191082803e-06, | |
| "loss": 0.7492, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.8471337579617835, | |
| "grad_norm": 6.75, | |
| "learning_rate": 4.318471337579619e-06, | |
| "loss": 0.747, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.853503184713376, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 4.3057324840764335e-06, | |
| "loss": 0.7899, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.859872611464968, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 4.292993630573249e-06, | |
| "loss": 0.8435, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 4.280254777070064e-06, | |
| "loss": 0.9187, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.872611464968153, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.26751592356688e-06, | |
| "loss": 0.8621, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.8789808917197455, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.254777070063695e-06, | |
| "loss": 0.8495, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.8853503184713376, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 4.2420382165605104e-06, | |
| "loss": 0.9509, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.8917197452229297, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.229299363057325e-06, | |
| "loss": 1.1276, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.8980891719745223, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.216560509554141e-06, | |
| "loss": 0.8085, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.904458598726115, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 4.203821656050956e-06, | |
| "loss": 0.7185, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.910828025477707, | |
| "grad_norm": 6.75, | |
| "learning_rate": 4.191082802547772e-06, | |
| "loss": 0.9432, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.917197452229299, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 4.1783439490445865e-06, | |
| "loss": 0.8644, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.9235668789808917, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 4.165605095541402e-06, | |
| "loss": 0.9468, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.9299363057324843, | |
| "grad_norm": 6.625, | |
| "learning_rate": 4.152866242038217e-06, | |
| "loss": 0.745, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.9363057324840764, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 4.140127388535032e-06, | |
| "loss": 0.8089, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.9426751592356686, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 4.127388535031848e-06, | |
| "loss": 0.95, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.949044585987261, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.1146496815286626e-06, | |
| "loss": 0.825, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.9554140127388537, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 4.101910828025478e-06, | |
| "loss": 0.7344, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.961783439490446, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 4.089171974522293e-06, | |
| "loss": 0.8908, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.968152866242038, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 4.076433121019109e-06, | |
| "loss": 0.7336, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.9745222929936306, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.063694267515924e-06, | |
| "loss": 0.8678, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.980891719745223, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 4.0509554140127395e-06, | |
| "loss": 0.8376, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.9872611464968153, | |
| "grad_norm": 8.5, | |
| "learning_rate": 4.038216560509554e-06, | |
| "loss": 1.0185, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.9936305732484074, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 4.02547770700637e-06, | |
| "loss": 0.6899, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 15.1875, | |
| "learning_rate": 4.012738853503185e-06, | |
| "loss": 1.1406, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.590538501739502, | |
| "eval_runtime": 8.8903, | |
| "eval_samples_per_second": 112.482, | |
| "eval_steps_per_second": 3.599, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 3.0063694267515926, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9258, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 3.0127388535031847, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.9872611464968155e-06, | |
| "loss": 0.7916, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 3.0191082802547773, | |
| "grad_norm": 8.875, | |
| "learning_rate": 3.974522292993631e-06, | |
| "loss": 0.8027, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 3.0254777070063694, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.961783439490446e-06, | |
| "loss": 0.9065, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.031847133757962, | |
| "grad_norm": 9.125, | |
| "learning_rate": 3.949044585987262e-06, | |
| "loss": 0.7015, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 3.038216560509554, | |
| "grad_norm": 7.125, | |
| "learning_rate": 3.936305732484077e-06, | |
| "loss": 0.5953, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 3.0445859872611467, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 3.9235668789808925e-06, | |
| "loss": 0.7153, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 3.050955414012739, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 3.910828025477707e-06, | |
| "loss": 0.7931, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 3.0573248407643314, | |
| "grad_norm": 7.625, | |
| "learning_rate": 3.898089171974523e-06, | |
| "loss": 0.7491, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.0636942675159236, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 3.885350318471338e-06, | |
| "loss": 0.7937, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 3.070063694267516, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 3.872611464968154e-06, | |
| "loss": 0.9503, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 3.0764331210191083, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 3.8598726114649685e-06, | |
| "loss": 0.725, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 3.082802547770701, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 3.847133757961784e-06, | |
| "loss": 0.7869, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 3.089171974522293, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.834394904458599e-06, | |
| "loss": 1.0002, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 3.0955414012738856, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 3.821656050955415e-06, | |
| "loss": 0.9329, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 3.1019108280254777, | |
| "grad_norm": 8.875, | |
| "learning_rate": 3.8089171974522297e-06, | |
| "loss": 0.8777, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 3.1082802547770703, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 3.796178343949045e-06, | |
| "loss": 0.7212, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 3.1146496815286624, | |
| "grad_norm": 7.125, | |
| "learning_rate": 3.7834394904458603e-06, | |
| "loss": 0.7911, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 3.121019108280255, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.7707006369426756e-06, | |
| "loss": 0.7187, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.127388535031847, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 3.757961783439491e-06, | |
| "loss": 0.764, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 3.1337579617834397, | |
| "grad_norm": 8.0, | |
| "learning_rate": 3.7452229299363058e-06, | |
| "loss": 0.9536, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 3.140127388535032, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.732484076433121e-06, | |
| "loss": 0.939, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 3.1464968152866244, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.7197452229299364e-06, | |
| "loss": 0.9698, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 3.1528662420382165, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 3.7070063694267517e-06, | |
| "loss": 0.8298, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 3.159235668789809, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 3.694267515923567e-06, | |
| "loss": 0.8173, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 3.1656050955414012, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 3.6815286624203823e-06, | |
| "loss": 0.8397, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 3.171974522292994, | |
| "grad_norm": 6.375, | |
| "learning_rate": 3.6687898089171976e-06, | |
| "loss": 0.6754, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 3.178343949044586, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 3.656050955414013e-06, | |
| "loss": 0.8032, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.643312101910828e-06, | |
| "loss": 0.6529, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.1910828025477707, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 3.6305732484076435e-06, | |
| "loss": 0.5822, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 3.1974522292993632, | |
| "grad_norm": 7.0, | |
| "learning_rate": 3.6178343949044588e-06, | |
| "loss": 0.6683, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 3.2038216560509554, | |
| "grad_norm": 7.25, | |
| "learning_rate": 3.605095541401274e-06, | |
| "loss": 0.6626, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 3.210191082802548, | |
| "grad_norm": 8.0, | |
| "learning_rate": 3.5923566878980894e-06, | |
| "loss": 0.7932, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 3.21656050955414, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 3.5796178343949047e-06, | |
| "loss": 0.9472, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 3.2229299363057327, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.56687898089172e-06, | |
| "loss": 0.7957, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 3.229299363057325, | |
| "grad_norm": 8.5, | |
| "learning_rate": 3.5541401273885352e-06, | |
| "loss": 0.8499, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 3.2356687898089174, | |
| "grad_norm": 6.875, | |
| "learning_rate": 3.5414012738853505e-06, | |
| "loss": 0.6337, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 3.2420382165605095, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 3.528662420382166e-06, | |
| "loss": 0.799, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 3.248407643312102, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 3.515923566878981e-06, | |
| "loss": 0.6682, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.254777070063694, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 3.5031847133757964e-06, | |
| "loss": 0.7758, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 3.261146496815287, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 3.4904458598726117e-06, | |
| "loss": 0.853, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 3.267515923566879, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 3.4777070063694266e-06, | |
| "loss": 0.7178, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 3.2738853503184715, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 3.464968152866242e-06, | |
| "loss": 0.7499, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 3.2802547770700636, | |
| "grad_norm": 8.5, | |
| "learning_rate": 3.452229299363057e-06, | |
| "loss": 0.7318, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 3.286624203821656, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 3.4394904458598725e-06, | |
| "loss": 0.6524, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 3.2929936305732483, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 3.426751592356688e-06, | |
| "loss": 0.7072, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 3.299363057324841, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.414012738853503e-06, | |
| "loss": 0.7335, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 3.305732484076433, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 3.4012738853503184e-06, | |
| "loss": 0.7985, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 3.3121019108280256, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 3.3885350318471337e-06, | |
| "loss": 0.9649, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.3184713375796178, | |
| "grad_norm": 20.0, | |
| "learning_rate": 3.375796178343949e-06, | |
| "loss": 0.8484, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 3.3248407643312103, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 3.3630573248407643e-06, | |
| "loss": 0.7687, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 3.3312101910828025, | |
| "grad_norm": 8.375, | |
| "learning_rate": 3.3503184713375796e-06, | |
| "loss": 0.9385, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 3.337579617834395, | |
| "grad_norm": 8.0, | |
| "learning_rate": 3.337579617834395e-06, | |
| "loss": 0.8823, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 3.343949044585987, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 3.32484076433121e-06, | |
| "loss": 0.7453, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.3503184713375798, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.3121019108280255e-06, | |
| "loss": 0.6488, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 3.356687898089172, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 3.2993630573248408e-06, | |
| "loss": 0.9876, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 3.3630573248407645, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 3.286624203821656e-06, | |
| "loss": 0.7621, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 3.3694267515923566, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 3.2738853503184714e-06, | |
| "loss": 0.732, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 3.375796178343949, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 3.2611464968152867e-06, | |
| "loss": 0.7077, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.3821656050955413, | |
| "grad_norm": 8.25, | |
| "learning_rate": 3.248407643312102e-06, | |
| "loss": 0.7588, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 3.388535031847134, | |
| "grad_norm": 8.75, | |
| "learning_rate": 3.2356687898089173e-06, | |
| "loss": 0.7197, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 3.394904458598726, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 3.2229299363057326e-06, | |
| "loss": 0.7606, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 3.4012738853503186, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 3.210191082802548e-06, | |
| "loss": 0.5293, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 3.4076433121019107, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 3.197452229299363e-06, | |
| "loss": 0.9173, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 3.4140127388535033, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.1847133757961785e-06, | |
| "loss": 0.8, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 3.4203821656050954, | |
| "grad_norm": 8.375, | |
| "learning_rate": 3.1719745222929938e-06, | |
| "loss": 0.7233, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 3.426751592356688, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 3.159235668789809e-06, | |
| "loss": 0.7931, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 3.43312101910828, | |
| "grad_norm": 8.0, | |
| "learning_rate": 3.1464968152866244e-06, | |
| "loss": 0.866, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 3.4394904458598727, | |
| "grad_norm": 8.625, | |
| "learning_rate": 3.1337579617834397e-06, | |
| "loss": 0.8575, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.445859872611465, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 3.121019108280255e-06, | |
| "loss": 0.6249, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 3.4522292993630574, | |
| "grad_norm": 7.125, | |
| "learning_rate": 3.1082802547770703e-06, | |
| "loss": 0.565, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 3.4585987261146496, | |
| "grad_norm": 6.75, | |
| "learning_rate": 3.0955414012738856e-06, | |
| "loss": 0.6712, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 3.464968152866242, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 3.082802547770701e-06, | |
| "loss": 0.7053, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 3.4713375796178343, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 3.070063694267516e-06, | |
| "loss": 0.8104, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 3.477707006369427, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 3.0573248407643314e-06, | |
| "loss": 0.8713, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 3.484076433121019, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 3.0445859872611467e-06, | |
| "loss": 0.7784, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 3.4904458598726116, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 3.031847133757962e-06, | |
| "loss": 0.5982, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 3.4968152866242037, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 3.0191082802547773e-06, | |
| "loss": 0.7742, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 3.5031847133757963, | |
| "grad_norm": 8.25, | |
| "learning_rate": 3.0063694267515926e-06, | |
| "loss": 0.6583, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.5095541401273884, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 2.993630573248408e-06, | |
| "loss": 0.7583, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 3.515923566878981, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 2.9808917197452232e-06, | |
| "loss": 0.8666, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 3.522292993630573, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 2.9681528662420385e-06, | |
| "loss": 0.8369, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 3.5286624203821657, | |
| "grad_norm": 8.875, | |
| "learning_rate": 2.955414012738854e-06, | |
| "loss": 0.7843, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 3.535031847133758, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 2.942675159235669e-06, | |
| "loss": 0.8713, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 3.5414012738853504, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 2.9299363057324844e-06, | |
| "loss": 0.6351, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 3.5477707006369426, | |
| "grad_norm": 8.5, | |
| "learning_rate": 2.9171974522292993e-06, | |
| "loss": 0.7874, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 3.554140127388535, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 2.9044585987261146e-06, | |
| "loss": 0.9089, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 3.5605095541401273, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 2.89171974522293e-06, | |
| "loss": 0.8104, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 3.56687898089172, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 2.878980891719745e-06, | |
| "loss": 0.8349, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 3.573248407643312, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.8662420382165605e-06, | |
| "loss": 0.8424, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 3.5796178343949046, | |
| "grad_norm": 7.75, | |
| "learning_rate": 2.853503184713376e-06, | |
| "loss": 0.889, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 3.5859872611464967, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.840764331210191e-06, | |
| "loss": 0.7141, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 3.5923566878980893, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 2.8280254777070064e-06, | |
| "loss": 0.782, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 3.5987261146496814, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 2.8152866242038217e-06, | |
| "loss": 0.8647, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 3.605095541401274, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.802547770700637e-06, | |
| "loss": 0.6843, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 3.611464968152866, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 2.7898089171974523e-06, | |
| "loss": 0.9196, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 3.6178343949044587, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.7770700636942676e-06, | |
| "loss": 0.9584, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 3.624203821656051, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.764331210191083e-06, | |
| "loss": 0.698, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 3.6305732484076434, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 2.751592356687898e-06, | |
| "loss": 0.7946, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 3.6369426751592355, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 2.7388535031847135e-06, | |
| "loss": 0.7206, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 3.643312101910828, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 2.7261146496815288e-06, | |
| "loss": 0.6167, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 3.6496815286624202, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 2.713375796178344e-06, | |
| "loss": 0.7068, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 3.656050955414013, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 2.7006369426751594e-06, | |
| "loss": 0.8919, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 3.662420382165605, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.6878980891719747e-06, | |
| "loss": 0.68, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.6687898089171975, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 2.67515923566879e-06, | |
| "loss": 0.8711, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 3.6751592356687897, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 2.6624203821656053e-06, | |
| "loss": 0.88, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 3.6815286624203822, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 2.6496815286624206e-06, | |
| "loss": 0.8026, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 3.6878980891719744, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.636942675159236e-06, | |
| "loss": 0.8391, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 3.694267515923567, | |
| "grad_norm": 9.125, | |
| "learning_rate": 2.624203821656051e-06, | |
| "loss": 0.874, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.700636942675159, | |
| "grad_norm": 6.625, | |
| "learning_rate": 2.6114649681528665e-06, | |
| "loss": 0.7192, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 3.7070063694267517, | |
| "grad_norm": 7.375, | |
| "learning_rate": 2.5987261146496818e-06, | |
| "loss": 0.7316, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 3.713375796178344, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 2.585987261146497e-06, | |
| "loss": 0.7397, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 3.7197452229299364, | |
| "grad_norm": 8.25, | |
| "learning_rate": 2.5732484076433123e-06, | |
| "loss": 0.8095, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 3.7261146496815285, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 2.5605095541401276e-06, | |
| "loss": 0.6748, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 3.732484076433121, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 2.547770700636943e-06, | |
| "loss": 0.6194, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 3.738853503184713, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 2.5350318471337582e-06, | |
| "loss": 0.7418, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 3.745222929936306, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 2.5222929936305735e-06, | |
| "loss": 0.693, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 3.7515923566878984, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 2.509554140127389e-06, | |
| "loss": 0.6708, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 3.7579617834394905, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 2.496815286624204e-06, | |
| "loss": 0.6303, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.7643312101910826, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 2.4840764331210194e-06, | |
| "loss": 0.7191, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 3.770700636942675, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 2.4713375796178347e-06, | |
| "loss": 1.0134, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 3.777070063694268, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 2.45859872611465e-06, | |
| "loss": 0.8218, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 3.78343949044586, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 2.4458598726114653e-06, | |
| "loss": 0.6153, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 3.789808917197452, | |
| "grad_norm": 8.0, | |
| "learning_rate": 2.4331210191082806e-06, | |
| "loss": 0.8309, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 3.7961783439490446, | |
| "grad_norm": 7.5, | |
| "learning_rate": 2.420382165605096e-06, | |
| "loss": 0.8049, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 3.802547770700637, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 2.4076433121019112e-06, | |
| "loss": 0.8137, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 3.8089171974522293, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 2.3949044585987265e-06, | |
| "loss": 0.9029, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 3.8152866242038215, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 2.382165605095542e-06, | |
| "loss": 0.7567, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.369426751592357e-06, | |
| "loss": 0.8199, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.8280254777070066, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 2.356687898089172e-06, | |
| "loss": 0.716, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 3.8343949044585988, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 2.3439490445859873e-06, | |
| "loss": 0.6286, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 3.840764331210191, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.3312101910828026e-06, | |
| "loss": 0.6282, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 3.8471337579617835, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.318471337579618e-06, | |
| "loss": 0.7313, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 3.853503184713376, | |
| "grad_norm": 7.5, | |
| "learning_rate": 2.305732484076433e-06, | |
| "loss": 0.6827, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 3.859872611464968, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 2.2929936305732485e-06, | |
| "loss": 0.6923, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 3.8662420382165603, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 2.2802547770700638e-06, | |
| "loss": 0.6667, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 3.872611464968153, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.267515923566879e-06, | |
| "loss": 0.8422, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 3.8789808917197455, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 2.2547770700636944e-06, | |
| "loss": 0.7772, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 3.8853503184713376, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 2.2420382165605097e-06, | |
| "loss": 0.7392, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.8917197452229297, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 2.229299363057325e-06, | |
| "loss": 0.7047, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 3.8980891719745223, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.2165605095541403e-06, | |
| "loss": 0.7936, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 3.904458598726115, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 2.2038216560509556e-06, | |
| "loss": 0.5801, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 3.910828025477707, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 2.191082802547771e-06, | |
| "loss": 0.7394, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 3.917197452229299, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.178343949044586e-06, | |
| "loss": 0.8315, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 3.9235668789808917, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 2.1656050955414015e-06, | |
| "loss": 0.8811, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 3.9299363057324843, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.1528662420382168e-06, | |
| "loss": 0.6759, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 3.9363057324840764, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 2.140127388535032e-06, | |
| "loss": 0.8692, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 3.9426751592356686, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 2.1273885350318474e-06, | |
| "loss": 0.7069, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 3.949044585987261, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.1146496815286627e-06, | |
| "loss": 0.7112, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.9554140127388537, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 2.101910828025478e-06, | |
| "loss": 0.7849, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 3.961783439490446, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.0891719745222932e-06, | |
| "loss": 0.8135, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 3.968152866242038, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.0764331210191085e-06, | |
| "loss": 0.8131, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 3.9745222929936306, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.063694267515924e-06, | |
| "loss": 0.7923, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 3.980891719745223, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 2.050955414012739e-06, | |
| "loss": 0.6715, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.9872611464968153, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.0382165605095544e-06, | |
| "loss": 0.7208, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 3.9936305732484074, | |
| "grad_norm": 8.375, | |
| "learning_rate": 2.0254777070063697e-06, | |
| "loss": 0.7934, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 14.9375, | |
| "learning_rate": 2.012738853503185e-06, | |
| "loss": 0.8361, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.678133010864258, | |
| "eval_runtime": 8.8907, | |
| "eval_samples_per_second": 112.477, | |
| "eval_steps_per_second": 3.599, | |
| "step": 628 | |
| } | |
| ], | |
| "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": 2.465274515305536e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |