| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 307.6923076923077, | |
| "eval_steps": 500, | |
| "global_step": 20000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 134.7136688232422, | |
| "learning_rate": 9.981600000000001e-06, | |
| "loss": 5.6198, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 115.98611450195312, | |
| "learning_rate": 9.961600000000001e-06, | |
| "loss": 3.5959, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 114.23514556884766, | |
| "learning_rate": 9.941600000000002e-06, | |
| "loss": 3.3845, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 140.5648193359375, | |
| "learning_rate": 9.921600000000002e-06, | |
| "loss": 3.1494, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 117.4485855102539, | |
| "learning_rate": 9.901600000000002e-06, | |
| "loss": 3.0773, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 389.71490478515625, | |
| "learning_rate": 9.8816e-06, | |
| "loss": 3.2116, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.76923076923077, | |
| "grad_norm": 107.11251831054688, | |
| "learning_rate": 9.8616e-06, | |
| "loss": 3.0471, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.307692307692308, | |
| "grad_norm": 85.49571228027344, | |
| "learning_rate": 9.8416e-06, | |
| "loss": 3.0201, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 13.846153846153847, | |
| "grad_norm": 121.62274932861328, | |
| "learning_rate": 9.821600000000001e-06, | |
| "loss": 2.9355, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 15.384615384615385, | |
| "grad_norm": 64.66451263427734, | |
| "learning_rate": 9.801600000000001e-06, | |
| "loss": 2.9637, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 16.923076923076923, | |
| "grad_norm": 182.4657440185547, | |
| "learning_rate": 9.781600000000001e-06, | |
| "loss": 2.9819, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 18.46153846153846, | |
| "grad_norm": 97.80529022216797, | |
| "learning_rate": 9.761600000000002e-06, | |
| "loss": 2.9486, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 220.0562744140625, | |
| "learning_rate": 9.741600000000002e-06, | |
| "loss": 2.8608, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 21.53846153846154, | |
| "grad_norm": 95.53397369384766, | |
| "learning_rate": 9.7216e-06, | |
| "loss": 2.8322, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 23.076923076923077, | |
| "grad_norm": 67.54853057861328, | |
| "learning_rate": 9.7016e-06, | |
| "loss": 2.9429, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 24.615384615384617, | |
| "grad_norm": 214.53131103515625, | |
| "learning_rate": 9.6816e-06, | |
| "loss": 2.7927, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 26.153846153846153, | |
| "grad_norm": 293.3318786621094, | |
| "learning_rate": 9.6616e-06, | |
| "loss": 2.7665, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 27.692307692307693, | |
| "grad_norm": 216.3682861328125, | |
| "learning_rate": 9.641600000000001e-06, | |
| "loss": 2.8309, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 29.23076923076923, | |
| "grad_norm": 168.0605010986328, | |
| "learning_rate": 9.621600000000001e-06, | |
| "loss": 2.8433, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 30.76923076923077, | |
| "grad_norm": 103.49143981933594, | |
| "learning_rate": 9.601600000000001e-06, | |
| "loss": 2.7606, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 32.30769230769231, | |
| "grad_norm": 116.6761474609375, | |
| "learning_rate": 9.581600000000002e-06, | |
| "loss": 2.6595, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 33.84615384615385, | |
| "grad_norm": 176.0087432861328, | |
| "learning_rate": 9.5616e-06, | |
| "loss": 2.632, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 35.38461538461539, | |
| "grad_norm": 120.39679718017578, | |
| "learning_rate": 9.5416e-06, | |
| "loss": 2.5941, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 36.92307692307692, | |
| "grad_norm": 100.69256591796875, | |
| "learning_rate": 9.5216e-06, | |
| "loss": 2.6007, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 38.46153846153846, | |
| "grad_norm": 254.54440307617188, | |
| "learning_rate": 9.5016e-06, | |
| "loss": 2.5548, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 113.618896484375, | |
| "learning_rate": 9.4816e-06, | |
| "loss": 2.5157, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 41.53846153846154, | |
| "grad_norm": 825.5616455078125, | |
| "learning_rate": 9.461600000000001e-06, | |
| "loss": 2.4832, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 43.07692307692308, | |
| "grad_norm": 58.83940505981445, | |
| "learning_rate": 9.441600000000001e-06, | |
| "loss": 2.4393, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 44.61538461538461, | |
| "grad_norm": 89.66182708740234, | |
| "learning_rate": 9.421600000000001e-06, | |
| "loss": 2.4381, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 46.15384615384615, | |
| "grad_norm": 242.81410217285156, | |
| "learning_rate": 9.4016e-06, | |
| "loss": 2.3978, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 47.69230769230769, | |
| "grad_norm": 70.43364715576172, | |
| "learning_rate": 9.3816e-06, | |
| "loss": 2.3782, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 49.23076923076923, | |
| "grad_norm": 244.31356811523438, | |
| "learning_rate": 9.3616e-06, | |
| "loss": 2.3319, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 50.76923076923077, | |
| "grad_norm": 59.848697662353516, | |
| "learning_rate": 9.3416e-06, | |
| "loss": 2.3205, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 52.30769230769231, | |
| "grad_norm": 109.26374053955078, | |
| "learning_rate": 9.3216e-06, | |
| "loss": 2.2943, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 53.84615384615385, | |
| "grad_norm": 307.2849426269531, | |
| "learning_rate": 9.301600000000001e-06, | |
| "loss": 2.2913, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 55.38461538461539, | |
| "grad_norm": 69.93729400634766, | |
| "learning_rate": 9.281600000000001e-06, | |
| "loss": 2.2166, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 56.92307692307692, | |
| "grad_norm": 83.3631591796875, | |
| "learning_rate": 9.261600000000001e-06, | |
| "loss": 2.2718, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 58.46153846153846, | |
| "grad_norm": 48.52922821044922, | |
| "learning_rate": 9.2416e-06, | |
| "loss": 2.2645, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 51.951324462890625, | |
| "learning_rate": 9.2218e-06, | |
| "loss": 2.2416, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 61.53846153846154, | |
| "grad_norm": 254.99429321289062, | |
| "learning_rate": 9.2018e-06, | |
| "loss": 2.2104, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 63.07692307692308, | |
| "grad_norm": 109.68157958984375, | |
| "learning_rate": 9.1818e-06, | |
| "loss": 2.1574, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 64.61538461538461, | |
| "grad_norm": 85.5855941772461, | |
| "learning_rate": 9.161800000000001e-06, | |
| "loss": 2.1513, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 66.15384615384616, | |
| "grad_norm": 63.89781951904297, | |
| "learning_rate": 9.141800000000001e-06, | |
| "loss": 2.1331, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 67.6923076923077, | |
| "grad_norm": 184.29612731933594, | |
| "learning_rate": 9.121800000000001e-06, | |
| "loss": 2.1866, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 69.23076923076923, | |
| "grad_norm": 108.59600830078125, | |
| "learning_rate": 9.1018e-06, | |
| "loss": 2.1453, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 70.76923076923077, | |
| "grad_norm": 80.51258087158203, | |
| "learning_rate": 9.0818e-06, | |
| "loss": 2.1125, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 72.3076923076923, | |
| "grad_norm": 172.7328643798828, | |
| "learning_rate": 9.0618e-06, | |
| "loss": 2.141, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 73.84615384615384, | |
| "grad_norm": 58.20785903930664, | |
| "learning_rate": 9.0418e-06, | |
| "loss": 2.0786, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 75.38461538461539, | |
| "grad_norm": 72.46143341064453, | |
| "learning_rate": 9.0218e-06, | |
| "loss": 2.0914, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 76.92307692307692, | |
| "grad_norm": 48.18838119506836, | |
| "learning_rate": 9.0018e-06, | |
| "loss": 2.0439, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 78.46153846153847, | |
| "grad_norm": 87.95365905761719, | |
| "learning_rate": 8.981800000000001e-06, | |
| "loss": 2.0331, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 99.82858276367188, | |
| "learning_rate": 8.961800000000001e-06, | |
| "loss": 2.1268, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 81.53846153846153, | |
| "grad_norm": 85.33229064941406, | |
| "learning_rate": 8.9418e-06, | |
| "loss": 2.028, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 83.07692307692308, | |
| "grad_norm": 83.01776885986328, | |
| "learning_rate": 8.9218e-06, | |
| "loss": 2.0179, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 84.61538461538461, | |
| "grad_norm": 95.0901107788086, | |
| "learning_rate": 8.9018e-06, | |
| "loss": 2.0825, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 86.15384615384616, | |
| "grad_norm": 85.66802978515625, | |
| "learning_rate": 8.8818e-06, | |
| "loss": 2.0588, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 87.6923076923077, | |
| "grad_norm": 69.63407897949219, | |
| "learning_rate": 8.8618e-06, | |
| "loss": 2.0298, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 89.23076923076923, | |
| "grad_norm": 174.88063049316406, | |
| "learning_rate": 8.8418e-06, | |
| "loss": 1.9867, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 90.76923076923077, | |
| "grad_norm": 65.64617156982422, | |
| "learning_rate": 8.8218e-06, | |
| "loss": 2.0674, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 92.3076923076923, | |
| "grad_norm": 152.2218780517578, | |
| "learning_rate": 8.802e-06, | |
| "loss": 1.992, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 93.84615384615384, | |
| "grad_norm": 74.4027328491211, | |
| "learning_rate": 8.782e-06, | |
| "loss": 2.0016, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 95.38461538461539, | |
| "grad_norm": 67.4207992553711, | |
| "learning_rate": 8.762e-06, | |
| "loss": 2.0001, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 96.92307692307692, | |
| "grad_norm": 182.99061584472656, | |
| "learning_rate": 8.742e-06, | |
| "loss": 1.9897, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 98.46153846153847, | |
| "grad_norm": 70.14286804199219, | |
| "learning_rate": 8.722e-06, | |
| "loss": 2.0146, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 117.57647705078125, | |
| "learning_rate": 8.702e-06, | |
| "loss": 1.9298, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 101.53846153846153, | |
| "grad_norm": 80.27694702148438, | |
| "learning_rate": 8.682000000000001e-06, | |
| "loss": 1.9364, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 103.07692307692308, | |
| "grad_norm": 219.95655822753906, | |
| "learning_rate": 8.662000000000001e-06, | |
| "loss": 1.9289, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 104.61538461538461, | |
| "grad_norm": 176.79177856445312, | |
| "learning_rate": 8.642e-06, | |
| "loss": 1.9142, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 106.15384615384616, | |
| "grad_norm": 60.430049896240234, | |
| "learning_rate": 8.622e-06, | |
| "loss": 1.8986, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 107.6923076923077, | |
| "grad_norm": 53.74516296386719, | |
| "learning_rate": 8.602e-06, | |
| "loss": 1.9029, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 109.23076923076923, | |
| "grad_norm": 70.89128112792969, | |
| "learning_rate": 8.582e-06, | |
| "loss": 1.9313, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 110.76923076923077, | |
| "grad_norm": 137.73828125, | |
| "learning_rate": 8.562e-06, | |
| "loss": 1.9328, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 112.3076923076923, | |
| "grad_norm": 86.80708312988281, | |
| "learning_rate": 8.542e-06, | |
| "loss": 1.8831, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 113.84615384615384, | |
| "grad_norm": 52.322608947753906, | |
| "learning_rate": 8.522200000000001e-06, | |
| "loss": 1.8679, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 115.38461538461539, | |
| "grad_norm": 215.9288787841797, | |
| "learning_rate": 8.5022e-06, | |
| "loss": 1.8743, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 116.92307692307692, | |
| "grad_norm": 74.13555908203125, | |
| "learning_rate": 8.4822e-06, | |
| "loss": 1.8575, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 118.46153846153847, | |
| "grad_norm": 138.6452178955078, | |
| "learning_rate": 8.4622e-06, | |
| "loss": 1.8942, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 120.0, | |
| "grad_norm": 57.42049789428711, | |
| "learning_rate": 8.4422e-06, | |
| "loss": 1.8777, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 121.53846153846153, | |
| "grad_norm": 41.08803939819336, | |
| "learning_rate": 8.4222e-06, | |
| "loss": 1.8409, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 123.07692307692308, | |
| "grad_norm": 66.80331420898438, | |
| "learning_rate": 8.4022e-06, | |
| "loss": 1.8325, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 124.61538461538461, | |
| "grad_norm": 69.8762435913086, | |
| "learning_rate": 8.3822e-06, | |
| "loss": 1.8685, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 126.15384615384616, | |
| "grad_norm": 109.0595703125, | |
| "learning_rate": 8.362200000000001e-06, | |
| "loss": 1.8423, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 127.6923076923077, | |
| "grad_norm": 75.782470703125, | |
| "learning_rate": 8.3422e-06, | |
| "loss": 1.8278, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 129.23076923076923, | |
| "grad_norm": 495.26025390625, | |
| "learning_rate": 8.3222e-06, | |
| "loss": 1.823, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 130.76923076923077, | |
| "grad_norm": 132.37982177734375, | |
| "learning_rate": 8.3022e-06, | |
| "loss": 1.8122, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 132.30769230769232, | |
| "grad_norm": 110.41214752197266, | |
| "learning_rate": 8.2822e-06, | |
| "loss": 1.798, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 133.84615384615384, | |
| "grad_norm": 155.9139404296875, | |
| "learning_rate": 8.2622e-06, | |
| "loss": 1.8158, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 135.3846153846154, | |
| "grad_norm": 61.789390563964844, | |
| "learning_rate": 8.2422e-06, | |
| "loss": 1.7863, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 136.92307692307693, | |
| "grad_norm": 86.55162048339844, | |
| "learning_rate": 8.2222e-06, | |
| "loss": 1.7971, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 138.46153846153845, | |
| "grad_norm": 52.087581634521484, | |
| "learning_rate": 8.202200000000001e-06, | |
| "loss": 1.7676, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 140.0, | |
| "grad_norm": 78.67654418945312, | |
| "learning_rate": 8.1822e-06, | |
| "loss": 1.8188, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 141.53846153846155, | |
| "grad_norm": 102.5003890991211, | |
| "learning_rate": 8.1622e-06, | |
| "loss": 1.8154, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 143.07692307692307, | |
| "grad_norm": 112.49629974365234, | |
| "learning_rate": 8.142200000000001e-06, | |
| "loss": 1.7687, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 144.6153846153846, | |
| "grad_norm": 64.63585662841797, | |
| "learning_rate": 8.122200000000002e-06, | |
| "loss": 1.7824, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 146.15384615384616, | |
| "grad_norm": 74.50164794921875, | |
| "learning_rate": 8.102200000000002e-06, | |
| "loss": 1.745, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 147.69230769230768, | |
| "grad_norm": 70.59980010986328, | |
| "learning_rate": 8.082200000000002e-06, | |
| "loss": 1.8179, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 149.23076923076923, | |
| "grad_norm": 275.9650573730469, | |
| "learning_rate": 8.0622e-06, | |
| "loss": 1.7744, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 150.76923076923077, | |
| "grad_norm": 54.21783447265625, | |
| "learning_rate": 8.0422e-06, | |
| "loss": 1.7608, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 152.30769230769232, | |
| "grad_norm": 58.89350128173828, | |
| "learning_rate": 8.0224e-06, | |
| "loss": 1.7835, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 153.84615384615384, | |
| "grad_norm": 90.88015747070312, | |
| "learning_rate": 8.0024e-06, | |
| "loss": 1.7567, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 155.3846153846154, | |
| "grad_norm": 45.44640350341797, | |
| "learning_rate": 7.9824e-06, | |
| "loss": 1.7399, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 156.92307692307693, | |
| "grad_norm": 58.44306945800781, | |
| "learning_rate": 7.9624e-06, | |
| "loss": 1.7803, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 158.46153846153845, | |
| "grad_norm": 80.27843475341797, | |
| "learning_rate": 7.9424e-06, | |
| "loss": 1.7106, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 160.0, | |
| "grad_norm": 111.12615203857422, | |
| "learning_rate": 7.9224e-06, | |
| "loss": 1.7134, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 161.53846153846155, | |
| "grad_norm": 63.302520751953125, | |
| "learning_rate": 7.9024e-06, | |
| "loss": 1.7031, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 163.07692307692307, | |
| "grad_norm": 41.779808044433594, | |
| "learning_rate": 7.882400000000001e-06, | |
| "loss": 1.7071, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 164.6153846153846, | |
| "grad_norm": 58.44318389892578, | |
| "learning_rate": 7.862400000000001e-06, | |
| "loss": 1.6759, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 166.15384615384616, | |
| "grad_norm": 36.30861282348633, | |
| "learning_rate": 7.842400000000001e-06, | |
| "loss": 1.7012, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 167.69230769230768, | |
| "grad_norm": 68.7526626586914, | |
| "learning_rate": 7.822400000000002e-06, | |
| "loss": 1.6963, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 169.23076923076923, | |
| "grad_norm": 82.03331756591797, | |
| "learning_rate": 7.802400000000002e-06, | |
| "loss": 1.7456, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 170.76923076923077, | |
| "grad_norm": 53.351959228515625, | |
| "learning_rate": 7.7824e-06, | |
| "loss": 1.7284, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 172.30769230769232, | |
| "grad_norm": 58.400909423828125, | |
| "learning_rate": 7.7624e-06, | |
| "loss": 1.7266, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 173.84615384615384, | |
| "grad_norm": 47.95368576049805, | |
| "learning_rate": 7.7424e-06, | |
| "loss": 1.6671, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 175.3846153846154, | |
| "grad_norm": 54.79304885864258, | |
| "learning_rate": 7.7224e-06, | |
| "loss": 1.7092, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 176.92307692307693, | |
| "grad_norm": 34.120758056640625, | |
| "learning_rate": 7.702400000000001e-06, | |
| "loss": 1.724, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 178.46153846153845, | |
| "grad_norm": 80.98869323730469, | |
| "learning_rate": 7.682400000000001e-06, | |
| "loss": 1.6693, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 180.0, | |
| "grad_norm": 65.80884552001953, | |
| "learning_rate": 7.662400000000001e-06, | |
| "loss": 1.6864, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 181.53846153846155, | |
| "grad_norm": 86.81282806396484, | |
| "learning_rate": 7.642400000000002e-06, | |
| "loss": 1.6715, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 183.07692307692307, | |
| "grad_norm": 63.2209358215332, | |
| "learning_rate": 7.622400000000001e-06, | |
| "loss": 1.6884, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 184.6153846153846, | |
| "grad_norm": 80.72114562988281, | |
| "learning_rate": 7.602400000000001e-06, | |
| "loss": 1.6781, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 186.15384615384616, | |
| "grad_norm": 65.09835815429688, | |
| "learning_rate": 7.5824000000000005e-06, | |
| "loss": 1.6451, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 187.69230769230768, | |
| "grad_norm": 51.8597297668457, | |
| "learning_rate": 7.562400000000001e-06, | |
| "loss": 1.6425, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 189.23076923076923, | |
| "grad_norm": 55.06999969482422, | |
| "learning_rate": 7.542400000000001e-06, | |
| "loss": 1.6475, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 190.76923076923077, | |
| "grad_norm": 221.47671508789062, | |
| "learning_rate": 7.522400000000001e-06, | |
| "loss": 1.6459, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 192.30769230769232, | |
| "grad_norm": 64.53801727294922, | |
| "learning_rate": 7.5024e-06, | |
| "loss": 1.6149, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 193.84615384615384, | |
| "grad_norm": 96.24422454833984, | |
| "learning_rate": 7.4824000000000006e-06, | |
| "loss": 1.6204, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 195.3846153846154, | |
| "grad_norm": 44.547306060791016, | |
| "learning_rate": 7.462400000000001e-06, | |
| "loss": 1.6151, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 196.92307692307693, | |
| "grad_norm": 125.61988830566406, | |
| "learning_rate": 7.442400000000001e-06, | |
| "loss": 1.6244, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 198.46153846153845, | |
| "grad_norm": 98.02783966064453, | |
| "learning_rate": 7.422400000000001e-06, | |
| "loss": 1.625, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "grad_norm": 72.0836181640625, | |
| "learning_rate": 7.4024000000000005e-06, | |
| "loss": 1.6201, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 201.53846153846155, | |
| "grad_norm": 157.2918701171875, | |
| "learning_rate": 7.382400000000001e-06, | |
| "loss": 1.6065, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 203.07692307692307, | |
| "grad_norm": 51.04494094848633, | |
| "learning_rate": 7.362400000000001e-06, | |
| "loss": 1.6227, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 204.6153846153846, | |
| "grad_norm": 53.116188049316406, | |
| "learning_rate": 7.342400000000001e-06, | |
| "loss": 1.6284, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 206.15384615384616, | |
| "grad_norm": 80.10623168945312, | |
| "learning_rate": 7.3224e-06, | |
| "loss": 1.619, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 207.69230769230768, | |
| "grad_norm": 249.86141967773438, | |
| "learning_rate": 7.302400000000001e-06, | |
| "loss": 1.6018, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 209.23076923076923, | |
| "grad_norm": 77.1427001953125, | |
| "learning_rate": 7.282400000000001e-06, | |
| "loss": 1.6329, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 210.76923076923077, | |
| "grad_norm": 86.5726318359375, | |
| "learning_rate": 7.2626000000000006e-06, | |
| "loss": 1.6112, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 212.30769230769232, | |
| "grad_norm": 66.48815155029297, | |
| "learning_rate": 7.242600000000001e-06, | |
| "loss": 1.6128, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 213.84615384615384, | |
| "grad_norm": 99.87622833251953, | |
| "learning_rate": 7.222600000000001e-06, | |
| "loss": 1.5902, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 215.3846153846154, | |
| "grad_norm": 78.00489807128906, | |
| "learning_rate": 7.2026e-06, | |
| "loss": 1.5964, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 216.92307692307693, | |
| "grad_norm": 726.4310913085938, | |
| "learning_rate": 7.1826000000000005e-06, | |
| "loss": 1.6301, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 218.46153846153845, | |
| "grad_norm": 60.30306625366211, | |
| "learning_rate": 7.162600000000001e-06, | |
| "loss": 1.5653, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 220.0, | |
| "grad_norm": 88.68881225585938, | |
| "learning_rate": 7.142600000000001e-06, | |
| "loss": 1.5743, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 221.53846153846155, | |
| "grad_norm": 101.28644561767578, | |
| "learning_rate": 7.1226e-06, | |
| "loss": 1.5589, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 223.07692307692307, | |
| "grad_norm": 248.6482696533203, | |
| "learning_rate": 7.1026e-06, | |
| "loss": 1.5706, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 224.6153846153846, | |
| "grad_norm": 110.11705780029297, | |
| "learning_rate": 7.082600000000001e-06, | |
| "loss": 1.5843, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 226.15384615384616, | |
| "grad_norm": 67.07064819335938, | |
| "learning_rate": 7.062600000000001e-06, | |
| "loss": 1.5643, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 227.69230769230768, | |
| "grad_norm": 97.37787628173828, | |
| "learning_rate": 7.0426e-06, | |
| "loss": 1.5424, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 229.23076923076923, | |
| "grad_norm": 53.19269561767578, | |
| "learning_rate": 7.0226e-06, | |
| "loss": 1.5644, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 230.76923076923077, | |
| "grad_norm": 213.32289123535156, | |
| "learning_rate": 7.0026000000000005e-06, | |
| "loss": 1.5541, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 232.30769230769232, | |
| "grad_norm": 105.63107299804688, | |
| "learning_rate": 6.982600000000001e-06, | |
| "loss": 1.5518, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 233.84615384615384, | |
| "grad_norm": 43.85015106201172, | |
| "learning_rate": 6.9626e-06, | |
| "loss": 1.5385, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 235.3846153846154, | |
| "grad_norm": 135.48651123046875, | |
| "learning_rate": 6.9426e-06, | |
| "loss": 1.5093, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 236.92307692307693, | |
| "grad_norm": 73.94020080566406, | |
| "learning_rate": 6.9226000000000004e-06, | |
| "loss": 1.537, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 238.46153846153845, | |
| "grad_norm": 157.08612060546875, | |
| "learning_rate": 6.902600000000001e-06, | |
| "loss": 1.5757, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 240.0, | |
| "grad_norm": 79.48292541503906, | |
| "learning_rate": 6.8826e-06, | |
| "loss": 1.5214, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 241.53846153846155, | |
| "grad_norm": 59.73866653442383, | |
| "learning_rate": 6.8626e-06, | |
| "loss": 1.5166, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 243.07692307692307, | |
| "grad_norm": 97.34918212890625, | |
| "learning_rate": 6.8426e-06, | |
| "loss": 1.5145, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 244.6153846153846, | |
| "grad_norm": 198.27203369140625, | |
| "learning_rate": 6.8226000000000005e-06, | |
| "loss": 1.5237, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 246.15384615384616, | |
| "grad_norm": 38.80288314819336, | |
| "learning_rate": 6.802600000000001e-06, | |
| "loss": 1.5118, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 247.69230769230768, | |
| "grad_norm": 89.98989868164062, | |
| "learning_rate": 6.7828000000000005e-06, | |
| "loss": 1.5171, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 249.23076923076923, | |
| "grad_norm": 55.13770294189453, | |
| "learning_rate": 6.762800000000001e-06, | |
| "loss": 1.5044, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 250.76923076923077, | |
| "grad_norm": 34.529293060302734, | |
| "learning_rate": 6.7428e-06, | |
| "loss": 1.4981, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 252.30769230769232, | |
| "grad_norm": 79.73957824707031, | |
| "learning_rate": 6.7228e-06, | |
| "loss": 1.501, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 253.84615384615384, | |
| "grad_norm": 51.99100875854492, | |
| "learning_rate": 6.7028000000000004e-06, | |
| "loss": 1.4968, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 255.3846153846154, | |
| "grad_norm": 47.11521530151367, | |
| "learning_rate": 6.682800000000001e-06, | |
| "loss": 1.4961, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 256.9230769230769, | |
| "grad_norm": 70.4249496459961, | |
| "learning_rate": 6.6628e-06, | |
| "loss": 1.5162, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 258.46153846153845, | |
| "grad_norm": 56.895957946777344, | |
| "learning_rate": 6.6428e-06, | |
| "loss": 1.5187, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 260.0, | |
| "grad_norm": 34.502071380615234, | |
| "learning_rate": 6.623000000000001e-06, | |
| "loss": 1.5325, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 261.53846153846155, | |
| "grad_norm": 66.03074645996094, | |
| "learning_rate": 6.603e-06, | |
| "loss": 1.5103, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 263.0769230769231, | |
| "grad_norm": 110.48493957519531, | |
| "learning_rate": 6.583e-06, | |
| "loss": 1.4571, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 264.61538461538464, | |
| "grad_norm": 55.59686279296875, | |
| "learning_rate": 6.5630000000000005e-06, | |
| "loss": 1.5147, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 266.15384615384613, | |
| "grad_norm": 72.83611297607422, | |
| "learning_rate": 6.543000000000001e-06, | |
| "loss": 1.48, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 267.6923076923077, | |
| "grad_norm": 54.794437408447266, | |
| "learning_rate": 6.523e-06, | |
| "loss": 1.461, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 269.2307692307692, | |
| "grad_norm": 41.280731201171875, | |
| "learning_rate": 6.503e-06, | |
| "loss": 1.4188, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 270.7692307692308, | |
| "grad_norm": 81.90428924560547, | |
| "learning_rate": 6.4830000000000004e-06, | |
| "loss": 1.4256, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 272.3076923076923, | |
| "grad_norm": 67.94799041748047, | |
| "learning_rate": 6.463000000000001e-06, | |
| "loss": 1.4295, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 273.84615384615387, | |
| "grad_norm": 77.92097473144531, | |
| "learning_rate": 6.443e-06, | |
| "loss": 1.4575, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 275.38461538461536, | |
| "grad_norm": 83.06784057617188, | |
| "learning_rate": 6.423e-06, | |
| "loss": 1.434, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 276.9230769230769, | |
| "grad_norm": 60.48369598388672, | |
| "learning_rate": 6.403e-06, | |
| "loss": 1.4496, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 278.46153846153845, | |
| "grad_norm": 47.32282257080078, | |
| "learning_rate": 6.3830000000000006e-06, | |
| "loss": 1.4149, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 280.0, | |
| "grad_norm": 134.77822875976562, | |
| "learning_rate": 6.363e-06, | |
| "loss": 1.3949, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 281.53846153846155, | |
| "grad_norm": 72.09223937988281, | |
| "learning_rate": 6.343e-06, | |
| "loss": 1.4298, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 283.0769230769231, | |
| "grad_norm": 95.94608306884766, | |
| "learning_rate": 6.323e-06, | |
| "loss": 1.4189, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 284.61538461538464, | |
| "grad_norm": 79.95153045654297, | |
| "learning_rate": 6.3030000000000005e-06, | |
| "loss": 1.4066, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 286.15384615384613, | |
| "grad_norm": 165.74986267089844, | |
| "learning_rate": 6.283e-06, | |
| "loss": 1.4231, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 287.6923076923077, | |
| "grad_norm": 105.84443664550781, | |
| "learning_rate": 6.263e-06, | |
| "loss": 1.3851, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 289.2307692307692, | |
| "grad_norm": 48.660987854003906, | |
| "learning_rate": 6.243e-06, | |
| "loss": 1.4234, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 290.7692307692308, | |
| "grad_norm": 53.28395462036133, | |
| "learning_rate": 6.223e-06, | |
| "loss": 1.3861, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 292.3076923076923, | |
| "grad_norm": 54.91928482055664, | |
| "learning_rate": 6.203e-06, | |
| "loss": 1.3799, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 293.84615384615387, | |
| "grad_norm": 69.98273468017578, | |
| "learning_rate": 6.183e-06, | |
| "loss": 1.4223, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 295.38461538461536, | |
| "grad_norm": 55.819786071777344, | |
| "learning_rate": 6.163e-06, | |
| "loss": 1.3747, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 296.9230769230769, | |
| "grad_norm": 64.86967468261719, | |
| "learning_rate": 6.143e-06, | |
| "loss": 1.4014, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 298.46153846153845, | |
| "grad_norm": 96.05943298339844, | |
| "learning_rate": 6.123e-06, | |
| "loss": 1.3517, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "grad_norm": 37.66103744506836, | |
| "learning_rate": 6.103e-06, | |
| "loss": 1.3625, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 301.53846153846155, | |
| "grad_norm": 38.60346603393555, | |
| "learning_rate": 6.083e-06, | |
| "loss": 1.4039, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 303.0769230769231, | |
| "grad_norm": 51.319610595703125, | |
| "learning_rate": 6.063e-06, | |
| "loss": 1.3722, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 304.61538461538464, | |
| "grad_norm": 63.50655746459961, | |
| "learning_rate": 6.0429999999999996e-06, | |
| "loss": 1.3789, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 306.15384615384613, | |
| "grad_norm": 1452.8902587890625, | |
| "learning_rate": 6.023e-06, | |
| "loss": 1.352, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 307.6923076923077, | |
| "grad_norm": 45.927669525146484, | |
| "learning_rate": 6.003e-06, | |
| "loss": 1.3539, | |
| "step": 20000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 50000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 770, | |
| "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": 7.64486332416e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |