Download trainer_state.json from datasetsANDmodels/time-extraction: direct link, hf CLI and curl.
- Browser
- Download file 408 kB
-
https://huggingface.co/datasetsANDmodels/time-extraction/resolve/main/trainer_state.json
- Command line
-
hf download hf://datasetsANDmodels/time-extraction/trainer_state.json
-
curl -L -o trainer_state.json https://huggingface.co/datasetsANDmodels/time-extraction/resolve/main/trainer_state.json
408 kB
| { | |
| "best_metric": 0.051427390426397324, | |
| "best_model_checkpoint": "time_base/checkpoint-2340", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2340, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008547008547008548, | |
| "grad_norm": 221.6373748779297, | |
| "learning_rate": 9.995726495726496e-06, | |
| "loss": 37.5765, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017094017094017096, | |
| "grad_norm": 219.50563049316406, | |
| "learning_rate": 9.991452991452993e-06, | |
| "loss": 38.6173, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02564102564102564, | |
| "grad_norm": 180.23829650878906, | |
| "learning_rate": 9.987179487179488e-06, | |
| "loss": 40.3853, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03418803418803419, | |
| "grad_norm": 166.3365478515625, | |
| "learning_rate": 9.982905982905984e-06, | |
| "loss": 35.9724, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.042735042735042736, | |
| "grad_norm": 199.6571044921875, | |
| "learning_rate": 9.97863247863248e-06, | |
| "loss": 35.0186, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05128205128205128, | |
| "grad_norm": 180.9748992919922, | |
| "learning_rate": 9.974358974358974e-06, | |
| "loss": 39.3679, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05982905982905983, | |
| "grad_norm": 200.05496215820312, | |
| "learning_rate": 9.970085470085471e-06, | |
| "loss": 37.1519, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06837606837606838, | |
| "grad_norm": 154.3177032470703, | |
| "learning_rate": 9.965811965811966e-06, | |
| "loss": 33.9309, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 198.05914306640625, | |
| "learning_rate": 9.961538461538463e-06, | |
| "loss": 34.8814, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08547008547008547, | |
| "grad_norm": 168.3035430908203, | |
| "learning_rate": 9.957264957264958e-06, | |
| "loss": 33.184, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09401709401709402, | |
| "grad_norm": 201.83705139160156, | |
| "learning_rate": 9.952991452991455e-06, | |
| "loss": 35.4025, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10256410256410256, | |
| "grad_norm": 224.4587860107422, | |
| "learning_rate": 9.94871794871795e-06, | |
| "loss": 39.222, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 192.1949005126953, | |
| "learning_rate": 9.944444444444445e-06, | |
| "loss": 37.1982, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11965811965811966, | |
| "grad_norm": 193.05662536621094, | |
| "learning_rate": 9.940170940170942e-06, | |
| "loss": 38.1325, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1282051282051282, | |
| "grad_norm": 150.61575317382812, | |
| "learning_rate": 9.935897435897437e-06, | |
| "loss": 34.8682, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.13675213675213677, | |
| "grad_norm": 170.1510772705078, | |
| "learning_rate": 9.931623931623933e-06, | |
| "loss": 33.3652, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1452991452991453, | |
| "grad_norm": 193.86875915527344, | |
| "learning_rate": 9.927350427350428e-06, | |
| "loss": 35.0785, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 164.41986083984375, | |
| "learning_rate": 9.923076923076923e-06, | |
| "loss": 31.9719, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1623931623931624, | |
| "grad_norm": 166.08953857421875, | |
| "learning_rate": 9.91880341880342e-06, | |
| "loss": 34.5398, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 152.2139892578125, | |
| "learning_rate": 9.914529914529915e-06, | |
| "loss": 36.9092, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1794871794871795, | |
| "grad_norm": 198.23095703125, | |
| "learning_rate": 9.910256410256412e-06, | |
| "loss": 35.6744, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.18803418803418803, | |
| "grad_norm": 174.7784881591797, | |
| "learning_rate": 9.905982905982907e-06, | |
| "loss": 32.8258, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.19658119658119658, | |
| "grad_norm": 133.69859313964844, | |
| "learning_rate": 9.901709401709402e-06, | |
| "loss": 31.431, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.20512820512820512, | |
| "grad_norm": 217.17169189453125, | |
| "learning_rate": 9.897435897435899e-06, | |
| "loss": 38.5649, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.21367521367521367, | |
| "grad_norm": 172.4914093017578, | |
| "learning_rate": 9.893162393162394e-06, | |
| "loss": 33.9858, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 186.39654541015625, | |
| "learning_rate": 9.88888888888889e-06, | |
| "loss": 32.8029, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 183.65159606933594, | |
| "learning_rate": 9.884615384615386e-06, | |
| "loss": 35.8633, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.23931623931623933, | |
| "grad_norm": 228.352294921875, | |
| "learning_rate": 9.880341880341882e-06, | |
| "loss": 35.0285, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.24786324786324787, | |
| "grad_norm": 156.77906799316406, | |
| "learning_rate": 9.876068376068377e-06, | |
| "loss": 29.2608, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 232.8336181640625, | |
| "learning_rate": 9.871794871794872e-06, | |
| "loss": 35.0349, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.26495726495726496, | |
| "grad_norm": 248.63247680664062, | |
| "learning_rate": 9.86752136752137e-06, | |
| "loss": 34.5067, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.27350427350427353, | |
| "grad_norm": 183.5840606689453, | |
| "learning_rate": 9.863247863247864e-06, | |
| "loss": 30.4758, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.28205128205128205, | |
| "grad_norm": 160.54530334472656, | |
| "learning_rate": 9.858974358974361e-06, | |
| "loss": 31.7959, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2905982905982906, | |
| "grad_norm": 199.88156127929688, | |
| "learning_rate": 9.854700854700856e-06, | |
| "loss": 35.6482, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.29914529914529914, | |
| "grad_norm": 272.9530029296875, | |
| "learning_rate": 9.850427350427351e-06, | |
| "loss": 33.0804, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 200.0990447998047, | |
| "learning_rate": 9.846153846153848e-06, | |
| "loss": 33.2675, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3162393162393162, | |
| "grad_norm": 202.014404296875, | |
| "learning_rate": 9.841880341880343e-06, | |
| "loss": 30.8991, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3247863247863248, | |
| "grad_norm": 181.14865112304688, | |
| "learning_rate": 9.837606837606838e-06, | |
| "loss": 32.3643, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 134.43423461914062, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 30.8094, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 155.96640014648438, | |
| "learning_rate": 9.82905982905983e-06, | |
| "loss": 31.7564, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3504273504273504, | |
| "grad_norm": 146.9285888671875, | |
| "learning_rate": 9.824786324786325e-06, | |
| "loss": 31.9905, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.358974358974359, | |
| "grad_norm": 159.67974853515625, | |
| "learning_rate": 9.820512820512821e-06, | |
| "loss": 32.5029, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.36752136752136755, | |
| "grad_norm": 172.4975128173828, | |
| "learning_rate": 9.816239316239316e-06, | |
| "loss": 31.2049, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.37606837606837606, | |
| "grad_norm": 148.97573852539062, | |
| "learning_rate": 9.811965811965812e-06, | |
| "loss": 27.1673, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 115.93009185791016, | |
| "learning_rate": 9.807692307692308e-06, | |
| "loss": 30.3342, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.39316239316239315, | |
| "grad_norm": 184.13145446777344, | |
| "learning_rate": 9.803418803418803e-06, | |
| "loss": 32.317, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4017094017094017, | |
| "grad_norm": 139.3995361328125, | |
| "learning_rate": 9.7991452991453e-06, | |
| "loss": 29.9643, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.41025641025641024, | |
| "grad_norm": 184.97996520996094, | |
| "learning_rate": 9.794871794871795e-06, | |
| "loss": 30.6427, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4188034188034188, | |
| "grad_norm": 120.04417419433594, | |
| "learning_rate": 9.790598290598292e-06, | |
| "loss": 26.9772, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "grad_norm": 183.2873077392578, | |
| "learning_rate": 9.786324786324787e-06, | |
| "loss": 31.6688, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4358974358974359, | |
| "grad_norm": 206.44898986816406, | |
| "learning_rate": 9.782051282051282e-06, | |
| "loss": 32.0574, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 180.7601318359375, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 31.2178, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.452991452991453, | |
| "grad_norm": 150.44012451171875, | |
| "learning_rate": 9.773504273504274e-06, | |
| "loss": 29.9826, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 119.02840423583984, | |
| "learning_rate": 9.76923076923077e-06, | |
| "loss": 26.876, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4700854700854701, | |
| "grad_norm": 164.58209228515625, | |
| "learning_rate": 9.764957264957265e-06, | |
| "loss": 28.1059, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.47863247863247865, | |
| "grad_norm": 160.416259765625, | |
| "learning_rate": 9.76068376068376e-06, | |
| "loss": 28.7022, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.48717948717948717, | |
| "grad_norm": 177.29747009277344, | |
| "learning_rate": 9.756410256410257e-06, | |
| "loss": 30.7275, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.49572649572649574, | |
| "grad_norm": 153.59686279296875, | |
| "learning_rate": 9.752136752136752e-06, | |
| "loss": 28.5575, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5042735042735043, | |
| "grad_norm": 155.79617309570312, | |
| "learning_rate": 9.747863247863249e-06, | |
| "loss": 28.1139, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 173.02581787109375, | |
| "learning_rate": 9.743589743589744e-06, | |
| "loss": 30.4744, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5213675213675214, | |
| "grad_norm": 125.31639862060547, | |
| "learning_rate": 9.739316239316239e-06, | |
| "loss": 26.5559, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5299145299145299, | |
| "grad_norm": 149.00302124023438, | |
| "learning_rate": 9.735042735042736e-06, | |
| "loss": 30.4065, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 101.76395416259766, | |
| "learning_rate": 9.730769230769231e-06, | |
| "loss": 25.8895, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5470085470085471, | |
| "grad_norm": 134.40159606933594, | |
| "learning_rate": 9.726495726495728e-06, | |
| "loss": 26.9317, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 151.01914978027344, | |
| "learning_rate": 9.722222222222223e-06, | |
| "loss": 27.9913, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5641025641025641, | |
| "grad_norm": 124.92068481445312, | |
| "learning_rate": 9.71794871794872e-06, | |
| "loss": 26.7874, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5726495726495726, | |
| "grad_norm": 131.29762268066406, | |
| "learning_rate": 9.713675213675214e-06, | |
| "loss": 27.4047, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5811965811965812, | |
| "grad_norm": 154.37120056152344, | |
| "learning_rate": 9.70940170940171e-06, | |
| "loss": 26.6812, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5897435897435898, | |
| "grad_norm": 86.31095886230469, | |
| "learning_rate": 9.705128205128206e-06, | |
| "loss": 22.9869, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5982905982905983, | |
| "grad_norm": 224.42613220214844, | |
| "learning_rate": 9.700854700854701e-06, | |
| "loss": 28.4812, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6068376068376068, | |
| "grad_norm": 156.15228271484375, | |
| "learning_rate": 9.696581196581198e-06, | |
| "loss": 26.1761, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 117.7806167602539, | |
| "learning_rate": 9.692307692307693e-06, | |
| "loss": 20.7307, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6239316239316239, | |
| "grad_norm": 169.99154663085938, | |
| "learning_rate": 9.688034188034188e-06, | |
| "loss": 27.6369, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6324786324786325, | |
| "grad_norm": 98.81549072265625, | |
| "learning_rate": 9.683760683760685e-06, | |
| "loss": 24.5898, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6410256410256411, | |
| "grad_norm": 199.0179443359375, | |
| "learning_rate": 9.67948717948718e-06, | |
| "loss": 27.664, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6495726495726496, | |
| "grad_norm": 129.81033325195312, | |
| "learning_rate": 9.675213675213677e-06, | |
| "loss": 25.2547, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6581196581196581, | |
| "grad_norm": 140.1121826171875, | |
| "learning_rate": 9.670940170940172e-06, | |
| "loss": 27.4914, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 139.8365478515625, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 24.0178, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6752136752136753, | |
| "grad_norm": 147.24945068359375, | |
| "learning_rate": 9.662393162393163e-06, | |
| "loss": 27.1404, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 165.67242431640625, | |
| "learning_rate": 9.658119658119659e-06, | |
| "loss": 25.6604, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 114.36772918701172, | |
| "learning_rate": 9.653846153846155e-06, | |
| "loss": 24.3695, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7008547008547008, | |
| "grad_norm": 149.76258850097656, | |
| "learning_rate": 9.64957264957265e-06, | |
| "loss": 26.5265, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7094017094017094, | |
| "grad_norm": 121.9085693359375, | |
| "learning_rate": 9.645299145299147e-06, | |
| "loss": 25.7008, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.717948717948718, | |
| "grad_norm": 106.49151611328125, | |
| "learning_rate": 9.641025641025642e-06, | |
| "loss": 20.5777, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7264957264957265, | |
| "grad_norm": 114.2357406616211, | |
| "learning_rate": 9.636752136752137e-06, | |
| "loss": 23.3429, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7350427350427351, | |
| "grad_norm": 107.38651275634766, | |
| "learning_rate": 9.632478632478634e-06, | |
| "loss": 24.6408, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7435897435897436, | |
| "grad_norm": 120.4283218383789, | |
| "learning_rate": 9.628205128205129e-06, | |
| "loss": 23.4563, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7521367521367521, | |
| "grad_norm": 165.21783447265625, | |
| "learning_rate": 9.623931623931626e-06, | |
| "loss": 25.878, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7606837606837606, | |
| "grad_norm": 105.8712387084961, | |
| "learning_rate": 9.61965811965812e-06, | |
| "loss": 23.605, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 147.31253051757812, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 24.537, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 127.37718963623047, | |
| "learning_rate": 9.611111111111112e-06, | |
| "loss": 24.6762, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7863247863247863, | |
| "grad_norm": 139.40553283691406, | |
| "learning_rate": 9.606837606837607e-06, | |
| "loss": 23.6076, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7948717948717948, | |
| "grad_norm": 218.39170837402344, | |
| "learning_rate": 9.602564102564104e-06, | |
| "loss": 25.2559, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8034188034188035, | |
| "grad_norm": 115.83401489257812, | |
| "learning_rate": 9.5982905982906e-06, | |
| "loss": 23.6758, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.811965811965812, | |
| "grad_norm": 115.8538818359375, | |
| "learning_rate": 9.594017094017094e-06, | |
| "loss": 24.2789, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8205128205128205, | |
| "grad_norm": 122.31534576416016, | |
| "learning_rate": 9.589743589743591e-06, | |
| "loss": 23.5114, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8290598290598291, | |
| "grad_norm": 171.58558654785156, | |
| "learning_rate": 9.585470085470086e-06, | |
| "loss": 24.7028, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8376068376068376, | |
| "grad_norm": 113.29806518554688, | |
| "learning_rate": 9.581196581196583e-06, | |
| "loss": 24.9667, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 183.74928283691406, | |
| "learning_rate": 9.576923076923078e-06, | |
| "loss": 24.7776, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 139.84701538085938, | |
| "learning_rate": 9.572649572649575e-06, | |
| "loss": 22.1558, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8632478632478633, | |
| "grad_norm": 145.9014129638672, | |
| "learning_rate": 9.56837606837607e-06, | |
| "loss": 23.0282, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8717948717948718, | |
| "grad_norm": 195.9859619140625, | |
| "learning_rate": 9.564102564102565e-06, | |
| "loss": 23.7194, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8803418803418803, | |
| "grad_norm": 70.51985168457031, | |
| "learning_rate": 9.559829059829061e-06, | |
| "loss": 16.9605, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 184.04209899902344, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 23.4229, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8974358974358975, | |
| "grad_norm": 177.86727905273438, | |
| "learning_rate": 9.551282051282053e-06, | |
| "loss": 23.6004, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.905982905982906, | |
| "grad_norm": 154.30784606933594, | |
| "learning_rate": 9.547008547008548e-06, | |
| "loss": 21.6725, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9145299145299145, | |
| "grad_norm": 104.27069854736328, | |
| "learning_rate": 9.542735042735043e-06, | |
| "loss": 22.856, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 157.4270477294922, | |
| "learning_rate": 9.53846153846154e-06, | |
| "loss": 24.398, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9316239316239316, | |
| "grad_norm": 123.56739807128906, | |
| "learning_rate": 9.534188034188035e-06, | |
| "loss": 20.6925, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9401709401709402, | |
| "grad_norm": 106.64054870605469, | |
| "learning_rate": 9.52991452991453e-06, | |
| "loss": 23.5794, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9487179487179487, | |
| "grad_norm": 88.68234252929688, | |
| "learning_rate": 9.525641025641025e-06, | |
| "loss": 20.729, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9572649572649573, | |
| "grad_norm": 91.86422729492188, | |
| "learning_rate": 9.521367521367522e-06, | |
| "loss": 18.7701, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9658119658119658, | |
| "grad_norm": 118.74354553222656, | |
| "learning_rate": 9.517094017094017e-06, | |
| "loss": 20.8439, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.9743589743589743, | |
| "grad_norm": 120.72904968261719, | |
| "learning_rate": 9.512820512820514e-06, | |
| "loss": 21.1903, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.9829059829059829, | |
| "grad_norm": 107.36665344238281, | |
| "learning_rate": 9.508547008547009e-06, | |
| "loss": 19.3457, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9914529914529915, | |
| "grad_norm": 131.74441528320312, | |
| "learning_rate": 9.504273504273504e-06, | |
| "loss": 21.4035, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 161.97703552246094, | |
| "learning_rate": 9.5e-06, | |
| "loss": 22.3831, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 17.230430603027344, | |
| "eval_runtime": 9.9187, | |
| "eval_samples_per_second": 46.982, | |
| "eval_steps_per_second": 5.948, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0085470085470085, | |
| "grad_norm": 109.44770050048828, | |
| "learning_rate": 9.495726495726496e-06, | |
| "loss": 20.3406, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.017094017094017, | |
| "grad_norm": 96.50030517578125, | |
| "learning_rate": 9.491452991452992e-06, | |
| "loss": 19.8086, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 91.27509307861328, | |
| "learning_rate": 9.487179487179487e-06, | |
| "loss": 18.8737, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0341880341880343, | |
| "grad_norm": 123.94478607177734, | |
| "learning_rate": 9.482905982905984e-06, | |
| "loss": 20.1785, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0427350427350428, | |
| "grad_norm": 109.29426574707031, | |
| "learning_rate": 9.478632478632479e-06, | |
| "loss": 18.8151, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0512820512820513, | |
| "grad_norm": 104.0233383178711, | |
| "learning_rate": 9.474358974358974e-06, | |
| "loss": 19.6281, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0598290598290598, | |
| "grad_norm": 75.7523193359375, | |
| "learning_rate": 9.470085470085471e-06, | |
| "loss": 18.5031, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.0683760683760684, | |
| "grad_norm": 103.1374740600586, | |
| "learning_rate": 9.465811965811966e-06, | |
| "loss": 19.6443, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 92.68035888671875, | |
| "learning_rate": 9.461538461538463e-06, | |
| "loss": 19.7327, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.0854700854700854, | |
| "grad_norm": 88.10079193115234, | |
| "learning_rate": 9.457264957264958e-06, | |
| "loss": 17.8832, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0940170940170941, | |
| "grad_norm": 80.04244232177734, | |
| "learning_rate": 9.452991452991453e-06, | |
| "loss": 16.4485, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.1025641025641026, | |
| "grad_norm": 81.02445983886719, | |
| "learning_rate": 9.44871794871795e-06, | |
| "loss": 17.3035, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 98.95979309082031, | |
| "learning_rate": 9.444444444444445e-06, | |
| "loss": 17.5734, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1196581196581197, | |
| "grad_norm": 109.76984405517578, | |
| "learning_rate": 9.440170940170941e-06, | |
| "loss": 20.3985, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1282051282051282, | |
| "grad_norm": 98.52857208251953, | |
| "learning_rate": 9.435897435897436e-06, | |
| "loss": 17.7275, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1367521367521367, | |
| "grad_norm": 91.28802490234375, | |
| "learning_rate": 9.431623931623931e-06, | |
| "loss": 17.9107, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1452991452991452, | |
| "grad_norm": 92.89081573486328, | |
| "learning_rate": 9.427350427350428e-06, | |
| "loss": 18.2876, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 78.9795150756836, | |
| "learning_rate": 9.423076923076923e-06, | |
| "loss": 15.5738, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1623931623931625, | |
| "grad_norm": 83.77166748046875, | |
| "learning_rate": 9.41880341880342e-06, | |
| "loss": 16.0825, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.170940170940171, | |
| "grad_norm": 129.62966918945312, | |
| "learning_rate": 9.414529914529915e-06, | |
| "loss": 18.4077, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.1794871794871795, | |
| "grad_norm": 110.26199340820312, | |
| "learning_rate": 9.410256410256412e-06, | |
| "loss": 17.6436, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.188034188034188, | |
| "grad_norm": 95.36865997314453, | |
| "learning_rate": 9.405982905982907e-06, | |
| "loss": 19.0424, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1965811965811965, | |
| "grad_norm": 98.36263275146484, | |
| "learning_rate": 9.401709401709402e-06, | |
| "loss": 16.6122, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.205128205128205, | |
| "grad_norm": 83.68401336669922, | |
| "learning_rate": 9.397435897435899e-06, | |
| "loss": 14.9218, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2136752136752136, | |
| "grad_norm": 92.4602279663086, | |
| "learning_rate": 9.393162393162394e-06, | |
| "loss": 16.3563, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 106.41629791259766, | |
| "learning_rate": 9.38888888888889e-06, | |
| "loss": 16.4447, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 97.70237731933594, | |
| "learning_rate": 9.384615384615385e-06, | |
| "loss": 16.8154, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.2393162393162394, | |
| "grad_norm": 76.88361358642578, | |
| "learning_rate": 9.38034188034188e-06, | |
| "loss": 15.7116, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2478632478632479, | |
| "grad_norm": 104.20966339111328, | |
| "learning_rate": 9.376068376068377e-06, | |
| "loss": 15.2283, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2564102564102564, | |
| "grad_norm": 80.29965209960938, | |
| "learning_rate": 9.371794871794872e-06, | |
| "loss": 15.3238, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.264957264957265, | |
| "grad_norm": 72.6979751586914, | |
| "learning_rate": 9.367521367521369e-06, | |
| "loss": 14.2293, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.2735042735042734, | |
| "grad_norm": 80.29464721679688, | |
| "learning_rate": 9.363247863247864e-06, | |
| "loss": 11.9706, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.282051282051282, | |
| "grad_norm": 97.80663299560547, | |
| "learning_rate": 9.358974358974359e-06, | |
| "loss": 14.3517, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2905982905982907, | |
| "grad_norm": 75.88921356201172, | |
| "learning_rate": 9.354700854700856e-06, | |
| "loss": 12.8289, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.2991452991452992, | |
| "grad_norm": 75.71963500976562, | |
| "learning_rate": 9.35042735042735e-06, | |
| "loss": 15.2496, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 84.5454330444336, | |
| "learning_rate": 9.346153846153847e-06, | |
| "loss": 15.7946, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.3162393162393162, | |
| "grad_norm": 92.24919128417969, | |
| "learning_rate": 9.341880341880343e-06, | |
| "loss": 13.2751, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.3247863247863247, | |
| "grad_norm": 76.51255798339844, | |
| "learning_rate": 9.33760683760684e-06, | |
| "loss": 14.1861, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 74.59149169921875, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 12.1881, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.341880341880342, | |
| "grad_norm": 69.84959411621094, | |
| "learning_rate": 9.32905982905983e-06, | |
| "loss": 13.1244, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3504273504273505, | |
| "grad_norm": 82.09815979003906, | |
| "learning_rate": 9.324786324786326e-06, | |
| "loss": 12.7492, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.358974358974359, | |
| "grad_norm": 87.25080108642578, | |
| "learning_rate": 9.320512820512821e-06, | |
| "loss": 15.5268, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.3675213675213675, | |
| "grad_norm": 51.60975646972656, | |
| "learning_rate": 9.316239316239318e-06, | |
| "loss": 10.9868, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.376068376068376, | |
| "grad_norm": 65.10023498535156, | |
| "learning_rate": 9.311965811965813e-06, | |
| "loss": 13.2106, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 86.3865737915039, | |
| "learning_rate": 9.307692307692308e-06, | |
| "loss": 12.4873, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.393162393162393, | |
| "grad_norm": 89.5868911743164, | |
| "learning_rate": 9.303418803418805e-06, | |
| "loss": 12.3125, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4017094017094016, | |
| "grad_norm": 87.308837890625, | |
| "learning_rate": 9.2991452991453e-06, | |
| "loss": 13.1855, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4102564102564101, | |
| "grad_norm": 79.86372375488281, | |
| "learning_rate": 9.294871794871796e-06, | |
| "loss": 11.2756, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4188034188034189, | |
| "grad_norm": 64.42597961425781, | |
| "learning_rate": 9.290598290598292e-06, | |
| "loss": 11.7395, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4273504273504274, | |
| "grad_norm": 64.65245056152344, | |
| "learning_rate": 9.286324786324787e-06, | |
| "loss": 10.2739, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.435897435897436, | |
| "grad_norm": 49.57310104370117, | |
| "learning_rate": 9.282051282051283e-06, | |
| "loss": 11.4798, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 89.93653106689453, | |
| "learning_rate": 9.277777777777778e-06, | |
| "loss": 13.8041, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.452991452991453, | |
| "grad_norm": 59.6973876953125, | |
| "learning_rate": 9.273504273504275e-06, | |
| "loss": 11.0414, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 63.07640838623047, | |
| "learning_rate": 9.26923076923077e-06, | |
| "loss": 10.2649, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.4700854700854702, | |
| "grad_norm": 121.3633041381836, | |
| "learning_rate": 9.264957264957267e-06, | |
| "loss": 11.9233, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.4786324786324787, | |
| "grad_norm": 50.96989822387695, | |
| "learning_rate": 9.260683760683762e-06, | |
| "loss": 8.3527, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.4871794871794872, | |
| "grad_norm": 71.61744689941406, | |
| "learning_rate": 9.256410256410257e-06, | |
| "loss": 11.4237, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.4957264957264957, | |
| "grad_norm": 69.43048858642578, | |
| "learning_rate": 9.252136752136754e-06, | |
| "loss": 9.9193, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.5042735042735043, | |
| "grad_norm": 130.2714385986328, | |
| "learning_rate": 9.247863247863249e-06, | |
| "loss": 12.0676, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5128205128205128, | |
| "grad_norm": 51.40456008911133, | |
| "learning_rate": 9.243589743589745e-06, | |
| "loss": 9.2348, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5213675213675213, | |
| "grad_norm": 48.94670486450195, | |
| "learning_rate": 9.23931623931624e-06, | |
| "loss": 8.8217, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5299145299145298, | |
| "grad_norm": 54.54533386230469, | |
| "learning_rate": 9.235042735042736e-06, | |
| "loss": 9.2478, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 46.581939697265625, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 8.746, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.547008547008547, | |
| "grad_norm": 49.31954574584961, | |
| "learning_rate": 9.226495726495727e-06, | |
| "loss": 8.7889, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 48.5145378112793, | |
| "learning_rate": 9.222222222222224e-06, | |
| "loss": 8.4478, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.564102564102564, | |
| "grad_norm": 49.587825775146484, | |
| "learning_rate": 9.217948717948717e-06, | |
| "loss": 10.5022, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.5726495726495726, | |
| "grad_norm": 47.89423751831055, | |
| "learning_rate": 9.213675213675214e-06, | |
| "loss": 8.7681, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.5811965811965814, | |
| "grad_norm": 59.971920013427734, | |
| "learning_rate": 9.20940170940171e-06, | |
| "loss": 9.6469, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5897435897435899, | |
| "grad_norm": 41.139957427978516, | |
| "learning_rate": 9.205128205128206e-06, | |
| "loss": 8.5196, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.5982905982905984, | |
| "grad_norm": 36.8078498840332, | |
| "learning_rate": 9.200854700854701e-06, | |
| "loss": 8.2513, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.606837606837607, | |
| "grad_norm": 62.23011016845703, | |
| "learning_rate": 9.196581196581196e-06, | |
| "loss": 9.239, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 41.35377502441406, | |
| "learning_rate": 9.192307692307693e-06, | |
| "loss": 8.6788, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.623931623931624, | |
| "grad_norm": 53.734134674072266, | |
| "learning_rate": 9.188034188034188e-06, | |
| "loss": 8.2624, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6324786324786325, | |
| "grad_norm": 60.738887786865234, | |
| "learning_rate": 9.183760683760685e-06, | |
| "loss": 9.2777, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.641025641025641, | |
| "grad_norm": 26.411643981933594, | |
| "learning_rate": 9.17948717948718e-06, | |
| "loss": 7.6894, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.6495726495726495, | |
| "grad_norm": 37.81135940551758, | |
| "learning_rate": 9.175213675213676e-06, | |
| "loss": 8.009, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.658119658119658, | |
| "grad_norm": 42.451080322265625, | |
| "learning_rate": 9.170940170940171e-06, | |
| "loss": 8.309, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 54.87519073486328, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 8.3505, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6752136752136753, | |
| "grad_norm": 47.997737884521484, | |
| "learning_rate": 9.162393162393163e-06, | |
| "loss": 8.9444, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.6837606837606838, | |
| "grad_norm": 33.1911506652832, | |
| "learning_rate": 9.158119658119658e-06, | |
| "loss": 6.8856, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 28.42953872680664, | |
| "learning_rate": 9.153846153846155e-06, | |
| "loss": 7.0575, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7008547008547008, | |
| "grad_norm": 34.74330520629883, | |
| "learning_rate": 9.14957264957265e-06, | |
| "loss": 7.6837, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7094017094017095, | |
| "grad_norm": 27.730812072753906, | |
| "learning_rate": 9.145299145299145e-06, | |
| "loss": 7.2591, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.717948717948718, | |
| "grad_norm": 36.658966064453125, | |
| "learning_rate": 9.141025641025642e-06, | |
| "loss": 7.6744, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.7264957264957266, | |
| "grad_norm": 52.580074310302734, | |
| "learning_rate": 9.136752136752137e-06, | |
| "loss": 8.9746, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.735042735042735, | |
| "grad_norm": 26.30430030822754, | |
| "learning_rate": 9.132478632478634e-06, | |
| "loss": 7.0829, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.7435897435897436, | |
| "grad_norm": 35.77456283569336, | |
| "learning_rate": 9.128205128205129e-06, | |
| "loss": 7.46, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.7521367521367521, | |
| "grad_norm": 46.80126953125, | |
| "learning_rate": 9.123931623931624e-06, | |
| "loss": 8.0331, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.7606837606837606, | |
| "grad_norm": 26.510988235473633, | |
| "learning_rate": 9.11965811965812e-06, | |
| "loss": 7.0434, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 30.846357345581055, | |
| "learning_rate": 9.115384615384615e-06, | |
| "loss": 6.9022, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 45.06099319458008, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 7.108, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.7863247863247862, | |
| "grad_norm": 40.050079345703125, | |
| "learning_rate": 9.106837606837607e-06, | |
| "loss": 7.3628, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.7948717948717947, | |
| "grad_norm": 32.066261291503906, | |
| "learning_rate": 9.102564102564104e-06, | |
| "loss": 7.3292, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8034188034188035, | |
| "grad_norm": 29.196252822875977, | |
| "learning_rate": 9.098290598290599e-06, | |
| "loss": 6.6194, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.811965811965812, | |
| "grad_norm": 34.54549026489258, | |
| "learning_rate": 9.094017094017094e-06, | |
| "loss": 7.224, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.8205128205128205, | |
| "grad_norm": 31.863550186157227, | |
| "learning_rate": 9.08974358974359e-06, | |
| "loss": 7.141, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.8290598290598292, | |
| "grad_norm": 36.79090118408203, | |
| "learning_rate": 9.085470085470086e-06, | |
| "loss": 6.9572, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.8376068376068377, | |
| "grad_norm": 24.298635482788086, | |
| "learning_rate": 9.081196581196583e-06, | |
| "loss": 6.6881, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 16.75456428527832, | |
| "learning_rate": 9.076923076923078e-06, | |
| "loss": 6.4055, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.8547008547008548, | |
| "grad_norm": 20.152400970458984, | |
| "learning_rate": 9.072649572649573e-06, | |
| "loss": 6.9078, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.8632478632478633, | |
| "grad_norm": 34.73337173461914, | |
| "learning_rate": 9.06837606837607e-06, | |
| "loss": 6.7923, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.8717948717948718, | |
| "grad_norm": 28.418310165405273, | |
| "learning_rate": 9.064102564102564e-06, | |
| "loss": 6.9382, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.8803418803418803, | |
| "grad_norm": 13.454174995422363, | |
| "learning_rate": 9.059829059829061e-06, | |
| "loss": 4.5504, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 20.746938705444336, | |
| "learning_rate": 9.055555555555556e-06, | |
| "loss": 6.4711, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.8974358974358974, | |
| "grad_norm": 23.29437828063965, | |
| "learning_rate": 9.051282051282051e-06, | |
| "loss": 6.1381, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.9059829059829059, | |
| "grad_norm": 31.720672607421875, | |
| "learning_rate": 9.047008547008548e-06, | |
| "loss": 6.7716, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.9145299145299144, | |
| "grad_norm": 16.971572875976562, | |
| "learning_rate": 9.042735042735043e-06, | |
| "loss": 6.4734, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 25.185396194458008, | |
| "learning_rate": 9.03846153846154e-06, | |
| "loss": 6.2505, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9316239316239316, | |
| "grad_norm": 42.373863220214844, | |
| "learning_rate": 9.034188034188035e-06, | |
| "loss": 7.1968, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.9401709401709402, | |
| "grad_norm": 21.06004524230957, | |
| "learning_rate": 9.029914529914532e-06, | |
| "loss": 6.082, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.9487179487179487, | |
| "grad_norm": 21.413599014282227, | |
| "learning_rate": 9.025641025641027e-06, | |
| "loss": 6.2279, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.9572649572649574, | |
| "grad_norm": 18.379974365234375, | |
| "learning_rate": 9.021367521367522e-06, | |
| "loss": 6.6032, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.965811965811966, | |
| "grad_norm": 28.239042282104492, | |
| "learning_rate": 9.017094017094018e-06, | |
| "loss": 6.5428, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9743589743589745, | |
| "grad_norm": 17.92879867553711, | |
| "learning_rate": 9.012820512820513e-06, | |
| "loss": 5.986, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.982905982905983, | |
| "grad_norm": 15.501392364501953, | |
| "learning_rate": 9.00854700854701e-06, | |
| "loss": 5.9526, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.9914529914529915, | |
| "grad_norm": 23.742633819580078, | |
| "learning_rate": 9.004273504273505e-06, | |
| "loss": 6.2462, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 28.22560691833496, | |
| "learning_rate": 9e-06, | |
| "loss": 5.8705, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 5.379393577575684, | |
| "eval_runtime": 9.2791, | |
| "eval_samples_per_second": 50.22, | |
| "eval_steps_per_second": 6.358, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.0085470085470085, | |
| "grad_norm": 21.7072696685791, | |
| "learning_rate": 8.995726495726497e-06, | |
| "loss": 6.2757, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.017094017094017, | |
| "grad_norm": 20.955190658569336, | |
| "learning_rate": 8.991452991452992e-06, | |
| "loss": 5.7265, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.0256410256410255, | |
| "grad_norm": 15.186567306518555, | |
| "learning_rate": 8.987179487179489e-06, | |
| "loss": 6.1958, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.034188034188034, | |
| "grad_norm": 20.938766479492188, | |
| "learning_rate": 8.982905982905984e-06, | |
| "loss": 6.2317, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.0427350427350426, | |
| "grad_norm": 18.457494735717773, | |
| "learning_rate": 8.978632478632479e-06, | |
| "loss": 6.4711, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 43.505149841308594, | |
| "learning_rate": 8.974358974358976e-06, | |
| "loss": 5.9632, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0598290598290596, | |
| "grad_norm": 15.558544158935547, | |
| "learning_rate": 8.97008547008547e-06, | |
| "loss": 5.8099, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.0683760683760686, | |
| "grad_norm": 22.20660400390625, | |
| "learning_rate": 8.965811965811967e-06, | |
| "loss": 5.7939, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 15.866617202758789, | |
| "learning_rate": 8.961538461538462e-06, | |
| "loss": 5.9473, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.0854700854700856, | |
| "grad_norm": 20.30729103088379, | |
| "learning_rate": 8.957264957264959e-06, | |
| "loss": 6.2028, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.094017094017094, | |
| "grad_norm": 15.517614364624023, | |
| "learning_rate": 8.952991452991454e-06, | |
| "loss": 5.906, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.1025641025641026, | |
| "grad_norm": 21.30764389038086, | |
| "learning_rate": 8.94871794871795e-06, | |
| "loss": 6.1907, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 19.973115921020508, | |
| "learning_rate": 8.944444444444446e-06, | |
| "loss": 5.6895, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.1196581196581197, | |
| "grad_norm": 17.40595817565918, | |
| "learning_rate": 8.940170940170941e-06, | |
| "loss": 5.4836, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.128205128205128, | |
| "grad_norm": 27.667421340942383, | |
| "learning_rate": 8.935897435897438e-06, | |
| "loss": 5.9082, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.1367521367521367, | |
| "grad_norm": 18.151315689086914, | |
| "learning_rate": 8.931623931623933e-06, | |
| "loss": 5.8102, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.1452991452991452, | |
| "grad_norm": 15.390297889709473, | |
| "learning_rate": 8.927350427350428e-06, | |
| "loss": 5.5504, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 17.257841110229492, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 5.9043, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.1623931623931623, | |
| "grad_norm": 19.2503604888916, | |
| "learning_rate": 8.91880341880342e-06, | |
| "loss": 5.8349, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.1709401709401708, | |
| "grad_norm": 25.236759185791016, | |
| "learning_rate": 8.914529914529916e-06, | |
| "loss": 5.2908, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.1794871794871793, | |
| "grad_norm": 13.771193504333496, | |
| "learning_rate": 8.910256410256411e-06, | |
| "loss": 5.4743, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.1880341880341883, | |
| "grad_norm": 17.406471252441406, | |
| "learning_rate": 8.905982905982906e-06, | |
| "loss": 5.6856, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.1965811965811968, | |
| "grad_norm": 14.727091789245605, | |
| "learning_rate": 8.901709401709401e-06, | |
| "loss": 5.7937, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.2051282051282053, | |
| "grad_norm": 18.193246841430664, | |
| "learning_rate": 8.897435897435898e-06, | |
| "loss": 5.5704, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.213675213675214, | |
| "grad_norm": 21.573726654052734, | |
| "learning_rate": 8.893162393162393e-06, | |
| "loss": 5.479, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 28.72640037536621, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 5.5096, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 15.4992094039917, | |
| "learning_rate": 8.884615384615385e-06, | |
| "loss": 5.217, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.2393162393162394, | |
| "grad_norm": 17.753416061401367, | |
| "learning_rate": 8.88034188034188e-06, | |
| "loss": 5.8173, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.247863247863248, | |
| "grad_norm": 15.91961669921875, | |
| "learning_rate": 8.876068376068377e-06, | |
| "loss": 5.7171, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.2564102564102564, | |
| "grad_norm": 23.30504035949707, | |
| "learning_rate": 8.871794871794872e-06, | |
| "loss": 5.6214, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.264957264957265, | |
| "grad_norm": 15.583686828613281, | |
| "learning_rate": 8.867521367521369e-06, | |
| "loss": 5.2343, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.2735042735042734, | |
| "grad_norm": 24.482046127319336, | |
| "learning_rate": 8.863247863247864e-06, | |
| "loss": 5.0747, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.282051282051282, | |
| "grad_norm": 16.17924690246582, | |
| "learning_rate": 8.858974358974359e-06, | |
| "loss": 5.2645, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.2905982905982905, | |
| "grad_norm": 19.538314819335938, | |
| "learning_rate": 8.854700854700855e-06, | |
| "loss": 5.3484, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.299145299145299, | |
| "grad_norm": 14.472186088562012, | |
| "learning_rate": 8.85042735042735e-06, | |
| "loss": 5.8159, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 16.797805786132812, | |
| "learning_rate": 8.846153846153847e-06, | |
| "loss": 5.4466, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.316239316239316, | |
| "grad_norm": 13.237580299377441, | |
| "learning_rate": 8.841880341880342e-06, | |
| "loss": 5.2189, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.324786324786325, | |
| "grad_norm": 16.685317993164062, | |
| "learning_rate": 8.837606837606837e-06, | |
| "loss": 5.7098, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 16.63880729675293, | |
| "learning_rate": 8.833333333333334e-06, | |
| "loss": 5.0714, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.341880341880342, | |
| "grad_norm": 20.871978759765625, | |
| "learning_rate": 8.829059829059829e-06, | |
| "loss": 4.9509, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.3504273504273505, | |
| "grad_norm": 16.95268440246582, | |
| "learning_rate": 8.824786324786326e-06, | |
| "loss": 5.4166, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.358974358974359, | |
| "grad_norm": 15.446279525756836, | |
| "learning_rate": 8.820512820512821e-06, | |
| "loss": 4.5967, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.3675213675213675, | |
| "grad_norm": 17.148235321044922, | |
| "learning_rate": 8.816239316239316e-06, | |
| "loss": 5.2542, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.376068376068376, | |
| "grad_norm": 17.014827728271484, | |
| "learning_rate": 8.811965811965813e-06, | |
| "loss": 5.4702, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 15.313383102416992, | |
| "learning_rate": 8.807692307692308e-06, | |
| "loss": 5.2119, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.393162393162393, | |
| "grad_norm": 20.2298641204834, | |
| "learning_rate": 8.803418803418804e-06, | |
| "loss": 5.4064, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.4017094017094016, | |
| "grad_norm": 14.982254981994629, | |
| "learning_rate": 8.7991452991453e-06, | |
| "loss": 5.2545, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.41025641025641, | |
| "grad_norm": 16.258047103881836, | |
| "learning_rate": 8.794871794871796e-06, | |
| "loss": 5.0141, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.4188034188034186, | |
| "grad_norm": 22.5199031829834, | |
| "learning_rate": 8.790598290598291e-06, | |
| "loss": 5.3486, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.427350427350427, | |
| "grad_norm": 17.546480178833008, | |
| "learning_rate": 8.786324786324786e-06, | |
| "loss": 5.2785, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.435897435897436, | |
| "grad_norm": 22.07866668701172, | |
| "learning_rate": 8.782051282051283e-06, | |
| "loss": 5.4471, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 409.2532043457031, | |
| "learning_rate": 8.777777777777778e-06, | |
| "loss": 6.0948, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.452991452991453, | |
| "grad_norm": 185.7334747314453, | |
| "learning_rate": 8.773504273504275e-06, | |
| "loss": 5.5538, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 30.8182430267334, | |
| "learning_rate": 8.76923076923077e-06, | |
| "loss": 4.9661, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.47008547008547, | |
| "grad_norm": 18.584409713745117, | |
| "learning_rate": 8.764957264957265e-06, | |
| "loss": 5.0947, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.4786324786324787, | |
| "grad_norm": 18.128522872924805, | |
| "learning_rate": 8.760683760683762e-06, | |
| "loss": 4.8816, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4871794871794872, | |
| "grad_norm": 18.800090789794922, | |
| "learning_rate": 8.756410256410257e-06, | |
| "loss": 5.0952, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.4957264957264957, | |
| "grad_norm": 22.140430450439453, | |
| "learning_rate": 8.752136752136753e-06, | |
| "loss": 4.5408, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.5042735042735043, | |
| "grad_norm": 19.867111206054688, | |
| "learning_rate": 8.747863247863248e-06, | |
| "loss": 4.7435, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.5128205128205128, | |
| "grad_norm": 19.437868118286133, | |
| "learning_rate": 8.743589743589743e-06, | |
| "loss": 5.2643, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.5213675213675213, | |
| "grad_norm": 18.256561279296875, | |
| "learning_rate": 8.73931623931624e-06, | |
| "loss": 5.2531, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.52991452991453, | |
| "grad_norm": 18.65209197998047, | |
| "learning_rate": 8.735042735042735e-06, | |
| "loss": 4.8646, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 14.704927444458008, | |
| "learning_rate": 8.730769230769232e-06, | |
| "loss": 4.8343, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.547008547008547, | |
| "grad_norm": 15.522851943969727, | |
| "learning_rate": 8.726495726495727e-06, | |
| "loss": 4.898, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 21.7825927734375, | |
| "learning_rate": 8.722222222222224e-06, | |
| "loss": 5.0732, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.564102564102564, | |
| "grad_norm": 17.963552474975586, | |
| "learning_rate": 8.717948717948719e-06, | |
| "loss": 4.9684, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5726495726495724, | |
| "grad_norm": 16.14459991455078, | |
| "learning_rate": 8.713675213675214e-06, | |
| "loss": 4.8802, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.5811965811965814, | |
| "grad_norm": 18.386646270751953, | |
| "learning_rate": 8.70940170940171e-06, | |
| "loss": 4.8837, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.58974358974359, | |
| "grad_norm": 19.471376419067383, | |
| "learning_rate": 8.705128205128206e-06, | |
| "loss": 4.6325, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.5982905982905984, | |
| "grad_norm": 17.839717864990234, | |
| "learning_rate": 8.700854700854702e-06, | |
| "loss": 4.7851, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.606837606837607, | |
| "grad_norm": 26.519363403320312, | |
| "learning_rate": 8.696581196581197e-06, | |
| "loss": 5.0576, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 14.135244369506836, | |
| "learning_rate": 8.692307692307692e-06, | |
| "loss": 4.7719, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.623931623931624, | |
| "grad_norm": 16.5241641998291, | |
| "learning_rate": 8.68803418803419e-06, | |
| "loss": 4.5826, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.6324786324786325, | |
| "grad_norm": 23.982437133789062, | |
| "learning_rate": 8.683760683760684e-06, | |
| "loss": 4.4878, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.641025641025641, | |
| "grad_norm": 16.036361694335938, | |
| "learning_rate": 8.679487179487181e-06, | |
| "loss": 4.3867, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.6495726495726495, | |
| "grad_norm": 16.19298553466797, | |
| "learning_rate": 8.675213675213676e-06, | |
| "loss": 4.763, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.658119658119658, | |
| "grad_norm": 19.32802963256836, | |
| "learning_rate": 8.670940170940171e-06, | |
| "loss": 4.4083, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 21.75898551940918, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 4.8782, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.6752136752136755, | |
| "grad_norm": 905.6954956054688, | |
| "learning_rate": 8.662393162393163e-06, | |
| "loss": 5.7901, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.683760683760684, | |
| "grad_norm": 21.126985549926758, | |
| "learning_rate": 8.65811965811966e-06, | |
| "loss": 4.918, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 22.190237045288086, | |
| "learning_rate": 8.653846153846155e-06, | |
| "loss": 4.4327, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.700854700854701, | |
| "grad_norm": 90.69184875488281, | |
| "learning_rate": 8.649572649572651e-06, | |
| "loss": 5.1477, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.7094017094017095, | |
| "grad_norm": 43.43864059448242, | |
| "learning_rate": 8.645299145299146e-06, | |
| "loss": 4.5476, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.717948717948718, | |
| "grad_norm": 19.24538230895996, | |
| "learning_rate": 8.641025641025641e-06, | |
| "loss": 4.4304, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.7264957264957266, | |
| "grad_norm": 21.809600830078125, | |
| "learning_rate": 8.636752136752138e-06, | |
| "loss": 4.4215, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.735042735042735, | |
| "grad_norm": 21.406156539916992, | |
| "learning_rate": 8.632478632478633e-06, | |
| "loss": 4.5411, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.7435897435897436, | |
| "grad_norm": 17.57236099243164, | |
| "learning_rate": 8.62820512820513e-06, | |
| "loss": 4.7952, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.752136752136752, | |
| "grad_norm": 21.049169540405273, | |
| "learning_rate": 8.623931623931625e-06, | |
| "loss": 4.4596, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.7606837606837606, | |
| "grad_norm": 20.04981803894043, | |
| "learning_rate": 8.61965811965812e-06, | |
| "loss": 4.4705, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 21.146499633789062, | |
| "learning_rate": 8.615384615384617e-06, | |
| "loss": 4.6081, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 20.9805908203125, | |
| "learning_rate": 8.611111111111112e-06, | |
| "loss": 4.8387, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.786324786324786, | |
| "grad_norm": 17.708343505859375, | |
| "learning_rate": 8.606837606837609e-06, | |
| "loss": 4.3455, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.7948717948717947, | |
| "grad_norm": 25.657032012939453, | |
| "learning_rate": 8.602564102564104e-06, | |
| "loss": 4.3119, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.8034188034188032, | |
| "grad_norm": 17.713972091674805, | |
| "learning_rate": 8.598290598290599e-06, | |
| "loss": 4.5597, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.8119658119658117, | |
| "grad_norm": 22.297082901000977, | |
| "learning_rate": 8.594017094017095e-06, | |
| "loss": 3.8398, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.8205128205128203, | |
| "grad_norm": 16.11454200744629, | |
| "learning_rate": 8.58974358974359e-06, | |
| "loss": 3.2049, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.8290598290598292, | |
| "grad_norm": 27.323585510253906, | |
| "learning_rate": 8.585470085470086e-06, | |
| "loss": 4.0371, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.8376068376068377, | |
| "grad_norm": 21.090797424316406, | |
| "learning_rate": 8.58119658119658e-06, | |
| "loss": 4.5193, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 39.087432861328125, | |
| "learning_rate": 8.576923076923077e-06, | |
| "loss": 4.3537, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.8547008547008548, | |
| "grad_norm": 18.49846839904785, | |
| "learning_rate": 8.572649572649572e-06, | |
| "loss": 4.614, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.8632478632478633, | |
| "grad_norm": 26.671632766723633, | |
| "learning_rate": 8.568376068376069e-06, | |
| "loss": 4.2224, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.871794871794872, | |
| "grad_norm": 25.799545288085938, | |
| "learning_rate": 8.564102564102564e-06, | |
| "loss": 4.2209, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.8803418803418803, | |
| "grad_norm": 20.131961822509766, | |
| "learning_rate": 8.559829059829061e-06, | |
| "loss": 4.5194, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 20.193859100341797, | |
| "learning_rate": 8.555555555555556e-06, | |
| "loss": 3.9966, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.8974358974358974, | |
| "grad_norm": 20.06737518310547, | |
| "learning_rate": 8.551282051282051e-06, | |
| "loss": 3.7394, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.905982905982906, | |
| "grad_norm": 438.34429931640625, | |
| "learning_rate": 8.547008547008548e-06, | |
| "loss": 5.1558, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.9145299145299144, | |
| "grad_norm": 22.152528762817383, | |
| "learning_rate": 8.542735042735043e-06, | |
| "loss": 3.9014, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 29.279739379882812, | |
| "learning_rate": 8.53846153846154e-06, | |
| "loss": 4.0479, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.931623931623932, | |
| "grad_norm": 26.182645797729492, | |
| "learning_rate": 8.534188034188035e-06, | |
| "loss": 4.2022, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.9401709401709404, | |
| "grad_norm": 22.329736709594727, | |
| "learning_rate": 8.52991452991453e-06, | |
| "loss": 3.8777, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.948717948717949, | |
| "grad_norm": 20.62833023071289, | |
| "learning_rate": 8.525641025641026e-06, | |
| "loss": 4.2189, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.9572649572649574, | |
| "grad_norm": 20.176612854003906, | |
| "learning_rate": 8.521367521367521e-06, | |
| "loss": 4.0124, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.965811965811966, | |
| "grad_norm": 18.77017593383789, | |
| "learning_rate": 8.517094017094018e-06, | |
| "loss": 3.3286, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.9743589743589745, | |
| "grad_norm": 226.93701171875, | |
| "learning_rate": 8.512820512820513e-06, | |
| "loss": 4.6969, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.982905982905983, | |
| "grad_norm": 675.1133422851562, | |
| "learning_rate": 8.508547008547008e-06, | |
| "loss": 4.6717, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.9914529914529915, | |
| "grad_norm": 19.938486099243164, | |
| "learning_rate": 8.504273504273505e-06, | |
| "loss": 4.0103, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 15.917003631591797, | |
| "learning_rate": 8.5e-06, | |
| "loss": 3.1643, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 3.4197537899017334, | |
| "eval_runtime": 9.289, | |
| "eval_samples_per_second": 50.167, | |
| "eval_steps_per_second": 6.352, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.0085470085470085, | |
| "grad_norm": 22.22833251953125, | |
| "learning_rate": 8.495726495726497e-06, | |
| "loss": 4.3458, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.017094017094017, | |
| "grad_norm": 16.4627685546875, | |
| "learning_rate": 8.491452991452992e-06, | |
| "loss": 3.5374, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.0256410256410255, | |
| "grad_norm": 16.389379501342773, | |
| "learning_rate": 8.487179487179488e-06, | |
| "loss": 4.1384, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.034188034188034, | |
| "grad_norm": 19.589706420898438, | |
| "learning_rate": 8.482905982905983e-06, | |
| "loss": 3.9522, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.0427350427350426, | |
| "grad_norm": 21.66250228881836, | |
| "learning_rate": 8.478632478632479e-06, | |
| "loss": 4.0197, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.051282051282051, | |
| "grad_norm": 42.1422119140625, | |
| "learning_rate": 8.474358974358975e-06, | |
| "loss": 3.9432, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.0598290598290596, | |
| "grad_norm": 23.0153751373291, | |
| "learning_rate": 8.47008547008547e-06, | |
| "loss": 3.9146, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.0683760683760686, | |
| "grad_norm": 20.847400665283203, | |
| "learning_rate": 8.465811965811967e-06, | |
| "loss": 3.9736, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 23.553855895996094, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 3.646, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.0854700854700856, | |
| "grad_norm": 18.651151657104492, | |
| "learning_rate": 8.457264957264957e-06, | |
| "loss": 3.761, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.094017094017094, | |
| "grad_norm": 23.437379837036133, | |
| "learning_rate": 8.452991452991454e-06, | |
| "loss": 3.9258, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.1025641025641026, | |
| "grad_norm": 19.025928497314453, | |
| "learning_rate": 8.448717948717949e-06, | |
| "loss": 3.4911, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 25.955963134765625, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 3.7231, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.1196581196581197, | |
| "grad_norm": 19.691673278808594, | |
| "learning_rate": 8.44017094017094e-06, | |
| "loss": 3.9225, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.128205128205128, | |
| "grad_norm": 19.47168731689453, | |
| "learning_rate": 8.435897435897436e-06, | |
| "loss": 3.6261, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.1367521367521367, | |
| "grad_norm": 20.50010108947754, | |
| "learning_rate": 8.431623931623932e-06, | |
| "loss": 3.3306, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.1452991452991452, | |
| "grad_norm": 21.198938369750977, | |
| "learning_rate": 8.427350427350428e-06, | |
| "loss": 3.6388, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 16.93203353881836, | |
| "learning_rate": 8.423076923076924e-06, | |
| "loss": 3.9556, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.1623931623931623, | |
| "grad_norm": 15.074128150939941, | |
| "learning_rate": 8.41880341880342e-06, | |
| "loss": 2.9899, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.1709401709401708, | |
| "grad_norm": 23.041452407836914, | |
| "learning_rate": 8.414529914529916e-06, | |
| "loss": 3.291, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.1794871794871793, | |
| "grad_norm": 24.146419525146484, | |
| "learning_rate": 8.410256410256411e-06, | |
| "loss": 4.0683, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.1880341880341883, | |
| "grad_norm": 27.864879608154297, | |
| "learning_rate": 8.405982905982906e-06, | |
| "loss": 3.6171, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.1965811965811968, | |
| "grad_norm": 33.83136749267578, | |
| "learning_rate": 8.401709401709403e-06, | |
| "loss": 3.7324, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.2051282051282053, | |
| "grad_norm": 21.020702362060547, | |
| "learning_rate": 8.397435897435898e-06, | |
| "loss": 3.5688, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.213675213675214, | |
| "grad_norm": 23.521453857421875, | |
| "learning_rate": 8.393162393162395e-06, | |
| "loss": 3.6917, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.2222222222222223, | |
| "grad_norm": 35.85578536987305, | |
| "learning_rate": 8.38888888888889e-06, | |
| "loss": 3.6532, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 26.080968856811523, | |
| "learning_rate": 8.384615384615385e-06, | |
| "loss": 3.8828, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.2393162393162394, | |
| "grad_norm": 20.829381942749023, | |
| "learning_rate": 8.380341880341881e-06, | |
| "loss": 3.8374, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.247863247863248, | |
| "grad_norm": 20.85077476501465, | |
| "learning_rate": 8.376068376068377e-06, | |
| "loss": 3.2896, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.2564102564102564, | |
| "grad_norm": 19.036088943481445, | |
| "learning_rate": 8.371794871794873e-06, | |
| "loss": 3.4996, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.264957264957265, | |
| "grad_norm": 23.725513458251953, | |
| "learning_rate": 8.367521367521368e-06, | |
| "loss": 3.7686, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.2735042735042734, | |
| "grad_norm": 22.553386688232422, | |
| "learning_rate": 8.363247863247865e-06, | |
| "loss": 3.8476, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.282051282051282, | |
| "grad_norm": 20.263992309570312, | |
| "learning_rate": 8.35897435897436e-06, | |
| "loss": 3.3278, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.2905982905982905, | |
| "grad_norm": 22.47858238220215, | |
| "learning_rate": 8.354700854700855e-06, | |
| "loss": 3.5437, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.299145299145299, | |
| "grad_norm": 24.14532470703125, | |
| "learning_rate": 8.350427350427352e-06, | |
| "loss": 3.696, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 31.457847595214844, | |
| "learning_rate": 8.346153846153847e-06, | |
| "loss": 4.3065, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.316239316239316, | |
| "grad_norm": 24.503095626831055, | |
| "learning_rate": 8.341880341880344e-06, | |
| "loss": 3.4798, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.324786324786325, | |
| "grad_norm": 19.798818588256836, | |
| "learning_rate": 8.337606837606839e-06, | |
| "loss": 3.5323, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 22.023189544677734, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 3.4088, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.341880341880342, | |
| "grad_norm": 17.314960479736328, | |
| "learning_rate": 8.32905982905983e-06, | |
| "loss": 3.2462, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.3504273504273505, | |
| "grad_norm": 22.714536666870117, | |
| "learning_rate": 8.324786324786326e-06, | |
| "loss": 3.7863, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.358974358974359, | |
| "grad_norm": 27.710514068603516, | |
| "learning_rate": 8.320512820512822e-06, | |
| "loss": 3.6032, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.3675213675213675, | |
| "grad_norm": 23.35419464111328, | |
| "learning_rate": 8.316239316239317e-06, | |
| "loss": 3.5599, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.376068376068376, | |
| "grad_norm": 24.0956974029541, | |
| "learning_rate": 8.311965811965812e-06, | |
| "loss": 3.5186, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 22.09107780456543, | |
| "learning_rate": 8.307692307692309e-06, | |
| "loss": 3.4843, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.393162393162393, | |
| "grad_norm": 23.956623077392578, | |
| "learning_rate": 8.303418803418804e-06, | |
| "loss": 3.1625, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.4017094017094016, | |
| "grad_norm": 18.875917434692383, | |
| "learning_rate": 8.299145299145301e-06, | |
| "loss": 3.3494, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.41025641025641, | |
| "grad_norm": 33.475467681884766, | |
| "learning_rate": 8.294871794871796e-06, | |
| "loss": 3.9247, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.4188034188034186, | |
| "grad_norm": 16.28295135498047, | |
| "learning_rate": 8.290598290598293e-06, | |
| "loss": 3.7446, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.427350427350427, | |
| "grad_norm": 24.205049514770508, | |
| "learning_rate": 8.286324786324788e-06, | |
| "loss": 3.343, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.435897435897436, | |
| "grad_norm": 21.21460723876953, | |
| "learning_rate": 8.282051282051283e-06, | |
| "loss": 3.2437, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.4444444444444446, | |
| "grad_norm": 36.8713264465332, | |
| "learning_rate": 8.277777777777778e-06, | |
| "loss": 3.5009, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.452991452991453, | |
| "grad_norm": 26.85513687133789, | |
| "learning_rate": 8.273504273504273e-06, | |
| "loss": 3.7271, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 18.184600830078125, | |
| "learning_rate": 8.26923076923077e-06, | |
| "loss": 3.2216, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.47008547008547, | |
| "grad_norm": 27.03692054748535, | |
| "learning_rate": 8.264957264957265e-06, | |
| "loss": 3.516, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.4786324786324787, | |
| "grad_norm": 20.63736915588379, | |
| "learning_rate": 8.260683760683761e-06, | |
| "loss": 3.1349, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.4871794871794872, | |
| "grad_norm": 22.467845916748047, | |
| "learning_rate": 8.256410256410256e-06, | |
| "loss": 3.3878, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.4957264957264957, | |
| "grad_norm": 21.25887107849121, | |
| "learning_rate": 8.252136752136753e-06, | |
| "loss": 3.8298, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.5042735042735043, | |
| "grad_norm": 47.3256721496582, | |
| "learning_rate": 8.247863247863248e-06, | |
| "loss": 3.5321, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.5128205128205128, | |
| "grad_norm": 22.103790283203125, | |
| "learning_rate": 8.243589743589743e-06, | |
| "loss": 3.335, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.5213675213675213, | |
| "grad_norm": 25.779077529907227, | |
| "learning_rate": 8.23931623931624e-06, | |
| "loss": 3.5047, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.52991452991453, | |
| "grad_norm": 22.78207778930664, | |
| "learning_rate": 8.235042735042735e-06, | |
| "loss": 3.3827, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 22.41836166381836, | |
| "learning_rate": 8.230769230769232e-06, | |
| "loss": 3.4521, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.547008547008547, | |
| "grad_norm": 60.29216384887695, | |
| "learning_rate": 8.226495726495727e-06, | |
| "loss": 3.4598, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 25.27474021911621, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 3.7443, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.564102564102564, | |
| "grad_norm": 25.297466278076172, | |
| "learning_rate": 8.217948717948719e-06, | |
| "loss": 3.3123, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.5726495726495724, | |
| "grad_norm": 28.5858154296875, | |
| "learning_rate": 8.213675213675214e-06, | |
| "loss": 3.1801, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.5811965811965814, | |
| "grad_norm": 20.05567741394043, | |
| "learning_rate": 8.20940170940171e-06, | |
| "loss": 3.7242, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.58974358974359, | |
| "grad_norm": 32.33693313598633, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 3.3587, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.5982905982905984, | |
| "grad_norm": 36.1716194152832, | |
| "learning_rate": 8.200854700854702e-06, | |
| "loss": 3.1573, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.606837606837607, | |
| "grad_norm": 33.39027404785156, | |
| "learning_rate": 8.196581196581197e-06, | |
| "loss": 3.098, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 28.4794864654541, | |
| "learning_rate": 8.192307692307692e-06, | |
| "loss": 3.6403, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.623931623931624, | |
| "grad_norm": 29.702611923217773, | |
| "learning_rate": 8.188034188034189e-06, | |
| "loss": 3.2569, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 3.6324786324786325, | |
| "grad_norm": 24.73663902282715, | |
| "learning_rate": 8.183760683760684e-06, | |
| "loss": 3.0508, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.641025641025641, | |
| "grad_norm": 29.606807708740234, | |
| "learning_rate": 8.17948717948718e-06, | |
| "loss": 3.2524, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 3.6495726495726495, | |
| "grad_norm": 22.721933364868164, | |
| "learning_rate": 8.175213675213676e-06, | |
| "loss": 3.2583, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 3.658119658119658, | |
| "grad_norm": 25.009403228759766, | |
| "learning_rate": 8.17094017094017e-06, | |
| "loss": 3.0678, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 25.776636123657227, | |
| "learning_rate": 8.166666666666668e-06, | |
| "loss": 3.1676, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 3.6752136752136755, | |
| "grad_norm": 28.210241317749023, | |
| "learning_rate": 8.162393162393163e-06, | |
| "loss": 3.2869, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.683760683760684, | |
| "grad_norm": 26.29328155517578, | |
| "learning_rate": 8.15811965811966e-06, | |
| "loss": 3.3618, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 19.813465118408203, | |
| "learning_rate": 8.153846153846154e-06, | |
| "loss": 3.0655, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.700854700854701, | |
| "grad_norm": 29.718812942504883, | |
| "learning_rate": 8.14957264957265e-06, | |
| "loss": 3.1538, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 3.7094017094017095, | |
| "grad_norm": 30.629135131835938, | |
| "learning_rate": 8.145299145299146e-06, | |
| "loss": 3.3252, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 3.717948717948718, | |
| "grad_norm": 27.716825485229492, | |
| "learning_rate": 8.141025641025641e-06, | |
| "loss": 3.4083, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.7264957264957266, | |
| "grad_norm": 39.23820877075195, | |
| "learning_rate": 8.136752136752138e-06, | |
| "loss": 3.3074, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 3.735042735042735, | |
| "grad_norm": 34.516422271728516, | |
| "learning_rate": 8.132478632478633e-06, | |
| "loss": 3.3529, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 3.7435897435897436, | |
| "grad_norm": 41.98606872558594, | |
| "learning_rate": 8.12820512820513e-06, | |
| "loss": 3.248, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.752136752136752, | |
| "grad_norm": 27.99711799621582, | |
| "learning_rate": 8.123931623931625e-06, | |
| "loss": 3.3054, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 3.7606837606837606, | |
| "grad_norm": 25.21969985961914, | |
| "learning_rate": 8.11965811965812e-06, | |
| "loss": 2.8518, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 29.14298439025879, | |
| "learning_rate": 8.115384615384617e-06, | |
| "loss": 3.0063, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 3.7777777777777777, | |
| "grad_norm": 27.040063858032227, | |
| "learning_rate": 8.111111111111112e-06, | |
| "loss": 3.3066, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 3.786324786324786, | |
| "grad_norm": 365.3290100097656, | |
| "learning_rate": 8.106837606837608e-06, | |
| "loss": 3.8057, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 3.7948717948717947, | |
| "grad_norm": 32.89745330810547, | |
| "learning_rate": 8.102564102564103e-06, | |
| "loss": 3.0903, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 3.8034188034188032, | |
| "grad_norm": 29.448022842407227, | |
| "learning_rate": 8.098290598290598e-06, | |
| "loss": 3.2723, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.8119658119658117, | |
| "grad_norm": 27.838123321533203, | |
| "learning_rate": 8.094017094017095e-06, | |
| "loss": 3.2903, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 3.8205128205128203, | |
| "grad_norm": 29.047847747802734, | |
| "learning_rate": 8.08974358974359e-06, | |
| "loss": 2.9048, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 3.8290598290598292, | |
| "grad_norm": 28.666589736938477, | |
| "learning_rate": 8.085470085470087e-06, | |
| "loss": 3.2186, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 3.8376068376068377, | |
| "grad_norm": 31.796804428100586, | |
| "learning_rate": 8.081196581196582e-06, | |
| "loss": 3.2668, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 22.665220260620117, | |
| "learning_rate": 8.076923076923077e-06, | |
| "loss": 3.0965, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.8547008547008548, | |
| "grad_norm": 32.7353630065918, | |
| "learning_rate": 8.072649572649574e-06, | |
| "loss": 3.1759, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 3.8632478632478633, | |
| "grad_norm": 32.95683670043945, | |
| "learning_rate": 8.068376068376069e-06, | |
| "loss": 2.9589, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 3.871794871794872, | |
| "grad_norm": 30.04659652709961, | |
| "learning_rate": 8.064102564102566e-06, | |
| "loss": 3.4709, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 3.8803418803418803, | |
| "grad_norm": 30.41158676147461, | |
| "learning_rate": 8.05982905982906e-06, | |
| "loss": 2.9385, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 3.888888888888889, | |
| "grad_norm": 30.059635162353516, | |
| "learning_rate": 8.055555555555557e-06, | |
| "loss": 3.0099, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 3.8974358974358974, | |
| "grad_norm": 24.83198356628418, | |
| "learning_rate": 8.051282051282052e-06, | |
| "loss": 2.9783, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 3.905982905982906, | |
| "grad_norm": 25.38758087158203, | |
| "learning_rate": 8.047008547008547e-06, | |
| "loss": 3.0275, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 3.9145299145299144, | |
| "grad_norm": 25.21868133544922, | |
| "learning_rate": 8.042735042735044e-06, | |
| "loss": 2.9096, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 32.02922058105469, | |
| "learning_rate": 8.03846153846154e-06, | |
| "loss": 3.059, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 3.931623931623932, | |
| "grad_norm": 22.240680694580078, | |
| "learning_rate": 8.034188034188036e-06, | |
| "loss": 2.9473, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.9401709401709404, | |
| "grad_norm": 27.61838150024414, | |
| "learning_rate": 8.029914529914531e-06, | |
| "loss": 2.4506, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 3.948717948717949, | |
| "grad_norm": 27.742216110229492, | |
| "learning_rate": 8.025641025641026e-06, | |
| "loss": 2.9082, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 3.9572649572649574, | |
| "grad_norm": 29.965059280395508, | |
| "learning_rate": 8.021367521367523e-06, | |
| "loss": 2.8268, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 3.965811965811966, | |
| "grad_norm": 31.429990768432617, | |
| "learning_rate": 8.017094017094018e-06, | |
| "loss": 3.1805, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 3.9743589743589745, | |
| "grad_norm": 31.162532806396484, | |
| "learning_rate": 8.012820512820515e-06, | |
| "loss": 2.64, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.982905982905983, | |
| "grad_norm": 28.240577697753906, | |
| "learning_rate": 8.00854700854701e-06, | |
| "loss": 3.249, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 3.9914529914529915, | |
| "grad_norm": 48.52914810180664, | |
| "learning_rate": 8.004273504273505e-06, | |
| "loss": 3.1619, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 36.80685806274414, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 3.5337, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.1340389251708984, | |
| "eval_runtime": 9.2211, | |
| "eval_samples_per_second": 50.536, | |
| "eval_steps_per_second": 6.398, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.0085470085470085, | |
| "grad_norm": 45.45211410522461, | |
| "learning_rate": 7.995726495726496e-06, | |
| "loss": 3.5596, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.017094017094017, | |
| "grad_norm": 32.711669921875, | |
| "learning_rate": 7.991452991452993e-06, | |
| "loss": 2.9362, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.0256410256410255, | |
| "grad_norm": 26.151872634887695, | |
| "learning_rate": 7.987179487179488e-06, | |
| "loss": 2.6796, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.034188034188034, | |
| "grad_norm": 33.02329635620117, | |
| "learning_rate": 7.982905982905985e-06, | |
| "loss": 2.7147, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.042735042735043, | |
| "grad_norm": 31.1684513092041, | |
| "learning_rate": 7.97863247863248e-06, | |
| "loss": 3.2356, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.051282051282051, | |
| "grad_norm": 37.0435905456543, | |
| "learning_rate": 7.974358974358975e-06, | |
| "loss": 2.9954, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.05982905982906, | |
| "grad_norm": 25.989973068237305, | |
| "learning_rate": 7.970085470085472e-06, | |
| "loss": 3.2143, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.068376068376068, | |
| "grad_norm": 27.048690795898438, | |
| "learning_rate": 7.965811965811967e-06, | |
| "loss": 2.5087, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 26.857696533203125, | |
| "learning_rate": 7.961538461538462e-06, | |
| "loss": 2.6466, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.085470085470085, | |
| "grad_norm": 33.342193603515625, | |
| "learning_rate": 7.957264957264957e-06, | |
| "loss": 2.6591, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.094017094017094, | |
| "grad_norm": 64.21253967285156, | |
| "learning_rate": 7.952991452991454e-06, | |
| "loss": 3.0295, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.102564102564102, | |
| "grad_norm": 31.240161895751953, | |
| "learning_rate": 7.948717948717949e-06, | |
| "loss": 2.9374, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.111111111111111, | |
| "grad_norm": 29.338851928710938, | |
| "learning_rate": 7.944444444444445e-06, | |
| "loss": 2.5019, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.119658119658119, | |
| "grad_norm": 36.79518127441406, | |
| "learning_rate": 7.94017094017094e-06, | |
| "loss": 2.7649, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.128205128205128, | |
| "grad_norm": 37.036739349365234, | |
| "learning_rate": 7.935897435897435e-06, | |
| "loss": 2.5182, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 4.136752136752137, | |
| "grad_norm": 42.571163177490234, | |
| "learning_rate": 7.931623931623932e-06, | |
| "loss": 2.767, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.145299145299146, | |
| "grad_norm": 33.72893524169922, | |
| "learning_rate": 7.927350427350427e-06, | |
| "loss": 3.1404, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 27.06032943725586, | |
| "learning_rate": 7.923076923076924e-06, | |
| "loss": 2.6825, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.162393162393163, | |
| "grad_norm": 31.8147029876709, | |
| "learning_rate": 7.918803418803419e-06, | |
| "loss": 2.5129, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 4.170940170940171, | |
| "grad_norm": 35.681793212890625, | |
| "learning_rate": 7.914529914529914e-06, | |
| "loss": 2.4793, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.17948717948718, | |
| "grad_norm": 159.4467315673828, | |
| "learning_rate": 7.91025641025641e-06, | |
| "loss": 3.5531, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 4.188034188034188, | |
| "grad_norm": 40.12252426147461, | |
| "learning_rate": 7.905982905982906e-06, | |
| "loss": 2.7095, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.196581196581197, | |
| "grad_norm": 27.05786895751953, | |
| "learning_rate": 7.901709401709403e-06, | |
| "loss": 2.5984, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 4.205128205128205, | |
| "grad_norm": 24.31035614013672, | |
| "learning_rate": 7.897435897435898e-06, | |
| "loss": 2.89, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.213675213675214, | |
| "grad_norm": 277.16156005859375, | |
| "learning_rate": 7.893162393162394e-06, | |
| "loss": 3.8076, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 4.222222222222222, | |
| "grad_norm": 29.722867965698242, | |
| "learning_rate": 7.88888888888889e-06, | |
| "loss": 2.4189, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 40.47605514526367, | |
| "learning_rate": 7.884615384615384e-06, | |
| "loss": 2.6225, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.239316239316239, | |
| "grad_norm": 29.136499404907227, | |
| "learning_rate": 7.880341880341881e-06, | |
| "loss": 2.5223, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.247863247863248, | |
| "grad_norm": 78.86258697509766, | |
| "learning_rate": 7.876068376068376e-06, | |
| "loss": 2.6587, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 4.256410256410256, | |
| "grad_norm": 24.473243713378906, | |
| "learning_rate": 7.871794871794873e-06, | |
| "loss": 2.456, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.264957264957265, | |
| "grad_norm": 80.45248413085938, | |
| "learning_rate": 7.867521367521368e-06, | |
| "loss": 3.1893, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 4.273504273504273, | |
| "grad_norm": 194.2708282470703, | |
| "learning_rate": 7.863247863247863e-06, | |
| "loss": 3.8294, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.282051282051282, | |
| "grad_norm": 27.74302101135254, | |
| "learning_rate": 7.85897435897436e-06, | |
| "loss": 2.2506, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 4.2905982905982905, | |
| "grad_norm": 21.90385627746582, | |
| "learning_rate": 7.854700854700855e-06, | |
| "loss": 3.0985, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 4.299145299145299, | |
| "grad_norm": 50.30342102050781, | |
| "learning_rate": 7.850427350427352e-06, | |
| "loss": 2.526, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 28.666881561279297, | |
| "learning_rate": 7.846153846153847e-06, | |
| "loss": 2.4213, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 4.316239316239316, | |
| "grad_norm": 27.927257537841797, | |
| "learning_rate": 7.841880341880342e-06, | |
| "loss": 2.6731, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.3247863247863245, | |
| "grad_norm": 36.12032699584961, | |
| "learning_rate": 7.837606837606838e-06, | |
| "loss": 2.3323, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 4.333333333333333, | |
| "grad_norm": 31.632287979125977, | |
| "learning_rate": 7.833333333333333e-06, | |
| "loss": 2.2966, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 4.3418803418803416, | |
| "grad_norm": 26.511537551879883, | |
| "learning_rate": 7.82905982905983e-06, | |
| "loss": 2.3422, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 4.35042735042735, | |
| "grad_norm": 31.429107666015625, | |
| "learning_rate": 7.824786324786325e-06, | |
| "loss": 2.6764, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 4.358974358974359, | |
| "grad_norm": 29.8817138671875, | |
| "learning_rate": 7.820512820512822e-06, | |
| "loss": 2.4358, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.367521367521368, | |
| "grad_norm": 29.293964385986328, | |
| "learning_rate": 7.816239316239317e-06, | |
| "loss": 2.504, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 4.3760683760683765, | |
| "grad_norm": 23.624290466308594, | |
| "learning_rate": 7.811965811965812e-06, | |
| "loss": 2.0312, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 25.336505889892578, | |
| "learning_rate": 7.807692307692309e-06, | |
| "loss": 2.1045, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 4.3931623931623935, | |
| "grad_norm": 24.755443572998047, | |
| "learning_rate": 7.803418803418804e-06, | |
| "loss": 2.5754, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 4.401709401709402, | |
| "grad_norm": 29.29696273803711, | |
| "learning_rate": 7.7991452991453e-06, | |
| "loss": 2.562, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.410256410256411, | |
| "grad_norm": 28.054868698120117, | |
| "learning_rate": 7.794871794871796e-06, | |
| "loss": 1.9815, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 4.418803418803419, | |
| "grad_norm": 20.894853591918945, | |
| "learning_rate": 7.79059829059829e-06, | |
| "loss": 2.5668, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 4.427350427350428, | |
| "grad_norm": 19.532094955444336, | |
| "learning_rate": 7.786324786324787e-06, | |
| "loss": 2.2314, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 4.435897435897436, | |
| "grad_norm": 27.919715881347656, | |
| "learning_rate": 7.782051282051282e-06, | |
| "loss": 1.9523, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 21.91543960571289, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 2.559, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.452991452991453, | |
| "grad_norm": 26.20106315612793, | |
| "learning_rate": 7.773504273504274e-06, | |
| "loss": 2.367, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 23.455419540405273, | |
| "learning_rate": 7.76923076923077e-06, | |
| "loss": 2.4132, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 4.47008547008547, | |
| "grad_norm": 49.62391662597656, | |
| "learning_rate": 7.764957264957266e-06, | |
| "loss": 1.8896, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 4.478632478632479, | |
| "grad_norm": 25.721101760864258, | |
| "learning_rate": 7.760683760683761e-06, | |
| "loss": 1.9918, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 4.487179487179487, | |
| "grad_norm": 22.906694412231445, | |
| "learning_rate": 7.756410256410258e-06, | |
| "loss": 2.1819, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.495726495726496, | |
| "grad_norm": 28.5809268951416, | |
| "learning_rate": 7.752136752136753e-06, | |
| "loss": 2.0516, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 4.504273504273504, | |
| "grad_norm": 26.47665023803711, | |
| "learning_rate": 7.74786324786325e-06, | |
| "loss": 2.0081, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 4.512820512820513, | |
| "grad_norm": 27.221372604370117, | |
| "learning_rate": 7.743589743589745e-06, | |
| "loss": 2.0414, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 4.521367521367521, | |
| "grad_norm": 27.931568145751953, | |
| "learning_rate": 7.73931623931624e-06, | |
| "loss": 2.0335, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 4.52991452991453, | |
| "grad_norm": 25.567049026489258, | |
| "learning_rate": 7.735042735042736e-06, | |
| "loss": 2.0129, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 30.897083282470703, | |
| "learning_rate": 7.730769230769231e-06, | |
| "loss": 2.3941, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 4.547008547008547, | |
| "grad_norm": 21.92133903503418, | |
| "learning_rate": 7.726495726495728e-06, | |
| "loss": 2.2563, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 4.555555555555555, | |
| "grad_norm": 27.053892135620117, | |
| "learning_rate": 7.722222222222223e-06, | |
| "loss": 2.2463, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 4.564102564102564, | |
| "grad_norm": 29.3230037689209, | |
| "learning_rate": 7.717948717948718e-06, | |
| "loss": 1.9167, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 4.572649572649572, | |
| "grad_norm": 36.06028747558594, | |
| "learning_rate": 7.713675213675215e-06, | |
| "loss": 1.9106, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.581196581196581, | |
| "grad_norm": 24.622135162353516, | |
| "learning_rate": 7.70940170940171e-06, | |
| "loss": 2.2899, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 4.589743589743589, | |
| "grad_norm": 21.3137264251709, | |
| "learning_rate": 7.705128205128207e-06, | |
| "loss": 2.0166, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 4.598290598290598, | |
| "grad_norm": 21.939279556274414, | |
| "learning_rate": 7.700854700854702e-06, | |
| "loss": 2.3319, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 4.6068376068376065, | |
| "grad_norm": 25.496994018554688, | |
| "learning_rate": 7.696581196581197e-06, | |
| "loss": 2.6162, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 24.095666885375977, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 2.2863, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.6239316239316235, | |
| "grad_norm": 31.96511459350586, | |
| "learning_rate": 7.688034188034189e-06, | |
| "loss": 2.0261, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 4.632478632478632, | |
| "grad_norm": 22.66115379333496, | |
| "learning_rate": 7.683760683760685e-06, | |
| "loss": 2.2786, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 4.641025641025641, | |
| "grad_norm": 23.661611557006836, | |
| "learning_rate": 7.67948717948718e-06, | |
| "loss": 1.7113, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 4.64957264957265, | |
| "grad_norm": 18.64708709716797, | |
| "learning_rate": 7.675213675213677e-06, | |
| "loss": 2.1389, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 4.6581196581196584, | |
| "grad_norm": 20.55480194091797, | |
| "learning_rate": 7.670940170940172e-06, | |
| "loss": 2.0831, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 27.876964569091797, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 2.0358, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 4.6752136752136755, | |
| "grad_norm": 20.236507415771484, | |
| "learning_rate": 7.662393162393164e-06, | |
| "loss": 1.5596, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 4.683760683760684, | |
| "grad_norm": 23.360782623291016, | |
| "learning_rate": 7.658119658119659e-06, | |
| "loss": 1.9623, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 41.7568359375, | |
| "learning_rate": 7.653846153846154e-06, | |
| "loss": 1.9884, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 4.700854700854701, | |
| "grad_norm": 28.651065826416016, | |
| "learning_rate": 7.649572649572649e-06, | |
| "loss": 2.1491, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.7094017094017095, | |
| "grad_norm": 23.636432647705078, | |
| "learning_rate": 7.645299145299146e-06, | |
| "loss": 1.9352, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 4.717948717948718, | |
| "grad_norm": 25.313966751098633, | |
| "learning_rate": 7.641025641025641e-06, | |
| "loss": 2.4112, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 4.726495726495727, | |
| "grad_norm": 32.4974479675293, | |
| "learning_rate": 7.636752136752138e-06, | |
| "loss": 1.7017, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 4.735042735042735, | |
| "grad_norm": 20.644481658935547, | |
| "learning_rate": 7.632478632478633e-06, | |
| "loss": 1.6904, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 4.743589743589744, | |
| "grad_norm": 26.526721954345703, | |
| "learning_rate": 7.6282051282051286e-06, | |
| "loss": 2.1666, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 4.752136752136752, | |
| "grad_norm": 23.375839233398438, | |
| "learning_rate": 7.6239316239316244e-06, | |
| "loss": 1.5555, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 4.760683760683761, | |
| "grad_norm": 29.890501022338867, | |
| "learning_rate": 7.6196581196581195e-06, | |
| "loss": 2.0195, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 687.5745239257812, | |
| "learning_rate": 7.615384615384615e-06, | |
| "loss": 2.4286, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 4.777777777777778, | |
| "grad_norm": 22.844587326049805, | |
| "learning_rate": 7.611111111111111e-06, | |
| "loss": 2.2335, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 4.786324786324786, | |
| "grad_norm": 29.633562088012695, | |
| "learning_rate": 7.606837606837607e-06, | |
| "loss": 1.7579, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.794871794871795, | |
| "grad_norm": 48.04582977294922, | |
| "learning_rate": 7.602564102564103e-06, | |
| "loss": 2.3846, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 4.803418803418803, | |
| "grad_norm": 27.2290096282959, | |
| "learning_rate": 7.598290598290599e-06, | |
| "loss": 2.2234, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 4.811965811965812, | |
| "grad_norm": 29.782209396362305, | |
| "learning_rate": 7.594017094017094e-06, | |
| "loss": 2.0365, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 4.82051282051282, | |
| "grad_norm": 32.457061767578125, | |
| "learning_rate": 7.58974358974359e-06, | |
| "loss": 2.0451, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 4.829059829059829, | |
| "grad_norm": 22.089427947998047, | |
| "learning_rate": 7.585470085470086e-06, | |
| "loss": 1.7105, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 4.837606837606837, | |
| "grad_norm": 23.105140686035156, | |
| "learning_rate": 7.581196581196582e-06, | |
| "loss": 1.6817, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 24.513713836669922, | |
| "learning_rate": 7.5769230769230775e-06, | |
| "loss": 1.9553, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 4.854700854700854, | |
| "grad_norm": 22.187759399414062, | |
| "learning_rate": 7.572649572649573e-06, | |
| "loss": 2.0309, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 4.863247863247864, | |
| "grad_norm": 53.56728744506836, | |
| "learning_rate": 7.5683760683760685e-06, | |
| "loss": 2.6508, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 4.871794871794872, | |
| "grad_norm": 27.983978271484375, | |
| "learning_rate": 7.564102564102564e-06, | |
| "loss": 2.1942, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.880341880341881, | |
| "grad_norm": 25.610252380371094, | |
| "learning_rate": 7.55982905982906e-06, | |
| "loss": 1.4151, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 19.856618881225586, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 1.6968, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 4.897435897435898, | |
| "grad_norm": 20.288606643676758, | |
| "learning_rate": 7.551282051282052e-06, | |
| "loss": 1.7494, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 4.905982905982906, | |
| "grad_norm": 23.206768035888672, | |
| "learning_rate": 7.547008547008547e-06, | |
| "loss": 2.1255, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 4.914529914529915, | |
| "grad_norm": 21.275257110595703, | |
| "learning_rate": 7.542735042735043e-06, | |
| "loss": 1.7442, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 22.635417938232422, | |
| "learning_rate": 7.538461538461539e-06, | |
| "loss": 1.9129, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 4.931623931623932, | |
| "grad_norm": 21.440109252929688, | |
| "learning_rate": 7.534188034188035e-06, | |
| "loss": 2.0056, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 4.94017094017094, | |
| "grad_norm": 20.939407348632812, | |
| "learning_rate": 7.529914529914531e-06, | |
| "loss": 1.7231, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 4.948717948717949, | |
| "grad_norm": 16.189861297607422, | |
| "learning_rate": 7.5256410256410265e-06, | |
| "loss": 1.4255, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 4.957264957264957, | |
| "grad_norm": 23.6302547454834, | |
| "learning_rate": 7.521367521367522e-06, | |
| "loss": 1.6748, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.965811965811966, | |
| "grad_norm": 22.29713249206543, | |
| "learning_rate": 7.5170940170940175e-06, | |
| "loss": 1.5285, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 4.9743589743589745, | |
| "grad_norm": 22.831275939941406, | |
| "learning_rate": 7.512820512820513e-06, | |
| "loss": 1.7742, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 4.982905982905983, | |
| "grad_norm": 630.5899658203125, | |
| "learning_rate": 7.508547008547009e-06, | |
| "loss": 2.8598, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 4.9914529914529915, | |
| "grad_norm": 22.880647659301758, | |
| "learning_rate": 7.504273504273505e-06, | |
| "loss": 1.6231, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 21.379072189331055, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.3506, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.8325614333152771, | |
| "eval_runtime": 9.2303, | |
| "eval_samples_per_second": 50.486, | |
| "eval_steps_per_second": 6.392, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 5.0085470085470085, | |
| "grad_norm": 23.968698501586914, | |
| "learning_rate": 7.495726495726496e-06, | |
| "loss": 1.4263, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 5.017094017094017, | |
| "grad_norm": 24.880769729614258, | |
| "learning_rate": 7.491452991452992e-06, | |
| "loss": 1.4994, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 5.0256410256410255, | |
| "grad_norm": 23.4547176361084, | |
| "learning_rate": 7.487179487179488e-06, | |
| "loss": 1.671, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 5.034188034188034, | |
| "grad_norm": 17.382152557373047, | |
| "learning_rate": 7.482905982905984e-06, | |
| "loss": 1.3935, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 5.042735042735043, | |
| "grad_norm": 19.607717514038086, | |
| "learning_rate": 7.47863247863248e-06, | |
| "loss": 1.5652, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.051282051282051, | |
| "grad_norm": 27.735240936279297, | |
| "learning_rate": 7.474358974358975e-06, | |
| "loss": 1.5491, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 5.05982905982906, | |
| "grad_norm": 20.493412017822266, | |
| "learning_rate": 7.4700854700854706e-06, | |
| "loss": 1.9229, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 5.068376068376068, | |
| "grad_norm": 20.492137908935547, | |
| "learning_rate": 7.4658119658119665e-06, | |
| "loss": 1.5066, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 5.076923076923077, | |
| "grad_norm": 27.650495529174805, | |
| "learning_rate": 7.461538461538462e-06, | |
| "loss": 1.4228, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 5.085470085470085, | |
| "grad_norm": 22.38190269470215, | |
| "learning_rate": 7.457264957264958e-06, | |
| "loss": 1.6243, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 5.094017094017094, | |
| "grad_norm": 22.862489700317383, | |
| "learning_rate": 7.452991452991454e-06, | |
| "loss": 1.9224, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 5.102564102564102, | |
| "grad_norm": 17.368051528930664, | |
| "learning_rate": 7.448717948717949e-06, | |
| "loss": 1.3642, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 5.111111111111111, | |
| "grad_norm": 20.587018966674805, | |
| "learning_rate": 7.444444444444445e-06, | |
| "loss": 1.471, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 5.119658119658119, | |
| "grad_norm": 18.502887725830078, | |
| "learning_rate": 7.440170940170941e-06, | |
| "loss": 1.9841, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 5.128205128205128, | |
| "grad_norm": 21.305294036865234, | |
| "learning_rate": 7.435897435897437e-06, | |
| "loss": 1.8564, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.136752136752137, | |
| "grad_norm": 20.61264419555664, | |
| "learning_rate": 7.431623931623933e-06, | |
| "loss": 1.3554, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 5.145299145299146, | |
| "grad_norm": 19.05555534362793, | |
| "learning_rate": 7.427350427350429e-06, | |
| "loss": 1.6612, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 5.153846153846154, | |
| "grad_norm": 20.392446517944336, | |
| "learning_rate": 7.423076923076924e-06, | |
| "loss": 1.5071, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 5.162393162393163, | |
| "grad_norm": 22.007591247558594, | |
| "learning_rate": 7.4188034188034196e-06, | |
| "loss": 1.3356, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 5.170940170940171, | |
| "grad_norm": 18.928104400634766, | |
| "learning_rate": 7.4145299145299155e-06, | |
| "loss": 1.6214, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 5.17948717948718, | |
| "grad_norm": 21.151193618774414, | |
| "learning_rate": 7.410256410256411e-06, | |
| "loss": 1.5275, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 5.188034188034188, | |
| "grad_norm": 16.272262573242188, | |
| "learning_rate": 7.405982905982907e-06, | |
| "loss": 1.2773, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 5.196581196581197, | |
| "grad_norm": 21.59275245666504, | |
| "learning_rate": 7.401709401709402e-06, | |
| "loss": 1.3503, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 5.205128205128205, | |
| "grad_norm": 84.31806182861328, | |
| "learning_rate": 7.397435897435898e-06, | |
| "loss": 1.8618, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 5.213675213675214, | |
| "grad_norm": 20.374465942382812, | |
| "learning_rate": 7.393162393162394e-06, | |
| "loss": 1.6153, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.222222222222222, | |
| "grad_norm": 18.569623947143555, | |
| "learning_rate": 7.38888888888889e-06, | |
| "loss": 1.7101, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 5.230769230769231, | |
| "grad_norm": 19.51409339904785, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 1.5801, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 5.239316239316239, | |
| "grad_norm": 19.45322608947754, | |
| "learning_rate": 7.380341880341882e-06, | |
| "loss": 1.1376, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 5.247863247863248, | |
| "grad_norm": 23.474557876586914, | |
| "learning_rate": 7.376068376068377e-06, | |
| "loss": 1.442, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 5.256410256410256, | |
| "grad_norm": 21.458847045898438, | |
| "learning_rate": 7.371794871794873e-06, | |
| "loss": 1.2769, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 5.264957264957265, | |
| "grad_norm": 25.741121292114258, | |
| "learning_rate": 7.3675213675213686e-06, | |
| "loss": 1.3321, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 5.273504273504273, | |
| "grad_norm": 15.394718170166016, | |
| "learning_rate": 7.3632478632478645e-06, | |
| "loss": 1.2335, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 5.282051282051282, | |
| "grad_norm": 20.938871383666992, | |
| "learning_rate": 7.35897435897436e-06, | |
| "loss": 1.5741, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 5.2905982905982905, | |
| "grad_norm": 19.348268508911133, | |
| "learning_rate": 7.354700854700856e-06, | |
| "loss": 1.2493, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 5.299145299145299, | |
| "grad_norm": 25.26751708984375, | |
| "learning_rate": 7.350427350427351e-06, | |
| "loss": 1.5167, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.3076923076923075, | |
| "grad_norm": 22.099227905273438, | |
| "learning_rate": 7.346153846153847e-06, | |
| "loss": 1.3269, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 5.316239316239316, | |
| "grad_norm": 21.483428955078125, | |
| "learning_rate": 7.341880341880342e-06, | |
| "loss": 1.4249, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 5.3247863247863245, | |
| "grad_norm": 20.089691162109375, | |
| "learning_rate": 7.337606837606837e-06, | |
| "loss": 1.351, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 138.9898223876953, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 1.5682, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 5.3418803418803416, | |
| "grad_norm": 16.808000564575195, | |
| "learning_rate": 7.329059829059829e-06, | |
| "loss": 1.4794, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 5.35042735042735, | |
| "grad_norm": 18.58464813232422, | |
| "learning_rate": 7.324786324786325e-06, | |
| "loss": 1.4486, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 5.358974358974359, | |
| "grad_norm": 15.074477195739746, | |
| "learning_rate": 7.320512820512821e-06, | |
| "loss": 1.3124, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 5.367521367521368, | |
| "grad_norm": 15.800148963928223, | |
| "learning_rate": 7.316239316239317e-06, | |
| "loss": 1.7055, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 5.3760683760683765, | |
| "grad_norm": 19.166179656982422, | |
| "learning_rate": 7.311965811965812e-06, | |
| "loss": 1.7306, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 55.91648864746094, | |
| "learning_rate": 7.307692307692308e-06, | |
| "loss": 1.2376, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.3931623931623935, | |
| "grad_norm": 16.606033325195312, | |
| "learning_rate": 7.3034188034188035e-06, | |
| "loss": 1.1159, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 5.401709401709402, | |
| "grad_norm": 17.0134220123291, | |
| "learning_rate": 7.299145299145299e-06, | |
| "loss": 1.2124, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 5.410256410256411, | |
| "grad_norm": 17.511932373046875, | |
| "learning_rate": 7.294871794871795e-06, | |
| "loss": 1.4221, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 5.418803418803419, | |
| "grad_norm": 44.53416061401367, | |
| "learning_rate": 7.290598290598291e-06, | |
| "loss": 1.9583, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 5.427350427350428, | |
| "grad_norm": 16.546630859375, | |
| "learning_rate": 7.286324786324786e-06, | |
| "loss": 1.1722, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 5.435897435897436, | |
| "grad_norm": 39.90822982788086, | |
| "learning_rate": 7.282051282051282e-06, | |
| "loss": 1.7482, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 5.444444444444445, | |
| "grad_norm": 16.186573028564453, | |
| "learning_rate": 7.277777777777778e-06, | |
| "loss": 1.3422, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 5.452991452991453, | |
| "grad_norm": 18.84516143798828, | |
| "learning_rate": 7.273504273504274e-06, | |
| "loss": 1.3299, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 5.461538461538462, | |
| "grad_norm": 14.620058059692383, | |
| "learning_rate": 7.26923076923077e-06, | |
| "loss": 1.0604, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 5.47008547008547, | |
| "grad_norm": 16.5911865234375, | |
| "learning_rate": 7.264957264957266e-06, | |
| "loss": 1.1138, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.478632478632479, | |
| "grad_norm": 15.44485092163086, | |
| "learning_rate": 7.260683760683761e-06, | |
| "loss": 1.435, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 5.487179487179487, | |
| "grad_norm": 121.76724243164062, | |
| "learning_rate": 7.256410256410257e-06, | |
| "loss": 1.7167, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 5.495726495726496, | |
| "grad_norm": 1996.141357421875, | |
| "learning_rate": 7.2521367521367525e-06, | |
| "loss": 4.0296, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 5.504273504273504, | |
| "grad_norm": 15.072067260742188, | |
| "learning_rate": 7.247863247863248e-06, | |
| "loss": 1.0455, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 5.512820512820513, | |
| "grad_norm": 16.684345245361328, | |
| "learning_rate": 7.243589743589744e-06, | |
| "loss": 1.7565, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 5.521367521367521, | |
| "grad_norm": 15.515148162841797, | |
| "learning_rate": 7.239316239316239e-06, | |
| "loss": 1.4601, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 5.52991452991453, | |
| "grad_norm": 20.1015625, | |
| "learning_rate": 7.235042735042735e-06, | |
| "loss": 1.073, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 5.538461538461538, | |
| "grad_norm": 67.10873413085938, | |
| "learning_rate": 7.230769230769231e-06, | |
| "loss": 1.8586, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 5.547008547008547, | |
| "grad_norm": 13.775193214416504, | |
| "learning_rate": 7.226495726495727e-06, | |
| "loss": 1.2891, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 14.612048149108887, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 1.033, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.564102564102564, | |
| "grad_norm": 14.512042999267578, | |
| "learning_rate": 7.217948717948719e-06, | |
| "loss": 1.1446, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 5.572649572649572, | |
| "grad_norm": 13.720820426940918, | |
| "learning_rate": 7.213675213675214e-06, | |
| "loss": 1.1246, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 5.581196581196581, | |
| "grad_norm": 16.548046112060547, | |
| "learning_rate": 7.20940170940171e-06, | |
| "loss": 1.3162, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 5.589743589743589, | |
| "grad_norm": 20.535181045532227, | |
| "learning_rate": 7.205128205128206e-06, | |
| "loss": 1.3019, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 5.598290598290598, | |
| "grad_norm": 14.317465782165527, | |
| "learning_rate": 7.2008547008547015e-06, | |
| "loss": 1.5447, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 5.6068376068376065, | |
| "grad_norm": 16.23088836669922, | |
| "learning_rate": 7.196581196581197e-06, | |
| "loss": 1.2701, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 5.615384615384615, | |
| "grad_norm": 13.754173278808594, | |
| "learning_rate": 7.192307692307693e-06, | |
| "loss": 1.2218, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 5.6239316239316235, | |
| "grad_norm": 75.77688598632812, | |
| "learning_rate": 7.188034188034188e-06, | |
| "loss": 1.7547, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 5.632478632478632, | |
| "grad_norm": 19.452077865600586, | |
| "learning_rate": 7.183760683760684e-06, | |
| "loss": 1.1446, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 5.641025641025641, | |
| "grad_norm": 14.513677597045898, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 1.0527, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.64957264957265, | |
| "grad_norm": 27.67446517944336, | |
| "learning_rate": 7.175213675213676e-06, | |
| "loss": 1.1953, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 5.6581196581196584, | |
| "grad_norm": 12.137639999389648, | |
| "learning_rate": 7.170940170940172e-06, | |
| "loss": 1.1127, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 5.666666666666667, | |
| "grad_norm": 17.2878475189209, | |
| "learning_rate": 7.166666666666667e-06, | |
| "loss": 1.0475, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 5.6752136752136755, | |
| "grad_norm": 28.070842742919922, | |
| "learning_rate": 7.162393162393163e-06, | |
| "loss": 1.6271, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 5.683760683760684, | |
| "grad_norm": 17.74942398071289, | |
| "learning_rate": 7.158119658119659e-06, | |
| "loss": 1.1759, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 5.6923076923076925, | |
| "grad_norm": 19.545486450195312, | |
| "learning_rate": 7.153846153846155e-06, | |
| "loss": 0.9753, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 5.700854700854701, | |
| "grad_norm": 24.34153938293457, | |
| "learning_rate": 7.1495726495726505e-06, | |
| "loss": 1.0905, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 5.7094017094017095, | |
| "grad_norm": 211.7845001220703, | |
| "learning_rate": 7.145299145299146e-06, | |
| "loss": 1.6455, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 5.717948717948718, | |
| "grad_norm": 14.03074836730957, | |
| "learning_rate": 7.1410256410256414e-06, | |
| "loss": 1.3728, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 5.726495726495727, | |
| "grad_norm": 27.600345611572266, | |
| "learning_rate": 7.136752136752137e-06, | |
| "loss": 1.4212, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.735042735042735, | |
| "grad_norm": 15.755846977233887, | |
| "learning_rate": 7.132478632478633e-06, | |
| "loss": 1.148, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 5.743589743589744, | |
| "grad_norm": 12.816133499145508, | |
| "learning_rate": 7.128205128205129e-06, | |
| "loss": 1.0053, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 5.752136752136752, | |
| "grad_norm": 25.097660064697266, | |
| "learning_rate": 7.123931623931625e-06, | |
| "loss": 1.1561, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 5.760683760683761, | |
| "grad_norm": 19.249279022216797, | |
| "learning_rate": 7.119658119658121e-06, | |
| "loss": 1.2582, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "grad_norm": 18.606924057006836, | |
| "learning_rate": 7.115384615384616e-06, | |
| "loss": 0.8569, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 5.777777777777778, | |
| "grad_norm": 20.2148380279541, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 1.1126, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 5.786324786324786, | |
| "grad_norm": 18.623268127441406, | |
| "learning_rate": 7.106837606837608e-06, | |
| "loss": 1.6129, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 5.794871794871795, | |
| "grad_norm": 14.888258934020996, | |
| "learning_rate": 7.102564102564104e-06, | |
| "loss": 1.2533, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 5.803418803418803, | |
| "grad_norm": 15.351551055908203, | |
| "learning_rate": 7.0982905982905995e-06, | |
| "loss": 1.2392, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 5.811965811965812, | |
| "grad_norm": 23.243993759155273, | |
| "learning_rate": 7.0940170940170945e-06, | |
| "loss": 1.3136, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.82051282051282, | |
| "grad_norm": 18.346277236938477, | |
| "learning_rate": 7.0897435897435904e-06, | |
| "loss": 1.5691, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 5.829059829059829, | |
| "grad_norm": 12.904829025268555, | |
| "learning_rate": 7.085470085470086e-06, | |
| "loss": 0.9248, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 5.837606837606837, | |
| "grad_norm": 13.263056755065918, | |
| "learning_rate": 7.081196581196582e-06, | |
| "loss": 1.0555, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 5.846153846153846, | |
| "grad_norm": 19.311899185180664, | |
| "learning_rate": 7.076923076923078e-06, | |
| "loss": 1.4341, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 5.854700854700854, | |
| "grad_norm": 282.1452331542969, | |
| "learning_rate": 7.072649572649574e-06, | |
| "loss": 1.9797, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 5.863247863247864, | |
| "grad_norm": 14.317438125610352, | |
| "learning_rate": 7.068376068376069e-06, | |
| "loss": 0.839, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 5.871794871794872, | |
| "grad_norm": 13.549150466918945, | |
| "learning_rate": 7.064102564102565e-06, | |
| "loss": 1.1003, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 5.880341880341881, | |
| "grad_norm": 14.283610343933105, | |
| "learning_rate": 7.059829059829061e-06, | |
| "loss": 1.0297, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 5.888888888888889, | |
| "grad_norm": 18.737884521484375, | |
| "learning_rate": 7.055555555555557e-06, | |
| "loss": 0.9817, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 5.897435897435898, | |
| "grad_norm": 24.12625503540039, | |
| "learning_rate": 7.051282051282053e-06, | |
| "loss": 1.1837, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.905982905982906, | |
| "grad_norm": 11.760732650756836, | |
| "learning_rate": 7.0470085470085485e-06, | |
| "loss": 1.5131, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 5.914529914529915, | |
| "grad_norm": 16.138668060302734, | |
| "learning_rate": 7.0427350427350435e-06, | |
| "loss": 0.9569, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 5.923076923076923, | |
| "grad_norm": 17.727285385131836, | |
| "learning_rate": 7.038461538461539e-06, | |
| "loss": 0.9834, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 5.931623931623932, | |
| "grad_norm": 13.434252738952637, | |
| "learning_rate": 7.034188034188035e-06, | |
| "loss": 1.3635, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 5.94017094017094, | |
| "grad_norm": 15.587186813354492, | |
| "learning_rate": 7.02991452991453e-06, | |
| "loss": 1.4814, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 5.948717948717949, | |
| "grad_norm": 31.379039764404297, | |
| "learning_rate": 7.025641025641025e-06, | |
| "loss": 0.8792, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 5.957264957264957, | |
| "grad_norm": 14.575559616088867, | |
| "learning_rate": 7.021367521367521e-06, | |
| "loss": 0.8865, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 5.965811965811966, | |
| "grad_norm": 13.55718994140625, | |
| "learning_rate": 7.017094017094017e-06, | |
| "loss": 0.9564, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 5.9743589743589745, | |
| "grad_norm": 13.288110733032227, | |
| "learning_rate": 7.012820512820513e-06, | |
| "loss": 0.8117, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 5.982905982905983, | |
| "grad_norm": 14.522254943847656, | |
| "learning_rate": 7.008547008547009e-06, | |
| "loss": 1.2037, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.9914529914529915, | |
| "grad_norm": 14.575456619262695, | |
| "learning_rate": 7.004273504273504e-06, | |
| "loss": 1.028, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 13.18249225616455, | |
| "learning_rate": 7e-06, | |
| "loss": 0.6528, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.4769609868526459, | |
| "eval_runtime": 9.253, | |
| "eval_samples_per_second": 50.362, | |
| "eval_steps_per_second": 6.376, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 6.0085470085470085, | |
| "grad_norm": 17.034433364868164, | |
| "learning_rate": 6.995726495726496e-06, | |
| "loss": 0.847, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 6.017094017094017, | |
| "grad_norm": 13.455194473266602, | |
| "learning_rate": 6.991452991452992e-06, | |
| "loss": 0.8545, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 6.0256410256410255, | |
| "grad_norm": 14.511704444885254, | |
| "learning_rate": 6.9871794871794876e-06, | |
| "loss": 0.9365, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 6.034188034188034, | |
| "grad_norm": 14.325255393981934, | |
| "learning_rate": 6.9829059829059835e-06, | |
| "loss": 0.869, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 6.042735042735043, | |
| "grad_norm": 12.944524765014648, | |
| "learning_rate": 6.9786324786324785e-06, | |
| "loss": 1.1417, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 6.051282051282051, | |
| "grad_norm": 14.992669105529785, | |
| "learning_rate": 6.974358974358974e-06, | |
| "loss": 1.4935, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 6.05982905982906, | |
| "grad_norm": 15.394392013549805, | |
| "learning_rate": 6.97008547008547e-06, | |
| "loss": 1.519, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 6.068376068376068, | |
| "grad_norm": 12.605085372924805, | |
| "learning_rate": 6.965811965811966e-06, | |
| "loss": 1.4419, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 6.076923076923077, | |
| "grad_norm": 16.47636604309082, | |
| "learning_rate": 6.961538461538462e-06, | |
| "loss": 0.9552, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 6.085470085470085, | |
| "grad_norm": 17.04586410522461, | |
| "learning_rate": 6.957264957264958e-06, | |
| "loss": 0.9847, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 6.094017094017094, | |
| "grad_norm": 15.464738845825195, | |
| "learning_rate": 6.952991452991453e-06, | |
| "loss": 0.9272, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 6.102564102564102, | |
| "grad_norm": 11.837206840515137, | |
| "learning_rate": 6.948717948717949e-06, | |
| "loss": 1.1682, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 6.111111111111111, | |
| "grad_norm": 11.013447761535645, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 1.222, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 6.119658119658119, | |
| "grad_norm": 15.37415885925293, | |
| "learning_rate": 6.940170940170941e-06, | |
| "loss": 0.9668, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 6.128205128205128, | |
| "grad_norm": 14.077155113220215, | |
| "learning_rate": 6.9358974358974366e-06, | |
| "loss": 0.8448, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 6.136752136752137, | |
| "grad_norm": 13.440519332885742, | |
| "learning_rate": 6.931623931623932e-06, | |
| "loss": 0.891, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 6.145299145299146, | |
| "grad_norm": 13.059304237365723, | |
| "learning_rate": 6.9273504273504275e-06, | |
| "loss": 0.655, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 12.96674633026123, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.7755, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.162393162393163, | |
| "grad_norm": 10.921567916870117, | |
| "learning_rate": 6.918803418803419e-06, | |
| "loss": 0.8533, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 6.170940170940171, | |
| "grad_norm": 10.439260482788086, | |
| "learning_rate": 6.914529914529915e-06, | |
| "loss": 0.8294, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 6.17948717948718, | |
| "grad_norm": 14.948200225830078, | |
| "learning_rate": 6.910256410256411e-06, | |
| "loss": 0.7326, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 6.188034188034188, | |
| "grad_norm": 12.733176231384277, | |
| "learning_rate": 6.905982905982906e-06, | |
| "loss": 1.0244, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 6.196581196581197, | |
| "grad_norm": 12.432938575744629, | |
| "learning_rate": 6.901709401709402e-06, | |
| "loss": 0.7375, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 6.205128205128205, | |
| "grad_norm": 12.047768592834473, | |
| "learning_rate": 6.897435897435898e-06, | |
| "loss": 0.8348, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 6.213675213675214, | |
| "grad_norm": 19.029287338256836, | |
| "learning_rate": 6.893162393162394e-06, | |
| "loss": 0.6091, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 6.222222222222222, | |
| "grad_norm": 11.650983810424805, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 0.9925, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 6.230769230769231, | |
| "grad_norm": 12.12030029296875, | |
| "learning_rate": 6.8846153846153855e-06, | |
| "loss": 1.0205, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 6.239316239316239, | |
| "grad_norm": 10.283143997192383, | |
| "learning_rate": 6.880341880341881e-06, | |
| "loss": 0.7726, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.247863247863248, | |
| "grad_norm": 12.965302467346191, | |
| "learning_rate": 6.8760683760683765e-06, | |
| "loss": 1.1761, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 6.256410256410256, | |
| "grad_norm": 9.0562105178833, | |
| "learning_rate": 6.871794871794872e-06, | |
| "loss": 0.9769, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.264957264957265, | |
| "grad_norm": 13.647340774536133, | |
| "learning_rate": 6.867521367521368e-06, | |
| "loss": 0.7613, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 6.273504273504273, | |
| "grad_norm": 11.598361015319824, | |
| "learning_rate": 6.863247863247864e-06, | |
| "loss": 0.6236, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 6.282051282051282, | |
| "grad_norm": 10.453935623168945, | |
| "learning_rate": 6.858974358974359e-06, | |
| "loss": 0.9752, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 6.2905982905982905, | |
| "grad_norm": 14.108942985534668, | |
| "learning_rate": 6.854700854700855e-06, | |
| "loss": 0.9212, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 6.299145299145299, | |
| "grad_norm": 21.230859756469727, | |
| "learning_rate": 6.850427350427351e-06, | |
| "loss": 0.9213, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 6.3076923076923075, | |
| "grad_norm": 11.801465034484863, | |
| "learning_rate": 6.846153846153847e-06, | |
| "loss": 0.8182, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 6.316239316239316, | |
| "grad_norm": 18.9310302734375, | |
| "learning_rate": 6.841880341880343e-06, | |
| "loss": 0.6214, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 6.3247863247863245, | |
| "grad_norm": 11.773117065429688, | |
| "learning_rate": 6.837606837606839e-06, | |
| "loss": 0.6221, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 6.333333333333333, | |
| "grad_norm": 187.00250244140625, | |
| "learning_rate": 6.833333333333334e-06, | |
| "loss": 1.5211, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 6.3418803418803416, | |
| "grad_norm": 70.96250915527344, | |
| "learning_rate": 6.82905982905983e-06, | |
| "loss": 1.3472, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 6.35042735042735, | |
| "grad_norm": 11.787941932678223, | |
| "learning_rate": 6.8247863247863255e-06, | |
| "loss": 0.8831, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 6.358974358974359, | |
| "grad_norm": 11.33661937713623, | |
| "learning_rate": 6.820512820512821e-06, | |
| "loss": 1.0555, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 6.367521367521368, | |
| "grad_norm": 14.255888938903809, | |
| "learning_rate": 6.816239316239317e-06, | |
| "loss": 0.8246, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 6.3760683760683765, | |
| "grad_norm": 10.89616870880127, | |
| "learning_rate": 6.811965811965813e-06, | |
| "loss": 1.0179, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 6.384615384615385, | |
| "grad_norm": 9.160380363464355, | |
| "learning_rate": 6.807692307692308e-06, | |
| "loss": 0.9019, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 6.3931623931623935, | |
| "grad_norm": 12.984644889831543, | |
| "learning_rate": 6.803418803418804e-06, | |
| "loss": 0.649, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 6.401709401709402, | |
| "grad_norm": 14.073376655578613, | |
| "learning_rate": 6.7991452991453e-06, | |
| "loss": 0.608, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 6.410256410256411, | |
| "grad_norm": 10.354485511779785, | |
| "learning_rate": 6.794871794871796e-06, | |
| "loss": 0.8812, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.418803418803419, | |
| "grad_norm": 9.121294975280762, | |
| "learning_rate": 6.790598290598292e-06, | |
| "loss": 0.768, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 6.427350427350428, | |
| "grad_norm": 10.909361839294434, | |
| "learning_rate": 6.786324786324787e-06, | |
| "loss": 0.8697, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 6.435897435897436, | |
| "grad_norm": 26.324186325073242, | |
| "learning_rate": 6.782051282051283e-06, | |
| "loss": 1.2437, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 6.444444444444445, | |
| "grad_norm": 11.972411155700684, | |
| "learning_rate": 6.777777777777779e-06, | |
| "loss": 0.6366, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 6.452991452991453, | |
| "grad_norm": 25.042150497436523, | |
| "learning_rate": 6.7735042735042745e-06, | |
| "loss": 1.0371, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 6.461538461538462, | |
| "grad_norm": 10.331900596618652, | |
| "learning_rate": 6.76923076923077e-06, | |
| "loss": 0.5618, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 6.47008547008547, | |
| "grad_norm": 11.925344467163086, | |
| "learning_rate": 6.764957264957266e-06, | |
| "loss": 0.629, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 6.478632478632479, | |
| "grad_norm": 10.309441566467285, | |
| "learning_rate": 6.760683760683761e-06, | |
| "loss": 0.7158, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 6.487179487179487, | |
| "grad_norm": 11.374105453491211, | |
| "learning_rate": 6.756410256410257e-06, | |
| "loss": 0.6909, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 6.495726495726496, | |
| "grad_norm": 11.613142967224121, | |
| "learning_rate": 6.752136752136753e-06, | |
| "loss": 0.6139, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.504273504273504, | |
| "grad_norm": 14.499147415161133, | |
| "learning_rate": 6.747863247863249e-06, | |
| "loss": 0.7242, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 6.512820512820513, | |
| "grad_norm": 13.683001518249512, | |
| "learning_rate": 6.743589743589745e-06, | |
| "loss": 0.9246, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 6.521367521367521, | |
| "grad_norm": 11.068865776062012, | |
| "learning_rate": 6.739316239316241e-06, | |
| "loss": 0.8866, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 6.52991452991453, | |
| "grad_norm": 13.0232572555542, | |
| "learning_rate": 6.735042735042736e-06, | |
| "loss": 0.86, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 6.538461538461538, | |
| "grad_norm": 10.639331817626953, | |
| "learning_rate": 6.730769230769232e-06, | |
| "loss": 0.6928, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 6.547008547008547, | |
| "grad_norm": 11.792994499206543, | |
| "learning_rate": 6.7264957264957276e-06, | |
| "loss": 0.6571, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 6.555555555555555, | |
| "grad_norm": 15.907414436340332, | |
| "learning_rate": 6.7222222222222235e-06, | |
| "loss": 1.1426, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 6.564102564102564, | |
| "grad_norm": 12.207514762878418, | |
| "learning_rate": 6.717948717948718e-06, | |
| "loss": 1.0932, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 6.572649572649572, | |
| "grad_norm": 20.145288467407227, | |
| "learning_rate": 6.7136752136752135e-06, | |
| "loss": 0.9706, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 6.581196581196581, | |
| "grad_norm": 9.820805549621582, | |
| "learning_rate": 6.7094017094017094e-06, | |
| "loss": 0.4955, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.589743589743589, | |
| "grad_norm": 10.385655403137207, | |
| "learning_rate": 6.705128205128205e-06, | |
| "loss": 1.0172, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 6.598290598290598, | |
| "grad_norm": 11.708373069763184, | |
| "learning_rate": 6.700854700854701e-06, | |
| "loss": 0.8048, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 6.6068376068376065, | |
| "grad_norm": 9.812984466552734, | |
| "learning_rate": 6.696581196581196e-06, | |
| "loss": 0.4831, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 6.615384615384615, | |
| "grad_norm": 9.146960258483887, | |
| "learning_rate": 6.692307692307692e-06, | |
| "loss": 0.6178, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 6.6239316239316235, | |
| "grad_norm": 13.61231517791748, | |
| "learning_rate": 6.688034188034188e-06, | |
| "loss": 0.7812, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 6.632478632478632, | |
| "grad_norm": 10.349262237548828, | |
| "learning_rate": 6.683760683760684e-06, | |
| "loss": 0.819, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 6.641025641025641, | |
| "grad_norm": 48.387847900390625, | |
| "learning_rate": 6.67948717948718e-06, | |
| "loss": 1.5294, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 6.64957264957265, | |
| "grad_norm": 9.540630340576172, | |
| "learning_rate": 6.675213675213676e-06, | |
| "loss": 0.6564, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 6.6581196581196584, | |
| "grad_norm": 10.83983039855957, | |
| "learning_rate": 6.670940170940171e-06, | |
| "loss": 0.5109, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 15.380743026733398, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.6504, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.6752136752136755, | |
| "grad_norm": 16.796918869018555, | |
| "learning_rate": 6.6623931623931625e-06, | |
| "loss": 0.7944, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 6.683760683760684, | |
| "grad_norm": 39.64078140258789, | |
| "learning_rate": 6.6581196581196584e-06, | |
| "loss": 0.6929, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 6.6923076923076925, | |
| "grad_norm": 7.730568885803223, | |
| "learning_rate": 6.653846153846154e-06, | |
| "loss": 0.6284, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 6.700854700854701, | |
| "grad_norm": 7.840725898742676, | |
| "learning_rate": 6.64957264957265e-06, | |
| "loss": 0.5113, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 6.7094017094017095, | |
| "grad_norm": 13.925577163696289, | |
| "learning_rate": 6.645299145299145e-06, | |
| "loss": 0.6846, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 6.717948717948718, | |
| "grad_norm": 10.926531791687012, | |
| "learning_rate": 6.641025641025641e-06, | |
| "loss": 1.3245, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 6.726495726495727, | |
| "grad_norm": 10.698541641235352, | |
| "learning_rate": 6.636752136752137e-06, | |
| "loss": 0.6025, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 6.735042735042735, | |
| "grad_norm": 7.572136878967285, | |
| "learning_rate": 6.632478632478633e-06, | |
| "loss": 0.5473, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 6.743589743589744, | |
| "grad_norm": 26.242990493774414, | |
| "learning_rate": 6.628205128205129e-06, | |
| "loss": 0.5637, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 6.752136752136752, | |
| "grad_norm": 8.79776668548584, | |
| "learning_rate": 6.623931623931624e-06, | |
| "loss": 0.7595, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.760683760683761, | |
| "grad_norm": 8.951017379760742, | |
| "learning_rate": 6.61965811965812e-06, | |
| "loss": 1.0365, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 6.769230769230769, | |
| "grad_norm": 13.799118041992188, | |
| "learning_rate": 6.615384615384616e-06, | |
| "loss": 1.4206, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 6.777777777777778, | |
| "grad_norm": 674.3671875, | |
| "learning_rate": 6.6111111111111115e-06, | |
| "loss": 1.1752, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 6.786324786324786, | |
| "grad_norm": 8.110879898071289, | |
| "learning_rate": 6.606837606837607e-06, | |
| "loss": 0.4668, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 6.794871794871795, | |
| "grad_norm": 8.119854927062988, | |
| "learning_rate": 6.602564102564103e-06, | |
| "loss": 0.7689, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 6.803418803418803, | |
| "grad_norm": 11.039762496948242, | |
| "learning_rate": 6.598290598290598e-06, | |
| "loss": 0.5636, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 6.811965811965812, | |
| "grad_norm": 12.724084854125977, | |
| "learning_rate": 6.594017094017094e-06, | |
| "loss": 0.5072, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 6.82051282051282, | |
| "grad_norm": 12.196049690246582, | |
| "learning_rate": 6.58974358974359e-06, | |
| "loss": 0.5073, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 6.829059829059829, | |
| "grad_norm": 9.072951316833496, | |
| "learning_rate": 6.585470085470086e-06, | |
| "loss": 0.4855, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 6.837606837606837, | |
| "grad_norm": 10.53836441040039, | |
| "learning_rate": 6.581196581196582e-06, | |
| "loss": 1.0017, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.846153846153846, | |
| "grad_norm": 7.728690147399902, | |
| "learning_rate": 6.576923076923078e-06, | |
| "loss": 0.5784, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 6.854700854700854, | |
| "grad_norm": 28.362455368041992, | |
| "learning_rate": 6.572649572649573e-06, | |
| "loss": 1.0295, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 6.863247863247864, | |
| "grad_norm": 7.291123390197754, | |
| "learning_rate": 6.568376068376069e-06, | |
| "loss": 0.7836, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 6.871794871794872, | |
| "grad_norm": 9.566614151000977, | |
| "learning_rate": 6.564102564102565e-06, | |
| "loss": 0.9979, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 6.880341880341881, | |
| "grad_norm": 13.544408798217773, | |
| "learning_rate": 6.5598290598290605e-06, | |
| "loss": 0.5354, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 6.888888888888889, | |
| "grad_norm": 8.546881675720215, | |
| "learning_rate": 6.555555555555556e-06, | |
| "loss": 0.4689, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 6.897435897435898, | |
| "grad_norm": 8.94822883605957, | |
| "learning_rate": 6.5512820512820515e-06, | |
| "loss": 0.4432, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 6.905982905982906, | |
| "grad_norm": 6.5176544189453125, | |
| "learning_rate": 6.547008547008547e-06, | |
| "loss": 0.6747, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 6.914529914529915, | |
| "grad_norm": 9.48947811126709, | |
| "learning_rate": 6.542735042735043e-06, | |
| "loss": 0.4268, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 11.432586669921875, | |
| "learning_rate": 6.538461538461539e-06, | |
| "loss": 0.5486, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 6.931623931623932, | |
| "grad_norm": 7.585604667663574, | |
| "learning_rate": 6.534188034188035e-06, | |
| "loss": 0.4412, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 6.94017094017094, | |
| "grad_norm": 7.860292911529541, | |
| "learning_rate": 6.529914529914531e-06, | |
| "loss": 0.6428, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 6.948717948717949, | |
| "grad_norm": 27.83890151977539, | |
| "learning_rate": 6.525641025641026e-06, | |
| "loss": 0.6735, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 6.957264957264957, | |
| "grad_norm": 10.266451835632324, | |
| "learning_rate": 6.521367521367522e-06, | |
| "loss": 0.6757, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 6.965811965811966, | |
| "grad_norm": 8.839099884033203, | |
| "learning_rate": 6.517094017094018e-06, | |
| "loss": 0.7897, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 6.9743589743589745, | |
| "grad_norm": 10.037760734558105, | |
| "learning_rate": 6.512820512820514e-06, | |
| "loss": 0.7133, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 6.982905982905983, | |
| "grad_norm": 14.50278377532959, | |
| "learning_rate": 6.5085470085470095e-06, | |
| "loss": 1.0051, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 6.9914529914529915, | |
| "grad_norm": 8.775527000427246, | |
| "learning_rate": 6.504273504273505e-06, | |
| "loss": 0.8769, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 8.891378402709961, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.9586, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.23673956096172333, | |
| "eval_runtime": 9.3447, | |
| "eval_samples_per_second": 49.868, | |
| "eval_steps_per_second": 6.314, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 7.0085470085470085, | |
| "grad_norm": 8.925857543945312, | |
| "learning_rate": 6.495726495726496e-06, | |
| "loss": 1.0406, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 7.017094017094017, | |
| "grad_norm": 8.222796440124512, | |
| "learning_rate": 6.491452991452992e-06, | |
| "loss": 0.4911, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 7.0256410256410255, | |
| "grad_norm": 11.528886795043945, | |
| "learning_rate": 6.487179487179488e-06, | |
| "loss": 0.8292, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 7.034188034188034, | |
| "grad_norm": 7.9031524658203125, | |
| "learning_rate": 6.482905982905984e-06, | |
| "loss": 0.5319, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 7.042735042735043, | |
| "grad_norm": 6.788857936859131, | |
| "learning_rate": 6.478632478632479e-06, | |
| "loss": 0.431, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 7.051282051282051, | |
| "grad_norm": 8.84765911102295, | |
| "learning_rate": 6.474358974358975e-06, | |
| "loss": 0.6417, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 7.05982905982906, | |
| "grad_norm": 7.517561435699463, | |
| "learning_rate": 6.470085470085471e-06, | |
| "loss": 0.5828, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 7.068376068376068, | |
| "grad_norm": 9.86832332611084, | |
| "learning_rate": 6.465811965811967e-06, | |
| "loss": 0.5851, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 7.076923076923077, | |
| "grad_norm": 9.632494926452637, | |
| "learning_rate": 6.461538461538463e-06, | |
| "loss": 0.769, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 7.085470085470085, | |
| "grad_norm": 9.874857902526855, | |
| "learning_rate": 6.4572649572649585e-06, | |
| "loss": 0.4393, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 7.094017094017094, | |
| "grad_norm": 11.78085994720459, | |
| "learning_rate": 6.4529914529914535e-06, | |
| "loss": 0.8784, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 7.102564102564102, | |
| "grad_norm": 8.85053825378418, | |
| "learning_rate": 6.4487179487179494e-06, | |
| "loss": 0.5911, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 7.111111111111111, | |
| "grad_norm": 12.405013084411621, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 0.4941, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 7.119658119658119, | |
| "grad_norm": 12.237760543823242, | |
| "learning_rate": 6.440170940170941e-06, | |
| "loss": 0.4468, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 7.128205128205128, | |
| "grad_norm": 7.945899486541748, | |
| "learning_rate": 6.435897435897437e-06, | |
| "loss": 0.4101, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 7.136752136752137, | |
| "grad_norm": 10.743217468261719, | |
| "learning_rate": 6.431623931623933e-06, | |
| "loss": 0.679, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 7.145299145299146, | |
| "grad_norm": 7.700406551361084, | |
| "learning_rate": 6.427350427350428e-06, | |
| "loss": 0.5067, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 7.153846153846154, | |
| "grad_norm": 8.401918411254883, | |
| "learning_rate": 6.423076923076924e-06, | |
| "loss": 0.5893, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 7.162393162393163, | |
| "grad_norm": 23.065881729125977, | |
| "learning_rate": 6.41880341880342e-06, | |
| "loss": 0.6768, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 7.170940170940171, | |
| "grad_norm": 38.71855545043945, | |
| "learning_rate": 6.414529914529916e-06, | |
| "loss": 0.8828, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 7.17948717948718, | |
| "grad_norm": 12.142110824584961, | |
| "learning_rate": 6.410256410256412e-06, | |
| "loss": 0.5444, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.188034188034188, | |
| "grad_norm": 69.4731674194336, | |
| "learning_rate": 6.405982905982906e-06, | |
| "loss": 0.7768, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 7.196581196581197, | |
| "grad_norm": 15.926841735839844, | |
| "learning_rate": 6.401709401709402e-06, | |
| "loss": 0.4348, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 7.205128205128205, | |
| "grad_norm": 6.8418965339660645, | |
| "learning_rate": 6.397435897435898e-06, | |
| "loss": 0.3821, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 7.213675213675214, | |
| "grad_norm": 6.716574192047119, | |
| "learning_rate": 6.3931623931623935e-06, | |
| "loss": 0.3621, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 7.222222222222222, | |
| "grad_norm": 7.452919006347656, | |
| "learning_rate": 6.3888888888888885e-06, | |
| "loss": 0.4997, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 7.230769230769231, | |
| "grad_norm": 11.502019882202148, | |
| "learning_rate": 6.384615384615384e-06, | |
| "loss": 0.8017, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 7.239316239316239, | |
| "grad_norm": 7.349746227264404, | |
| "learning_rate": 6.38034188034188e-06, | |
| "loss": 0.2745, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 7.247863247863248, | |
| "grad_norm": 6.269787311553955, | |
| "learning_rate": 6.376068376068376e-06, | |
| "loss": 0.4131, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 7.256410256410256, | |
| "grad_norm": 9.56203842163086, | |
| "learning_rate": 6.371794871794872e-06, | |
| "loss": 0.8147, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 7.264957264957265, | |
| "grad_norm": 7.358108043670654, | |
| "learning_rate": 6.367521367521368e-06, | |
| "loss": 0.3552, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.273504273504273, | |
| "grad_norm": 7.6359782218933105, | |
| "learning_rate": 6.363247863247863e-06, | |
| "loss": 0.3302, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 7.282051282051282, | |
| "grad_norm": 7.356925010681152, | |
| "learning_rate": 6.358974358974359e-06, | |
| "loss": 0.2927, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 7.2905982905982905, | |
| "grad_norm": 11.097757339477539, | |
| "learning_rate": 6.354700854700855e-06, | |
| "loss": 0.8117, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 7.299145299145299, | |
| "grad_norm": 10.301170349121094, | |
| "learning_rate": 6.350427350427351e-06, | |
| "loss": 0.4044, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.3076923076923075, | |
| "grad_norm": 7.116042613983154, | |
| "learning_rate": 6.3461538461538466e-06, | |
| "loss": 0.289, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 7.316239316239316, | |
| "grad_norm": 7.453964710235596, | |
| "learning_rate": 6.3418803418803425e-06, | |
| "loss": 0.4652, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 7.3247863247863245, | |
| "grad_norm": 11.864774703979492, | |
| "learning_rate": 6.3376068376068375e-06, | |
| "loss": 0.4667, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 7.333333333333333, | |
| "grad_norm": 8.79547119140625, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 0.2874, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 7.3418803418803416, | |
| "grad_norm": 10.173043251037598, | |
| "learning_rate": 6.329059829059829e-06, | |
| "loss": 0.6844, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 7.35042735042735, | |
| "grad_norm": 9.26555061340332, | |
| "learning_rate": 6.324786324786325e-06, | |
| "loss": 0.2903, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 7.358974358974359, | |
| "grad_norm": 10.274518013000488, | |
| "learning_rate": 6.320512820512821e-06, | |
| "loss": 0.7824, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 7.367521367521368, | |
| "grad_norm": 7.104451656341553, | |
| "learning_rate": 6.316239316239316e-06, | |
| "loss": 0.3024, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 7.3760683760683765, | |
| "grad_norm": 9.522738456726074, | |
| "learning_rate": 6.311965811965812e-06, | |
| "loss": 0.3219, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 7.384615384615385, | |
| "grad_norm": 10.145588874816895, | |
| "learning_rate": 6.307692307692308e-06, | |
| "loss": 0.5319, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 7.3931623931623935, | |
| "grad_norm": 8.828988075256348, | |
| "learning_rate": 6.303418803418804e-06, | |
| "loss": 0.3286, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 7.401709401709402, | |
| "grad_norm": 7.314462661743164, | |
| "learning_rate": 6.2991452991453e-06, | |
| "loss": 0.2951, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 7.410256410256411, | |
| "grad_norm": 13.465666770935059, | |
| "learning_rate": 6.2948717948717956e-06, | |
| "loss": 0.4046, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 7.418803418803419, | |
| "grad_norm": 12.40607738494873, | |
| "learning_rate": 6.290598290598291e-06, | |
| "loss": 0.71, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 7.427350427350428, | |
| "grad_norm": 9.282904624938965, | |
| "learning_rate": 6.2863247863247865e-06, | |
| "loss": 0.4083, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 7.435897435897436, | |
| "grad_norm": 5.755247116088867, | |
| "learning_rate": 6.282051282051282e-06, | |
| "loss": 0.3858, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 7.444444444444445, | |
| "grad_norm": 6.996497631072998, | |
| "learning_rate": 6.277777777777778e-06, | |
| "loss": 0.2692, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 7.452991452991453, | |
| "grad_norm": 7.235395431518555, | |
| "learning_rate": 6.273504273504274e-06, | |
| "loss": 0.3936, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 7.461538461538462, | |
| "grad_norm": 14.275704383850098, | |
| "learning_rate": 6.26923076923077e-06, | |
| "loss": 0.4022, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 7.47008547008547, | |
| "grad_norm": 10.365689277648926, | |
| "learning_rate": 6.264957264957265e-06, | |
| "loss": 1.0508, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 7.478632478632479, | |
| "grad_norm": 5.840590000152588, | |
| "learning_rate": 6.260683760683761e-06, | |
| "loss": 0.2511, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 7.487179487179487, | |
| "grad_norm": 10.25346851348877, | |
| "learning_rate": 6.256410256410257e-06, | |
| "loss": 0.5836, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 7.495726495726496, | |
| "grad_norm": 27.662694931030273, | |
| "learning_rate": 6.252136752136753e-06, | |
| "loss": 0.7677, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 7.504273504273504, | |
| "grad_norm": 5.840217590332031, | |
| "learning_rate": 6.247863247863249e-06, | |
| "loss": 0.3889, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 7.512820512820513, | |
| "grad_norm": 9.813179016113281, | |
| "learning_rate": 6.243589743589744e-06, | |
| "loss": 0.8929, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 7.521367521367521, | |
| "grad_norm": 5.49755334854126, | |
| "learning_rate": 6.23931623931624e-06, | |
| "loss": 0.2712, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 7.52991452991453, | |
| "grad_norm": 7.17311429977417, | |
| "learning_rate": 6.2350427350427355e-06, | |
| "loss": 0.3071, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 7.538461538461538, | |
| "grad_norm": 7.706870079040527, | |
| "learning_rate": 6.230769230769231e-06, | |
| "loss": 0.3797, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 7.547008547008547, | |
| "grad_norm": 7.891415596008301, | |
| "learning_rate": 6.226495726495727e-06, | |
| "loss": 0.5352, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 7.555555555555555, | |
| "grad_norm": 8.746044158935547, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 0.263, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 7.564102564102564, | |
| "grad_norm": 9.096441268920898, | |
| "learning_rate": 6.217948717948718e-06, | |
| "loss": 0.2736, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 7.572649572649572, | |
| "grad_norm": 7.031003475189209, | |
| "learning_rate": 6.213675213675214e-06, | |
| "loss": 0.4705, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 7.581196581196581, | |
| "grad_norm": 6.6503143310546875, | |
| "learning_rate": 6.20940170940171e-06, | |
| "loss": 0.3285, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 7.589743589743589, | |
| "grad_norm": 5.398913383483887, | |
| "learning_rate": 6.205128205128206e-06, | |
| "loss": 0.41, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 7.598290598290598, | |
| "grad_norm": 7.47569465637207, | |
| "learning_rate": 6.200854700854702e-06, | |
| "loss": 0.4005, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 7.6068376068376065, | |
| "grad_norm": 8.79906940460205, | |
| "learning_rate": 6.196581196581198e-06, | |
| "loss": 0.2608, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.615384615384615, | |
| "grad_norm": 7.604002475738525, | |
| "learning_rate": 6.192307692307693e-06, | |
| "loss": 0.577, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 7.6239316239316235, | |
| "grad_norm": 12.666848182678223, | |
| "learning_rate": 6.188034188034189e-06, | |
| "loss": 0.7296, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 7.632478632478632, | |
| "grad_norm": 20.92390251159668, | |
| "learning_rate": 6.1837606837606845e-06, | |
| "loss": 0.9276, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 7.641025641025641, | |
| "grad_norm": 6.779317855834961, | |
| "learning_rate": 6.17948717948718e-06, | |
| "loss": 0.818, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 7.64957264957265, | |
| "grad_norm": 5.249539852142334, | |
| "learning_rate": 6.175213675213676e-06, | |
| "loss": 0.2117, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 7.6581196581196584, | |
| "grad_norm": 23.55508041381836, | |
| "learning_rate": 6.170940170940171e-06, | |
| "loss": 0.5239, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 7.666666666666667, | |
| "grad_norm": 11.711256980895996, | |
| "learning_rate": 6.166666666666667e-06, | |
| "loss": 0.6595, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 7.6752136752136755, | |
| "grad_norm": 6.641115188598633, | |
| "learning_rate": 6.162393162393163e-06, | |
| "loss": 0.4888, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 7.683760683760684, | |
| "grad_norm": 7.913390159606934, | |
| "learning_rate": 6.158119658119659e-06, | |
| "loss": 0.66, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 17.927574157714844, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9603, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.700854700854701, | |
| "grad_norm": 4.567203998565674, | |
| "learning_rate": 6.149572649572651e-06, | |
| "loss": 0.1638, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 7.7094017094017095, | |
| "grad_norm": 5.995935440063477, | |
| "learning_rate": 6.145299145299146e-06, | |
| "loss": 0.6852, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 7.717948717948718, | |
| "grad_norm": 8.323802947998047, | |
| "learning_rate": 6.141025641025642e-06, | |
| "loss": 0.5293, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 7.726495726495727, | |
| "grad_norm": 6.8586859703063965, | |
| "learning_rate": 6.136752136752138e-06, | |
| "loss": 0.3265, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 7.735042735042735, | |
| "grad_norm": 6.507427215576172, | |
| "learning_rate": 6.1324786324786335e-06, | |
| "loss": 0.2841, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 7.743589743589744, | |
| "grad_norm": 6.789999485015869, | |
| "learning_rate": 6.128205128205129e-06, | |
| "loss": 0.4236, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 7.752136752136752, | |
| "grad_norm": 19.444454193115234, | |
| "learning_rate": 6.123931623931625e-06, | |
| "loss": 0.2829, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 7.760683760683761, | |
| "grad_norm": 31.564800262451172, | |
| "learning_rate": 6.11965811965812e-06, | |
| "loss": 1.093, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 7.769230769230769, | |
| "grad_norm": 9.956007957458496, | |
| "learning_rate": 6.115384615384616e-06, | |
| "loss": 0.6749, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 7.777777777777778, | |
| "grad_norm": 5.193087577819824, | |
| "learning_rate": 6.111111111111112e-06, | |
| "loss": 0.1986, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 7.786324786324786, | |
| "grad_norm": 4.792945384979248, | |
| "learning_rate": 6.106837606837608e-06, | |
| "loss": 0.5179, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 7.794871794871795, | |
| "grad_norm": 20.602317810058594, | |
| "learning_rate": 6.102564102564104e-06, | |
| "loss": 1.0343, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 7.803418803418803, | |
| "grad_norm": 22.205543518066406, | |
| "learning_rate": 6.098290598290599e-06, | |
| "loss": 0.4921, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 7.811965811965812, | |
| "grad_norm": 13.392712593078613, | |
| "learning_rate": 6.094017094017095e-06, | |
| "loss": 0.9058, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 7.82051282051282, | |
| "grad_norm": 6.262679100036621, | |
| "learning_rate": 6.08974358974359e-06, | |
| "loss": 0.3877, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 7.829059829059829, | |
| "grad_norm": 12.727428436279297, | |
| "learning_rate": 6.085470085470086e-06, | |
| "loss": 0.4477, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 7.837606837606837, | |
| "grad_norm": 6.595224380493164, | |
| "learning_rate": 6.081196581196581e-06, | |
| "loss": 0.5553, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 7.846153846153846, | |
| "grad_norm": 6.815043926239014, | |
| "learning_rate": 6.076923076923077e-06, | |
| "loss": 0.2978, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 7.854700854700854, | |
| "grad_norm": 11.751949310302734, | |
| "learning_rate": 6.0726495726495726e-06, | |
| "loss": 0.5509, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 7.863247863247864, | |
| "grad_norm": 6.067570209503174, | |
| "learning_rate": 6.0683760683760684e-06, | |
| "loss": 0.475, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 7.871794871794872, | |
| "grad_norm": 7.4297919273376465, | |
| "learning_rate": 6.064102564102564e-06, | |
| "loss": 0.5073, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 7.880341880341881, | |
| "grad_norm": 6.778268337249756, | |
| "learning_rate": 6.05982905982906e-06, | |
| "loss": 0.4718, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 7.888888888888889, | |
| "grad_norm": 9.401915550231934, | |
| "learning_rate": 6.055555555555555e-06, | |
| "loss": 0.7151, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 7.897435897435898, | |
| "grad_norm": 6.359888553619385, | |
| "learning_rate": 6.051282051282051e-06, | |
| "loss": 0.3175, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 7.905982905982906, | |
| "grad_norm": 7.036016464233398, | |
| "learning_rate": 6.047008547008547e-06, | |
| "loss": 0.3172, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 7.914529914529915, | |
| "grad_norm": 5.980124473571777, | |
| "learning_rate": 6.042735042735043e-06, | |
| "loss": 0.2949, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 7.923076923076923, | |
| "grad_norm": 5.738795280456543, | |
| "learning_rate": 6.038461538461539e-06, | |
| "loss": 0.2454, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 7.931623931623932, | |
| "grad_norm": 4.688748359680176, | |
| "learning_rate": 6.034188034188035e-06, | |
| "loss": 0.1949, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 7.94017094017094, | |
| "grad_norm": 7.2333984375, | |
| "learning_rate": 6.02991452991453e-06, | |
| "loss": 0.2174, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 7.948717948717949, | |
| "grad_norm": 6.005523204803467, | |
| "learning_rate": 6.025641025641026e-06, | |
| "loss": 0.4216, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 7.957264957264957, | |
| "grad_norm": 6.017541885375977, | |
| "learning_rate": 6.0213675213675215e-06, | |
| "loss": 0.4904, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 7.965811965811966, | |
| "grad_norm": 19.559003829956055, | |
| "learning_rate": 6.0170940170940174e-06, | |
| "loss": 0.2616, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 7.9743589743589745, | |
| "grad_norm": 5.360724449157715, | |
| "learning_rate": 6.012820512820513e-06, | |
| "loss": 0.3629, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 7.982905982905983, | |
| "grad_norm": 9.472721099853516, | |
| "learning_rate": 6.008547008547008e-06, | |
| "loss": 0.5044, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 7.9914529914529915, | |
| "grad_norm": 6.453597068786621, | |
| "learning_rate": 6.004273504273504e-06, | |
| "loss": 0.4742, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 7.647386074066162, | |
| "learning_rate": 6e-06, | |
| "loss": 0.402, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.1672903448343277, | |
| "eval_runtime": 9.3047, | |
| "eval_samples_per_second": 50.082, | |
| "eval_steps_per_second": 6.341, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 8.008547008547009, | |
| "grad_norm": 5.8361663818359375, | |
| "learning_rate": 5.995726495726496e-06, | |
| "loss": 0.164, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 8.017094017094017, | |
| "grad_norm": 5.801360130310059, | |
| "learning_rate": 5.991452991452992e-06, | |
| "loss": 0.2858, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 8.025641025641026, | |
| "grad_norm": 4.43051290512085, | |
| "learning_rate": 5.987179487179488e-06, | |
| "loss": 0.2068, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 8.034188034188034, | |
| "grad_norm": 6.544061660766602, | |
| "learning_rate": 5.982905982905983e-06, | |
| "loss": 0.3499, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 8.042735042735043, | |
| "grad_norm": 5.500844955444336, | |
| "learning_rate": 5.978632478632479e-06, | |
| "loss": 0.3134, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 8.051282051282051, | |
| "grad_norm": 4.286651611328125, | |
| "learning_rate": 5.974358974358975e-06, | |
| "loss": 0.1767, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 8.05982905982906, | |
| "grad_norm": 13.860437393188477, | |
| "learning_rate": 5.9700854700854705e-06, | |
| "loss": 0.3913, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 8.068376068376068, | |
| "grad_norm": 5.998767852783203, | |
| "learning_rate": 5.9658119658119664e-06, | |
| "loss": 0.2275, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 8.076923076923077, | |
| "grad_norm": 9.01196002960205, | |
| "learning_rate": 5.961538461538462e-06, | |
| "loss": 0.5202, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 8.085470085470085, | |
| "grad_norm": 6.81577730178833, | |
| "learning_rate": 5.957264957264957e-06, | |
| "loss": 0.5923, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 8.094017094017094, | |
| "grad_norm": 7.400684833526611, | |
| "learning_rate": 5.952991452991453e-06, | |
| "loss": 0.2883, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 8.102564102564102, | |
| "grad_norm": 16.18587875366211, | |
| "learning_rate": 5.948717948717949e-06, | |
| "loss": 0.3377, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 8.11111111111111, | |
| "grad_norm": 5.017345428466797, | |
| "learning_rate": 5.944444444444445e-06, | |
| "loss": 0.3912, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 8.11965811965812, | |
| "grad_norm": 5.300196647644043, | |
| "learning_rate": 5.940170940170941e-06, | |
| "loss": 0.4056, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 8.128205128205128, | |
| "grad_norm": 6.3473405838012695, | |
| "learning_rate": 5.935897435897436e-06, | |
| "loss": 0.2559, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 8.136752136752136, | |
| "grad_norm": 12.37689208984375, | |
| "learning_rate": 5.931623931623932e-06, | |
| "loss": 0.2216, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 8.145299145299145, | |
| "grad_norm": 5.573046684265137, | |
| "learning_rate": 5.927350427350428e-06, | |
| "loss": 0.2047, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 8.153846153846153, | |
| "grad_norm": 5.033559322357178, | |
| "learning_rate": 5.923076923076924e-06, | |
| "loss": 0.3661, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 8.162393162393162, | |
| "grad_norm": 5.341614246368408, | |
| "learning_rate": 5.9188034188034195e-06, | |
| "loss": 0.2597, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 8.17094017094017, | |
| "grad_norm": 8.67937183380127, | |
| "learning_rate": 5.914529914529915e-06, | |
| "loss": 0.4098, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 8.179487179487179, | |
| "grad_norm": 3.957489252090454, | |
| "learning_rate": 5.9102564102564105e-06, | |
| "loss": 0.18, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 8.188034188034187, | |
| "grad_norm": 6.377108573913574, | |
| "learning_rate": 5.905982905982906e-06, | |
| "loss": 0.3414, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 8.196581196581196, | |
| "grad_norm": 8.621227264404297, | |
| "learning_rate": 5.901709401709402e-06, | |
| "loss": 1.1625, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 8.205128205128204, | |
| "grad_norm": 5.775392532348633, | |
| "learning_rate": 5.897435897435898e-06, | |
| "loss": 0.4283, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.213675213675213, | |
| "grad_norm": 4.522337913513184, | |
| "learning_rate": 5.893162393162394e-06, | |
| "loss": 0.3432, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 8.222222222222221, | |
| "grad_norm": 5.594667434692383, | |
| "learning_rate": 5.88888888888889e-06, | |
| "loss": 0.5212, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 8.23076923076923, | |
| "grad_norm": 5.478531837463379, | |
| "learning_rate": 5.884615384615385e-06, | |
| "loss": 0.2273, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 8.239316239316238, | |
| "grad_norm": 6.08770751953125, | |
| "learning_rate": 5.880341880341881e-06, | |
| "loss": 0.2673, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 8.247863247863247, | |
| "grad_norm": 7.962898254394531, | |
| "learning_rate": 5.876068376068377e-06, | |
| "loss": 0.2654, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 8.256410256410255, | |
| "grad_norm": 6.443154335021973, | |
| "learning_rate": 5.871794871794873e-06, | |
| "loss": 0.2982, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 8.264957264957266, | |
| "grad_norm": 4.689123153686523, | |
| "learning_rate": 5.8675213675213685e-06, | |
| "loss": 0.3459, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 8.273504273504274, | |
| "grad_norm": 5.446859359741211, | |
| "learning_rate": 5.863247863247864e-06, | |
| "loss": 0.2792, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 8.282051282051283, | |
| "grad_norm": 5.562478542327881, | |
| "learning_rate": 5.8589743589743595e-06, | |
| "loss": 0.1939, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 8.290598290598291, | |
| "grad_norm": 4.726650714874268, | |
| "learning_rate": 5.854700854700855e-06, | |
| "loss": 0.1368, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 8.2991452991453, | |
| "grad_norm": 17.44293785095215, | |
| "learning_rate": 5.850427350427351e-06, | |
| "loss": 0.3836, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 8.307692307692308, | |
| "grad_norm": 5.568243980407715, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 0.3674, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 8.316239316239317, | |
| "grad_norm": 3.488147258758545, | |
| "learning_rate": 5.841880341880343e-06, | |
| "loss": 0.197, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 8.324786324786325, | |
| "grad_norm": 15.902129173278809, | |
| "learning_rate": 5.837606837606838e-06, | |
| "loss": 0.4199, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 8.055335998535156, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.277, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 8.341880341880342, | |
| "grad_norm": 8.122756004333496, | |
| "learning_rate": 5.82905982905983e-06, | |
| "loss": 0.5572, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.350427350427351, | |
| "grad_norm": 5.7439961433410645, | |
| "learning_rate": 5.824786324786326e-06, | |
| "loss": 0.2031, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 8.35897435897436, | |
| "grad_norm": 4.329511642456055, | |
| "learning_rate": 5.820512820512822e-06, | |
| "loss": 0.4405, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 8.367521367521368, | |
| "grad_norm": 10.946788787841797, | |
| "learning_rate": 5.8162393162393175e-06, | |
| "loss": 0.4619, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 8.376068376068377, | |
| "grad_norm": 6.0579352378845215, | |
| "learning_rate": 5.8119658119658126e-06, | |
| "loss": 0.4679, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.384615384615385, | |
| "grad_norm": 5.656944751739502, | |
| "learning_rate": 5.8076923076923084e-06, | |
| "loss": 0.2395, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 8.393162393162394, | |
| "grad_norm": 5.344303607940674, | |
| "learning_rate": 5.803418803418804e-06, | |
| "loss": 0.2516, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 8.401709401709402, | |
| "grad_norm": 7.070309638977051, | |
| "learning_rate": 5.7991452991453e-06, | |
| "loss": 0.3169, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 8.41025641025641, | |
| "grad_norm": 5.168705940246582, | |
| "learning_rate": 5.794871794871796e-06, | |
| "loss": 0.3007, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 8.418803418803419, | |
| "grad_norm": 3.556293249130249, | |
| "learning_rate": 5.790598290598292e-06, | |
| "loss": 0.2089, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 8.427350427350428, | |
| "grad_norm": 4.943065166473389, | |
| "learning_rate": 5.786324786324787e-06, | |
| "loss": 0.2093, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 8.435897435897436, | |
| "grad_norm": 6.991105556488037, | |
| "learning_rate": 5.782051282051283e-06, | |
| "loss": 0.4671, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 8.444444444444445, | |
| "grad_norm": 5.276190280914307, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 0.2092, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 8.452991452991453, | |
| "grad_norm": 77.91864776611328, | |
| "learning_rate": 5.773504273504273e-06, | |
| "loss": 1.7536, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 8.461538461538462, | |
| "grad_norm": 4.864828109741211, | |
| "learning_rate": 5.769230769230769e-06, | |
| "loss": 0.1669, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 8.47008547008547, | |
| "grad_norm": 4.416967391967773, | |
| "learning_rate": 5.764957264957265e-06, | |
| "loss": 0.2705, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 8.478632478632479, | |
| "grad_norm": 4.558652400970459, | |
| "learning_rate": 5.760683760683761e-06, | |
| "loss": 0.4332, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 8.487179487179487, | |
| "grad_norm": 8.17482852935791, | |
| "learning_rate": 5.756410256410257e-06, | |
| "loss": 0.7286, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 8.495726495726496, | |
| "grad_norm": 7.322425365447998, | |
| "learning_rate": 5.7521367521367525e-06, | |
| "loss": 0.8554, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 8.504273504273504, | |
| "grad_norm": 4.249075889587402, | |
| "learning_rate": 5.7478632478632475e-06, | |
| "loss": 0.2442, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 8.512820512820513, | |
| "grad_norm": 4.157267093658447, | |
| "learning_rate": 5.743589743589743e-06, | |
| "loss": 0.4207, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 8.521367521367521, | |
| "grad_norm": 4.118504047393799, | |
| "learning_rate": 5.739316239316239e-06, | |
| "loss": 0.1411, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 8.52991452991453, | |
| "grad_norm": 7.273322105407715, | |
| "learning_rate": 5.735042735042735e-06, | |
| "loss": 0.6269, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 8.538461538461538, | |
| "grad_norm": 4.7668633460998535, | |
| "learning_rate": 5.730769230769231e-06, | |
| "loss": 0.1894, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 8.547008547008547, | |
| "grad_norm": 5.869007110595703, | |
| "learning_rate": 5.726495726495727e-06, | |
| "loss": 0.7301, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.555555555555555, | |
| "grad_norm": 5.987617015838623, | |
| "learning_rate": 5.722222222222222e-06, | |
| "loss": 0.29, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 8.564102564102564, | |
| "grad_norm": 5.445812702178955, | |
| "learning_rate": 5.717948717948718e-06, | |
| "loss": 0.4278, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 8.572649572649572, | |
| "grad_norm": 4.7509002685546875, | |
| "learning_rate": 5.713675213675214e-06, | |
| "loss": 0.3396, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 8.581196581196581, | |
| "grad_norm": 5.584397315979004, | |
| "learning_rate": 5.70940170940171e-06, | |
| "loss": 0.1329, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 8.58974358974359, | |
| "grad_norm": 4.627229690551758, | |
| "learning_rate": 5.705128205128206e-06, | |
| "loss": 0.3012, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 8.598290598290598, | |
| "grad_norm": 7.724045276641846, | |
| "learning_rate": 5.7008547008547015e-06, | |
| "loss": 0.4876, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 8.606837606837606, | |
| "grad_norm": 3.488499164581299, | |
| "learning_rate": 5.6965811965811965e-06, | |
| "loss": 0.2025, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 8.615384615384615, | |
| "grad_norm": 14.487537384033203, | |
| "learning_rate": 5.692307692307692e-06, | |
| "loss": 0.6795, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 8.623931623931623, | |
| "grad_norm": 4.03059196472168, | |
| "learning_rate": 5.688034188034188e-06, | |
| "loss": 0.2121, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 8.632478632478632, | |
| "grad_norm": 3.278873920440674, | |
| "learning_rate": 5.683760683760684e-06, | |
| "loss": 0.3475, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.64102564102564, | |
| "grad_norm": 4.599937915802002, | |
| "learning_rate": 5.67948717948718e-06, | |
| "loss": 0.2355, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 8.649572649572649, | |
| "grad_norm": 6.314788818359375, | |
| "learning_rate": 5.675213675213675e-06, | |
| "loss": 0.2402, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 8.658119658119658, | |
| "grad_norm": 3.4483532905578613, | |
| "learning_rate": 5.670940170940171e-06, | |
| "loss": 0.2189, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 8.666666666666666, | |
| "grad_norm": 299.8923645019531, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 1.0473, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 8.675213675213675, | |
| "grad_norm": 13.14855670928955, | |
| "learning_rate": 5.662393162393163e-06, | |
| "loss": 0.3723, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 8.683760683760683, | |
| "grad_norm": 6.513180732727051, | |
| "learning_rate": 5.658119658119659e-06, | |
| "loss": 0.483, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 8.692307692307692, | |
| "grad_norm": 5.026037693023682, | |
| "learning_rate": 5.6538461538461546e-06, | |
| "loss": 0.4417, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 8.7008547008547, | |
| "grad_norm": 176.535888671875, | |
| "learning_rate": 5.64957264957265e-06, | |
| "loss": 0.5256, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 8.709401709401709, | |
| "grad_norm": 6.023639678955078, | |
| "learning_rate": 5.6452991452991455e-06, | |
| "loss": 0.3708, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 8.717948717948717, | |
| "grad_norm": 16.64018440246582, | |
| "learning_rate": 5.641025641025641e-06, | |
| "loss": 0.8908, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 8.726495726495726, | |
| "grad_norm": 2.9167582988739014, | |
| "learning_rate": 5.636752136752137e-06, | |
| "loss": 0.077, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 8.735042735042736, | |
| "grad_norm": 3.368325710296631, | |
| "learning_rate": 5.632478632478633e-06, | |
| "loss": 0.2495, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 8.743589743589745, | |
| "grad_norm": 3.7961905002593994, | |
| "learning_rate": 5.628205128205129e-06, | |
| "loss": 0.4427, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 8.752136752136753, | |
| "grad_norm": 4.661024570465088, | |
| "learning_rate": 5.623931623931624e-06, | |
| "loss": 0.3092, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 8.760683760683762, | |
| "grad_norm": 5.1971588134765625, | |
| "learning_rate": 5.61965811965812e-06, | |
| "loss": 0.2213, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 8.76923076923077, | |
| "grad_norm": 4.427041530609131, | |
| "learning_rate": 5.615384615384616e-06, | |
| "loss": 0.2885, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 8.777777777777779, | |
| "grad_norm": 7.352906703948975, | |
| "learning_rate": 5.611111111111112e-06, | |
| "loss": 0.2689, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 8.786324786324787, | |
| "grad_norm": 5.306934833526611, | |
| "learning_rate": 5.606837606837608e-06, | |
| "loss": 0.3758, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 8.794871794871796, | |
| "grad_norm": 4.502418041229248, | |
| "learning_rate": 5.602564102564103e-06, | |
| "loss": 0.4655, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 8.803418803418804, | |
| "grad_norm": 3.427734851837158, | |
| "learning_rate": 5.598290598290599e-06, | |
| "loss": 0.1145, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 8.811965811965813, | |
| "grad_norm": 4.047433376312256, | |
| "learning_rate": 5.5940170940170945e-06, | |
| "loss": 0.1482, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 8.820512820512821, | |
| "grad_norm": 3.6860435009002686, | |
| "learning_rate": 5.58974358974359e-06, | |
| "loss": 0.1152, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 8.82905982905983, | |
| "grad_norm": 6.792733669281006, | |
| "learning_rate": 5.585470085470086e-06, | |
| "loss": 0.1732, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 8.837606837606838, | |
| "grad_norm": 4.222206115722656, | |
| "learning_rate": 5.581196581196582e-06, | |
| "loss": 0.1259, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 8.846153846153847, | |
| "grad_norm": 4.376220703125, | |
| "learning_rate": 5.576923076923077e-06, | |
| "loss": 0.2403, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 8.854700854700855, | |
| "grad_norm": 3.459076166152954, | |
| "learning_rate": 5.572649572649573e-06, | |
| "loss": 0.2064, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 8.863247863247864, | |
| "grad_norm": 6.312697410583496, | |
| "learning_rate": 5.568376068376069e-06, | |
| "loss": 0.5076, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 8.871794871794872, | |
| "grad_norm": 10.137848854064941, | |
| "learning_rate": 5.564102564102565e-06, | |
| "loss": 0.1649, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 8.88034188034188, | |
| "grad_norm": 6.605007171630859, | |
| "learning_rate": 5.559829059829061e-06, | |
| "loss": 0.4233, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 3.9786465167999268, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.1801, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 8.897435897435898, | |
| "grad_norm": 4.40491247177124, | |
| "learning_rate": 5.551282051282052e-06, | |
| "loss": 0.169, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 8.905982905982906, | |
| "grad_norm": 4.719818592071533, | |
| "learning_rate": 5.547008547008548e-06, | |
| "loss": 0.1454, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 8.914529914529915, | |
| "grad_norm": 2.384941577911377, | |
| "learning_rate": 5.5427350427350435e-06, | |
| "loss": 0.0723, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 8.923076923076923, | |
| "grad_norm": 3.258315324783325, | |
| "learning_rate": 5.538461538461539e-06, | |
| "loss": 0.1023, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 8.931623931623932, | |
| "grad_norm": 18.745052337646484, | |
| "learning_rate": 5.534188034188035e-06, | |
| "loss": 0.2673, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 8.94017094017094, | |
| "grad_norm": 3.788177967071533, | |
| "learning_rate": 5.52991452991453e-06, | |
| "loss": 0.3173, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 8.948717948717949, | |
| "grad_norm": 2.734895944595337, | |
| "learning_rate": 5.525641025641026e-06, | |
| "loss": 0.0834, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 8.957264957264957, | |
| "grad_norm": 4.158284664154053, | |
| "learning_rate": 5.521367521367522e-06, | |
| "loss": 0.3414, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 8.965811965811966, | |
| "grad_norm": 4.875148296356201, | |
| "learning_rate": 5.517094017094018e-06, | |
| "loss": 0.2729, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 8.974358974358974, | |
| "grad_norm": 5.2556352615356445, | |
| "learning_rate": 5.512820512820514e-06, | |
| "loss": 0.1422, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 8.982905982905983, | |
| "grad_norm": 3.817049980163574, | |
| "learning_rate": 5.50854700854701e-06, | |
| "loss": 0.2514, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 8.991452991452991, | |
| "grad_norm": 2.247227668762207, | |
| "learning_rate": 5.504273504273505e-06, | |
| "loss": 0.0703, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 34.36362838745117, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.7433, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.12675683200359344, | |
| "eval_runtime": 9.3141, | |
| "eval_samples_per_second": 50.032, | |
| "eval_steps_per_second": 6.334, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 9.008547008547009, | |
| "grad_norm": 5.314228057861328, | |
| "learning_rate": 5.495726495726497e-06, | |
| "loss": 0.2576, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 9.017094017094017, | |
| "grad_norm": 34.33782958984375, | |
| "learning_rate": 5.4914529914529925e-06, | |
| "loss": 0.3833, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 9.025641025641026, | |
| "grad_norm": 5.440598964691162, | |
| "learning_rate": 5.487179487179488e-06, | |
| "loss": 0.3898, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 9.034188034188034, | |
| "grad_norm": 3.561518907546997, | |
| "learning_rate": 5.482905982905984e-06, | |
| "loss": 0.2197, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 9.042735042735043, | |
| "grad_norm": 4.7679762840271, | |
| "learning_rate": 5.478632478632479e-06, | |
| "loss": 0.3885, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 9.051282051282051, | |
| "grad_norm": 4.694134712219238, | |
| "learning_rate": 5.474358974358975e-06, | |
| "loss": 0.2532, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 9.05982905982906, | |
| "grad_norm": 4.347025394439697, | |
| "learning_rate": 5.470085470085471e-06, | |
| "loss": 0.1949, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 9.068376068376068, | |
| "grad_norm": 4.064525127410889, | |
| "learning_rate": 5.465811965811966e-06, | |
| "loss": 0.1597, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 9.076923076923077, | |
| "grad_norm": 3.78560471534729, | |
| "learning_rate": 5.461538461538461e-06, | |
| "loss": 0.18, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 9.085470085470085, | |
| "grad_norm": 7.843743324279785, | |
| "learning_rate": 5.457264957264957e-06, | |
| "loss": 0.3146, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 9.094017094017094, | |
| "grad_norm": 8.152037620544434, | |
| "learning_rate": 5.452991452991453e-06, | |
| "loss": 0.3384, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 9.102564102564102, | |
| "grad_norm": 3.987872838973999, | |
| "learning_rate": 5.448717948717949e-06, | |
| "loss": 0.2071, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 9.11111111111111, | |
| "grad_norm": 3.478532552719116, | |
| "learning_rate": 5.444444444444445e-06, | |
| "loss": 0.1788, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 9.11965811965812, | |
| "grad_norm": 3.6598286628723145, | |
| "learning_rate": 5.44017094017094e-06, | |
| "loss": 0.2459, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 9.128205128205128, | |
| "grad_norm": 9.528829574584961, | |
| "learning_rate": 5.435897435897436e-06, | |
| "loss": 0.2046, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 9.136752136752136, | |
| "grad_norm": 3.3274407386779785, | |
| "learning_rate": 5.4316239316239316e-06, | |
| "loss": 0.1414, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 9.145299145299145, | |
| "grad_norm": 5.117324352264404, | |
| "learning_rate": 5.4273504273504275e-06, | |
| "loss": 0.3636, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 9.153846153846153, | |
| "grad_norm": 8.604976654052734, | |
| "learning_rate": 5.423076923076923e-06, | |
| "loss": 0.2723, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 9.162393162393162, | |
| "grad_norm": 72.67993927001953, | |
| "learning_rate": 5.418803418803419e-06, | |
| "loss": 0.5863, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 9.17094017094017, | |
| "grad_norm": 3.8609094619750977, | |
| "learning_rate": 5.414529914529914e-06, | |
| "loss": 0.1778, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 9.179487179487179, | |
| "grad_norm": 21.24209976196289, | |
| "learning_rate": 5.41025641025641e-06, | |
| "loss": 0.2062, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 9.188034188034187, | |
| "grad_norm": 5.552285194396973, | |
| "learning_rate": 5.405982905982906e-06, | |
| "loss": 0.4685, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 9.196581196581196, | |
| "grad_norm": 12.241254806518555, | |
| "learning_rate": 5.401709401709402e-06, | |
| "loss": 0.4309, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 9.205128205128204, | |
| "grad_norm": 3.6276049613952637, | |
| "learning_rate": 5.397435897435898e-06, | |
| "loss": 0.0924, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 9.213675213675213, | |
| "grad_norm": 10.98838996887207, | |
| "learning_rate": 5.393162393162394e-06, | |
| "loss": 0.7616, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 9.222222222222221, | |
| "grad_norm": 4.689146041870117, | |
| "learning_rate": 5.388888888888889e-06, | |
| "loss": 0.346, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 6.385439872741699, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.2945, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.239316239316238, | |
| "grad_norm": 2.4931023120880127, | |
| "learning_rate": 5.3803418803418806e-06, | |
| "loss": 0.172, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 9.247863247863247, | |
| "grad_norm": 3.797539472579956, | |
| "learning_rate": 5.3760683760683764e-06, | |
| "loss": 0.0927, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 9.256410256410255, | |
| "grad_norm": 2.7136716842651367, | |
| "learning_rate": 5.371794871794872e-06, | |
| "loss": 0.0932, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 9.264957264957266, | |
| "grad_norm": 5.207858085632324, | |
| "learning_rate": 5.367521367521367e-06, | |
| "loss": 0.1176, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 9.273504273504274, | |
| "grad_norm": 3.95009183883667, | |
| "learning_rate": 5.363247863247863e-06, | |
| "loss": 0.3045, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 9.282051282051283, | |
| "grad_norm": 1.9097685813903809, | |
| "learning_rate": 5.358974358974359e-06, | |
| "loss": 0.1793, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 9.290598290598291, | |
| "grad_norm": 3.205216407775879, | |
| "learning_rate": 5.354700854700855e-06, | |
| "loss": 0.1071, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 9.2991452991453, | |
| "grad_norm": 3.481822967529297, | |
| "learning_rate": 5.350427350427351e-06, | |
| "loss": 0.3885, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 9.307692307692308, | |
| "grad_norm": 11.802562713623047, | |
| "learning_rate": 5.346153846153847e-06, | |
| "loss": 0.1769, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 9.316239316239317, | |
| "grad_norm": 3.101505994796753, | |
| "learning_rate": 5.341880341880342e-06, | |
| "loss": 0.1265, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 9.324786324786325, | |
| "grad_norm": 5.163032054901123, | |
| "learning_rate": 5.337606837606838e-06, | |
| "loss": 0.4768, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 1.8217605352401733, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.053, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 9.341880341880342, | |
| "grad_norm": 2.6139562129974365, | |
| "learning_rate": 5.3290598290598295e-06, | |
| "loss": 0.0848, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 9.350427350427351, | |
| "grad_norm": 3.1172311305999756, | |
| "learning_rate": 5.3247863247863254e-06, | |
| "loss": 0.1076, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 9.35897435897436, | |
| "grad_norm": 5.907342433929443, | |
| "learning_rate": 5.320512820512821e-06, | |
| "loss": 0.1737, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 9.367521367521368, | |
| "grad_norm": 45.74967575073242, | |
| "learning_rate": 5.316239316239316e-06, | |
| "loss": 0.2455, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 9.376068376068377, | |
| "grad_norm": 3.1865549087524414, | |
| "learning_rate": 5.311965811965812e-06, | |
| "loss": 0.2236, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 9.384615384615385, | |
| "grad_norm": 4.028379917144775, | |
| "learning_rate": 5.307692307692308e-06, | |
| "loss": 0.1065, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.393162393162394, | |
| "grad_norm": 5.388605117797852, | |
| "learning_rate": 5.303418803418804e-06, | |
| "loss": 0.2967, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 9.401709401709402, | |
| "grad_norm": 3.661736249923706, | |
| "learning_rate": 5.2991452991453e-06, | |
| "loss": 0.1271, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.41025641025641, | |
| "grad_norm": 4.693649768829346, | |
| "learning_rate": 5.294871794871795e-06, | |
| "loss": 0.7891, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 9.418803418803419, | |
| "grad_norm": 14.75247573852539, | |
| "learning_rate": 5.290598290598291e-06, | |
| "loss": 0.707, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 9.427350427350428, | |
| "grad_norm": 5.123616695404053, | |
| "learning_rate": 5.286324786324787e-06, | |
| "loss": 0.2424, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 9.435897435897436, | |
| "grad_norm": 5.946259021759033, | |
| "learning_rate": 5.282051282051283e-06, | |
| "loss": 0.2558, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 9.444444444444445, | |
| "grad_norm": 3.3757872581481934, | |
| "learning_rate": 5.2777777777777785e-06, | |
| "loss": 0.072, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 9.452991452991453, | |
| "grad_norm": 4.639676094055176, | |
| "learning_rate": 5.2735042735042744e-06, | |
| "loss": 0.1483, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 9.461538461538462, | |
| "grad_norm": 5.552156925201416, | |
| "learning_rate": 5.2692307692307695e-06, | |
| "loss": 0.341, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 9.47008547008547, | |
| "grad_norm": 10.601661682128906, | |
| "learning_rate": 5.264957264957265e-06, | |
| "loss": 0.5964, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 9.478632478632479, | |
| "grad_norm": 4.391530513763428, | |
| "learning_rate": 5.260683760683761e-06, | |
| "loss": 0.2346, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 9.487179487179487, | |
| "grad_norm": 3.150240659713745, | |
| "learning_rate": 5.256410256410257e-06, | |
| "loss": 0.1, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 9.495726495726496, | |
| "grad_norm": 5.60894775390625, | |
| "learning_rate": 5.252136752136753e-06, | |
| "loss": 0.397, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 9.504273504273504, | |
| "grad_norm": 9.21768856048584, | |
| "learning_rate": 5.247863247863249e-06, | |
| "loss": 0.2292, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 9.512820512820513, | |
| "grad_norm": 8.351348876953125, | |
| "learning_rate": 5.243589743589744e-06, | |
| "loss": 0.3129, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 9.521367521367521, | |
| "grad_norm": 3.0813419818878174, | |
| "learning_rate": 5.23931623931624e-06, | |
| "loss": 0.2539, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 9.52991452991453, | |
| "grad_norm": 5.553039073944092, | |
| "learning_rate": 5.235042735042736e-06, | |
| "loss": 0.1121, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 9.538461538461538, | |
| "grad_norm": 3.973057746887207, | |
| "learning_rate": 5.230769230769232e-06, | |
| "loss": 0.4928, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 9.547008547008547, | |
| "grad_norm": 4.753414630889893, | |
| "learning_rate": 5.2264957264957275e-06, | |
| "loss": 0.2247, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 9.555555555555555, | |
| "grad_norm": 7.344094753265381, | |
| "learning_rate": 5.2222222222222226e-06, | |
| "loss": 0.1405, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 9.564102564102564, | |
| "grad_norm": 47.83219528198242, | |
| "learning_rate": 5.2179487179487185e-06, | |
| "loss": 0.3108, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 9.572649572649572, | |
| "grad_norm": 2.31591796875, | |
| "learning_rate": 5.213675213675214e-06, | |
| "loss": 0.1019, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.581196581196581, | |
| "grad_norm": 3.871413230895996, | |
| "learning_rate": 5.20940170940171e-06, | |
| "loss": 0.2562, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 9.58974358974359, | |
| "grad_norm": 2.1789255142211914, | |
| "learning_rate": 5.205128205128206e-06, | |
| "loss": 0.0571, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 9.598290598290598, | |
| "grad_norm": 4.119174957275391, | |
| "learning_rate": 5.200854700854702e-06, | |
| "loss": 0.2799, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 9.606837606837606, | |
| "grad_norm": 7.873704433441162, | |
| "learning_rate": 5.196581196581197e-06, | |
| "loss": 0.2154, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 9.615384615384615, | |
| "grad_norm": 3.386780023574829, | |
| "learning_rate": 5.192307692307693e-06, | |
| "loss": 0.1607, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 9.623931623931623, | |
| "grad_norm": 3.3607964515686035, | |
| "learning_rate": 5.188034188034189e-06, | |
| "loss": 0.22, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 9.632478632478632, | |
| "grad_norm": 10.655082702636719, | |
| "learning_rate": 5.183760683760685e-06, | |
| "loss": 0.2102, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 9.64102564102564, | |
| "grad_norm": 5.550488471984863, | |
| "learning_rate": 5.179487179487181e-06, | |
| "loss": 0.347, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 9.649572649572649, | |
| "grad_norm": 4.184569835662842, | |
| "learning_rate": 5.1752136752136765e-06, | |
| "loss": 0.183, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 9.658119658119658, | |
| "grad_norm": 4.892969131469727, | |
| "learning_rate": 5.1709401709401716e-06, | |
| "loss": 0.2896, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 9.666666666666666, | |
| "grad_norm": 5.926670074462891, | |
| "learning_rate": 5.1666666666666675e-06, | |
| "loss": 0.3321, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 9.675213675213675, | |
| "grad_norm": 11.719461441040039, | |
| "learning_rate": 5.162393162393163e-06, | |
| "loss": 0.4055, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 9.683760683760683, | |
| "grad_norm": 3.5666840076446533, | |
| "learning_rate": 5.158119658119659e-06, | |
| "loss": 0.2318, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 9.692307692307692, | |
| "grad_norm": 6.800848484039307, | |
| "learning_rate": 5.1538461538461534e-06, | |
| "loss": 0.1202, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 9.7008547008547, | |
| "grad_norm": 4.50139856338501, | |
| "learning_rate": 5.149572649572649e-06, | |
| "loss": 0.1914, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 9.709401709401709, | |
| "grad_norm": 2.599607467651367, | |
| "learning_rate": 5.145299145299145e-06, | |
| "loss": 0.0833, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 9.717948717948717, | |
| "grad_norm": 6.084483623504639, | |
| "learning_rate": 5.141025641025641e-06, | |
| "loss": 0.0907, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 9.726495726495726, | |
| "grad_norm": 4.542915344238281, | |
| "learning_rate": 5.136752136752137e-06, | |
| "loss": 0.4554, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 9.735042735042736, | |
| "grad_norm": 3.871166229248047, | |
| "learning_rate": 5.132478632478632e-06, | |
| "loss": 0.3037, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 9.743589743589745, | |
| "grad_norm": 5.121057033538818, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.1751, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 9.752136752136753, | |
| "grad_norm": 3.7517125606536865, | |
| "learning_rate": 5.123931623931624e-06, | |
| "loss": 0.3144, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 9.760683760683762, | |
| "grad_norm": 1.7604278326034546, | |
| "learning_rate": 5.11965811965812e-06, | |
| "loss": 0.0649, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 9.76923076923077, | |
| "grad_norm": 13.68947982788086, | |
| "learning_rate": 5.115384615384616e-06, | |
| "loss": 0.2184, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 9.777777777777779, | |
| "grad_norm": 5.716836452484131, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 0.1876, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 9.786324786324787, | |
| "grad_norm": 8.21943187713623, | |
| "learning_rate": 5.1068376068376065e-06, | |
| "loss": 0.349, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 9.794871794871796, | |
| "grad_norm": 5.270402908325195, | |
| "learning_rate": 5.1025641025641024e-06, | |
| "loss": 0.4442, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 9.803418803418804, | |
| "grad_norm": 2.3825948238372803, | |
| "learning_rate": 5.098290598290598e-06, | |
| "loss": 0.2237, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 9.811965811965813, | |
| "grad_norm": 11.812047958374023, | |
| "learning_rate": 5.094017094017094e-06, | |
| "loss": 0.5122, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 9.820512820512821, | |
| "grad_norm": 9.14202880859375, | |
| "learning_rate": 5.08974358974359e-06, | |
| "loss": 0.3407, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 9.82905982905983, | |
| "grad_norm": 5.273305892944336, | |
| "learning_rate": 5.085470085470086e-06, | |
| "loss": 0.1702, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 9.837606837606838, | |
| "grad_norm": 2.995126485824585, | |
| "learning_rate": 5.081196581196581e-06, | |
| "loss": 0.228, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 9.846153846153847, | |
| "grad_norm": 4.077675819396973, | |
| "learning_rate": 5.076923076923077e-06, | |
| "loss": 0.4022, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 9.854700854700855, | |
| "grad_norm": 2.1732425689697266, | |
| "learning_rate": 5.072649572649573e-06, | |
| "loss": 0.1178, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 9.863247863247864, | |
| "grad_norm": 2.905172109603882, | |
| "learning_rate": 5.068376068376069e-06, | |
| "loss": 0.1718, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 9.871794871794872, | |
| "grad_norm": 2.702521324157715, | |
| "learning_rate": 5.064102564102565e-06, | |
| "loss": 0.1488, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 9.88034188034188, | |
| "grad_norm": 2.414088487625122, | |
| "learning_rate": 5.05982905982906e-06, | |
| "loss": 0.1034, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 9.88888888888889, | |
| "grad_norm": 2.618173360824585, | |
| "learning_rate": 5.0555555555555555e-06, | |
| "loss": 0.0783, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 9.897435897435898, | |
| "grad_norm": 5.002628803253174, | |
| "learning_rate": 5.051282051282051e-06, | |
| "loss": 0.1195, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 9.905982905982906, | |
| "grad_norm": 2.84708833694458, | |
| "learning_rate": 5.047008547008547e-06, | |
| "loss": 0.0906, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 9.914529914529915, | |
| "grad_norm": 5.564020156860352, | |
| "learning_rate": 5.042735042735043e-06, | |
| "loss": 0.2037, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 9.923076923076923, | |
| "grad_norm": 3.7763166427612305, | |
| "learning_rate": 5.038461538461539e-06, | |
| "loss": 0.2067, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 9.931623931623932, | |
| "grad_norm": 2.67268705368042, | |
| "learning_rate": 5.034188034188034e-06, | |
| "loss": 0.0557, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 9.94017094017094, | |
| "grad_norm": 2.4144680500030518, | |
| "learning_rate": 5.02991452991453e-06, | |
| "loss": 0.194, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 9.948717948717949, | |
| "grad_norm": 2.0716731548309326, | |
| "learning_rate": 5.025641025641026e-06, | |
| "loss": 0.1253, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 9.957264957264957, | |
| "grad_norm": 13.20478630065918, | |
| "learning_rate": 5.021367521367522e-06, | |
| "loss": 0.268, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 9.965811965811966, | |
| "grad_norm": 2.093698263168335, | |
| "learning_rate": 5.017094017094018e-06, | |
| "loss": 0.0738, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 9.974358974358974, | |
| "grad_norm": 2.2758119106292725, | |
| "learning_rate": 5.012820512820514e-06, | |
| "loss": 0.0804, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 9.982905982905983, | |
| "grad_norm": 21.843395233154297, | |
| "learning_rate": 5.008547008547009e-06, | |
| "loss": 0.3298, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 9.991452991452991, | |
| "grad_norm": 3.0435073375701904, | |
| "learning_rate": 5.0042735042735045e-06, | |
| "loss": 0.1318, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 8.449163436889648, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1725, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.10285739600658417, | |
| "eval_runtime": 9.2384, | |
| "eval_samples_per_second": 50.441, | |
| "eval_steps_per_second": 6.386, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 10.008547008547009, | |
| "grad_norm": 4.151456356048584, | |
| "learning_rate": 4.995726495726496e-06, | |
| "loss": 0.3336, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 10.017094017094017, | |
| "grad_norm": 2.38647723197937, | |
| "learning_rate": 4.991452991452992e-06, | |
| "loss": 0.1138, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 10.025641025641026, | |
| "grad_norm": 4.44817590713501, | |
| "learning_rate": 4.987179487179487e-06, | |
| "loss": 0.0954, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 10.034188034188034, | |
| "grad_norm": 2.6213347911834717, | |
| "learning_rate": 4.982905982905983e-06, | |
| "loss": 0.0695, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 10.042735042735043, | |
| "grad_norm": 4.664891719818115, | |
| "learning_rate": 4.978632478632479e-06, | |
| "loss": 0.1067, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 10.051282051282051, | |
| "grad_norm": 1.7059048414230347, | |
| "learning_rate": 4.974358974358975e-06, | |
| "loss": 0.0321, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 10.05982905982906, | |
| "grad_norm": 5.123709678649902, | |
| "learning_rate": 4.970085470085471e-06, | |
| "loss": 0.2117, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 10.068376068376068, | |
| "grad_norm": 2.2717695236206055, | |
| "learning_rate": 4.965811965811967e-06, | |
| "loss": 0.2187, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 10.076923076923077, | |
| "grad_norm": 4.669886112213135, | |
| "learning_rate": 4.961538461538462e-06, | |
| "loss": 0.4615, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 10.085470085470085, | |
| "grad_norm": 18.739727020263672, | |
| "learning_rate": 4.957264957264958e-06, | |
| "loss": 0.3431, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 10.094017094017094, | |
| "grad_norm": 7.798559188842773, | |
| "learning_rate": 4.9529914529914535e-06, | |
| "loss": 0.2483, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 10.102564102564102, | |
| "grad_norm": 22.59453773498535, | |
| "learning_rate": 4.948717948717949e-06, | |
| "loss": 0.15, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 10.11111111111111, | |
| "grad_norm": 2.5734364986419678, | |
| "learning_rate": 4.944444444444445e-06, | |
| "loss": 0.0465, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 10.11965811965812, | |
| "grad_norm": 3.1944875717163086, | |
| "learning_rate": 4.940170940170941e-06, | |
| "loss": 0.1429, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 10.128205128205128, | |
| "grad_norm": 1.6943906545639038, | |
| "learning_rate": 4.935897435897436e-06, | |
| "loss": 0.0685, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 10.136752136752136, | |
| "grad_norm": 4.497282981872559, | |
| "learning_rate": 4.931623931623932e-06, | |
| "loss": 0.2113, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 10.145299145299145, | |
| "grad_norm": 2.9377167224884033, | |
| "learning_rate": 4.927350427350428e-06, | |
| "loss": 0.1352, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 10.153846153846153, | |
| "grad_norm": 8.528215408325195, | |
| "learning_rate": 4.923076923076924e-06, | |
| "loss": 0.3268, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 10.162393162393162, | |
| "grad_norm": 2.143850803375244, | |
| "learning_rate": 4.918803418803419e-06, | |
| "loss": 0.0923, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 10.17094017094017, | |
| "grad_norm": 3.921250343322754, | |
| "learning_rate": 4.914529914529915e-06, | |
| "loss": 0.1451, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 10.179487179487179, | |
| "grad_norm": 10.713285446166992, | |
| "learning_rate": 4.910256410256411e-06, | |
| "loss": 0.17, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 10.188034188034187, | |
| "grad_norm": 2.450204849243164, | |
| "learning_rate": 4.905982905982906e-06, | |
| "loss": 0.0765, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 10.196581196581196, | |
| "grad_norm": 4.750647068023682, | |
| "learning_rate": 4.901709401709402e-06, | |
| "loss": 0.2829, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 10.205128205128204, | |
| "grad_norm": 12.714463233947754, | |
| "learning_rate": 4.8974358974358975e-06, | |
| "loss": 0.6767, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 10.213675213675213, | |
| "grad_norm": 6.759951591491699, | |
| "learning_rate": 4.8931623931623934e-06, | |
| "loss": 0.2369, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 10.222222222222221, | |
| "grad_norm": 8.592784881591797, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 0.4203, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 10.23076923076923, | |
| "grad_norm": 5.04047155380249, | |
| "learning_rate": 4.884615384615385e-06, | |
| "loss": 0.1023, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 10.239316239316238, | |
| "grad_norm": 38.112152099609375, | |
| "learning_rate": 4.88034188034188e-06, | |
| "loss": 0.4686, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 10.247863247863247, | |
| "grad_norm": 6.751104354858398, | |
| "learning_rate": 4.876068376068376e-06, | |
| "loss": 0.085, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 10.256410256410255, | |
| "grad_norm": 4.3117594718933105, | |
| "learning_rate": 4.871794871794872e-06, | |
| "loss": 0.1504, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.264957264957266, | |
| "grad_norm": 2.251265287399292, | |
| "learning_rate": 4.867521367521368e-06, | |
| "loss": 0.1664, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 10.273504273504274, | |
| "grad_norm": 2.1650373935699463, | |
| "learning_rate": 4.863247863247864e-06, | |
| "loss": 0.0959, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 10.282051282051283, | |
| "grad_norm": 2.5863089561462402, | |
| "learning_rate": 4.85897435897436e-06, | |
| "loss": 0.1148, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 10.290598290598291, | |
| "grad_norm": 1.974357008934021, | |
| "learning_rate": 4.854700854700855e-06, | |
| "loss": 0.0663, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 10.2991452991453, | |
| "grad_norm": 2.3226940631866455, | |
| "learning_rate": 4.850427350427351e-06, | |
| "loss": 0.1363, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 10.307692307692308, | |
| "grad_norm": 4.034085750579834, | |
| "learning_rate": 4.8461538461538465e-06, | |
| "loss": 0.3473, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 10.316239316239317, | |
| "grad_norm": 2.492307186126709, | |
| "learning_rate": 4.8418803418803424e-06, | |
| "loss": 0.1742, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 10.324786324786325, | |
| "grad_norm": 2.886432409286499, | |
| "learning_rate": 4.837606837606838e-06, | |
| "loss": 0.1382, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 10.333333333333334, | |
| "grad_norm": 3.6314749717712402, | |
| "learning_rate": 4.833333333333333e-06, | |
| "loss": 0.1556, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 10.341880341880342, | |
| "grad_norm": 2.2757928371429443, | |
| "learning_rate": 4.829059829059829e-06, | |
| "loss": 0.0434, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 10.350427350427351, | |
| "grad_norm": 3.4152615070343018, | |
| "learning_rate": 4.824786324786325e-06, | |
| "loss": 0.2903, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 10.35897435897436, | |
| "grad_norm": 3.873960256576538, | |
| "learning_rate": 4.820512820512821e-06, | |
| "loss": 0.2611, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 10.367521367521368, | |
| "grad_norm": 4.2241291999816895, | |
| "learning_rate": 4.816239316239317e-06, | |
| "loss": 0.0954, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 10.376068376068377, | |
| "grad_norm": 5.454725742340088, | |
| "learning_rate": 4.811965811965813e-06, | |
| "loss": 0.1361, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 10.384615384615385, | |
| "grad_norm": 3.482558012008667, | |
| "learning_rate": 4.807692307692308e-06, | |
| "loss": 0.0861, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 10.393162393162394, | |
| "grad_norm": 2.301254987716675, | |
| "learning_rate": 4.803418803418804e-06, | |
| "loss": 0.1571, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 10.401709401709402, | |
| "grad_norm": 6.0665602684021, | |
| "learning_rate": 4.7991452991453e-06, | |
| "loss": 0.5323, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 10.41025641025641, | |
| "grad_norm": 3.6052770614624023, | |
| "learning_rate": 4.7948717948717955e-06, | |
| "loss": 0.3789, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 10.418803418803419, | |
| "grad_norm": 3.9434757232666016, | |
| "learning_rate": 4.790598290598291e-06, | |
| "loss": 0.0605, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 10.427350427350428, | |
| "grad_norm": 5.260069847106934, | |
| "learning_rate": 4.786324786324787e-06, | |
| "loss": 0.3163, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.435897435897436, | |
| "grad_norm": 5.219394207000732, | |
| "learning_rate": 4.782051282051282e-06, | |
| "loss": 0.4339, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 10.444444444444445, | |
| "grad_norm": 2.7057230472564697, | |
| "learning_rate": 4.777777777777778e-06, | |
| "loss": 0.0787, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 10.452991452991453, | |
| "grad_norm": 11.005247116088867, | |
| "learning_rate": 4.773504273504274e-06, | |
| "loss": 0.255, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 10.461538461538462, | |
| "grad_norm": 1.7238801717758179, | |
| "learning_rate": 4.76923076923077e-06, | |
| "loss": 0.0605, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 10.47008547008547, | |
| "grad_norm": 6.509312629699707, | |
| "learning_rate": 4.764957264957265e-06, | |
| "loss": 0.2899, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 10.478632478632479, | |
| "grad_norm": 7.1476359367370605, | |
| "learning_rate": 4.760683760683761e-06, | |
| "loss": 0.336, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 10.487179487179487, | |
| "grad_norm": 15.92902660369873, | |
| "learning_rate": 4.756410256410257e-06, | |
| "loss": 0.4864, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 10.495726495726496, | |
| "grad_norm": 5.545684337615967, | |
| "learning_rate": 4.752136752136752e-06, | |
| "loss": 0.4741, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 10.504273504273504, | |
| "grad_norm": 3.2521066665649414, | |
| "learning_rate": 4.747863247863248e-06, | |
| "loss": 0.0894, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 10.512820512820513, | |
| "grad_norm": 2.696866512298584, | |
| "learning_rate": 4.743589743589744e-06, | |
| "loss": 0.111, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 10.521367521367521, | |
| "grad_norm": 1.8362340927124023, | |
| "learning_rate": 4.7393162393162396e-06, | |
| "loss": 0.0579, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 10.52991452991453, | |
| "grad_norm": 2.96872878074646, | |
| "learning_rate": 4.7350427350427355e-06, | |
| "loss": 0.0781, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 10.538461538461538, | |
| "grad_norm": 1.5503445863723755, | |
| "learning_rate": 4.730769230769231e-06, | |
| "loss": 0.0451, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 10.547008547008547, | |
| "grad_norm": 3.9600377082824707, | |
| "learning_rate": 4.726495726495726e-06, | |
| "loss": 0.1721, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 10.555555555555555, | |
| "grad_norm": 3.3868823051452637, | |
| "learning_rate": 4.722222222222222e-06, | |
| "loss": 0.1803, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 10.564102564102564, | |
| "grad_norm": 2.528111219406128, | |
| "learning_rate": 4.717948717948718e-06, | |
| "loss": 0.238, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 10.572649572649572, | |
| "grad_norm": 6.960350036621094, | |
| "learning_rate": 4.713675213675214e-06, | |
| "loss": 0.4353, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 10.581196581196581, | |
| "grad_norm": 2.3169686794281006, | |
| "learning_rate": 4.70940170940171e-06, | |
| "loss": 0.1891, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 10.58974358974359, | |
| "grad_norm": 2.021212577819824, | |
| "learning_rate": 4.705128205128206e-06, | |
| "loss": 0.0865, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 10.598290598290598, | |
| "grad_norm": 2.445462942123413, | |
| "learning_rate": 4.700854700854701e-06, | |
| "loss": 0.0973, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 10.606837606837606, | |
| "grad_norm": 3.4490067958831787, | |
| "learning_rate": 4.696581196581197e-06, | |
| "loss": 0.1419, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 10.615384615384615, | |
| "grad_norm": 3.2859914302825928, | |
| "learning_rate": 4.692307692307693e-06, | |
| "loss": 0.1587, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 10.623931623931623, | |
| "grad_norm": 4.754831790924072, | |
| "learning_rate": 4.6880341880341886e-06, | |
| "loss": 0.2537, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 10.632478632478632, | |
| "grad_norm": 3.220867156982422, | |
| "learning_rate": 4.6837606837606844e-06, | |
| "loss": 0.0941, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 10.64102564102564, | |
| "grad_norm": 5.699328422546387, | |
| "learning_rate": 4.6794871794871795e-06, | |
| "loss": 0.255, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 10.649572649572649, | |
| "grad_norm": 1.5174522399902344, | |
| "learning_rate": 4.675213675213675e-06, | |
| "loss": 0.048, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 10.658119658119658, | |
| "grad_norm": 2.4277050495147705, | |
| "learning_rate": 4.670940170940171e-06, | |
| "loss": 0.1127, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 10.666666666666666, | |
| "grad_norm": 2.079031229019165, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.1038, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 10.675213675213675, | |
| "grad_norm": 953.4605102539062, | |
| "learning_rate": 4.662393162393163e-06, | |
| "loss": 1.1892, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 10.683760683760683, | |
| "grad_norm": 9.190105438232422, | |
| "learning_rate": 4.658119658119659e-06, | |
| "loss": 0.3541, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 10.692307692307692, | |
| "grad_norm": 2.3222947120666504, | |
| "learning_rate": 4.653846153846154e-06, | |
| "loss": 0.0842, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 10.7008547008547, | |
| "grad_norm": 2.2312700748443604, | |
| "learning_rate": 4.64957264957265e-06, | |
| "loss": 0.088, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 10.709401709401709, | |
| "grad_norm": 3.987630844116211, | |
| "learning_rate": 4.645299145299146e-06, | |
| "loss": 0.1667, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 10.717948717948717, | |
| "grad_norm": 5.108981609344482, | |
| "learning_rate": 4.641025641025642e-06, | |
| "loss": 0.4291, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 10.726495726495726, | |
| "grad_norm": 2.8597464561462402, | |
| "learning_rate": 4.6367521367521375e-06, | |
| "loss": 0.0564, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 10.735042735042736, | |
| "grad_norm": 2.3642940521240234, | |
| "learning_rate": 4.6324786324786334e-06, | |
| "loss": 0.0909, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 10.743589743589745, | |
| "grad_norm": 1.5703462362289429, | |
| "learning_rate": 4.6282051282051285e-06, | |
| "loss": 0.0395, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 10.752136752136753, | |
| "grad_norm": 2.952786922454834, | |
| "learning_rate": 4.623931623931624e-06, | |
| "loss": 0.1824, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 10.760683760683762, | |
| "grad_norm": 2.9027185440063477, | |
| "learning_rate": 4.61965811965812e-06, | |
| "loss": 0.0765, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 10.76923076923077, | |
| "grad_norm": 2.4386038780212402, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.2761, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 10.777777777777779, | |
| "grad_norm": 7.146468639373779, | |
| "learning_rate": 4.611111111111112e-06, | |
| "loss": 0.4427, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 10.786324786324787, | |
| "grad_norm": 2.002096652984619, | |
| "learning_rate": 4.606837606837607e-06, | |
| "loss": 0.0879, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 10.794871794871796, | |
| "grad_norm": 6.504697322845459, | |
| "learning_rate": 4.602564102564103e-06, | |
| "loss": 0.1805, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 10.803418803418804, | |
| "grad_norm": 9.748340606689453, | |
| "learning_rate": 4.598290598290598e-06, | |
| "loss": 0.5813, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 10.811965811965813, | |
| "grad_norm": 3.67153000831604, | |
| "learning_rate": 4.594017094017094e-06, | |
| "loss": 0.4175, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 10.820512820512821, | |
| "grad_norm": 9.109044075012207, | |
| "learning_rate": 4.58974358974359e-06, | |
| "loss": 0.4505, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 10.82905982905983, | |
| "grad_norm": 5.419683933258057, | |
| "learning_rate": 4.585470085470086e-06, | |
| "loss": 0.2316, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 10.837606837606838, | |
| "grad_norm": 2.901182174682617, | |
| "learning_rate": 4.581196581196582e-06, | |
| "loss": 0.0583, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 10.846153846153847, | |
| "grad_norm": 4.579897403717041, | |
| "learning_rate": 4.5769230769230775e-06, | |
| "loss": 0.0536, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 10.854700854700855, | |
| "grad_norm": 4.232446670532227, | |
| "learning_rate": 4.5726495726495725e-06, | |
| "loss": 0.17, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 10.863247863247864, | |
| "grad_norm": 8.059329986572266, | |
| "learning_rate": 4.568376068376068e-06, | |
| "loss": 0.256, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 10.871794871794872, | |
| "grad_norm": 1.5736984014511108, | |
| "learning_rate": 4.564102564102564e-06, | |
| "loss": 0.058, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 10.88034188034188, | |
| "grad_norm": 5.397885799407959, | |
| "learning_rate": 4.55982905982906e-06, | |
| "loss": 0.1299, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 10.88888888888889, | |
| "grad_norm": 3.9831533432006836, | |
| "learning_rate": 4.555555555555556e-06, | |
| "loss": 0.1762, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 10.897435897435898, | |
| "grad_norm": 2.170370101928711, | |
| "learning_rate": 4.551282051282052e-06, | |
| "loss": 0.1355, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 10.905982905982906, | |
| "grad_norm": 5.151463508605957, | |
| "learning_rate": 4.547008547008547e-06, | |
| "loss": 0.3151, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 10.914529914529915, | |
| "grad_norm": 2.215559482574463, | |
| "learning_rate": 4.542735042735043e-06, | |
| "loss": 0.1054, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 10.923076923076923, | |
| "grad_norm": 3.62188458442688, | |
| "learning_rate": 4.538461538461539e-06, | |
| "loss": 0.3839, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 10.931623931623932, | |
| "grad_norm": 1.8855514526367188, | |
| "learning_rate": 4.534188034188035e-06, | |
| "loss": 0.0639, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 10.94017094017094, | |
| "grad_norm": 3.0260651111602783, | |
| "learning_rate": 4.5299145299145306e-06, | |
| "loss": 0.1216, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 10.948717948717949, | |
| "grad_norm": 13.30820083618164, | |
| "learning_rate": 4.525641025641026e-06, | |
| "loss": 0.3337, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 10.957264957264957, | |
| "grad_norm": 4.356720447540283, | |
| "learning_rate": 4.5213675213675215e-06, | |
| "loss": 0.2692, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 10.965811965811966, | |
| "grad_norm": 2.077742576599121, | |
| "learning_rate": 4.517094017094017e-06, | |
| "loss": 0.1181, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 10.974358974358974, | |
| "grad_norm": 6.6224284172058105, | |
| "learning_rate": 4.512820512820513e-06, | |
| "loss": 0.1526, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 10.982905982905983, | |
| "grad_norm": 4.072678565979004, | |
| "learning_rate": 4.508547008547009e-06, | |
| "loss": 0.1804, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 10.991452991452991, | |
| "grad_norm": 3.430922269821167, | |
| "learning_rate": 4.504273504273505e-06, | |
| "loss": 0.1316, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 1.6371959447860718, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0596, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.08654214441776276, | |
| "eval_runtime": 9.3013, | |
| "eval_samples_per_second": 50.1, | |
| "eval_steps_per_second": 6.343, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 11.008547008547009, | |
| "grad_norm": 5.072701454162598, | |
| "learning_rate": 4.495726495726496e-06, | |
| "loss": 0.2195, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 11.017094017094017, | |
| "grad_norm": 6.791895389556885, | |
| "learning_rate": 4.491452991452992e-06, | |
| "loss": 0.5354, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 11.025641025641026, | |
| "grad_norm": 12.475218772888184, | |
| "learning_rate": 4.487179487179488e-06, | |
| "loss": 0.1828, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 11.034188034188034, | |
| "grad_norm": 5.892624855041504, | |
| "learning_rate": 4.482905982905984e-06, | |
| "loss": 0.1617, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 11.042735042735043, | |
| "grad_norm": 1.742074728012085, | |
| "learning_rate": 4.4786324786324796e-06, | |
| "loss": 0.0508, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 11.051282051282051, | |
| "grad_norm": 2.389373302459717, | |
| "learning_rate": 4.474358974358975e-06, | |
| "loss": 0.1009, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 11.05982905982906, | |
| "grad_norm": 3.7152106761932373, | |
| "learning_rate": 4.4700854700854705e-06, | |
| "loss": 0.2157, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 11.068376068376068, | |
| "grad_norm": 7.217955112457275, | |
| "learning_rate": 4.465811965811966e-06, | |
| "loss": 0.2737, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 11.076923076923077, | |
| "grad_norm": 2.0971977710723877, | |
| "learning_rate": 4.461538461538462e-06, | |
| "loss": 0.1273, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 11.085470085470085, | |
| "grad_norm": 1.1616859436035156, | |
| "learning_rate": 4.457264957264958e-06, | |
| "loss": 0.0325, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 11.094017094017094, | |
| "grad_norm": 3.4287424087524414, | |
| "learning_rate": 4.452991452991453e-06, | |
| "loss": 0.1136, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 11.102564102564102, | |
| "grad_norm": 1.6207005977630615, | |
| "learning_rate": 4.448717948717949e-06, | |
| "loss": 0.0344, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 11.11111111111111, | |
| "grad_norm": 3.009976863861084, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1532, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.11965811965812, | |
| "grad_norm": 2.9768505096435547, | |
| "learning_rate": 4.44017094017094e-06, | |
| "loss": 0.0874, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 11.128205128205128, | |
| "grad_norm": 3.622715473175049, | |
| "learning_rate": 4.435897435897436e-06, | |
| "loss": 0.3132, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 11.136752136752136, | |
| "grad_norm": 3.5741326808929443, | |
| "learning_rate": 4.431623931623932e-06, | |
| "loss": 0.0914, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 11.145299145299145, | |
| "grad_norm": 7.436197280883789, | |
| "learning_rate": 4.427350427350428e-06, | |
| "loss": 0.329, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 11.153846153846153, | |
| "grad_norm": 2.390066146850586, | |
| "learning_rate": 4.423076923076924e-06, | |
| "loss": 0.0867, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 11.162393162393162, | |
| "grad_norm": 1.928227424621582, | |
| "learning_rate": 4.418803418803419e-06, | |
| "loss": 0.0294, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 11.17094017094017, | |
| "grad_norm": 4.40464448928833, | |
| "learning_rate": 4.4145299145299145e-06, | |
| "loss": 0.3704, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 11.179487179487179, | |
| "grad_norm": 22.183835983276367, | |
| "learning_rate": 4.4102564102564104e-06, | |
| "loss": 0.6011, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 11.188034188034187, | |
| "grad_norm": 2.496633768081665, | |
| "learning_rate": 4.405982905982906e-06, | |
| "loss": 0.0494, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 11.196581196581196, | |
| "grad_norm": 1.142687201499939, | |
| "learning_rate": 4.401709401709402e-06, | |
| "loss": 0.0292, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 11.205128205128204, | |
| "grad_norm": 2.0762455463409424, | |
| "learning_rate": 4.397435897435898e-06, | |
| "loss": 0.1123, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 11.213675213675213, | |
| "grad_norm": 1.5389565229415894, | |
| "learning_rate": 4.393162393162393e-06, | |
| "loss": 0.0316, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 11.222222222222221, | |
| "grad_norm": 4.252040386199951, | |
| "learning_rate": 4.388888888888889e-06, | |
| "loss": 0.0832, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 11.23076923076923, | |
| "grad_norm": 2.1999545097351074, | |
| "learning_rate": 4.384615384615385e-06, | |
| "loss": 0.1121, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 11.239316239316238, | |
| "grad_norm": 3.3256099224090576, | |
| "learning_rate": 4.380341880341881e-06, | |
| "loss": 0.1288, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 11.247863247863247, | |
| "grad_norm": 2.6664986610412598, | |
| "learning_rate": 4.376068376068377e-06, | |
| "loss": 0.1044, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 11.256410256410255, | |
| "grad_norm": 4.103114604949951, | |
| "learning_rate": 4.371794871794872e-06, | |
| "loss": 0.3115, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 11.264957264957266, | |
| "grad_norm": 2.717532157897949, | |
| "learning_rate": 4.367521367521368e-06, | |
| "loss": 0.1144, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 11.273504273504274, | |
| "grad_norm": 2.7918317317962646, | |
| "learning_rate": 4.3632478632478635e-06, | |
| "loss": 0.1205, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 11.282051282051283, | |
| "grad_norm": 2.439854383468628, | |
| "learning_rate": 4.358974358974359e-06, | |
| "loss": 0.05, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 11.290598290598291, | |
| "grad_norm": 1.3528865575790405, | |
| "learning_rate": 4.354700854700855e-06, | |
| "loss": 0.0437, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 11.2991452991453, | |
| "grad_norm": 3.3273401260375977, | |
| "learning_rate": 4.350427350427351e-06, | |
| "loss": 0.1417, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 11.307692307692308, | |
| "grad_norm": 4.022815704345703, | |
| "learning_rate": 4.346153846153846e-06, | |
| "loss": 0.0845, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 11.316239316239317, | |
| "grad_norm": 5.169338703155518, | |
| "learning_rate": 4.341880341880342e-06, | |
| "loss": 0.5235, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 11.324786324786325, | |
| "grad_norm": 1.8199687004089355, | |
| "learning_rate": 4.337606837606838e-06, | |
| "loss": 0.0399, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 11.333333333333334, | |
| "grad_norm": 3.3616087436676025, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 0.1428, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 11.341880341880342, | |
| "grad_norm": 14.056232452392578, | |
| "learning_rate": 4.32905982905983e-06, | |
| "loss": 0.2921, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 11.350427350427351, | |
| "grad_norm": 2.3905317783355713, | |
| "learning_rate": 4.324786324786326e-06, | |
| "loss": 0.0478, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 11.35897435897436, | |
| "grad_norm": 9.876815795898438, | |
| "learning_rate": 4.320512820512821e-06, | |
| "loss": 0.1926, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 11.367521367521368, | |
| "grad_norm": 1.3726049661636353, | |
| "learning_rate": 4.316239316239317e-06, | |
| "loss": 0.0416, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 11.376068376068377, | |
| "grad_norm": 3.0890841484069824, | |
| "learning_rate": 4.3119658119658125e-06, | |
| "loss": 0.0614, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 11.384615384615385, | |
| "grad_norm": 2.858560562133789, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 0.2068, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 11.393162393162394, | |
| "grad_norm": 4.6819658279418945, | |
| "learning_rate": 4.303418803418804e-06, | |
| "loss": 0.5773, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 11.401709401709402, | |
| "grad_norm": 1.741450548171997, | |
| "learning_rate": 4.299145299145299e-06, | |
| "loss": 0.0505, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 11.41025641025641, | |
| "grad_norm": 3.5882327556610107, | |
| "learning_rate": 4.294871794871795e-06, | |
| "loss": 0.1797, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 11.418803418803419, | |
| "grad_norm": 3.59714937210083, | |
| "learning_rate": 4.29059829059829e-06, | |
| "loss": 0.1531, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 11.427350427350428, | |
| "grad_norm": 3.619572877883911, | |
| "learning_rate": 4.286324786324786e-06, | |
| "loss": 0.1028, | |
| "step": 1337 | |
| }, | |
| { | |
| "epoch": 11.435897435897436, | |
| "grad_norm": 3.9230782985687256, | |
| "learning_rate": 4.282051282051282e-06, | |
| "loss": 0.2404, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 11.444444444444445, | |
| "grad_norm": 3.6987717151641846, | |
| "learning_rate": 4.277777777777778e-06, | |
| "loss": 0.1795, | |
| "step": 1339 | |
| }, | |
| { | |
| "epoch": 11.452991452991453, | |
| "grad_norm": 3.322707176208496, | |
| "learning_rate": 4.273504273504274e-06, | |
| "loss": 0.0968, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 11.461538461538462, | |
| "grad_norm": 1.2378501892089844, | |
| "learning_rate": 4.26923076923077e-06, | |
| "loss": 0.0387, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 11.47008547008547, | |
| "grad_norm": 2.6801578998565674, | |
| "learning_rate": 4.264957264957265e-06, | |
| "loss": 0.0475, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.478632478632479, | |
| "grad_norm": 2.2003352642059326, | |
| "learning_rate": 4.260683760683761e-06, | |
| "loss": 0.0505, | |
| "step": 1343 | |
| }, | |
| { | |
| "epoch": 11.487179487179487, | |
| "grad_norm": 1.701341152191162, | |
| "learning_rate": 4.2564102564102566e-06, | |
| "loss": 0.064, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 11.495726495726496, | |
| "grad_norm": 9.939803123474121, | |
| "learning_rate": 4.2521367521367524e-06, | |
| "loss": 0.461, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 11.504273504273504, | |
| "grad_norm": 3.2999305725097656, | |
| "learning_rate": 4.247863247863248e-06, | |
| "loss": 0.1653, | |
| "step": 1346 | |
| }, | |
| { | |
| "epoch": 11.512820512820513, | |
| "grad_norm": 3.9968252182006836, | |
| "learning_rate": 4.243589743589744e-06, | |
| "loss": 0.123, | |
| "step": 1347 | |
| }, | |
| { | |
| "epoch": 11.521367521367521, | |
| "grad_norm": 2.846968173980713, | |
| "learning_rate": 4.239316239316239e-06, | |
| "loss": 0.1161, | |
| "step": 1348 | |
| }, | |
| { | |
| "epoch": 11.52991452991453, | |
| "grad_norm": 4.328092575073242, | |
| "learning_rate": 4.235042735042735e-06, | |
| "loss": 0.065, | |
| "step": 1349 | |
| }, | |
| { | |
| "epoch": 11.538461538461538, | |
| "grad_norm": 3.649003267288208, | |
| "learning_rate": 4.230769230769231e-06, | |
| "loss": 0.1919, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 11.547008547008547, | |
| "grad_norm": 4.094634056091309, | |
| "learning_rate": 4.226495726495727e-06, | |
| "loss": 0.1728, | |
| "step": 1351 | |
| }, | |
| { | |
| "epoch": 11.555555555555555, | |
| "grad_norm": 2.3904240131378174, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 0.105, | |
| "step": 1352 | |
| }, | |
| { | |
| "epoch": 11.564102564102564, | |
| "grad_norm": 1.8493746519088745, | |
| "learning_rate": 4.217948717948718e-06, | |
| "loss": 0.0373, | |
| "step": 1353 | |
| }, | |
| { | |
| "epoch": 11.572649572649572, | |
| "grad_norm": 4.690928936004639, | |
| "learning_rate": 4.213675213675214e-06, | |
| "loss": 0.3405, | |
| "step": 1354 | |
| }, | |
| { | |
| "epoch": 11.581196581196581, | |
| "grad_norm": 6.808948516845703, | |
| "learning_rate": 4.20940170940171e-06, | |
| "loss": 0.1308, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 11.58974358974359, | |
| "grad_norm": 6.060946464538574, | |
| "learning_rate": 4.2051282051282055e-06, | |
| "loss": 0.1494, | |
| "step": 1356 | |
| }, | |
| { | |
| "epoch": 11.598290598290598, | |
| "grad_norm": 1.5923279523849487, | |
| "learning_rate": 4.2008547008547014e-06, | |
| "loss": 0.044, | |
| "step": 1357 | |
| }, | |
| { | |
| "epoch": 11.606837606837606, | |
| "grad_norm": 1.7796354293823242, | |
| "learning_rate": 4.196581196581197e-06, | |
| "loss": 0.0558, | |
| "step": 1358 | |
| }, | |
| { | |
| "epoch": 11.615384615384615, | |
| "grad_norm": 1.2209490537643433, | |
| "learning_rate": 4.192307692307692e-06, | |
| "loss": 0.0492, | |
| "step": 1359 | |
| }, | |
| { | |
| "epoch": 11.623931623931623, | |
| "grad_norm": 4.0859880447387695, | |
| "learning_rate": 4.188034188034188e-06, | |
| "loss": 0.0759, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 11.632478632478632, | |
| "grad_norm": 3.5021755695343018, | |
| "learning_rate": 4.183760683760684e-06, | |
| "loss": 0.1263, | |
| "step": 1361 | |
| }, | |
| { | |
| "epoch": 11.64102564102564, | |
| "grad_norm": 2.5915517807006836, | |
| "learning_rate": 4.17948717948718e-06, | |
| "loss": 0.1949, | |
| "step": 1362 | |
| }, | |
| { | |
| "epoch": 11.649572649572649, | |
| "grad_norm": 2.8024656772613525, | |
| "learning_rate": 4.175213675213676e-06, | |
| "loss": 0.2325, | |
| "step": 1363 | |
| }, | |
| { | |
| "epoch": 11.658119658119658, | |
| "grad_norm": 5.795172691345215, | |
| "learning_rate": 4.170940170940172e-06, | |
| "loss": 0.3253, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 11.666666666666666, | |
| "grad_norm": 5.056031227111816, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.102, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 11.675213675213675, | |
| "grad_norm": 6.092950820922852, | |
| "learning_rate": 4.162393162393163e-06, | |
| "loss": 0.1938, | |
| "step": 1366 | |
| }, | |
| { | |
| "epoch": 11.683760683760683, | |
| "grad_norm": 4.44755744934082, | |
| "learning_rate": 4.158119658119659e-06, | |
| "loss": 0.1588, | |
| "step": 1367 | |
| }, | |
| { | |
| "epoch": 11.692307692307692, | |
| "grad_norm": 171.19509887695312, | |
| "learning_rate": 4.1538461538461545e-06, | |
| "loss": 0.3077, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 11.7008547008547, | |
| "grad_norm": 13.992602348327637, | |
| "learning_rate": 4.1495726495726504e-06, | |
| "loss": 0.4401, | |
| "step": 1369 | |
| }, | |
| { | |
| "epoch": 11.709401709401709, | |
| "grad_norm": 2.2174923419952393, | |
| "learning_rate": 4.145299145299146e-06, | |
| "loss": 0.1751, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 11.717948717948717, | |
| "grad_norm": 2.031663179397583, | |
| "learning_rate": 4.141025641025641e-06, | |
| "loss": 0.049, | |
| "step": 1371 | |
| }, | |
| { | |
| "epoch": 11.726495726495726, | |
| "grad_norm": 4.201449394226074, | |
| "learning_rate": 4.136752136752136e-06, | |
| "loss": 0.1016, | |
| "step": 1372 | |
| }, | |
| { | |
| "epoch": 11.735042735042736, | |
| "grad_norm": 3.953226089477539, | |
| "learning_rate": 4.132478632478632e-06, | |
| "loss": 0.1336, | |
| "step": 1373 | |
| }, | |
| { | |
| "epoch": 11.743589743589745, | |
| "grad_norm": 1.4856081008911133, | |
| "learning_rate": 4.128205128205128e-06, | |
| "loss": 0.0537, | |
| "step": 1374 | |
| }, | |
| { | |
| "epoch": 11.752136752136753, | |
| "grad_norm": 1.2989288568496704, | |
| "learning_rate": 4.123931623931624e-06, | |
| "loss": 0.0351, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 11.760683760683762, | |
| "grad_norm": 4.335974216461182, | |
| "learning_rate": 4.11965811965812e-06, | |
| "loss": 0.0722, | |
| "step": 1376 | |
| }, | |
| { | |
| "epoch": 11.76923076923077, | |
| "grad_norm": 6.298306941986084, | |
| "learning_rate": 4.115384615384616e-06, | |
| "loss": 0.2359, | |
| "step": 1377 | |
| }, | |
| { | |
| "epoch": 11.777777777777779, | |
| "grad_norm": 0.7119566798210144, | |
| "learning_rate": 4.111111111111111e-06, | |
| "loss": 0.0192, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 11.786324786324787, | |
| "grad_norm": 2.7993624210357666, | |
| "learning_rate": 4.106837606837607e-06, | |
| "loss": 0.0605, | |
| "step": 1379 | |
| }, | |
| { | |
| "epoch": 11.794871794871796, | |
| "grad_norm": 6.566782474517822, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.3883, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 11.803418803418804, | |
| "grad_norm": 8.177830696105957, | |
| "learning_rate": 4.0982905982905986e-06, | |
| "loss": 0.257, | |
| "step": 1381 | |
| }, | |
| { | |
| "epoch": 11.811965811965813, | |
| "grad_norm": 4.04230260848999, | |
| "learning_rate": 4.0940170940170945e-06, | |
| "loss": 0.0943, | |
| "step": 1382 | |
| }, | |
| { | |
| "epoch": 11.820512820512821, | |
| "grad_norm": 3.595386505126953, | |
| "learning_rate": 4.08974358974359e-06, | |
| "loss": 0.0533, | |
| "step": 1383 | |
| }, | |
| { | |
| "epoch": 11.82905982905983, | |
| "grad_norm": 3.755312204360962, | |
| "learning_rate": 4.085470085470085e-06, | |
| "loss": 0.0468, | |
| "step": 1384 | |
| }, | |
| { | |
| "epoch": 11.837606837606838, | |
| "grad_norm": 2.0697362422943115, | |
| "learning_rate": 4.081196581196581e-06, | |
| "loss": 0.063, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 11.846153846153847, | |
| "grad_norm": 7.690021991729736, | |
| "learning_rate": 4.076923076923077e-06, | |
| "loss": 0.2415, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 11.854700854700855, | |
| "grad_norm": 3.0239031314849854, | |
| "learning_rate": 4.072649572649573e-06, | |
| "loss": 0.1257, | |
| "step": 1387 | |
| }, | |
| { | |
| "epoch": 11.863247863247864, | |
| "grad_norm": 2.263847589492798, | |
| "learning_rate": 4.068376068376069e-06, | |
| "loss": 0.132, | |
| "step": 1388 | |
| }, | |
| { | |
| "epoch": 11.871794871794872, | |
| "grad_norm": 2.9513261318206787, | |
| "learning_rate": 4.064102564102565e-06, | |
| "loss": 0.1229, | |
| "step": 1389 | |
| }, | |
| { | |
| "epoch": 11.88034188034188, | |
| "grad_norm": 3.03973388671875, | |
| "learning_rate": 4.05982905982906e-06, | |
| "loss": 0.0966, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 11.88888888888889, | |
| "grad_norm": 1.0075026750564575, | |
| "learning_rate": 4.055555555555556e-06, | |
| "loss": 0.0284, | |
| "step": 1391 | |
| }, | |
| { | |
| "epoch": 11.897435897435898, | |
| "grad_norm": 1.5330802202224731, | |
| "learning_rate": 4.051282051282052e-06, | |
| "loss": 0.0614, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 11.905982905982906, | |
| "grad_norm": 3.6498589515686035, | |
| "learning_rate": 4.0470085470085476e-06, | |
| "loss": 0.2236, | |
| "step": 1393 | |
| }, | |
| { | |
| "epoch": 11.914529914529915, | |
| "grad_norm": 4.659658908843994, | |
| "learning_rate": 4.0427350427350435e-06, | |
| "loss": 0.3245, | |
| "step": 1394 | |
| }, | |
| { | |
| "epoch": 11.923076923076923, | |
| "grad_norm": 3.921703815460205, | |
| "learning_rate": 4.0384615384615385e-06, | |
| "loss": 0.2981, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 11.931623931623932, | |
| "grad_norm": 5.816749572753906, | |
| "learning_rate": 4.034188034188034e-06, | |
| "loss": 0.1606, | |
| "step": 1396 | |
| }, | |
| { | |
| "epoch": 11.94017094017094, | |
| "grad_norm": 1.2831742763519287, | |
| "learning_rate": 4.02991452991453e-06, | |
| "loss": 0.0307, | |
| "step": 1397 | |
| }, | |
| { | |
| "epoch": 11.948717948717949, | |
| "grad_norm": 5.745227813720703, | |
| "learning_rate": 4.025641025641026e-06, | |
| "loss": 0.5323, | |
| "step": 1398 | |
| }, | |
| { | |
| "epoch": 11.957264957264957, | |
| "grad_norm": 2.4196462631225586, | |
| "learning_rate": 4.021367521367522e-06, | |
| "loss": 0.09, | |
| "step": 1399 | |
| }, | |
| { | |
| "epoch": 11.965811965811966, | |
| "grad_norm": 8.084505081176758, | |
| "learning_rate": 4.017094017094018e-06, | |
| "loss": 0.2991, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 11.974358974358974, | |
| "grad_norm": 3.786708116531372, | |
| "learning_rate": 4.012820512820513e-06, | |
| "loss": 0.2163, | |
| "step": 1401 | |
| }, | |
| { | |
| "epoch": 11.982905982905983, | |
| "grad_norm": 4.76535701751709, | |
| "learning_rate": 4.008547008547009e-06, | |
| "loss": 0.2453, | |
| "step": 1402 | |
| }, | |
| { | |
| "epoch": 11.991452991452991, | |
| "grad_norm": 7.380269527435303, | |
| "learning_rate": 4.004273504273505e-06, | |
| "loss": 0.3525, | |
| "step": 1403 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 41.21335983276367, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.2139, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.07730000466108322, | |
| "eval_runtime": 9.2426, | |
| "eval_samples_per_second": 50.419, | |
| "eval_steps_per_second": 6.383, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 12.008547008547009, | |
| "grad_norm": 2.3692574501037598, | |
| "learning_rate": 3.9957264957264966e-06, | |
| "loss": 0.0939, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 12.017094017094017, | |
| "grad_norm": 8.087658882141113, | |
| "learning_rate": 3.9914529914529924e-06, | |
| "loss": 0.2801, | |
| "step": 1406 | |
| }, | |
| { | |
| "epoch": 12.025641025641026, | |
| "grad_norm": 8.448614120483398, | |
| "learning_rate": 3.9871794871794875e-06, | |
| "loss": 0.2069, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 12.034188034188034, | |
| "grad_norm": 1.8581651449203491, | |
| "learning_rate": 3.982905982905983e-06, | |
| "loss": 0.0509, | |
| "step": 1408 | |
| }, | |
| { | |
| "epoch": 12.042735042735043, | |
| "grad_norm": 1.711654543876648, | |
| "learning_rate": 3.9786324786324784e-06, | |
| "loss": 0.0464, | |
| "step": 1409 | |
| }, | |
| { | |
| "epoch": 12.051282051282051, | |
| "grad_norm": 1.482553482055664, | |
| "learning_rate": 3.974358974358974e-06, | |
| "loss": 0.028, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 12.05982905982906, | |
| "grad_norm": 8.005542755126953, | |
| "learning_rate": 3.97008547008547e-06, | |
| "loss": 0.2587, | |
| "step": 1411 | |
| }, | |
| { | |
| "epoch": 12.068376068376068, | |
| "grad_norm": 2.1153948307037354, | |
| "learning_rate": 3.965811965811966e-06, | |
| "loss": 0.0563, | |
| "step": 1412 | |
| }, | |
| { | |
| "epoch": 12.076923076923077, | |
| "grad_norm": 7.791186809539795, | |
| "learning_rate": 3.961538461538462e-06, | |
| "loss": 0.0587, | |
| "step": 1413 | |
| }, | |
| { | |
| "epoch": 12.085470085470085, | |
| "grad_norm": 21.04537582397461, | |
| "learning_rate": 3.957264957264957e-06, | |
| "loss": 0.252, | |
| "step": 1414 | |
| }, | |
| { | |
| "epoch": 12.094017094017094, | |
| "grad_norm": 3.144742727279663, | |
| "learning_rate": 3.952991452991453e-06, | |
| "loss": 0.2207, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 12.102564102564102, | |
| "grad_norm": 2.23223614692688, | |
| "learning_rate": 3.948717948717949e-06, | |
| "loss": 0.0923, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 12.11111111111111, | |
| "grad_norm": 3.5652217864990234, | |
| "learning_rate": 3.944444444444445e-06, | |
| "loss": 0.2197, | |
| "step": 1417 | |
| }, | |
| { | |
| "epoch": 12.11965811965812, | |
| "grad_norm": 3.1105499267578125, | |
| "learning_rate": 3.940170940170941e-06, | |
| "loss": 0.071, | |
| "step": 1418 | |
| }, | |
| { | |
| "epoch": 12.128205128205128, | |
| "grad_norm": 2.525405168533325, | |
| "learning_rate": 3.9358974358974365e-06, | |
| "loss": 0.0874, | |
| "step": 1419 | |
| }, | |
| { | |
| "epoch": 12.136752136752136, | |
| "grad_norm": 4.479174613952637, | |
| "learning_rate": 3.9316239316239315e-06, | |
| "loss": 0.1872, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 12.145299145299145, | |
| "grad_norm": 2.0484113693237305, | |
| "learning_rate": 3.927350427350427e-06, | |
| "loss": 0.0739, | |
| "step": 1421 | |
| }, | |
| { | |
| "epoch": 12.153846153846153, | |
| "grad_norm": 2.014679431915283, | |
| "learning_rate": 3.923076923076923e-06, | |
| "loss": 0.1089, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 12.162393162393162, | |
| "grad_norm": 4.71014404296875, | |
| "learning_rate": 3.918803418803419e-06, | |
| "loss": 0.3136, | |
| "step": 1423 | |
| }, | |
| { | |
| "epoch": 12.17094017094017, | |
| "grad_norm": 2.1372437477111816, | |
| "learning_rate": 3.914529914529915e-06, | |
| "loss": 0.0458, | |
| "step": 1424 | |
| }, | |
| { | |
| "epoch": 12.179487179487179, | |
| "grad_norm": 1.4595564603805542, | |
| "learning_rate": 3.910256410256411e-06, | |
| "loss": 0.0601, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 12.188034188034187, | |
| "grad_norm": 4.45602560043335, | |
| "learning_rate": 3.905982905982906e-06, | |
| "loss": 0.091, | |
| "step": 1426 | |
| }, | |
| { | |
| "epoch": 12.196581196581196, | |
| "grad_norm": 1.473585844039917, | |
| "learning_rate": 3.901709401709402e-06, | |
| "loss": 0.0515, | |
| "step": 1427 | |
| }, | |
| { | |
| "epoch": 12.205128205128204, | |
| "grad_norm": 1.8761534690856934, | |
| "learning_rate": 3.897435897435898e-06, | |
| "loss": 0.055, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 12.213675213675213, | |
| "grad_norm": 0.7121579647064209, | |
| "learning_rate": 3.893162393162394e-06, | |
| "loss": 0.0197, | |
| "step": 1429 | |
| }, | |
| { | |
| "epoch": 12.222222222222221, | |
| "grad_norm": 2.0035219192504883, | |
| "learning_rate": 3.88888888888889e-06, | |
| "loss": 0.0904, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 12.23076923076923, | |
| "grad_norm": 3.820181369781494, | |
| "learning_rate": 3.884615384615385e-06, | |
| "loss": 0.2415, | |
| "step": 1431 | |
| }, | |
| { | |
| "epoch": 12.239316239316238, | |
| "grad_norm": 3.40633225440979, | |
| "learning_rate": 3.8803418803418805e-06, | |
| "loss": 0.0593, | |
| "step": 1432 | |
| }, | |
| { | |
| "epoch": 12.247863247863247, | |
| "grad_norm": 7.093897342681885, | |
| "learning_rate": 3.876068376068376e-06, | |
| "loss": 0.2504, | |
| "step": 1433 | |
| }, | |
| { | |
| "epoch": 12.256410256410255, | |
| "grad_norm": 2.1057517528533936, | |
| "learning_rate": 3.871794871794872e-06, | |
| "loss": 0.0573, | |
| "step": 1434 | |
| }, | |
| { | |
| "epoch": 12.264957264957266, | |
| "grad_norm": 4.797401428222656, | |
| "learning_rate": 3.867521367521368e-06, | |
| "loss": 0.338, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 12.273504273504274, | |
| "grad_norm": 20.711339950561523, | |
| "learning_rate": 3.863247863247864e-06, | |
| "loss": 0.1964, | |
| "step": 1436 | |
| }, | |
| { | |
| "epoch": 12.282051282051283, | |
| "grad_norm": 2.725280523300171, | |
| "learning_rate": 3.858974358974359e-06, | |
| "loss": 0.1837, | |
| "step": 1437 | |
| }, | |
| { | |
| "epoch": 12.290598290598291, | |
| "grad_norm": 0.9469479322433472, | |
| "learning_rate": 3.854700854700855e-06, | |
| "loss": 0.0231, | |
| "step": 1438 | |
| }, | |
| { | |
| "epoch": 12.2991452991453, | |
| "grad_norm": 2.0424935817718506, | |
| "learning_rate": 3.850427350427351e-06, | |
| "loss": 0.1373, | |
| "step": 1439 | |
| }, | |
| { | |
| "epoch": 12.307692307692308, | |
| "grad_norm": 1.4781558513641357, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.0393, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 12.316239316239317, | |
| "grad_norm": 3.7576427459716797, | |
| "learning_rate": 3.841880341880343e-06, | |
| "loss": 0.1134, | |
| "step": 1441 | |
| }, | |
| { | |
| "epoch": 12.324786324786325, | |
| "grad_norm": 299.5986633300781, | |
| "learning_rate": 3.8376068376068386e-06, | |
| "loss": 0.8017, | |
| "step": 1442 | |
| }, | |
| { | |
| "epoch": 12.333333333333334, | |
| "grad_norm": 3.109199047088623, | |
| "learning_rate": 3.833333333333334e-06, | |
| "loss": 0.1014, | |
| "step": 1443 | |
| }, | |
| { | |
| "epoch": 12.341880341880342, | |
| "grad_norm": 6.353960990905762, | |
| "learning_rate": 3.8290598290598295e-06, | |
| "loss": 0.3484, | |
| "step": 1444 | |
| }, | |
| { | |
| "epoch": 12.350427350427351, | |
| "grad_norm": 12.957517623901367, | |
| "learning_rate": 3.8247863247863246e-06, | |
| "loss": 0.5644, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 12.35897435897436, | |
| "grad_norm": 10.197676658630371, | |
| "learning_rate": 3.8205128205128204e-06, | |
| "loss": 0.1525, | |
| "step": 1446 | |
| }, | |
| { | |
| "epoch": 12.367521367521368, | |
| "grad_norm": 1.7754546403884888, | |
| "learning_rate": 3.816239316239316e-06, | |
| "loss": 0.0259, | |
| "step": 1447 | |
| }, | |
| { | |
| "epoch": 12.376068376068377, | |
| "grad_norm": 1.4237226247787476, | |
| "learning_rate": 3.8119658119658122e-06, | |
| "loss": 0.0307, | |
| "step": 1448 | |
| }, | |
| { | |
| "epoch": 12.384615384615385, | |
| "grad_norm": 2.94474458694458, | |
| "learning_rate": 3.8076923076923077e-06, | |
| "loss": 0.1447, | |
| "step": 1449 | |
| }, | |
| { | |
| "epoch": 12.393162393162394, | |
| "grad_norm": 3.7823615074157715, | |
| "learning_rate": 3.8034188034188036e-06, | |
| "loss": 0.071, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 12.401709401709402, | |
| "grad_norm": 7.5281081199646, | |
| "learning_rate": 3.7991452991452995e-06, | |
| "loss": 0.1805, | |
| "step": 1451 | |
| }, | |
| { | |
| "epoch": 12.41025641025641, | |
| "grad_norm": 2.523592233657837, | |
| "learning_rate": 3.794871794871795e-06, | |
| "loss": 0.0684, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 12.418803418803419, | |
| "grad_norm": 2.423443078994751, | |
| "learning_rate": 3.790598290598291e-06, | |
| "loss": 0.0726, | |
| "step": 1453 | |
| }, | |
| { | |
| "epoch": 12.427350427350428, | |
| "grad_norm": 6.3336005210876465, | |
| "learning_rate": 3.7863247863247863e-06, | |
| "loss": 0.1684, | |
| "step": 1454 | |
| }, | |
| { | |
| "epoch": 12.435897435897436, | |
| "grad_norm": 248.31146240234375, | |
| "learning_rate": 3.782051282051282e-06, | |
| "loss": 0.6863, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 12.444444444444445, | |
| "grad_norm": 3.0117695331573486, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 0.217, | |
| "step": 1456 | |
| }, | |
| { | |
| "epoch": 12.452991452991453, | |
| "grad_norm": 1.4753539562225342, | |
| "learning_rate": 3.7735042735042735e-06, | |
| "loss": 0.0623, | |
| "step": 1457 | |
| }, | |
| { | |
| "epoch": 12.461538461538462, | |
| "grad_norm": 2.095745325088501, | |
| "learning_rate": 3.7692307692307694e-06, | |
| "loss": 0.055, | |
| "step": 1458 | |
| }, | |
| { | |
| "epoch": 12.47008547008547, | |
| "grad_norm": 3.508305788040161, | |
| "learning_rate": 3.7649572649572653e-06, | |
| "loss": 0.1097, | |
| "step": 1459 | |
| }, | |
| { | |
| "epoch": 12.478632478632479, | |
| "grad_norm": 3.0965282917022705, | |
| "learning_rate": 3.760683760683761e-06, | |
| "loss": 0.3374, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 12.487179487179487, | |
| "grad_norm": 0.7286785244941711, | |
| "learning_rate": 3.7564102564102567e-06, | |
| "loss": 0.0182, | |
| "step": 1461 | |
| }, | |
| { | |
| "epoch": 12.495726495726496, | |
| "grad_norm": 5.957888126373291, | |
| "learning_rate": 3.7521367521367526e-06, | |
| "loss": 0.3498, | |
| "step": 1462 | |
| }, | |
| { | |
| "epoch": 12.504273504273504, | |
| "grad_norm": 10.433263778686523, | |
| "learning_rate": 3.747863247863248e-06, | |
| "loss": 0.446, | |
| "step": 1463 | |
| }, | |
| { | |
| "epoch": 12.512820512820513, | |
| "grad_norm": 4.565568923950195, | |
| "learning_rate": 3.743589743589744e-06, | |
| "loss": 0.1026, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.521367521367521, | |
| "grad_norm": 2.607106924057007, | |
| "learning_rate": 3.73931623931624e-06, | |
| "loss": 0.0912, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 12.52991452991453, | |
| "grad_norm": 2.415541410446167, | |
| "learning_rate": 3.7350427350427353e-06, | |
| "loss": 0.0594, | |
| "step": 1466 | |
| }, | |
| { | |
| "epoch": 12.538461538461538, | |
| "grad_norm": 7.978870868682861, | |
| "learning_rate": 3.730769230769231e-06, | |
| "loss": 0.2617, | |
| "step": 1467 | |
| }, | |
| { | |
| "epoch": 12.547008547008547, | |
| "grad_norm": 6.858293056488037, | |
| "learning_rate": 3.726495726495727e-06, | |
| "loss": 0.3642, | |
| "step": 1468 | |
| }, | |
| { | |
| "epoch": 12.555555555555555, | |
| "grad_norm": 1.3900551795959473, | |
| "learning_rate": 3.7222222222222225e-06, | |
| "loss": 0.0445, | |
| "step": 1469 | |
| }, | |
| { | |
| "epoch": 12.564102564102564, | |
| "grad_norm": 8.111970901489258, | |
| "learning_rate": 3.7179487179487184e-06, | |
| "loss": 0.1828, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 12.572649572649572, | |
| "grad_norm": 2.731841802597046, | |
| "learning_rate": 3.7136752136752143e-06, | |
| "loss": 0.2027, | |
| "step": 1471 | |
| }, | |
| { | |
| "epoch": 12.581196581196581, | |
| "grad_norm": 4.418527126312256, | |
| "learning_rate": 3.7094017094017098e-06, | |
| "loss": 0.1744, | |
| "step": 1472 | |
| }, | |
| { | |
| "epoch": 12.58974358974359, | |
| "grad_norm": 2.8263015747070312, | |
| "learning_rate": 3.7051282051282057e-06, | |
| "loss": 0.1123, | |
| "step": 1473 | |
| }, | |
| { | |
| "epoch": 12.598290598290598, | |
| "grad_norm": 2.3524725437164307, | |
| "learning_rate": 3.700854700854701e-06, | |
| "loss": 0.0999, | |
| "step": 1474 | |
| }, | |
| { | |
| "epoch": 12.606837606837606, | |
| "grad_norm": 9.863709449768066, | |
| "learning_rate": 3.696581196581197e-06, | |
| "loss": 0.4589, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 12.615384615384615, | |
| "grad_norm": 3.5506396293640137, | |
| "learning_rate": 3.692307692307693e-06, | |
| "loss": 0.2034, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 12.623931623931623, | |
| "grad_norm": 2.4352779388427734, | |
| "learning_rate": 3.6880341880341884e-06, | |
| "loss": 0.0806, | |
| "step": 1477 | |
| }, | |
| { | |
| "epoch": 12.632478632478632, | |
| "grad_norm": 1.8339797258377075, | |
| "learning_rate": 3.6837606837606843e-06, | |
| "loss": 0.0635, | |
| "step": 1478 | |
| }, | |
| { | |
| "epoch": 12.64102564102564, | |
| "grad_norm": 4.63474178314209, | |
| "learning_rate": 3.67948717948718e-06, | |
| "loss": 0.4568, | |
| "step": 1479 | |
| }, | |
| { | |
| "epoch": 12.649572649572649, | |
| "grad_norm": 7.696872711181641, | |
| "learning_rate": 3.6752136752136756e-06, | |
| "loss": 0.1769, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 12.658119658119658, | |
| "grad_norm": 1.3894271850585938, | |
| "learning_rate": 3.670940170940171e-06, | |
| "loss": 0.0747, | |
| "step": 1481 | |
| }, | |
| { | |
| "epoch": 12.666666666666666, | |
| "grad_norm": 5.607828140258789, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 0.1178, | |
| "step": 1482 | |
| }, | |
| { | |
| "epoch": 12.675213675213675, | |
| "grad_norm": 2.120594024658203, | |
| "learning_rate": 3.6623931623931625e-06, | |
| "loss": 0.0497, | |
| "step": 1483 | |
| }, | |
| { | |
| "epoch": 12.683760683760683, | |
| "grad_norm": 1.359381914138794, | |
| "learning_rate": 3.6581196581196584e-06, | |
| "loss": 0.035, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 12.692307692307692, | |
| "grad_norm": 2.8533923625946045, | |
| "learning_rate": 3.653846153846154e-06, | |
| "loss": 0.1048, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 12.7008547008547, | |
| "grad_norm": 6.021198749542236, | |
| "learning_rate": 3.6495726495726497e-06, | |
| "loss": 0.1604, | |
| "step": 1486 | |
| }, | |
| { | |
| "epoch": 12.709401709401709, | |
| "grad_norm": 7.198216915130615, | |
| "learning_rate": 3.6452991452991456e-06, | |
| "loss": 0.1656, | |
| "step": 1487 | |
| }, | |
| { | |
| "epoch": 12.717948717948717, | |
| "grad_norm": 1.4581981897354126, | |
| "learning_rate": 3.641025641025641e-06, | |
| "loss": 0.0398, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 12.726495726495726, | |
| "grad_norm": 30.704627990722656, | |
| "learning_rate": 3.636752136752137e-06, | |
| "loss": 0.3371, | |
| "step": 1489 | |
| }, | |
| { | |
| "epoch": 12.735042735042736, | |
| "grad_norm": 2.5204057693481445, | |
| "learning_rate": 3.632478632478633e-06, | |
| "loss": 0.0742, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 12.743589743589745, | |
| "grad_norm": 2.3917508125305176, | |
| "learning_rate": 3.6282051282051283e-06, | |
| "loss": 0.1681, | |
| "step": 1491 | |
| }, | |
| { | |
| "epoch": 12.752136752136753, | |
| "grad_norm": 1.4529337882995605, | |
| "learning_rate": 3.623931623931624e-06, | |
| "loss": 0.0247, | |
| "step": 1492 | |
| }, | |
| { | |
| "epoch": 12.760683760683762, | |
| "grad_norm": 31.894805908203125, | |
| "learning_rate": 3.6196581196581197e-06, | |
| "loss": 0.2222, | |
| "step": 1493 | |
| }, | |
| { | |
| "epoch": 12.76923076923077, | |
| "grad_norm": 3.4240164756774902, | |
| "learning_rate": 3.6153846153846156e-06, | |
| "loss": 0.1432, | |
| "step": 1494 | |
| }, | |
| { | |
| "epoch": 12.777777777777779, | |
| "grad_norm": 2.0000102519989014, | |
| "learning_rate": 3.6111111111111115e-06, | |
| "loss": 0.0383, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 12.786324786324787, | |
| "grad_norm": 3.7665908336639404, | |
| "learning_rate": 3.606837606837607e-06, | |
| "loss": 0.2719, | |
| "step": 1496 | |
| }, | |
| { | |
| "epoch": 12.794871794871796, | |
| "grad_norm": 2.0319290161132812, | |
| "learning_rate": 3.602564102564103e-06, | |
| "loss": 0.0741, | |
| "step": 1497 | |
| }, | |
| { | |
| "epoch": 12.803418803418804, | |
| "grad_norm": 2.3379619121551514, | |
| "learning_rate": 3.5982905982905987e-06, | |
| "loss": 0.1155, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 12.811965811965813, | |
| "grad_norm": 5.183985233306885, | |
| "learning_rate": 3.594017094017094e-06, | |
| "loss": 0.0815, | |
| "step": 1499 | |
| }, | |
| { | |
| "epoch": 12.820512820512821, | |
| "grad_norm": 3.1432502269744873, | |
| "learning_rate": 3.58974358974359e-06, | |
| "loss": 0.1855, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 12.82905982905983, | |
| "grad_norm": 4.5739946365356445, | |
| "learning_rate": 3.585470085470086e-06, | |
| "loss": 0.1424, | |
| "step": 1501 | |
| }, | |
| { | |
| "epoch": 12.837606837606838, | |
| "grad_norm": 1.6006520986557007, | |
| "learning_rate": 3.5811965811965814e-06, | |
| "loss": 0.0305, | |
| "step": 1502 | |
| }, | |
| { | |
| "epoch": 12.846153846153847, | |
| "grad_norm": 3.937011241912842, | |
| "learning_rate": 3.5769230769230773e-06, | |
| "loss": 0.2497, | |
| "step": 1503 | |
| }, | |
| { | |
| "epoch": 12.854700854700855, | |
| "grad_norm": 2.6159651279449463, | |
| "learning_rate": 3.572649572649573e-06, | |
| "loss": 0.1067, | |
| "step": 1504 | |
| }, | |
| { | |
| "epoch": 12.863247863247864, | |
| "grad_norm": 2.578547239303589, | |
| "learning_rate": 3.5683760683760687e-06, | |
| "loss": 0.0663, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 12.871794871794872, | |
| "grad_norm": 2.3777639865875244, | |
| "learning_rate": 3.5641025641025646e-06, | |
| "loss": 0.0558, | |
| "step": 1506 | |
| }, | |
| { | |
| "epoch": 12.88034188034188, | |
| "grad_norm": 7.5656561851501465, | |
| "learning_rate": 3.5598290598290604e-06, | |
| "loss": 0.2448, | |
| "step": 1507 | |
| }, | |
| { | |
| "epoch": 12.88888888888889, | |
| "grad_norm": 4.21798849105835, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 0.1916, | |
| "step": 1508 | |
| }, | |
| { | |
| "epoch": 12.897435897435898, | |
| "grad_norm": 1.318049669265747, | |
| "learning_rate": 3.551282051282052e-06, | |
| "loss": 0.0387, | |
| "step": 1509 | |
| }, | |
| { | |
| "epoch": 12.905982905982906, | |
| "grad_norm": 2.4345362186431885, | |
| "learning_rate": 3.5470085470085473e-06, | |
| "loss": 0.061, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 12.914529914529915, | |
| "grad_norm": 3.2767112255096436, | |
| "learning_rate": 3.542735042735043e-06, | |
| "loss": 0.1627, | |
| "step": 1511 | |
| }, | |
| { | |
| "epoch": 12.923076923076923, | |
| "grad_norm": 6.881056785583496, | |
| "learning_rate": 3.538461538461539e-06, | |
| "loss": 0.2452, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 12.931623931623932, | |
| "grad_norm": 8.017362594604492, | |
| "learning_rate": 3.5341880341880345e-06, | |
| "loss": 0.1972, | |
| "step": 1513 | |
| }, | |
| { | |
| "epoch": 12.94017094017094, | |
| "grad_norm": 1.1411398649215698, | |
| "learning_rate": 3.5299145299145304e-06, | |
| "loss": 0.0243, | |
| "step": 1514 | |
| }, | |
| { | |
| "epoch": 12.948717948717949, | |
| "grad_norm": 4.486563205718994, | |
| "learning_rate": 3.5256410256410263e-06, | |
| "loss": 0.1347, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 12.957264957264957, | |
| "grad_norm": 2.348222494125366, | |
| "learning_rate": 3.5213675213675218e-06, | |
| "loss": 0.1828, | |
| "step": 1516 | |
| }, | |
| { | |
| "epoch": 12.965811965811966, | |
| "grad_norm": 2.2855775356292725, | |
| "learning_rate": 3.5170940170940177e-06, | |
| "loss": 0.0465, | |
| "step": 1517 | |
| }, | |
| { | |
| "epoch": 12.974358974358974, | |
| "grad_norm": 10.313456535339355, | |
| "learning_rate": 3.5128205128205127e-06, | |
| "loss": 0.3033, | |
| "step": 1518 | |
| }, | |
| { | |
| "epoch": 12.982905982905983, | |
| "grad_norm": 12.115890502929688, | |
| "learning_rate": 3.5085470085470086e-06, | |
| "loss": 0.6762, | |
| "step": 1519 | |
| }, | |
| { | |
| "epoch": 12.991452991452991, | |
| "grad_norm": 2.746267557144165, | |
| "learning_rate": 3.5042735042735045e-06, | |
| "loss": 0.123, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 5.204991340637207, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.2086, | |
| "step": 1521 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.06878729909658432, | |
| "eval_runtime": 9.2334, | |
| "eval_samples_per_second": 50.469, | |
| "eval_steps_per_second": 6.39, | |
| "step": 1521 | |
| }, | |
| { | |
| "epoch": 13.008547008547009, | |
| "grad_norm": 1.8741862773895264, | |
| "learning_rate": 3.495726495726496e-06, | |
| "loss": 0.0594, | |
| "step": 1522 | |
| }, | |
| { | |
| "epoch": 13.017094017094017, | |
| "grad_norm": 1.6060154438018799, | |
| "learning_rate": 3.4914529914529917e-06, | |
| "loss": 0.0426, | |
| "step": 1523 | |
| }, | |
| { | |
| "epoch": 13.025641025641026, | |
| "grad_norm": 2.194714069366455, | |
| "learning_rate": 3.487179487179487e-06, | |
| "loss": 0.1907, | |
| "step": 1524 | |
| }, | |
| { | |
| "epoch": 13.034188034188034, | |
| "grad_norm": 0.716149628162384, | |
| "learning_rate": 3.482905982905983e-06, | |
| "loss": 0.0177, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 13.042735042735043, | |
| "grad_norm": 4.787989139556885, | |
| "learning_rate": 3.478632478632479e-06, | |
| "loss": 0.246, | |
| "step": 1526 | |
| }, | |
| { | |
| "epoch": 13.051282051282051, | |
| "grad_norm": 1.662338137626648, | |
| "learning_rate": 3.4743589743589744e-06, | |
| "loss": 0.0561, | |
| "step": 1527 | |
| }, | |
| { | |
| "epoch": 13.05982905982906, | |
| "grad_norm": 0.9663236737251282, | |
| "learning_rate": 3.4700854700854703e-06, | |
| "loss": 0.0392, | |
| "step": 1528 | |
| }, | |
| { | |
| "epoch": 13.068376068376068, | |
| "grad_norm": 0.8232766389846802, | |
| "learning_rate": 3.465811965811966e-06, | |
| "loss": 0.0221, | |
| "step": 1529 | |
| }, | |
| { | |
| "epoch": 13.076923076923077, | |
| "grad_norm": 2.434157609939575, | |
| "learning_rate": 3.4615384615384617e-06, | |
| "loss": 0.1777, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 13.085470085470085, | |
| "grad_norm": 2.768070936203003, | |
| "learning_rate": 3.4572649572649576e-06, | |
| "loss": 0.1101, | |
| "step": 1531 | |
| }, | |
| { | |
| "epoch": 13.094017094017094, | |
| "grad_norm": 2.061371088027954, | |
| "learning_rate": 3.452991452991453e-06, | |
| "loss": 0.0591, | |
| "step": 1532 | |
| }, | |
| { | |
| "epoch": 13.102564102564102, | |
| "grad_norm": 1.6127598285675049, | |
| "learning_rate": 3.448717948717949e-06, | |
| "loss": 0.3858, | |
| "step": 1533 | |
| }, | |
| { | |
| "epoch": 13.11111111111111, | |
| "grad_norm": 1.2561885118484497, | |
| "learning_rate": 3.444444444444445e-06, | |
| "loss": 0.0315, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 13.11965811965812, | |
| "grad_norm": 2.2859408855438232, | |
| "learning_rate": 3.4401709401709403e-06, | |
| "loss": 0.047, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 13.128205128205128, | |
| "grad_norm": 3.7528388500213623, | |
| "learning_rate": 3.435897435897436e-06, | |
| "loss": 0.1069, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 13.136752136752136, | |
| "grad_norm": 5.547614574432373, | |
| "learning_rate": 3.431623931623932e-06, | |
| "loss": 0.1411, | |
| "step": 1537 | |
| }, | |
| { | |
| "epoch": 13.145299145299145, | |
| "grad_norm": 1.6566565036773682, | |
| "learning_rate": 3.4273504273504275e-06, | |
| "loss": 0.0266, | |
| "step": 1538 | |
| }, | |
| { | |
| "epoch": 13.153846153846153, | |
| "grad_norm": 5.280163288116455, | |
| "learning_rate": 3.4230769230769234e-06, | |
| "loss": 0.0843, | |
| "step": 1539 | |
| }, | |
| { | |
| "epoch": 13.162393162393162, | |
| "grad_norm": 6.624744892120361, | |
| "learning_rate": 3.4188034188034193e-06, | |
| "loss": 0.1652, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 13.17094017094017, | |
| "grad_norm": 5.325616359710693, | |
| "learning_rate": 3.414529914529915e-06, | |
| "loss": 0.077, | |
| "step": 1541 | |
| }, | |
| { | |
| "epoch": 13.179487179487179, | |
| "grad_norm": 11.31779956817627, | |
| "learning_rate": 3.4102564102564107e-06, | |
| "loss": 0.4377, | |
| "step": 1542 | |
| }, | |
| { | |
| "epoch": 13.188034188034187, | |
| "grad_norm": 4.86885404586792, | |
| "learning_rate": 3.4059829059829066e-06, | |
| "loss": 0.2312, | |
| "step": 1543 | |
| }, | |
| { | |
| "epoch": 13.196581196581196, | |
| "grad_norm": 1.779068112373352, | |
| "learning_rate": 3.401709401709402e-06, | |
| "loss": 0.032, | |
| "step": 1544 | |
| }, | |
| { | |
| "epoch": 13.205128205128204, | |
| "grad_norm": 1.9934108257293701, | |
| "learning_rate": 3.397435897435898e-06, | |
| "loss": 0.0861, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 13.213675213675213, | |
| "grad_norm": 2.1829612255096436, | |
| "learning_rate": 3.3931623931623934e-06, | |
| "loss": 0.0855, | |
| "step": 1546 | |
| }, | |
| { | |
| "epoch": 13.222222222222221, | |
| "grad_norm": 31.108810424804688, | |
| "learning_rate": 3.3888888888888893e-06, | |
| "loss": 0.334, | |
| "step": 1547 | |
| }, | |
| { | |
| "epoch": 13.23076923076923, | |
| "grad_norm": 4.867705345153809, | |
| "learning_rate": 3.384615384615385e-06, | |
| "loss": 0.0808, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 13.239316239316238, | |
| "grad_norm": 3.226783275604248, | |
| "learning_rate": 3.3803418803418806e-06, | |
| "loss": 0.1806, | |
| "step": 1549 | |
| }, | |
| { | |
| "epoch": 13.247863247863247, | |
| "grad_norm": 1.4822824001312256, | |
| "learning_rate": 3.3760683760683765e-06, | |
| "loss": 0.0602, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 13.256410256410255, | |
| "grad_norm": 4.529379844665527, | |
| "learning_rate": 3.3717948717948724e-06, | |
| "loss": 0.318, | |
| "step": 1551 | |
| }, | |
| { | |
| "epoch": 13.264957264957266, | |
| "grad_norm": 3.2155706882476807, | |
| "learning_rate": 3.367521367521368e-06, | |
| "loss": 0.1006, | |
| "step": 1552 | |
| }, | |
| { | |
| "epoch": 13.273504273504274, | |
| "grad_norm": 2.2805707454681396, | |
| "learning_rate": 3.3632478632478638e-06, | |
| "loss": 0.0774, | |
| "step": 1553 | |
| }, | |
| { | |
| "epoch": 13.282051282051283, | |
| "grad_norm": 11.477370262145996, | |
| "learning_rate": 3.358974358974359e-06, | |
| "loss": 0.8342, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 13.290598290598291, | |
| "grad_norm": 3.8596534729003906, | |
| "learning_rate": 3.3547008547008547e-06, | |
| "loss": 0.1924, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 13.2991452991453, | |
| "grad_norm": 4.497336387634277, | |
| "learning_rate": 3.3504273504273506e-06, | |
| "loss": 0.2425, | |
| "step": 1556 | |
| }, | |
| { | |
| "epoch": 13.307692307692308, | |
| "grad_norm": 1.4496978521347046, | |
| "learning_rate": 3.346153846153846e-06, | |
| "loss": 0.0168, | |
| "step": 1557 | |
| }, | |
| { | |
| "epoch": 13.316239316239317, | |
| "grad_norm": 2.0277416706085205, | |
| "learning_rate": 3.341880341880342e-06, | |
| "loss": 0.0634, | |
| "step": 1558 | |
| }, | |
| { | |
| "epoch": 13.324786324786325, | |
| "grad_norm": 2.9120066165924072, | |
| "learning_rate": 3.337606837606838e-06, | |
| "loss": 0.1153, | |
| "step": 1559 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 4.949625015258789, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.1412, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 13.341880341880342, | |
| "grad_norm": 5.970853805541992, | |
| "learning_rate": 3.3290598290598292e-06, | |
| "loss": 0.1607, | |
| "step": 1561 | |
| }, | |
| { | |
| "epoch": 13.350427350427351, | |
| "grad_norm": 2.1988022327423096, | |
| "learning_rate": 3.324786324786325e-06, | |
| "loss": 0.0329, | |
| "step": 1562 | |
| }, | |
| { | |
| "epoch": 13.35897435897436, | |
| "grad_norm": 2.3578758239746094, | |
| "learning_rate": 3.3205128205128206e-06, | |
| "loss": 0.0711, | |
| "step": 1563 | |
| }, | |
| { | |
| "epoch": 13.367521367521368, | |
| "grad_norm": 4.554023742675781, | |
| "learning_rate": 3.3162393162393165e-06, | |
| "loss": 0.1929, | |
| "step": 1564 | |
| }, | |
| { | |
| "epoch": 13.376068376068377, | |
| "grad_norm": 3.577073335647583, | |
| "learning_rate": 3.311965811965812e-06, | |
| "loss": 0.0969, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 13.384615384615385, | |
| "grad_norm": 3.3863015174865723, | |
| "learning_rate": 3.307692307692308e-06, | |
| "loss": 0.2402, | |
| "step": 1566 | |
| }, | |
| { | |
| "epoch": 13.393162393162394, | |
| "grad_norm": 1.044550895690918, | |
| "learning_rate": 3.3034188034188037e-06, | |
| "loss": 0.026, | |
| "step": 1567 | |
| }, | |
| { | |
| "epoch": 13.401709401709402, | |
| "grad_norm": 3.1525843143463135, | |
| "learning_rate": 3.299145299145299e-06, | |
| "loss": 0.0619, | |
| "step": 1568 | |
| }, | |
| { | |
| "epoch": 13.41025641025641, | |
| "grad_norm": 2.0380606651306152, | |
| "learning_rate": 3.294871794871795e-06, | |
| "loss": 0.0477, | |
| "step": 1569 | |
| }, | |
| { | |
| "epoch": 13.418803418803419, | |
| "grad_norm": 2.4260973930358887, | |
| "learning_rate": 3.290598290598291e-06, | |
| "loss": 0.0709, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 13.427350427350428, | |
| "grad_norm": 20.958803176879883, | |
| "learning_rate": 3.2863247863247864e-06, | |
| "loss": 0.2297, | |
| "step": 1571 | |
| }, | |
| { | |
| "epoch": 13.435897435897436, | |
| "grad_norm": 2.847252368927002, | |
| "learning_rate": 3.2820512820512823e-06, | |
| "loss": 0.0565, | |
| "step": 1572 | |
| }, | |
| { | |
| "epoch": 13.444444444444445, | |
| "grad_norm": 3.646381139755249, | |
| "learning_rate": 3.277777777777778e-06, | |
| "loss": 0.3043, | |
| "step": 1573 | |
| }, | |
| { | |
| "epoch": 13.452991452991453, | |
| "grad_norm": 3.0526609420776367, | |
| "learning_rate": 3.2735042735042737e-06, | |
| "loss": 0.0941, | |
| "step": 1574 | |
| }, | |
| { | |
| "epoch": 13.461538461538462, | |
| "grad_norm": 1.6154388189315796, | |
| "learning_rate": 3.2692307692307696e-06, | |
| "loss": 0.0597, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 13.47008547008547, | |
| "grad_norm": 1.0825392007827759, | |
| "learning_rate": 3.2649572649572655e-06, | |
| "loss": 0.0325, | |
| "step": 1576 | |
| }, | |
| { | |
| "epoch": 13.478632478632479, | |
| "grad_norm": 6.045910358428955, | |
| "learning_rate": 3.260683760683761e-06, | |
| "loss": 0.2202, | |
| "step": 1577 | |
| }, | |
| { | |
| "epoch": 13.487179487179487, | |
| "grad_norm": 3.0401153564453125, | |
| "learning_rate": 3.256410256410257e-06, | |
| "loss": 0.0923, | |
| "step": 1578 | |
| }, | |
| { | |
| "epoch": 13.495726495726496, | |
| "grad_norm": 5.485551834106445, | |
| "learning_rate": 3.2521367521367527e-06, | |
| "loss": 0.3851, | |
| "step": 1579 | |
| }, | |
| { | |
| "epoch": 13.504273504273504, | |
| "grad_norm": 2.575057029724121, | |
| "learning_rate": 3.247863247863248e-06, | |
| "loss": 0.0307, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 13.512820512820513, | |
| "grad_norm": 2.7744545936584473, | |
| "learning_rate": 3.243589743589744e-06, | |
| "loss": 0.1791, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 13.521367521367521, | |
| "grad_norm": 2.430640459060669, | |
| "learning_rate": 3.2393162393162395e-06, | |
| "loss": 0.1128, | |
| "step": 1582 | |
| }, | |
| { | |
| "epoch": 13.52991452991453, | |
| "grad_norm": 4.902276992797852, | |
| "learning_rate": 3.2350427350427354e-06, | |
| "loss": 0.2661, | |
| "step": 1583 | |
| }, | |
| { | |
| "epoch": 13.538461538461538, | |
| "grad_norm": 2.601134777069092, | |
| "learning_rate": 3.2307692307692313e-06, | |
| "loss": 0.1311, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 13.547008547008547, | |
| "grad_norm": 6.309877395629883, | |
| "learning_rate": 3.2264957264957268e-06, | |
| "loss": 0.2621, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 13.555555555555555, | |
| "grad_norm": 2.079618215560913, | |
| "learning_rate": 3.2222222222222227e-06, | |
| "loss": 0.0702, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.564102564102564, | |
| "grad_norm": 2.309541702270508, | |
| "learning_rate": 3.2179487179487186e-06, | |
| "loss": 0.1577, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 13.572649572649572, | |
| "grad_norm": 4.723629951477051, | |
| "learning_rate": 3.213675213675214e-06, | |
| "loss": 0.142, | |
| "step": 1588 | |
| }, | |
| { | |
| "epoch": 13.581196581196581, | |
| "grad_norm": 2.557123899459839, | |
| "learning_rate": 3.20940170940171e-06, | |
| "loss": 0.1506, | |
| "step": 1589 | |
| }, | |
| { | |
| "epoch": 13.58974358974359, | |
| "grad_norm": 2.3154499530792236, | |
| "learning_rate": 3.205128205128206e-06, | |
| "loss": 0.1039, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 13.598290598290598, | |
| "grad_norm": 1.5464012622833252, | |
| "learning_rate": 3.200854700854701e-06, | |
| "loss": 0.0989, | |
| "step": 1591 | |
| }, | |
| { | |
| "epoch": 13.606837606837606, | |
| "grad_norm": 1.5885653495788574, | |
| "learning_rate": 3.1965811965811967e-06, | |
| "loss": 0.0278, | |
| "step": 1592 | |
| }, | |
| { | |
| "epoch": 13.615384615384615, | |
| "grad_norm": 2.7710390090942383, | |
| "learning_rate": 3.192307692307692e-06, | |
| "loss": 0.0521, | |
| "step": 1593 | |
| }, | |
| { | |
| "epoch": 13.623931623931623, | |
| "grad_norm": 4.587305545806885, | |
| "learning_rate": 3.188034188034188e-06, | |
| "loss": 0.2609, | |
| "step": 1594 | |
| }, | |
| { | |
| "epoch": 13.632478632478632, | |
| "grad_norm": 4.343963623046875, | |
| "learning_rate": 3.183760683760684e-06, | |
| "loss": 0.1079, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 13.64102564102564, | |
| "grad_norm": 2.7653536796569824, | |
| "learning_rate": 3.1794871794871795e-06, | |
| "loss": 0.1293, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 13.649572649572649, | |
| "grad_norm": 3.1731350421905518, | |
| "learning_rate": 3.1752136752136753e-06, | |
| "loss": 0.1279, | |
| "step": 1597 | |
| }, | |
| { | |
| "epoch": 13.658119658119658, | |
| "grad_norm": 8.032745361328125, | |
| "learning_rate": 3.1709401709401712e-06, | |
| "loss": 0.2114, | |
| "step": 1598 | |
| }, | |
| { | |
| "epoch": 13.666666666666666, | |
| "grad_norm": 5.6177263259887695, | |
| "learning_rate": 3.1666666666666667e-06, | |
| "loss": 0.0926, | |
| "step": 1599 | |
| }, | |
| { | |
| "epoch": 13.675213675213675, | |
| "grad_norm": 3.3568480014801025, | |
| "learning_rate": 3.1623931623931626e-06, | |
| "loss": 0.1299, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 13.683760683760683, | |
| "grad_norm": 5.182860374450684, | |
| "learning_rate": 3.158119658119658e-06, | |
| "loss": 0.1688, | |
| "step": 1601 | |
| }, | |
| { | |
| "epoch": 13.692307692307692, | |
| "grad_norm": 5.954287052154541, | |
| "learning_rate": 3.153846153846154e-06, | |
| "loss": 0.2634, | |
| "step": 1602 | |
| }, | |
| { | |
| "epoch": 13.7008547008547, | |
| "grad_norm": 2.8563358783721924, | |
| "learning_rate": 3.14957264957265e-06, | |
| "loss": 0.0469, | |
| "step": 1603 | |
| }, | |
| { | |
| "epoch": 13.709401709401709, | |
| "grad_norm": 1.6049034595489502, | |
| "learning_rate": 3.1452991452991453e-06, | |
| "loss": 0.0855, | |
| "step": 1604 | |
| }, | |
| { | |
| "epoch": 13.717948717948717, | |
| "grad_norm": 1.9734570980072021, | |
| "learning_rate": 3.141025641025641e-06, | |
| "loss": 0.0554, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 13.726495726495726, | |
| "grad_norm": 1.8398605585098267, | |
| "learning_rate": 3.136752136752137e-06, | |
| "loss": 0.1033, | |
| "step": 1606 | |
| }, | |
| { | |
| "epoch": 13.735042735042736, | |
| "grad_norm": 3.3013346195220947, | |
| "learning_rate": 3.1324786324786326e-06, | |
| "loss": 0.1476, | |
| "step": 1607 | |
| }, | |
| { | |
| "epoch": 13.743589743589745, | |
| "grad_norm": 1.2622041702270508, | |
| "learning_rate": 3.1282051282051284e-06, | |
| "loss": 0.0222, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 13.752136752136753, | |
| "grad_norm": 3.983888626098633, | |
| "learning_rate": 3.1239316239316243e-06, | |
| "loss": 0.0861, | |
| "step": 1609 | |
| }, | |
| { | |
| "epoch": 13.760683760683762, | |
| "grad_norm": 2.883335828781128, | |
| "learning_rate": 3.11965811965812e-06, | |
| "loss": 0.0737, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 13.76923076923077, | |
| "grad_norm": 0.9045059680938721, | |
| "learning_rate": 3.1153846153846157e-06, | |
| "loss": 0.0232, | |
| "step": 1611 | |
| }, | |
| { | |
| "epoch": 13.777777777777779, | |
| "grad_norm": 1.8752232789993286, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 0.0602, | |
| "step": 1612 | |
| }, | |
| { | |
| "epoch": 13.786324786324787, | |
| "grad_norm": 3.088440418243408, | |
| "learning_rate": 3.106837606837607e-06, | |
| "loss": 0.102, | |
| "step": 1613 | |
| }, | |
| { | |
| "epoch": 13.794871794871796, | |
| "grad_norm": 4.067224502563477, | |
| "learning_rate": 3.102564102564103e-06, | |
| "loss": 0.1461, | |
| "step": 1614 | |
| }, | |
| { | |
| "epoch": 13.803418803418804, | |
| "grad_norm": 6.9123148918151855, | |
| "learning_rate": 3.098290598290599e-06, | |
| "loss": 0.0752, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 13.811965811965813, | |
| "grad_norm": 17.15372657775879, | |
| "learning_rate": 3.0940170940170943e-06, | |
| "loss": 0.5163, | |
| "step": 1616 | |
| }, | |
| { | |
| "epoch": 13.820512820512821, | |
| "grad_norm": 2.4951720237731934, | |
| "learning_rate": 3.08974358974359e-06, | |
| "loss": 0.1326, | |
| "step": 1617 | |
| }, | |
| { | |
| "epoch": 13.82905982905983, | |
| "grad_norm": 2.1316449642181396, | |
| "learning_rate": 3.0854700854700857e-06, | |
| "loss": 0.0469, | |
| "step": 1618 | |
| }, | |
| { | |
| "epoch": 13.837606837606838, | |
| "grad_norm": 2.5955941677093506, | |
| "learning_rate": 3.0811965811965815e-06, | |
| "loss": 0.1056, | |
| "step": 1619 | |
| }, | |
| { | |
| "epoch": 13.846153846153847, | |
| "grad_norm": 14.360347747802734, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.4793, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 13.854700854700855, | |
| "grad_norm": 1.9134567975997925, | |
| "learning_rate": 3.072649572649573e-06, | |
| "loss": 0.054, | |
| "step": 1621 | |
| }, | |
| { | |
| "epoch": 13.863247863247864, | |
| "grad_norm": 3.1168692111968994, | |
| "learning_rate": 3.068376068376069e-06, | |
| "loss": 0.321, | |
| "step": 1622 | |
| }, | |
| { | |
| "epoch": 13.871794871794872, | |
| "grad_norm": 4.940008163452148, | |
| "learning_rate": 3.0641025641025647e-06, | |
| "loss": 0.1452, | |
| "step": 1623 | |
| }, | |
| { | |
| "epoch": 13.88034188034188, | |
| "grad_norm": 3.001660108566284, | |
| "learning_rate": 3.05982905982906e-06, | |
| "loss": 0.1094, | |
| "step": 1624 | |
| }, | |
| { | |
| "epoch": 13.88888888888889, | |
| "grad_norm": 1.3110100030899048, | |
| "learning_rate": 3.055555555555556e-06, | |
| "loss": 0.0305, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 13.897435897435898, | |
| "grad_norm": 269.3442077636719, | |
| "learning_rate": 3.051282051282052e-06, | |
| "loss": 0.8319, | |
| "step": 1626 | |
| }, | |
| { | |
| "epoch": 13.905982905982906, | |
| "grad_norm": 1.5236955881118774, | |
| "learning_rate": 3.0470085470085474e-06, | |
| "loss": 0.0294, | |
| "step": 1627 | |
| }, | |
| { | |
| "epoch": 13.914529914529915, | |
| "grad_norm": 1.8342583179473877, | |
| "learning_rate": 3.042735042735043e-06, | |
| "loss": 0.1122, | |
| "step": 1628 | |
| }, | |
| { | |
| "epoch": 13.923076923076923, | |
| "grad_norm": 1.7902953624725342, | |
| "learning_rate": 3.0384615384615383e-06, | |
| "loss": 0.0426, | |
| "step": 1629 | |
| }, | |
| { | |
| "epoch": 13.931623931623932, | |
| "grad_norm": 1.461769938468933, | |
| "learning_rate": 3.0341880341880342e-06, | |
| "loss": 0.0326, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 13.94017094017094, | |
| "grad_norm": 2.2590038776397705, | |
| "learning_rate": 3.02991452991453e-06, | |
| "loss": 0.067, | |
| "step": 1631 | |
| }, | |
| { | |
| "epoch": 13.948717948717949, | |
| "grad_norm": 0.8894402980804443, | |
| "learning_rate": 3.0256410256410256e-06, | |
| "loss": 0.0269, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 13.957264957264957, | |
| "grad_norm": 2.097757339477539, | |
| "learning_rate": 3.0213675213675215e-06, | |
| "loss": 0.1211, | |
| "step": 1633 | |
| }, | |
| { | |
| "epoch": 13.965811965811966, | |
| "grad_norm": 4.112930774688721, | |
| "learning_rate": 3.0170940170940174e-06, | |
| "loss": 0.1026, | |
| "step": 1634 | |
| }, | |
| { | |
| "epoch": 13.974358974358974, | |
| "grad_norm": 4.55318021774292, | |
| "learning_rate": 3.012820512820513e-06, | |
| "loss": 0.2808, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 13.982905982905983, | |
| "grad_norm": 2.1912014484405518, | |
| "learning_rate": 3.0085470085470087e-06, | |
| "loss": 0.0906, | |
| "step": 1636 | |
| }, | |
| { | |
| "epoch": 13.991452991452991, | |
| "grad_norm": 4.612771511077881, | |
| "learning_rate": 3.004273504273504e-06, | |
| "loss": 0.17, | |
| "step": 1637 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 7.162411212921143, | |
| "learning_rate": 3e-06, | |
| "loss": 0.131, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.06268326193094254, | |
| "eval_runtime": 9.262, | |
| "eval_samples_per_second": 50.313, | |
| "eval_steps_per_second": 6.37, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 14.008547008547009, | |
| "grad_norm": 4.41022253036499, | |
| "learning_rate": 2.995726495726496e-06, | |
| "loss": 0.1989, | |
| "step": 1639 | |
| }, | |
| { | |
| "epoch": 14.017094017094017, | |
| "grad_norm": 2.2863216400146484, | |
| "learning_rate": 2.9914529914529914e-06, | |
| "loss": 0.0612, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 14.025641025641026, | |
| "grad_norm": 1.5455230474472046, | |
| "learning_rate": 2.9871794871794873e-06, | |
| "loss": 0.0378, | |
| "step": 1641 | |
| }, | |
| { | |
| "epoch": 14.034188034188034, | |
| "grad_norm": 0.9546025991439819, | |
| "learning_rate": 2.9829059829059832e-06, | |
| "loss": 0.0214, | |
| "step": 1642 | |
| }, | |
| { | |
| "epoch": 14.042735042735043, | |
| "grad_norm": 5.546824932098389, | |
| "learning_rate": 2.9786324786324787e-06, | |
| "loss": 0.2502, | |
| "step": 1643 | |
| }, | |
| { | |
| "epoch": 14.051282051282051, | |
| "grad_norm": 1.6261364221572876, | |
| "learning_rate": 2.9743589743589746e-06, | |
| "loss": 0.0271, | |
| "step": 1644 | |
| }, | |
| { | |
| "epoch": 14.05982905982906, | |
| "grad_norm": 1.710256814956665, | |
| "learning_rate": 2.9700854700854705e-06, | |
| "loss": 0.0582, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 14.068376068376068, | |
| "grad_norm": 1.2083494663238525, | |
| "learning_rate": 2.965811965811966e-06, | |
| "loss": 0.026, | |
| "step": 1646 | |
| }, | |
| { | |
| "epoch": 14.076923076923077, | |
| "grad_norm": 3.6400561332702637, | |
| "learning_rate": 2.961538461538462e-06, | |
| "loss": 0.0896, | |
| "step": 1647 | |
| }, | |
| { | |
| "epoch": 14.085470085470085, | |
| "grad_norm": 2.1084742546081543, | |
| "learning_rate": 2.9572649572649577e-06, | |
| "loss": 0.0269, | |
| "step": 1648 | |
| }, | |
| { | |
| "epoch": 14.094017094017094, | |
| "grad_norm": 1.5661289691925049, | |
| "learning_rate": 2.952991452991453e-06, | |
| "loss": 0.0401, | |
| "step": 1649 | |
| }, | |
| { | |
| "epoch": 14.102564102564102, | |
| "grad_norm": 23.358585357666016, | |
| "learning_rate": 2.948717948717949e-06, | |
| "loss": 0.2069, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 14.11111111111111, | |
| "grad_norm": 9.171899795532227, | |
| "learning_rate": 2.944444444444445e-06, | |
| "loss": 0.2842, | |
| "step": 1651 | |
| }, | |
| { | |
| "epoch": 14.11965811965812, | |
| "grad_norm": 1.3189946413040161, | |
| "learning_rate": 2.9401709401709404e-06, | |
| "loss": 0.0331, | |
| "step": 1652 | |
| }, | |
| { | |
| "epoch": 14.128205128205128, | |
| "grad_norm": 3.6144192218780518, | |
| "learning_rate": 2.9358974358974363e-06, | |
| "loss": 0.2069, | |
| "step": 1653 | |
| }, | |
| { | |
| "epoch": 14.136752136752136, | |
| "grad_norm": 2.764681577682495, | |
| "learning_rate": 2.931623931623932e-06, | |
| "loss": 0.0646, | |
| "step": 1654 | |
| }, | |
| { | |
| "epoch": 14.145299145299145, | |
| "grad_norm": 2.073028564453125, | |
| "learning_rate": 2.9273504273504277e-06, | |
| "loss": 0.1223, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 14.153846153846153, | |
| "grad_norm": 12.209549903869629, | |
| "learning_rate": 2.9230769230769236e-06, | |
| "loss": 0.1922, | |
| "step": 1656 | |
| }, | |
| { | |
| "epoch": 14.162393162393162, | |
| "grad_norm": 3.1137638092041016, | |
| "learning_rate": 2.918803418803419e-06, | |
| "loss": 0.2586, | |
| "step": 1657 | |
| }, | |
| { | |
| "epoch": 14.17094017094017, | |
| "grad_norm": 5.130307674407959, | |
| "learning_rate": 2.914529914529915e-06, | |
| "loss": 0.2695, | |
| "step": 1658 | |
| }, | |
| { | |
| "epoch": 14.179487179487179, | |
| "grad_norm": 3.475097894668579, | |
| "learning_rate": 2.910256410256411e-06, | |
| "loss": 0.2131, | |
| "step": 1659 | |
| }, | |
| { | |
| "epoch": 14.188034188034187, | |
| "grad_norm": 0.5851498246192932, | |
| "learning_rate": 2.9059829059829063e-06, | |
| "loss": 0.0167, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 14.196581196581196, | |
| "grad_norm": 1.795509934425354, | |
| "learning_rate": 2.901709401709402e-06, | |
| "loss": 0.0857, | |
| "step": 1661 | |
| }, | |
| { | |
| "epoch": 14.205128205128204, | |
| "grad_norm": 1.7123979330062866, | |
| "learning_rate": 2.897435897435898e-06, | |
| "loss": 0.0599, | |
| "step": 1662 | |
| }, | |
| { | |
| "epoch": 14.213675213675213, | |
| "grad_norm": 1.230388879776001, | |
| "learning_rate": 2.8931623931623935e-06, | |
| "loss": 0.0255, | |
| "step": 1663 | |
| }, | |
| { | |
| "epoch": 14.222222222222221, | |
| "grad_norm": 3.8747615814208984, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 0.1412, | |
| "step": 1664 | |
| }, | |
| { | |
| "epoch": 14.23076923076923, | |
| "grad_norm": 2.233584403991699, | |
| "learning_rate": 2.8846153846153845e-06, | |
| "loss": 0.068, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 14.239316239316238, | |
| "grad_norm": 5.327254772186279, | |
| "learning_rate": 2.8803418803418804e-06, | |
| "loss": 0.2616, | |
| "step": 1666 | |
| }, | |
| { | |
| "epoch": 14.247863247863247, | |
| "grad_norm": 6.126563549041748, | |
| "learning_rate": 2.8760683760683762e-06, | |
| "loss": 0.0931, | |
| "step": 1667 | |
| }, | |
| { | |
| "epoch": 14.256410256410255, | |
| "grad_norm": 1.4305050373077393, | |
| "learning_rate": 2.8717948717948717e-06, | |
| "loss": 0.0221, | |
| "step": 1668 | |
| }, | |
| { | |
| "epoch": 14.264957264957266, | |
| "grad_norm": 3.0924506187438965, | |
| "learning_rate": 2.8675213675213676e-06, | |
| "loss": 0.0417, | |
| "step": 1669 | |
| }, | |
| { | |
| "epoch": 14.273504273504274, | |
| "grad_norm": 2.548558235168457, | |
| "learning_rate": 2.8632478632478635e-06, | |
| "loss": 0.0744, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 14.282051282051283, | |
| "grad_norm": 0.46632057428359985, | |
| "learning_rate": 2.858974358974359e-06, | |
| "loss": 0.0114, | |
| "step": 1671 | |
| }, | |
| { | |
| "epoch": 14.290598290598291, | |
| "grad_norm": 2.5199391841888428, | |
| "learning_rate": 2.854700854700855e-06, | |
| "loss": 0.0819, | |
| "step": 1672 | |
| }, | |
| { | |
| "epoch": 14.2991452991453, | |
| "grad_norm": 1.849133014678955, | |
| "learning_rate": 2.8504273504273507e-06, | |
| "loss": 0.0424, | |
| "step": 1673 | |
| }, | |
| { | |
| "epoch": 14.307692307692308, | |
| "grad_norm": 2.9396777153015137, | |
| "learning_rate": 2.846153846153846e-06, | |
| "loss": 0.0836, | |
| "step": 1674 | |
| }, | |
| { | |
| "epoch": 14.316239316239317, | |
| "grad_norm": 0.7128950953483582, | |
| "learning_rate": 2.841880341880342e-06, | |
| "loss": 0.0181, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 14.324786324786325, | |
| "grad_norm": 2.1387767791748047, | |
| "learning_rate": 2.8376068376068376e-06, | |
| "loss": 0.0432, | |
| "step": 1676 | |
| }, | |
| { | |
| "epoch": 14.333333333333334, | |
| "grad_norm": 7.104556083679199, | |
| "learning_rate": 2.8333333333333335e-06, | |
| "loss": 0.1277, | |
| "step": 1677 | |
| }, | |
| { | |
| "epoch": 14.341880341880342, | |
| "grad_norm": 3.718749761581421, | |
| "learning_rate": 2.8290598290598293e-06, | |
| "loss": 0.0738, | |
| "step": 1678 | |
| }, | |
| { | |
| "epoch": 14.350427350427351, | |
| "grad_norm": 3.9387831687927246, | |
| "learning_rate": 2.824786324786325e-06, | |
| "loss": 0.1374, | |
| "step": 1679 | |
| }, | |
| { | |
| "epoch": 14.35897435897436, | |
| "grad_norm": 2.1527843475341797, | |
| "learning_rate": 2.8205128205128207e-06, | |
| "loss": 0.1426, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 14.367521367521368, | |
| "grad_norm": 1.0589011907577515, | |
| "learning_rate": 2.8162393162393166e-06, | |
| "loss": 0.0343, | |
| "step": 1681 | |
| }, | |
| { | |
| "epoch": 14.376068376068377, | |
| "grad_norm": 3.55014967918396, | |
| "learning_rate": 2.811965811965812e-06, | |
| "loss": 0.2962, | |
| "step": 1682 | |
| }, | |
| { | |
| "epoch": 14.384615384615385, | |
| "grad_norm": 3.996713399887085, | |
| "learning_rate": 2.807692307692308e-06, | |
| "loss": 0.1458, | |
| "step": 1683 | |
| }, | |
| { | |
| "epoch": 14.393162393162394, | |
| "grad_norm": 73.28384399414062, | |
| "learning_rate": 2.803418803418804e-06, | |
| "loss": 0.6138, | |
| "step": 1684 | |
| }, | |
| { | |
| "epoch": 14.401709401709402, | |
| "grad_norm": 5.780628681182861, | |
| "learning_rate": 2.7991452991452993e-06, | |
| "loss": 0.2619, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 14.41025641025641, | |
| "grad_norm": 3.2047317028045654, | |
| "learning_rate": 2.794871794871795e-06, | |
| "loss": 0.1917, | |
| "step": 1686 | |
| }, | |
| { | |
| "epoch": 14.418803418803419, | |
| "grad_norm": 7.041647434234619, | |
| "learning_rate": 2.790598290598291e-06, | |
| "loss": 0.2136, | |
| "step": 1687 | |
| }, | |
| { | |
| "epoch": 14.427350427350428, | |
| "grad_norm": 3.391404867172241, | |
| "learning_rate": 2.7863247863247866e-06, | |
| "loss": 0.094, | |
| "step": 1688 | |
| }, | |
| { | |
| "epoch": 14.435897435897436, | |
| "grad_norm": 0.5430964231491089, | |
| "learning_rate": 2.7820512820512824e-06, | |
| "loss": 0.0139, | |
| "step": 1689 | |
| }, | |
| { | |
| "epoch": 14.444444444444445, | |
| "grad_norm": 5.696547985076904, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 0.5808, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 14.452991452991453, | |
| "grad_norm": 3.5785481929779053, | |
| "learning_rate": 2.773504273504274e-06, | |
| "loss": 0.219, | |
| "step": 1691 | |
| }, | |
| { | |
| "epoch": 14.461538461538462, | |
| "grad_norm": 6.63624906539917, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.2586, | |
| "step": 1692 | |
| }, | |
| { | |
| "epoch": 14.47008547008547, | |
| "grad_norm": 16.79705810546875, | |
| "learning_rate": 2.764957264957265e-06, | |
| "loss": 0.1762, | |
| "step": 1693 | |
| }, | |
| { | |
| "epoch": 14.478632478632479, | |
| "grad_norm": 4.069973468780518, | |
| "learning_rate": 2.760683760683761e-06, | |
| "loss": 0.1191, | |
| "step": 1694 | |
| }, | |
| { | |
| "epoch": 14.487179487179487, | |
| "grad_norm": 1.1191340684890747, | |
| "learning_rate": 2.756410256410257e-06, | |
| "loss": 0.0529, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 14.495726495726496, | |
| "grad_norm": 2.23835825920105, | |
| "learning_rate": 2.7521367521367524e-06, | |
| "loss": 0.0681, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 14.504273504273504, | |
| "grad_norm": 2.745694160461426, | |
| "learning_rate": 2.7478632478632483e-06, | |
| "loss": 0.1885, | |
| "step": 1697 | |
| }, | |
| { | |
| "epoch": 14.512820512820513, | |
| "grad_norm": 3.642946720123291, | |
| "learning_rate": 2.743589743589744e-06, | |
| "loss": 0.2061, | |
| "step": 1698 | |
| }, | |
| { | |
| "epoch": 14.521367521367521, | |
| "grad_norm": 2.7571651935577393, | |
| "learning_rate": 2.7393162393162397e-06, | |
| "loss": 0.074, | |
| "step": 1699 | |
| }, | |
| { | |
| "epoch": 14.52991452991453, | |
| "grad_norm": 0.889057457447052, | |
| "learning_rate": 2.7350427350427355e-06, | |
| "loss": 0.0342, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 14.538461538461538, | |
| "grad_norm": 0.5471668243408203, | |
| "learning_rate": 2.7307692307692306e-06, | |
| "loss": 0.0125, | |
| "step": 1701 | |
| }, | |
| { | |
| "epoch": 14.547008547008547, | |
| "grad_norm": 6.883024215698242, | |
| "learning_rate": 2.7264957264957265e-06, | |
| "loss": 0.4102, | |
| "step": 1702 | |
| }, | |
| { | |
| "epoch": 14.555555555555555, | |
| "grad_norm": 2.6678171157836914, | |
| "learning_rate": 2.7222222222222224e-06, | |
| "loss": 0.0872, | |
| "step": 1703 | |
| }, | |
| { | |
| "epoch": 14.564102564102564, | |
| "grad_norm": 5.825995445251465, | |
| "learning_rate": 2.717948717948718e-06, | |
| "loss": 0.1081, | |
| "step": 1704 | |
| }, | |
| { | |
| "epoch": 14.572649572649572, | |
| "grad_norm": 1.5447179079055786, | |
| "learning_rate": 2.7136752136752137e-06, | |
| "loss": 0.0838, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 14.581196581196581, | |
| "grad_norm": 17.58099937438965, | |
| "learning_rate": 2.7094017094017096e-06, | |
| "loss": 0.6379, | |
| "step": 1706 | |
| }, | |
| { | |
| "epoch": 14.58974358974359, | |
| "grad_norm": 0.9537908434867859, | |
| "learning_rate": 2.705128205128205e-06, | |
| "loss": 0.0221, | |
| "step": 1707 | |
| }, | |
| { | |
| "epoch": 14.598290598290598, | |
| "grad_norm": 3.264037847518921, | |
| "learning_rate": 2.700854700854701e-06, | |
| "loss": 0.1282, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.606837606837606, | |
| "grad_norm": 1.7752703428268433, | |
| "learning_rate": 2.696581196581197e-06, | |
| "loss": 0.0194, | |
| "step": 1709 | |
| }, | |
| { | |
| "epoch": 14.615384615384615, | |
| "grad_norm": 4.8417649269104, | |
| "learning_rate": 2.6923076923076923e-06, | |
| "loss": 0.2217, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 14.623931623931623, | |
| "grad_norm": 2.915694236755371, | |
| "learning_rate": 2.6880341880341882e-06, | |
| "loss": 0.1506, | |
| "step": 1711 | |
| }, | |
| { | |
| "epoch": 14.632478632478632, | |
| "grad_norm": 10.983115196228027, | |
| "learning_rate": 2.6837606837606837e-06, | |
| "loss": 0.4307, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 14.64102564102564, | |
| "grad_norm": 1.1121952533721924, | |
| "learning_rate": 2.6794871794871796e-06, | |
| "loss": 0.0211, | |
| "step": 1713 | |
| }, | |
| { | |
| "epoch": 14.649572649572649, | |
| "grad_norm": 2.6676313877105713, | |
| "learning_rate": 2.6752136752136755e-06, | |
| "loss": 0.0997, | |
| "step": 1714 | |
| }, | |
| { | |
| "epoch": 14.658119658119658, | |
| "grad_norm": 1.718767523765564, | |
| "learning_rate": 2.670940170940171e-06, | |
| "loss": 0.0533, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 14.666666666666666, | |
| "grad_norm": 1.567866563796997, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.0913, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 14.675213675213675, | |
| "grad_norm": 3.0697431564331055, | |
| "learning_rate": 2.6623931623931627e-06, | |
| "loss": 0.1133, | |
| "step": 1717 | |
| }, | |
| { | |
| "epoch": 14.683760683760683, | |
| "grad_norm": 2.2237489223480225, | |
| "learning_rate": 2.658119658119658e-06, | |
| "loss": 0.1091, | |
| "step": 1718 | |
| }, | |
| { | |
| "epoch": 14.692307692307692, | |
| "grad_norm": 6.050041198730469, | |
| "learning_rate": 2.653846153846154e-06, | |
| "loss": 0.5622, | |
| "step": 1719 | |
| }, | |
| { | |
| "epoch": 14.7008547008547, | |
| "grad_norm": 1.1796153783798218, | |
| "learning_rate": 2.64957264957265e-06, | |
| "loss": 0.0522, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 14.709401709401709, | |
| "grad_norm": 2.4849863052368164, | |
| "learning_rate": 2.6452991452991454e-06, | |
| "loss": 0.0332, | |
| "step": 1721 | |
| }, | |
| { | |
| "epoch": 14.717948717948717, | |
| "grad_norm": 1.771933674812317, | |
| "learning_rate": 2.6410256410256413e-06, | |
| "loss": 0.0692, | |
| "step": 1722 | |
| }, | |
| { | |
| "epoch": 14.726495726495726, | |
| "grad_norm": 4.174441337585449, | |
| "learning_rate": 2.6367521367521372e-06, | |
| "loss": 0.1419, | |
| "step": 1723 | |
| }, | |
| { | |
| "epoch": 14.735042735042736, | |
| "grad_norm": 4.145920276641846, | |
| "learning_rate": 2.6324786324786327e-06, | |
| "loss": 0.5196, | |
| "step": 1724 | |
| }, | |
| { | |
| "epoch": 14.743589743589745, | |
| "grad_norm": 3.363537073135376, | |
| "learning_rate": 2.6282051282051286e-06, | |
| "loss": 0.1187, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 14.752136752136753, | |
| "grad_norm": 1.9558751583099365, | |
| "learning_rate": 2.6239316239316245e-06, | |
| "loss": 0.0193, | |
| "step": 1726 | |
| }, | |
| { | |
| "epoch": 14.760683760683762, | |
| "grad_norm": 2.8293466567993164, | |
| "learning_rate": 2.61965811965812e-06, | |
| "loss": 0.0551, | |
| "step": 1727 | |
| }, | |
| { | |
| "epoch": 14.76923076923077, | |
| "grad_norm": 1.2654905319213867, | |
| "learning_rate": 2.615384615384616e-06, | |
| "loss": 0.0805, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 14.777777777777779, | |
| "grad_norm": 0.9344054460525513, | |
| "learning_rate": 2.6111111111111113e-06, | |
| "loss": 0.0177, | |
| "step": 1729 | |
| }, | |
| { | |
| "epoch": 14.786324786324787, | |
| "grad_norm": 1.268433690071106, | |
| "learning_rate": 2.606837606837607e-06, | |
| "loss": 0.0185, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 14.794871794871796, | |
| "grad_norm": 2.5544192790985107, | |
| "learning_rate": 2.602564102564103e-06, | |
| "loss": 0.063, | |
| "step": 1731 | |
| }, | |
| { | |
| "epoch": 14.803418803418804, | |
| "grad_norm": 2.1078386306762695, | |
| "learning_rate": 2.5982905982905985e-06, | |
| "loss": 0.1203, | |
| "step": 1732 | |
| }, | |
| { | |
| "epoch": 14.811965811965813, | |
| "grad_norm": 1.526848554611206, | |
| "learning_rate": 2.5940170940170944e-06, | |
| "loss": 0.0524, | |
| "step": 1733 | |
| }, | |
| { | |
| "epoch": 14.820512820512821, | |
| "grad_norm": 0.7479220628738403, | |
| "learning_rate": 2.5897435897435903e-06, | |
| "loss": 0.0197, | |
| "step": 1734 | |
| }, | |
| { | |
| "epoch": 14.82905982905983, | |
| "grad_norm": 2.937556266784668, | |
| "learning_rate": 2.5854700854700858e-06, | |
| "loss": 0.1406, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 14.837606837606838, | |
| "grad_norm": 2.3128576278686523, | |
| "learning_rate": 2.5811965811965817e-06, | |
| "loss": 0.056, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 14.846153846153847, | |
| "grad_norm": 2.1093039512634277, | |
| "learning_rate": 2.5769230769230767e-06, | |
| "loss": 0.0645, | |
| "step": 1737 | |
| }, | |
| { | |
| "epoch": 14.854700854700855, | |
| "grad_norm": 2.104214668273926, | |
| "learning_rate": 2.5726495726495726e-06, | |
| "loss": 0.1097, | |
| "step": 1738 | |
| }, | |
| { | |
| "epoch": 14.863247863247864, | |
| "grad_norm": 3.781390428543091, | |
| "learning_rate": 2.5683760683760685e-06, | |
| "loss": 0.1214, | |
| "step": 1739 | |
| }, | |
| { | |
| "epoch": 14.871794871794872, | |
| "grad_norm": 4.119661331176758, | |
| "learning_rate": 2.564102564102564e-06, | |
| "loss": 0.1797, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 14.88034188034188, | |
| "grad_norm": 6.488205909729004, | |
| "learning_rate": 2.55982905982906e-06, | |
| "loss": 0.0679, | |
| "step": 1741 | |
| }, | |
| { | |
| "epoch": 14.88888888888889, | |
| "grad_norm": 1.4211604595184326, | |
| "learning_rate": 2.5555555555555557e-06, | |
| "loss": 0.0375, | |
| "step": 1742 | |
| }, | |
| { | |
| "epoch": 14.897435897435898, | |
| "grad_norm": 3.577533721923828, | |
| "learning_rate": 2.5512820512820512e-06, | |
| "loss": 0.1914, | |
| "step": 1743 | |
| }, | |
| { | |
| "epoch": 14.905982905982906, | |
| "grad_norm": 8.697205543518066, | |
| "learning_rate": 2.547008547008547e-06, | |
| "loss": 0.5511, | |
| "step": 1744 | |
| }, | |
| { | |
| "epoch": 14.914529914529915, | |
| "grad_norm": 0.49716269969940186, | |
| "learning_rate": 2.542735042735043e-06, | |
| "loss": 0.0125, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 14.923076923076923, | |
| "grad_norm": 2.8563008308410645, | |
| "learning_rate": 2.5384615384615385e-06, | |
| "loss": 0.0901, | |
| "step": 1746 | |
| }, | |
| { | |
| "epoch": 14.931623931623932, | |
| "grad_norm": 3.6407926082611084, | |
| "learning_rate": 2.5341880341880344e-06, | |
| "loss": 0.0718, | |
| "step": 1747 | |
| }, | |
| { | |
| "epoch": 14.94017094017094, | |
| "grad_norm": 1.2601441144943237, | |
| "learning_rate": 2.52991452991453e-06, | |
| "loss": 0.0451, | |
| "step": 1748 | |
| }, | |
| { | |
| "epoch": 14.948717948717949, | |
| "grad_norm": 2.4402401447296143, | |
| "learning_rate": 2.5256410256410257e-06, | |
| "loss": 0.0771, | |
| "step": 1749 | |
| }, | |
| { | |
| "epoch": 14.957264957264957, | |
| "grad_norm": 0.6150484681129456, | |
| "learning_rate": 2.5213675213675216e-06, | |
| "loss": 0.0151, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 14.965811965811966, | |
| "grad_norm": 3.6569836139678955, | |
| "learning_rate": 2.517094017094017e-06, | |
| "loss": 0.0905, | |
| "step": 1751 | |
| }, | |
| { | |
| "epoch": 14.974358974358974, | |
| "grad_norm": 3.4421300888061523, | |
| "learning_rate": 2.512820512820513e-06, | |
| "loss": 0.0456, | |
| "step": 1752 | |
| }, | |
| { | |
| "epoch": 14.982905982905983, | |
| "grad_norm": 3.565871477127075, | |
| "learning_rate": 2.508547008547009e-06, | |
| "loss": 0.0491, | |
| "step": 1753 | |
| }, | |
| { | |
| "epoch": 14.991452991452991, | |
| "grad_norm": 37.519065856933594, | |
| "learning_rate": 2.5042735042735043e-06, | |
| "loss": 0.1348, | |
| "step": 1754 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 5.1902899742126465, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1099, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.05930963531136513, | |
| "eval_runtime": 9.2206, | |
| "eval_samples_per_second": 50.539, | |
| "eval_steps_per_second": 6.399, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 15.008547008547009, | |
| "grad_norm": 5.6569342613220215, | |
| "learning_rate": 2.495726495726496e-06, | |
| "loss": 0.1931, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 15.017094017094017, | |
| "grad_norm": 5.23728084564209, | |
| "learning_rate": 2.4914529914529916e-06, | |
| "loss": 0.2789, | |
| "step": 1757 | |
| }, | |
| { | |
| "epoch": 15.025641025641026, | |
| "grad_norm": 0.8648807406425476, | |
| "learning_rate": 2.4871794871794875e-06, | |
| "loss": 0.0227, | |
| "step": 1758 | |
| }, | |
| { | |
| "epoch": 15.034188034188034, | |
| "grad_norm": 3.0654587745666504, | |
| "learning_rate": 2.4829059829059833e-06, | |
| "loss": 0.0602, | |
| "step": 1759 | |
| }, | |
| { | |
| "epoch": 15.042735042735043, | |
| "grad_norm": 4.374608039855957, | |
| "learning_rate": 2.478632478632479e-06, | |
| "loss": 0.2133, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 15.051282051282051, | |
| "grad_norm": 1.2764301300048828, | |
| "learning_rate": 2.4743589743589747e-06, | |
| "loss": 0.0296, | |
| "step": 1761 | |
| }, | |
| { | |
| "epoch": 15.05982905982906, | |
| "grad_norm": 0.9672349095344543, | |
| "learning_rate": 2.4700854700854706e-06, | |
| "loss": 0.0224, | |
| "step": 1762 | |
| }, | |
| { | |
| "epoch": 15.068376068376068, | |
| "grad_norm": 8.807465553283691, | |
| "learning_rate": 2.465811965811966e-06, | |
| "loss": 0.0925, | |
| "step": 1763 | |
| }, | |
| { | |
| "epoch": 15.076923076923077, | |
| "grad_norm": 1.4733474254608154, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 0.0286, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 15.085470085470085, | |
| "grad_norm": 6.014289855957031, | |
| "learning_rate": 2.4572649572649574e-06, | |
| "loss": 0.1387, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 15.094017094017094, | |
| "grad_norm": 1.899086356163025, | |
| "learning_rate": 2.452991452991453e-06, | |
| "loss": 0.07, | |
| "step": 1766 | |
| }, | |
| { | |
| "epoch": 15.102564102564102, | |
| "grad_norm": 11.32197380065918, | |
| "learning_rate": 2.4487179487179488e-06, | |
| "loss": 0.2452, | |
| "step": 1767 | |
| }, | |
| { | |
| "epoch": 15.11111111111111, | |
| "grad_norm": 3.223996639251709, | |
| "learning_rate": 2.4444444444444447e-06, | |
| "loss": 0.139, | |
| "step": 1768 | |
| }, | |
| { | |
| "epoch": 15.11965811965812, | |
| "grad_norm": 2.8729913234710693, | |
| "learning_rate": 2.44017094017094e-06, | |
| "loss": 0.1386, | |
| "step": 1769 | |
| }, | |
| { | |
| "epoch": 15.128205128205128, | |
| "grad_norm": 1.9730579853057861, | |
| "learning_rate": 2.435897435897436e-06, | |
| "loss": 0.0882, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 15.136752136752136, | |
| "grad_norm": 5.556413650512695, | |
| "learning_rate": 2.431623931623932e-06, | |
| "loss": 0.1554, | |
| "step": 1771 | |
| }, | |
| { | |
| "epoch": 15.145299145299145, | |
| "grad_norm": 1.2356898784637451, | |
| "learning_rate": 2.4273504273504274e-06, | |
| "loss": 0.0217, | |
| "step": 1772 | |
| }, | |
| { | |
| "epoch": 15.153846153846153, | |
| "grad_norm": 7.849127769470215, | |
| "learning_rate": 2.4230769230769233e-06, | |
| "loss": 0.221, | |
| "step": 1773 | |
| }, | |
| { | |
| "epoch": 15.162393162393162, | |
| "grad_norm": 0.5792569518089294, | |
| "learning_rate": 2.418803418803419e-06, | |
| "loss": 0.017, | |
| "step": 1774 | |
| }, | |
| { | |
| "epoch": 15.17094017094017, | |
| "grad_norm": 2.2549376487731934, | |
| "learning_rate": 2.4145299145299146e-06, | |
| "loss": 0.0499, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 15.179487179487179, | |
| "grad_norm": 2.722200870513916, | |
| "learning_rate": 2.4102564102564105e-06, | |
| "loss": 0.0408, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 15.188034188034187, | |
| "grad_norm": 3.1140944957733154, | |
| "learning_rate": 2.4059829059829064e-06, | |
| "loss": 0.1001, | |
| "step": 1777 | |
| }, | |
| { | |
| "epoch": 15.196581196581196, | |
| "grad_norm": 4.461791515350342, | |
| "learning_rate": 2.401709401709402e-06, | |
| "loss": 0.3419, | |
| "step": 1778 | |
| }, | |
| { | |
| "epoch": 15.205128205128204, | |
| "grad_norm": 1.8562372922897339, | |
| "learning_rate": 2.3974358974358978e-06, | |
| "loss": 0.1092, | |
| "step": 1779 | |
| }, | |
| { | |
| "epoch": 15.213675213675213, | |
| "grad_norm": 5.2086181640625, | |
| "learning_rate": 2.3931623931623937e-06, | |
| "loss": 0.1767, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 15.222222222222221, | |
| "grad_norm": 1.6226582527160645, | |
| "learning_rate": 2.388888888888889e-06, | |
| "loss": 0.0347, | |
| "step": 1781 | |
| }, | |
| { | |
| "epoch": 15.23076923076923, | |
| "grad_norm": 2.8507306575775146, | |
| "learning_rate": 2.384615384615385e-06, | |
| "loss": 0.0934, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 15.239316239316238, | |
| "grad_norm": 2.74642276763916, | |
| "learning_rate": 2.3803418803418805e-06, | |
| "loss": 0.0857, | |
| "step": 1783 | |
| }, | |
| { | |
| "epoch": 15.247863247863247, | |
| "grad_norm": 3.4352660179138184, | |
| "learning_rate": 2.376068376068376e-06, | |
| "loss": 0.2336, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 15.256410256410255, | |
| "grad_norm": 3.4673473834991455, | |
| "learning_rate": 2.371794871794872e-06, | |
| "loss": 0.1974, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 15.264957264957266, | |
| "grad_norm": 21.467744827270508, | |
| "learning_rate": 2.3675213675213677e-06, | |
| "loss": 0.6836, | |
| "step": 1786 | |
| }, | |
| { | |
| "epoch": 15.273504273504274, | |
| "grad_norm": 2.832465887069702, | |
| "learning_rate": 2.363247863247863e-06, | |
| "loss": 0.245, | |
| "step": 1787 | |
| }, | |
| { | |
| "epoch": 15.282051282051283, | |
| "grad_norm": 9.717825889587402, | |
| "learning_rate": 2.358974358974359e-06, | |
| "loss": 0.5324, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 15.290598290598291, | |
| "grad_norm": 2.209528923034668, | |
| "learning_rate": 2.354700854700855e-06, | |
| "loss": 0.0854, | |
| "step": 1789 | |
| }, | |
| { | |
| "epoch": 15.2991452991453, | |
| "grad_norm": 4.554971218109131, | |
| "learning_rate": 2.3504273504273504e-06, | |
| "loss": 0.1271, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 15.307692307692308, | |
| "grad_norm": 3.1280457973480225, | |
| "learning_rate": 2.3461538461538463e-06, | |
| "loss": 0.1265, | |
| "step": 1791 | |
| }, | |
| { | |
| "epoch": 15.316239316239317, | |
| "grad_norm": 2.647224187850952, | |
| "learning_rate": 2.3418803418803422e-06, | |
| "loss": 0.1965, | |
| "step": 1792 | |
| }, | |
| { | |
| "epoch": 15.324786324786325, | |
| "grad_norm": 2.7695155143737793, | |
| "learning_rate": 2.3376068376068377e-06, | |
| "loss": 0.0528, | |
| "step": 1793 | |
| }, | |
| { | |
| "epoch": 15.333333333333334, | |
| "grad_norm": 20.151025772094727, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 0.2011, | |
| "step": 1794 | |
| }, | |
| { | |
| "epoch": 15.341880341880342, | |
| "grad_norm": 2.8718080520629883, | |
| "learning_rate": 2.3290598290598295e-06, | |
| "loss": 0.0502, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 15.350427350427351, | |
| "grad_norm": 2.17462158203125, | |
| "learning_rate": 2.324786324786325e-06, | |
| "loss": 0.0658, | |
| "step": 1796 | |
| }, | |
| { | |
| "epoch": 15.35897435897436, | |
| "grad_norm": 4.324810981750488, | |
| "learning_rate": 2.320512820512821e-06, | |
| "loss": 0.1429, | |
| "step": 1797 | |
| }, | |
| { | |
| "epoch": 15.367521367521368, | |
| "grad_norm": 184.52798461914062, | |
| "learning_rate": 2.3162393162393167e-06, | |
| "loss": 0.5155, | |
| "step": 1798 | |
| }, | |
| { | |
| "epoch": 15.376068376068377, | |
| "grad_norm": 2.6076488494873047, | |
| "learning_rate": 2.311965811965812e-06, | |
| "loss": 0.0708, | |
| "step": 1799 | |
| }, | |
| { | |
| "epoch": 15.384615384615385, | |
| "grad_norm": 3.0682790279388428, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 0.2662, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.393162393162394, | |
| "grad_norm": 1.3366855382919312, | |
| "learning_rate": 2.3034188034188035e-06, | |
| "loss": 0.0136, | |
| "step": 1801 | |
| }, | |
| { | |
| "epoch": 15.401709401709402, | |
| "grad_norm": 0.5489670634269714, | |
| "learning_rate": 2.299145299145299e-06, | |
| "loss": 0.0148, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 15.41025641025641, | |
| "grad_norm": 1.080804705619812, | |
| "learning_rate": 2.294871794871795e-06, | |
| "loss": 0.025, | |
| "step": 1803 | |
| }, | |
| { | |
| "epoch": 15.418803418803419, | |
| "grad_norm": 8.801629066467285, | |
| "learning_rate": 2.290598290598291e-06, | |
| "loss": 0.2038, | |
| "step": 1804 | |
| }, | |
| { | |
| "epoch": 15.427350427350428, | |
| "grad_norm": 66.96419525146484, | |
| "learning_rate": 2.2863247863247863e-06, | |
| "loss": 0.4094, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 15.435897435897436, | |
| "grad_norm": 1.3400782346725464, | |
| "learning_rate": 2.282051282051282e-06, | |
| "loss": 0.0452, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 15.444444444444445, | |
| "grad_norm": 3.5850300788879395, | |
| "learning_rate": 2.277777777777778e-06, | |
| "loss": 0.0919, | |
| "step": 1807 | |
| }, | |
| { | |
| "epoch": 15.452991452991453, | |
| "grad_norm": 8.670539855957031, | |
| "learning_rate": 2.2735042735042735e-06, | |
| "loss": 0.255, | |
| "step": 1808 | |
| }, | |
| { | |
| "epoch": 15.461538461538462, | |
| "grad_norm": 3.609617233276367, | |
| "learning_rate": 2.2692307692307694e-06, | |
| "loss": 0.1203, | |
| "step": 1809 | |
| }, | |
| { | |
| "epoch": 15.47008547008547, | |
| "grad_norm": 1.5857924222946167, | |
| "learning_rate": 2.2649572649572653e-06, | |
| "loss": 0.0371, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 15.478632478632479, | |
| "grad_norm": 1.386805534362793, | |
| "learning_rate": 2.2606837606837608e-06, | |
| "loss": 0.0385, | |
| "step": 1811 | |
| }, | |
| { | |
| "epoch": 15.487179487179487, | |
| "grad_norm": 4.130802631378174, | |
| "learning_rate": 2.2564102564102566e-06, | |
| "loss": 0.2261, | |
| "step": 1812 | |
| }, | |
| { | |
| "epoch": 15.495726495726496, | |
| "grad_norm": 2.974247455596924, | |
| "learning_rate": 2.2521367521367525e-06, | |
| "loss": 0.0651, | |
| "step": 1813 | |
| }, | |
| { | |
| "epoch": 15.504273504273504, | |
| "grad_norm": 1.2551554441452026, | |
| "learning_rate": 2.247863247863248e-06, | |
| "loss": 0.0229, | |
| "step": 1814 | |
| }, | |
| { | |
| "epoch": 15.512820512820513, | |
| "grad_norm": 3.1401453018188477, | |
| "learning_rate": 2.243589743589744e-06, | |
| "loss": 0.0409, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 15.521367521367521, | |
| "grad_norm": 1.3921948671340942, | |
| "learning_rate": 2.2393162393162398e-06, | |
| "loss": 0.0335, | |
| "step": 1816 | |
| }, | |
| { | |
| "epoch": 15.52991452991453, | |
| "grad_norm": 5.457981586456299, | |
| "learning_rate": 2.2350427350427353e-06, | |
| "loss": 0.22, | |
| "step": 1817 | |
| }, | |
| { | |
| "epoch": 15.538461538461538, | |
| "grad_norm": 0.9100427031517029, | |
| "learning_rate": 2.230769230769231e-06, | |
| "loss": 0.0217, | |
| "step": 1818 | |
| }, | |
| { | |
| "epoch": 15.547008547008547, | |
| "grad_norm": 3.5890519618988037, | |
| "learning_rate": 2.2264957264957266e-06, | |
| "loss": 0.2241, | |
| "step": 1819 | |
| }, | |
| { | |
| "epoch": 15.555555555555555, | |
| "grad_norm": 2.965954303741455, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.1453, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 15.564102564102564, | |
| "grad_norm": 8.436135292053223, | |
| "learning_rate": 2.217948717948718e-06, | |
| "loss": 0.2784, | |
| "step": 1821 | |
| }, | |
| { | |
| "epoch": 15.572649572649572, | |
| "grad_norm": 2.043687582015991, | |
| "learning_rate": 2.213675213675214e-06, | |
| "loss": 0.0755, | |
| "step": 1822 | |
| }, | |
| { | |
| "epoch": 15.581196581196581, | |
| "grad_norm": 2.380276918411255, | |
| "learning_rate": 2.2094017094017093e-06, | |
| "loss": 0.1867, | |
| "step": 1823 | |
| }, | |
| { | |
| "epoch": 15.58974358974359, | |
| "grad_norm": 2.5189390182495117, | |
| "learning_rate": 2.2051282051282052e-06, | |
| "loss": 0.0619, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 15.598290598290598, | |
| "grad_norm": 1.123610258102417, | |
| "learning_rate": 2.200854700854701e-06, | |
| "loss": 0.0286, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 15.606837606837606, | |
| "grad_norm": 3.0018534660339355, | |
| "learning_rate": 2.1965811965811966e-06, | |
| "loss": 0.1449, | |
| "step": 1826 | |
| }, | |
| { | |
| "epoch": 15.615384615384615, | |
| "grad_norm": 2.178926706314087, | |
| "learning_rate": 2.1923076923076925e-06, | |
| "loss": 0.0859, | |
| "step": 1827 | |
| }, | |
| { | |
| "epoch": 15.623931623931623, | |
| "grad_norm": 5.799438953399658, | |
| "learning_rate": 2.1880341880341884e-06, | |
| "loss": 0.2669, | |
| "step": 1828 | |
| }, | |
| { | |
| "epoch": 15.632478632478632, | |
| "grad_norm": 2.0338144302368164, | |
| "learning_rate": 2.183760683760684e-06, | |
| "loss": 0.0616, | |
| "step": 1829 | |
| }, | |
| { | |
| "epoch": 15.64102564102564, | |
| "grad_norm": 3.789525032043457, | |
| "learning_rate": 2.1794871794871797e-06, | |
| "loss": 0.0439, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.649572649572649, | |
| "grad_norm": 2.3695919513702393, | |
| "learning_rate": 2.1752136752136756e-06, | |
| "loss": 0.0979, | |
| "step": 1831 | |
| }, | |
| { | |
| "epoch": 15.658119658119658, | |
| "grad_norm": 0.8543546795845032, | |
| "learning_rate": 2.170940170940171e-06, | |
| "loss": 0.0171, | |
| "step": 1832 | |
| }, | |
| { | |
| "epoch": 15.666666666666666, | |
| "grad_norm": 3.7921054363250732, | |
| "learning_rate": 2.166666666666667e-06, | |
| "loss": 0.1094, | |
| "step": 1833 | |
| }, | |
| { | |
| "epoch": 15.675213675213675, | |
| "grad_norm": 1.9967904090881348, | |
| "learning_rate": 2.162393162393163e-06, | |
| "loss": 0.0382, | |
| "step": 1834 | |
| }, | |
| { | |
| "epoch": 15.683760683760683, | |
| "grad_norm": 2.5073959827423096, | |
| "learning_rate": 2.1581196581196583e-06, | |
| "loss": 0.0554, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 15.692307692307692, | |
| "grad_norm": 1.2741888761520386, | |
| "learning_rate": 2.153846153846154e-06, | |
| "loss": 0.056, | |
| "step": 1836 | |
| }, | |
| { | |
| "epoch": 15.7008547008547, | |
| "grad_norm": 1.992280125617981, | |
| "learning_rate": 2.1495726495726497e-06, | |
| "loss": 0.0206, | |
| "step": 1837 | |
| }, | |
| { | |
| "epoch": 15.709401709401709, | |
| "grad_norm": 1.0176990032196045, | |
| "learning_rate": 2.145299145299145e-06, | |
| "loss": 0.0276, | |
| "step": 1838 | |
| }, | |
| { | |
| "epoch": 15.717948717948717, | |
| "grad_norm": 1.6685941219329834, | |
| "learning_rate": 2.141025641025641e-06, | |
| "loss": 0.0222, | |
| "step": 1839 | |
| }, | |
| { | |
| "epoch": 15.726495726495726, | |
| "grad_norm": 3.171050548553467, | |
| "learning_rate": 2.136752136752137e-06, | |
| "loss": 0.1526, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 15.735042735042736, | |
| "grad_norm": 1.5068336725234985, | |
| "learning_rate": 2.1324786324786324e-06, | |
| "loss": 0.0271, | |
| "step": 1841 | |
| }, | |
| { | |
| "epoch": 15.743589743589745, | |
| "grad_norm": 3.171870708465576, | |
| "learning_rate": 2.1282051282051283e-06, | |
| "loss": 0.0628, | |
| "step": 1842 | |
| }, | |
| { | |
| "epoch": 15.752136752136753, | |
| "grad_norm": 1.9212791919708252, | |
| "learning_rate": 2.123931623931624e-06, | |
| "loss": 0.1018, | |
| "step": 1843 | |
| }, | |
| { | |
| "epoch": 15.760683760683762, | |
| "grad_norm": 4.073456287384033, | |
| "learning_rate": 2.1196581196581196e-06, | |
| "loss": 0.1144, | |
| "step": 1844 | |
| }, | |
| { | |
| "epoch": 15.76923076923077, | |
| "grad_norm": 1.8453985452651978, | |
| "learning_rate": 2.1153846153846155e-06, | |
| "loss": 0.0995, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 15.777777777777779, | |
| "grad_norm": 3.285759210586548, | |
| "learning_rate": 2.1111111111111114e-06, | |
| "loss": 0.1173, | |
| "step": 1846 | |
| }, | |
| { | |
| "epoch": 15.786324786324787, | |
| "grad_norm": 3.709202289581299, | |
| "learning_rate": 2.106837606837607e-06, | |
| "loss": 0.1906, | |
| "step": 1847 | |
| }, | |
| { | |
| "epoch": 15.794871794871796, | |
| "grad_norm": 1.951262354850769, | |
| "learning_rate": 2.1025641025641028e-06, | |
| "loss": 0.0954, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 15.803418803418804, | |
| "grad_norm": 3.249171257019043, | |
| "learning_rate": 2.0982905982905987e-06, | |
| "loss": 0.1258, | |
| "step": 1849 | |
| }, | |
| { | |
| "epoch": 15.811965811965813, | |
| "grad_norm": 0.5708752274513245, | |
| "learning_rate": 2.094017094017094e-06, | |
| "loss": 0.0128, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 15.820512820512821, | |
| "grad_norm": 3.2894484996795654, | |
| "learning_rate": 2.08974358974359e-06, | |
| "loss": 0.0621, | |
| "step": 1851 | |
| }, | |
| { | |
| "epoch": 15.82905982905983, | |
| "grad_norm": 0.8564540147781372, | |
| "learning_rate": 2.085470085470086e-06, | |
| "loss": 0.0194, | |
| "step": 1852 | |
| }, | |
| { | |
| "epoch": 15.837606837606838, | |
| "grad_norm": 3.319011926651001, | |
| "learning_rate": 2.0811965811965814e-06, | |
| "loss": 0.1413, | |
| "step": 1853 | |
| }, | |
| { | |
| "epoch": 15.846153846153847, | |
| "grad_norm": 1.5385066270828247, | |
| "learning_rate": 2.0769230769230773e-06, | |
| "loss": 0.0316, | |
| "step": 1854 | |
| }, | |
| { | |
| "epoch": 15.854700854700855, | |
| "grad_norm": 4.076297283172607, | |
| "learning_rate": 2.072649572649573e-06, | |
| "loss": 0.2257, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 15.863247863247864, | |
| "grad_norm": 4.738671779632568, | |
| "learning_rate": 2.068376068376068e-06, | |
| "loss": 0.1627, | |
| "step": 1856 | |
| }, | |
| { | |
| "epoch": 15.871794871794872, | |
| "grad_norm": 5.589550495147705, | |
| "learning_rate": 2.064102564102564e-06, | |
| "loss": 0.3182, | |
| "step": 1857 | |
| }, | |
| { | |
| "epoch": 15.88034188034188, | |
| "grad_norm": 1.6303757429122925, | |
| "learning_rate": 2.05982905982906e-06, | |
| "loss": 0.0384, | |
| "step": 1858 | |
| }, | |
| { | |
| "epoch": 15.88888888888889, | |
| "grad_norm": 3.0257458686828613, | |
| "learning_rate": 2.0555555555555555e-06, | |
| "loss": 0.0967, | |
| "step": 1859 | |
| }, | |
| { | |
| "epoch": 15.897435897435898, | |
| "grad_norm": 2.4926559925079346, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.0703, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 15.905982905982906, | |
| "grad_norm": 2.0784358978271484, | |
| "learning_rate": 2.0470085470085472e-06, | |
| "loss": 0.062, | |
| "step": 1861 | |
| }, | |
| { | |
| "epoch": 15.914529914529915, | |
| "grad_norm": 4.92131233215332, | |
| "learning_rate": 2.0427350427350427e-06, | |
| "loss": 0.0875, | |
| "step": 1862 | |
| }, | |
| { | |
| "epoch": 15.923076923076923, | |
| "grad_norm": 2.999511241912842, | |
| "learning_rate": 2.0384615384615386e-06, | |
| "loss": 0.0388, | |
| "step": 1863 | |
| }, | |
| { | |
| "epoch": 15.931623931623932, | |
| "grad_norm": 5.770095348358154, | |
| "learning_rate": 2.0341880341880345e-06, | |
| "loss": 0.1257, | |
| "step": 1864 | |
| }, | |
| { | |
| "epoch": 15.94017094017094, | |
| "grad_norm": 4.730950832366943, | |
| "learning_rate": 2.02991452991453e-06, | |
| "loss": 0.2386, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 15.948717948717949, | |
| "grad_norm": 1.8125661611557007, | |
| "learning_rate": 2.025641025641026e-06, | |
| "loss": 0.0433, | |
| "step": 1866 | |
| }, | |
| { | |
| "epoch": 15.957264957264957, | |
| "grad_norm": 5.433501243591309, | |
| "learning_rate": 2.0213675213675217e-06, | |
| "loss": 0.0536, | |
| "step": 1867 | |
| }, | |
| { | |
| "epoch": 15.965811965811966, | |
| "grad_norm": 1.2565219402313232, | |
| "learning_rate": 2.017094017094017e-06, | |
| "loss": 0.0263, | |
| "step": 1868 | |
| }, | |
| { | |
| "epoch": 15.974358974358974, | |
| "grad_norm": 1.5660192966461182, | |
| "learning_rate": 2.012820512820513e-06, | |
| "loss": 0.0387, | |
| "step": 1869 | |
| }, | |
| { | |
| "epoch": 15.982905982905983, | |
| "grad_norm": 5.742929935455322, | |
| "learning_rate": 2.008547008547009e-06, | |
| "loss": 0.2158, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 15.991452991452991, | |
| "grad_norm": 3.597506284713745, | |
| "learning_rate": 2.0042735042735044e-06, | |
| "loss": 0.0962, | |
| "step": 1871 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.753219485282898, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.0193, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.05589358136057854, | |
| "eval_runtime": 9.2203, | |
| "eval_samples_per_second": 50.541, | |
| "eval_steps_per_second": 6.399, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 16.00854700854701, | |
| "grad_norm": 9.627431869506836, | |
| "learning_rate": 1.9957264957264962e-06, | |
| "loss": 0.4748, | |
| "step": 1873 | |
| }, | |
| { | |
| "epoch": 16.017094017094017, | |
| "grad_norm": 7.770556926727295, | |
| "learning_rate": 1.9914529914529917e-06, | |
| "loss": 0.2615, | |
| "step": 1874 | |
| }, | |
| { | |
| "epoch": 16.025641025641026, | |
| "grad_norm": 1.7268822193145752, | |
| "learning_rate": 1.987179487179487e-06, | |
| "loss": 0.0808, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 16.034188034188034, | |
| "grad_norm": 1.7209370136260986, | |
| "learning_rate": 1.982905982905983e-06, | |
| "loss": 0.0575, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 16.042735042735043, | |
| "grad_norm": 2.6422786712646484, | |
| "learning_rate": 1.9786324786324785e-06, | |
| "loss": 0.0815, | |
| "step": 1877 | |
| }, | |
| { | |
| "epoch": 16.05128205128205, | |
| "grad_norm": 0.9057373404502869, | |
| "learning_rate": 1.9743589743589744e-06, | |
| "loss": 0.0359, | |
| "step": 1878 | |
| }, | |
| { | |
| "epoch": 16.05982905982906, | |
| "grad_norm": 1.4879076480865479, | |
| "learning_rate": 1.9700854700854703e-06, | |
| "loss": 0.0658, | |
| "step": 1879 | |
| }, | |
| { | |
| "epoch": 16.068376068376068, | |
| "grad_norm": 2.1336488723754883, | |
| "learning_rate": 1.9658119658119658e-06, | |
| "loss": 0.0434, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 16.076923076923077, | |
| "grad_norm": 2.642249822616577, | |
| "learning_rate": 1.9615384615384617e-06, | |
| "loss": 0.0768, | |
| "step": 1881 | |
| }, | |
| { | |
| "epoch": 16.085470085470085, | |
| "grad_norm": 398.1800842285156, | |
| "learning_rate": 1.9572649572649575e-06, | |
| "loss": 1.7061, | |
| "step": 1882 | |
| }, | |
| { | |
| "epoch": 16.094017094017094, | |
| "grad_norm": 1.6067556142807007, | |
| "learning_rate": 1.952991452991453e-06, | |
| "loss": 0.0492, | |
| "step": 1883 | |
| }, | |
| { | |
| "epoch": 16.102564102564102, | |
| "grad_norm": 45.67499542236328, | |
| "learning_rate": 1.948717948717949e-06, | |
| "loss": 0.2883, | |
| "step": 1884 | |
| }, | |
| { | |
| "epoch": 16.11111111111111, | |
| "grad_norm": 5.477624416351318, | |
| "learning_rate": 1.944444444444445e-06, | |
| "loss": 0.1107, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 16.11965811965812, | |
| "grad_norm": 2.2795376777648926, | |
| "learning_rate": 1.9401709401709403e-06, | |
| "loss": 0.0427, | |
| "step": 1886 | |
| }, | |
| { | |
| "epoch": 16.128205128205128, | |
| "grad_norm": 1.9572805166244507, | |
| "learning_rate": 1.935897435897436e-06, | |
| "loss": 0.04, | |
| "step": 1887 | |
| }, | |
| { | |
| "epoch": 16.136752136752136, | |
| "grad_norm": 1.9205402135849, | |
| "learning_rate": 1.931623931623932e-06, | |
| "loss": 0.0384, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 16.145299145299145, | |
| "grad_norm": 1.6124738454818726, | |
| "learning_rate": 1.9273504273504275e-06, | |
| "loss": 0.0322, | |
| "step": 1889 | |
| }, | |
| { | |
| "epoch": 16.153846153846153, | |
| "grad_norm": 3.3396270275115967, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 0.1302, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 16.162393162393162, | |
| "grad_norm": 2.4800124168395996, | |
| "learning_rate": 1.9188034188034193e-06, | |
| "loss": 0.1181, | |
| "step": 1891 | |
| }, | |
| { | |
| "epoch": 16.17094017094017, | |
| "grad_norm": 5.452153205871582, | |
| "learning_rate": 1.9145299145299148e-06, | |
| "loss": 0.2054, | |
| "step": 1892 | |
| }, | |
| { | |
| "epoch": 16.17948717948718, | |
| "grad_norm": 4.445066452026367, | |
| "learning_rate": 1.9102564102564102e-06, | |
| "loss": 0.1649, | |
| "step": 1893 | |
| }, | |
| { | |
| "epoch": 16.188034188034187, | |
| "grad_norm": 1.0402263402938843, | |
| "learning_rate": 1.9059829059829061e-06, | |
| "loss": 0.0285, | |
| "step": 1894 | |
| }, | |
| { | |
| "epoch": 16.196581196581196, | |
| "grad_norm": 1.8124594688415527, | |
| "learning_rate": 1.9017094017094018e-06, | |
| "loss": 0.0717, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 16.205128205128204, | |
| "grad_norm": 5.0620245933532715, | |
| "learning_rate": 1.8974358974358975e-06, | |
| "loss": 0.3833, | |
| "step": 1896 | |
| }, | |
| { | |
| "epoch": 16.213675213675213, | |
| "grad_norm": 3.201596975326538, | |
| "learning_rate": 1.8931623931623931e-06, | |
| "loss": 0.0687, | |
| "step": 1897 | |
| }, | |
| { | |
| "epoch": 16.22222222222222, | |
| "grad_norm": 0.9610732793807983, | |
| "learning_rate": 1.888888888888889e-06, | |
| "loss": 0.0165, | |
| "step": 1898 | |
| }, | |
| { | |
| "epoch": 16.23076923076923, | |
| "grad_norm": 1.3409554958343506, | |
| "learning_rate": 1.8846153846153847e-06, | |
| "loss": 0.024, | |
| "step": 1899 | |
| }, | |
| { | |
| "epoch": 16.23931623931624, | |
| "grad_norm": 1.2862681150436401, | |
| "learning_rate": 1.8803418803418804e-06, | |
| "loss": 0.042, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 16.247863247863247, | |
| "grad_norm": 6.403625011444092, | |
| "learning_rate": 1.8760683760683763e-06, | |
| "loss": 0.5536, | |
| "step": 1901 | |
| }, | |
| { | |
| "epoch": 16.256410256410255, | |
| "grad_norm": 3.241731882095337, | |
| "learning_rate": 1.871794871794872e-06, | |
| "loss": 0.1045, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 16.264957264957264, | |
| "grad_norm": 1.1206634044647217, | |
| "learning_rate": 1.8675213675213676e-06, | |
| "loss": 0.0383, | |
| "step": 1903 | |
| }, | |
| { | |
| "epoch": 16.273504273504273, | |
| "grad_norm": 3.3005762100219727, | |
| "learning_rate": 1.8632478632478635e-06, | |
| "loss": 0.0786, | |
| "step": 1904 | |
| }, | |
| { | |
| "epoch": 16.28205128205128, | |
| "grad_norm": 0.44867634773254395, | |
| "learning_rate": 1.8589743589743592e-06, | |
| "loss": 0.0104, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 16.29059829059829, | |
| "grad_norm": 2.7023422718048096, | |
| "learning_rate": 1.8547008547008549e-06, | |
| "loss": 0.1091, | |
| "step": 1906 | |
| }, | |
| { | |
| "epoch": 16.299145299145298, | |
| "grad_norm": 0.9612734317779541, | |
| "learning_rate": 1.8504273504273506e-06, | |
| "loss": 0.0165, | |
| "step": 1907 | |
| }, | |
| { | |
| "epoch": 16.307692307692307, | |
| "grad_norm": 3.0632894039154053, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 0.1118, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 16.316239316239315, | |
| "grad_norm": 3.932769775390625, | |
| "learning_rate": 1.8418803418803421e-06, | |
| "loss": 0.1084, | |
| "step": 1909 | |
| }, | |
| { | |
| "epoch": 16.324786324786324, | |
| "grad_norm": 7.795356273651123, | |
| "learning_rate": 1.8376068376068378e-06, | |
| "loss": 0.2923, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 16.333333333333332, | |
| "grad_norm": 1.4187766313552856, | |
| "learning_rate": 1.8333333333333333e-06, | |
| "loss": 0.0408, | |
| "step": 1911 | |
| }, | |
| { | |
| "epoch": 16.34188034188034, | |
| "grad_norm": 1.1020699739456177, | |
| "learning_rate": 1.8290598290598292e-06, | |
| "loss": 0.0168, | |
| "step": 1912 | |
| }, | |
| { | |
| "epoch": 16.35042735042735, | |
| "grad_norm": 0.9890375733375549, | |
| "learning_rate": 1.8247863247863249e-06, | |
| "loss": 0.0391, | |
| "step": 1913 | |
| }, | |
| { | |
| "epoch": 16.358974358974358, | |
| "grad_norm": 39.418235778808594, | |
| "learning_rate": 1.8205128205128205e-06, | |
| "loss": 0.2804, | |
| "step": 1914 | |
| }, | |
| { | |
| "epoch": 16.367521367521366, | |
| "grad_norm": 1.6613589525222778, | |
| "learning_rate": 1.8162393162393164e-06, | |
| "loss": 0.0475, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 16.376068376068375, | |
| "grad_norm": 4.359612464904785, | |
| "learning_rate": 1.811965811965812e-06, | |
| "loss": 0.2247, | |
| "step": 1916 | |
| }, | |
| { | |
| "epoch": 16.384615384615383, | |
| "grad_norm": 1.970078706741333, | |
| "learning_rate": 1.8076923076923078e-06, | |
| "loss": 0.03, | |
| "step": 1917 | |
| }, | |
| { | |
| "epoch": 16.39316239316239, | |
| "grad_norm": 2.046025037765503, | |
| "learning_rate": 1.8034188034188035e-06, | |
| "loss": 0.0277, | |
| "step": 1918 | |
| }, | |
| { | |
| "epoch": 16.4017094017094, | |
| "grad_norm": 1.5775028467178345, | |
| "learning_rate": 1.7991452991452994e-06, | |
| "loss": 0.0764, | |
| "step": 1919 | |
| }, | |
| { | |
| "epoch": 16.41025641025641, | |
| "grad_norm": 2.8837273120880127, | |
| "learning_rate": 1.794871794871795e-06, | |
| "loss": 0.0903, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 16.418803418803417, | |
| "grad_norm": 7.059972763061523, | |
| "learning_rate": 1.7905982905982907e-06, | |
| "loss": 0.0679, | |
| "step": 1921 | |
| }, | |
| { | |
| "epoch": 16.427350427350426, | |
| "grad_norm": 3.6101839542388916, | |
| "learning_rate": 1.7863247863247866e-06, | |
| "loss": 0.1402, | |
| "step": 1922 | |
| }, | |
| { | |
| "epoch": 16.435897435897434, | |
| "grad_norm": 2.3459484577178955, | |
| "learning_rate": 1.7820512820512823e-06, | |
| "loss": 0.0751, | |
| "step": 1923 | |
| }, | |
| { | |
| "epoch": 16.444444444444443, | |
| "grad_norm": 2.0556280612945557, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 0.0452, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 16.45299145299145, | |
| "grad_norm": 0.5339368581771851, | |
| "learning_rate": 1.7735042735042736e-06, | |
| "loss": 0.013, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 16.46153846153846, | |
| "grad_norm": 1.393329381942749, | |
| "learning_rate": 1.7692307692307695e-06, | |
| "loss": 0.038, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 16.47008547008547, | |
| "grad_norm": 0.9439583420753479, | |
| "learning_rate": 1.7649572649572652e-06, | |
| "loss": 0.0228, | |
| "step": 1927 | |
| }, | |
| { | |
| "epoch": 16.478632478632477, | |
| "grad_norm": 3.437713384628296, | |
| "learning_rate": 1.7606837606837609e-06, | |
| "loss": 0.2072, | |
| "step": 1928 | |
| }, | |
| { | |
| "epoch": 16.487179487179485, | |
| "grad_norm": 1.725557804107666, | |
| "learning_rate": 1.7564102564102563e-06, | |
| "loss": 0.0494, | |
| "step": 1929 | |
| }, | |
| { | |
| "epoch": 16.495726495726494, | |
| "grad_norm": 2.4226529598236084, | |
| "learning_rate": 1.7521367521367522e-06, | |
| "loss": 0.0796, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 16.504273504273506, | |
| "grad_norm": 36.0551643371582, | |
| "learning_rate": 1.747863247863248e-06, | |
| "loss": 0.1966, | |
| "step": 1931 | |
| }, | |
| { | |
| "epoch": 16.51282051282051, | |
| "grad_norm": 0.8370515704154968, | |
| "learning_rate": 1.7435897435897436e-06, | |
| "loss": 0.0346, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 16.521367521367523, | |
| "grad_norm": 2.486854314804077, | |
| "learning_rate": 1.7393162393162395e-06, | |
| "loss": 0.1423, | |
| "step": 1933 | |
| }, | |
| { | |
| "epoch": 16.52991452991453, | |
| "grad_norm": 3.2457993030548096, | |
| "learning_rate": 1.7350427350427352e-06, | |
| "loss": 0.1894, | |
| "step": 1934 | |
| }, | |
| { | |
| "epoch": 16.53846153846154, | |
| "grad_norm": 2.1744906902313232, | |
| "learning_rate": 1.7307692307692308e-06, | |
| "loss": 0.0889, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 16.54700854700855, | |
| "grad_norm": 1.9443250894546509, | |
| "learning_rate": 1.7264957264957265e-06, | |
| "loss": 0.0413, | |
| "step": 1936 | |
| }, | |
| { | |
| "epoch": 16.555555555555557, | |
| "grad_norm": 2.0389249324798584, | |
| "learning_rate": 1.7222222222222224e-06, | |
| "loss": 0.0798, | |
| "step": 1937 | |
| }, | |
| { | |
| "epoch": 16.564102564102566, | |
| "grad_norm": 4.600223064422607, | |
| "learning_rate": 1.717948717948718e-06, | |
| "loss": 0.0706, | |
| "step": 1938 | |
| }, | |
| { | |
| "epoch": 16.572649572649574, | |
| "grad_norm": 1.4231921434402466, | |
| "learning_rate": 1.7136752136752138e-06, | |
| "loss": 0.0856, | |
| "step": 1939 | |
| }, | |
| { | |
| "epoch": 16.581196581196583, | |
| "grad_norm": 4.8655290603637695, | |
| "learning_rate": 1.7094017094017097e-06, | |
| "loss": 0.2519, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 16.58974358974359, | |
| "grad_norm": 2.6834962368011475, | |
| "learning_rate": 1.7051282051282053e-06, | |
| "loss": 0.0328, | |
| "step": 1941 | |
| }, | |
| { | |
| "epoch": 16.5982905982906, | |
| "grad_norm": 0.625557541847229, | |
| "learning_rate": 1.700854700854701e-06, | |
| "loss": 0.0129, | |
| "step": 1942 | |
| }, | |
| { | |
| "epoch": 16.60683760683761, | |
| "grad_norm": 10.57834243774414, | |
| "learning_rate": 1.6965811965811967e-06, | |
| "loss": 0.2987, | |
| "step": 1943 | |
| }, | |
| { | |
| "epoch": 16.615384615384617, | |
| "grad_norm": 1.2357791662216187, | |
| "learning_rate": 1.6923076923076926e-06, | |
| "loss": 0.0294, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 16.623931623931625, | |
| "grad_norm": 1.8380581140518188, | |
| "learning_rate": 1.6880341880341883e-06, | |
| "loss": 0.0298, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 16.632478632478634, | |
| "grad_norm": 1.2370020151138306, | |
| "learning_rate": 1.683760683760684e-06, | |
| "loss": 0.0285, | |
| "step": 1946 | |
| }, | |
| { | |
| "epoch": 16.641025641025642, | |
| "grad_norm": 5.922267913818359, | |
| "learning_rate": 1.6794871794871794e-06, | |
| "loss": 0.24, | |
| "step": 1947 | |
| }, | |
| { | |
| "epoch": 16.64957264957265, | |
| "grad_norm": 2.439023494720459, | |
| "learning_rate": 1.6752136752136753e-06, | |
| "loss": 0.0988, | |
| "step": 1948 | |
| }, | |
| { | |
| "epoch": 16.65811965811966, | |
| "grad_norm": 0.8908723592758179, | |
| "learning_rate": 1.670940170940171e-06, | |
| "loss": 0.026, | |
| "step": 1949 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 0.8728394508361816, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.018, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 16.675213675213676, | |
| "grad_norm": 2.7304019927978516, | |
| "learning_rate": 1.6623931623931626e-06, | |
| "loss": 0.1567, | |
| "step": 1951 | |
| }, | |
| { | |
| "epoch": 16.683760683760685, | |
| "grad_norm": 2.8601150512695312, | |
| "learning_rate": 1.6581196581196582e-06, | |
| "loss": 0.0721, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.692307692307693, | |
| "grad_norm": 2.5990025997161865, | |
| "learning_rate": 1.653846153846154e-06, | |
| "loss": 0.2296, | |
| "step": 1953 | |
| }, | |
| { | |
| "epoch": 16.700854700854702, | |
| "grad_norm": 3.7956109046936035, | |
| "learning_rate": 1.6495726495726496e-06, | |
| "loss": 0.2565, | |
| "step": 1954 | |
| }, | |
| { | |
| "epoch": 16.70940170940171, | |
| "grad_norm": 5.933072566986084, | |
| "learning_rate": 1.6452991452991455e-06, | |
| "loss": 0.2712, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 16.71794871794872, | |
| "grad_norm": 0.5651862621307373, | |
| "learning_rate": 1.6410256410256412e-06, | |
| "loss": 0.0132, | |
| "step": 1956 | |
| }, | |
| { | |
| "epoch": 16.726495726495727, | |
| "grad_norm": 3.033231735229492, | |
| "learning_rate": 1.6367521367521368e-06, | |
| "loss": 0.074, | |
| "step": 1957 | |
| }, | |
| { | |
| "epoch": 16.735042735042736, | |
| "grad_norm": 1.3515870571136475, | |
| "learning_rate": 1.6324786324786327e-06, | |
| "loss": 0.0614, | |
| "step": 1958 | |
| }, | |
| { | |
| "epoch": 16.743589743589745, | |
| "grad_norm": 3.091700792312622, | |
| "learning_rate": 1.6282051282051284e-06, | |
| "loss": 0.1284, | |
| "step": 1959 | |
| }, | |
| { | |
| "epoch": 16.752136752136753, | |
| "grad_norm": 7.142216205596924, | |
| "learning_rate": 1.623931623931624e-06, | |
| "loss": 0.1965, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 16.76068376068376, | |
| "grad_norm": 7.488593578338623, | |
| "learning_rate": 1.6196581196581198e-06, | |
| "loss": 0.2498, | |
| "step": 1961 | |
| }, | |
| { | |
| "epoch": 16.76923076923077, | |
| "grad_norm": 3.943833351135254, | |
| "learning_rate": 1.6153846153846157e-06, | |
| "loss": 0.0967, | |
| "step": 1962 | |
| }, | |
| { | |
| "epoch": 16.77777777777778, | |
| "grad_norm": 1.8732318878173828, | |
| "learning_rate": 1.6111111111111113e-06, | |
| "loss": 0.029, | |
| "step": 1963 | |
| }, | |
| { | |
| "epoch": 16.786324786324787, | |
| "grad_norm": 2.5445902347564697, | |
| "learning_rate": 1.606837606837607e-06, | |
| "loss": 0.0808, | |
| "step": 1964 | |
| }, | |
| { | |
| "epoch": 16.794871794871796, | |
| "grad_norm": 4.969367504119873, | |
| "learning_rate": 1.602564102564103e-06, | |
| "loss": 0.164, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 16.803418803418804, | |
| "grad_norm": 1.6954468488693237, | |
| "learning_rate": 1.5982905982905984e-06, | |
| "loss": 0.0645, | |
| "step": 1966 | |
| }, | |
| { | |
| "epoch": 16.811965811965813, | |
| "grad_norm": 1.536352276802063, | |
| "learning_rate": 1.594017094017094e-06, | |
| "loss": 0.0595, | |
| "step": 1967 | |
| }, | |
| { | |
| "epoch": 16.82051282051282, | |
| "grad_norm": 0.7326592803001404, | |
| "learning_rate": 1.5897435897435897e-06, | |
| "loss": 0.0153, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 16.82905982905983, | |
| "grad_norm": 10.959025382995605, | |
| "learning_rate": 1.5854700854700856e-06, | |
| "loss": 0.3274, | |
| "step": 1969 | |
| }, | |
| { | |
| "epoch": 16.837606837606838, | |
| "grad_norm": 10.305845260620117, | |
| "learning_rate": 1.5811965811965813e-06, | |
| "loss": 0.1404, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 16.846153846153847, | |
| "grad_norm": 7.498697280883789, | |
| "learning_rate": 1.576923076923077e-06, | |
| "loss": 0.2269, | |
| "step": 1971 | |
| }, | |
| { | |
| "epoch": 16.854700854700855, | |
| "grad_norm": 0.29253125190734863, | |
| "learning_rate": 1.5726495726495727e-06, | |
| "loss": 0.0074, | |
| "step": 1972 | |
| }, | |
| { | |
| "epoch": 16.863247863247864, | |
| "grad_norm": 9.320234298706055, | |
| "learning_rate": 1.5683760683760685e-06, | |
| "loss": 0.067, | |
| "step": 1973 | |
| }, | |
| { | |
| "epoch": 16.871794871794872, | |
| "grad_norm": 6.572272300720215, | |
| "learning_rate": 1.5641025641025642e-06, | |
| "loss": 0.4577, | |
| "step": 1974 | |
| }, | |
| { | |
| "epoch": 16.88034188034188, | |
| "grad_norm": 5.368937969207764, | |
| "learning_rate": 1.55982905982906e-06, | |
| "loss": 0.2016, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 16.88888888888889, | |
| "grad_norm": 0.5891698598861694, | |
| "learning_rate": 1.5555555555555558e-06, | |
| "loss": 0.0174, | |
| "step": 1976 | |
| }, | |
| { | |
| "epoch": 16.897435897435898, | |
| "grad_norm": 3.045989751815796, | |
| "learning_rate": 1.5512820512820515e-06, | |
| "loss": 0.1748, | |
| "step": 1977 | |
| }, | |
| { | |
| "epoch": 16.905982905982906, | |
| "grad_norm": 3.013834238052368, | |
| "learning_rate": 1.5470085470085471e-06, | |
| "loss": 0.2283, | |
| "step": 1978 | |
| }, | |
| { | |
| "epoch": 16.914529914529915, | |
| "grad_norm": 1.2644447088241577, | |
| "learning_rate": 1.5427350427350428e-06, | |
| "loss": 0.0302, | |
| "step": 1979 | |
| }, | |
| { | |
| "epoch": 16.923076923076923, | |
| "grad_norm": 4.429958820343018, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.2458, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 16.931623931623932, | |
| "grad_norm": 1.1556981801986694, | |
| "learning_rate": 1.5341880341880344e-06, | |
| "loss": 0.0179, | |
| "step": 1981 | |
| }, | |
| { | |
| "epoch": 16.94017094017094, | |
| "grad_norm": 1.4588316679000854, | |
| "learning_rate": 1.52991452991453e-06, | |
| "loss": 0.1063, | |
| "step": 1982 | |
| }, | |
| { | |
| "epoch": 16.94871794871795, | |
| "grad_norm": 1.124496340751648, | |
| "learning_rate": 1.525641025641026e-06, | |
| "loss": 0.0278, | |
| "step": 1983 | |
| }, | |
| { | |
| "epoch": 16.957264957264957, | |
| "grad_norm": 0.7231981754302979, | |
| "learning_rate": 1.5213675213675214e-06, | |
| "loss": 0.0141, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 16.965811965811966, | |
| "grad_norm": 1.4819642305374146, | |
| "learning_rate": 1.5170940170940171e-06, | |
| "loss": 0.0601, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 16.974358974358974, | |
| "grad_norm": 0.7296791672706604, | |
| "learning_rate": 1.5128205128205128e-06, | |
| "loss": 0.0215, | |
| "step": 1986 | |
| }, | |
| { | |
| "epoch": 16.982905982905983, | |
| "grad_norm": 15.651564598083496, | |
| "learning_rate": 1.5085470085470087e-06, | |
| "loss": 0.2954, | |
| "step": 1987 | |
| }, | |
| { | |
| "epoch": 16.99145299145299, | |
| "grad_norm": 0.48891735076904297, | |
| "learning_rate": 1.5042735042735044e-06, | |
| "loss": 0.015, | |
| "step": 1988 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 7.363093376159668, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.2366, | |
| "step": 1989 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 0.05406723916530609, | |
| "eval_runtime": 9.389, | |
| "eval_samples_per_second": 49.633, | |
| "eval_steps_per_second": 6.284, | |
| "step": 1989 | |
| }, | |
| { | |
| "epoch": 17.00854700854701, | |
| "grad_norm": 2.8626017570495605, | |
| "learning_rate": 1.4957264957264957e-06, | |
| "loss": 0.0902, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 17.017094017094017, | |
| "grad_norm": 2.461879253387451, | |
| "learning_rate": 1.4914529914529916e-06, | |
| "loss": 0.0387, | |
| "step": 1991 | |
| }, | |
| { | |
| "epoch": 17.025641025641026, | |
| "grad_norm": 6.336863994598389, | |
| "learning_rate": 1.4871794871794873e-06, | |
| "loss": 0.196, | |
| "step": 1992 | |
| }, | |
| { | |
| "epoch": 17.034188034188034, | |
| "grad_norm": 1.1044467687606812, | |
| "learning_rate": 1.482905982905983e-06, | |
| "loss": 0.0352, | |
| "step": 1993 | |
| }, | |
| { | |
| "epoch": 17.042735042735043, | |
| "grad_norm": 3.3509342670440674, | |
| "learning_rate": 1.4786324786324789e-06, | |
| "loss": 0.1459, | |
| "step": 1994 | |
| }, | |
| { | |
| "epoch": 17.05128205128205, | |
| "grad_norm": 3.2349629402160645, | |
| "learning_rate": 1.4743589743589745e-06, | |
| "loss": 0.0179, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 17.05982905982906, | |
| "grad_norm": 3.650749921798706, | |
| "learning_rate": 1.4700854700854702e-06, | |
| "loss": 0.1549, | |
| "step": 1996 | |
| }, | |
| { | |
| "epoch": 17.068376068376068, | |
| "grad_norm": 1.6349891424179077, | |
| "learning_rate": 1.465811965811966e-06, | |
| "loss": 0.0713, | |
| "step": 1997 | |
| }, | |
| { | |
| "epoch": 17.076923076923077, | |
| "grad_norm": 8.602070808410645, | |
| "learning_rate": 1.4615384615384618e-06, | |
| "loss": 0.3582, | |
| "step": 1998 | |
| }, | |
| { | |
| "epoch": 17.085470085470085, | |
| "grad_norm": 3.1162590980529785, | |
| "learning_rate": 1.4572649572649575e-06, | |
| "loss": 0.2455, | |
| "step": 1999 | |
| }, | |
| { | |
| "epoch": 17.094017094017094, | |
| "grad_norm": 1.4878407716751099, | |
| "learning_rate": 1.4529914529914531e-06, | |
| "loss": 0.0195, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.102564102564102, | |
| "grad_norm": 2.565297842025757, | |
| "learning_rate": 1.448717948717949e-06, | |
| "loss": 0.1126, | |
| "step": 2001 | |
| }, | |
| { | |
| "epoch": 17.11111111111111, | |
| "grad_norm": 4.169450759887695, | |
| "learning_rate": 1.4444444444444445e-06, | |
| "loss": 0.1774, | |
| "step": 2002 | |
| }, | |
| { | |
| "epoch": 17.11965811965812, | |
| "grad_norm": 1.8476792573928833, | |
| "learning_rate": 1.4401709401709402e-06, | |
| "loss": 0.0288, | |
| "step": 2003 | |
| }, | |
| { | |
| "epoch": 17.128205128205128, | |
| "grad_norm": 0.7279506921768188, | |
| "learning_rate": 1.4358974358974359e-06, | |
| "loss": 0.0217, | |
| "step": 2004 | |
| }, | |
| { | |
| "epoch": 17.136752136752136, | |
| "grad_norm": 7.387227535247803, | |
| "learning_rate": 1.4316239316239317e-06, | |
| "loss": 0.248, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 17.145299145299145, | |
| "grad_norm": 2.9455361366271973, | |
| "learning_rate": 1.4273504273504274e-06, | |
| "loss": 0.0439, | |
| "step": 2006 | |
| }, | |
| { | |
| "epoch": 17.153846153846153, | |
| "grad_norm": 6.015694618225098, | |
| "learning_rate": 1.423076923076923e-06, | |
| "loss": 0.0656, | |
| "step": 2007 | |
| }, | |
| { | |
| "epoch": 17.162393162393162, | |
| "grad_norm": 1.741774320602417, | |
| "learning_rate": 1.4188034188034188e-06, | |
| "loss": 0.0344, | |
| "step": 2008 | |
| }, | |
| { | |
| "epoch": 17.17094017094017, | |
| "grad_norm": 0.5282659530639648, | |
| "learning_rate": 1.4145299145299147e-06, | |
| "loss": 0.0128, | |
| "step": 2009 | |
| }, | |
| { | |
| "epoch": 17.17948717948718, | |
| "grad_norm": 2.4927468299865723, | |
| "learning_rate": 1.4102564102564104e-06, | |
| "loss": 0.1839, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 17.188034188034187, | |
| "grad_norm": 0.7872166037559509, | |
| "learning_rate": 1.405982905982906e-06, | |
| "loss": 0.0204, | |
| "step": 2011 | |
| }, | |
| { | |
| "epoch": 17.196581196581196, | |
| "grad_norm": 0.7072253227233887, | |
| "learning_rate": 1.401709401709402e-06, | |
| "loss": 0.0206, | |
| "step": 2012 | |
| }, | |
| { | |
| "epoch": 17.205128205128204, | |
| "grad_norm": 1.0154236555099487, | |
| "learning_rate": 1.3974358974358976e-06, | |
| "loss": 0.0238, | |
| "step": 2013 | |
| }, | |
| { | |
| "epoch": 17.213675213675213, | |
| "grad_norm": 2.9798424243927, | |
| "learning_rate": 1.3931623931623933e-06, | |
| "loss": 0.0542, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 17.22222222222222, | |
| "grad_norm": 0.9568426012992859, | |
| "learning_rate": 1.3888888888888892e-06, | |
| "loss": 0.0239, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 17.23076923076923, | |
| "grad_norm": 10.525039672851562, | |
| "learning_rate": 1.3846153846153848e-06, | |
| "loss": 0.1768, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 17.23931623931624, | |
| "grad_norm": 1.697314977645874, | |
| "learning_rate": 1.3803418803418805e-06, | |
| "loss": 0.0453, | |
| "step": 2017 | |
| }, | |
| { | |
| "epoch": 17.247863247863247, | |
| "grad_norm": 0.6436419486999512, | |
| "learning_rate": 1.3760683760683762e-06, | |
| "loss": 0.0163, | |
| "step": 2018 | |
| }, | |
| { | |
| "epoch": 17.256410256410255, | |
| "grad_norm": 4.984555721282959, | |
| "learning_rate": 1.371794871794872e-06, | |
| "loss": 0.1157, | |
| "step": 2019 | |
| }, | |
| { | |
| "epoch": 17.264957264957264, | |
| "grad_norm": 9.088909149169922, | |
| "learning_rate": 1.3675213675213678e-06, | |
| "loss": 0.2842, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 17.273504273504273, | |
| "grad_norm": 10.398246765136719, | |
| "learning_rate": 1.3632478632478632e-06, | |
| "loss": 0.2528, | |
| "step": 2021 | |
| }, | |
| { | |
| "epoch": 17.28205128205128, | |
| "grad_norm": 3.60273814201355, | |
| "learning_rate": 1.358974358974359e-06, | |
| "loss": 0.1799, | |
| "step": 2022 | |
| }, | |
| { | |
| "epoch": 17.29059829059829, | |
| "grad_norm": 0.6845250129699707, | |
| "learning_rate": 1.3547008547008548e-06, | |
| "loss": 0.0196, | |
| "step": 2023 | |
| }, | |
| { | |
| "epoch": 17.299145299145298, | |
| "grad_norm": 0.5363795161247253, | |
| "learning_rate": 1.3504273504273505e-06, | |
| "loss": 0.0136, | |
| "step": 2024 | |
| }, | |
| { | |
| "epoch": 17.307692307692307, | |
| "grad_norm": 3.880434274673462, | |
| "learning_rate": 1.3461538461538462e-06, | |
| "loss": 0.3665, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 17.316239316239315, | |
| "grad_norm": 4.580989360809326, | |
| "learning_rate": 1.3418803418803418e-06, | |
| "loss": 0.2593, | |
| "step": 2026 | |
| }, | |
| { | |
| "epoch": 17.324786324786324, | |
| "grad_norm": 2.781501293182373, | |
| "learning_rate": 1.3376068376068377e-06, | |
| "loss": 0.1777, | |
| "step": 2027 | |
| }, | |
| { | |
| "epoch": 17.333333333333332, | |
| "grad_norm": 5.605004787445068, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.3633, | |
| "step": 2028 | |
| }, | |
| { | |
| "epoch": 17.34188034188034, | |
| "grad_norm": 1.696486473083496, | |
| "learning_rate": 1.329059829059829e-06, | |
| "loss": 0.0353, | |
| "step": 2029 | |
| }, | |
| { | |
| "epoch": 17.35042735042735, | |
| "grad_norm": 3.4415268898010254, | |
| "learning_rate": 1.324786324786325e-06, | |
| "loss": 0.0906, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 17.358974358974358, | |
| "grad_norm": 7.722592353820801, | |
| "learning_rate": 1.3205128205128207e-06, | |
| "loss": 0.1804, | |
| "step": 2031 | |
| }, | |
| { | |
| "epoch": 17.367521367521366, | |
| "grad_norm": 3.3161542415618896, | |
| "learning_rate": 1.3162393162393163e-06, | |
| "loss": 0.1336, | |
| "step": 2032 | |
| }, | |
| { | |
| "epoch": 17.376068376068375, | |
| "grad_norm": 2.568871021270752, | |
| "learning_rate": 1.3119658119658122e-06, | |
| "loss": 0.0658, | |
| "step": 2033 | |
| }, | |
| { | |
| "epoch": 17.384615384615383, | |
| "grad_norm": 3.5799806118011475, | |
| "learning_rate": 1.307692307692308e-06, | |
| "loss": 0.0652, | |
| "step": 2034 | |
| }, | |
| { | |
| "epoch": 17.39316239316239, | |
| "grad_norm": 1.1399949789047241, | |
| "learning_rate": 1.3034188034188036e-06, | |
| "loss": 0.0196, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 17.4017094017094, | |
| "grad_norm": 2.3688738346099854, | |
| "learning_rate": 1.2991452991452993e-06, | |
| "loss": 0.0706, | |
| "step": 2036 | |
| }, | |
| { | |
| "epoch": 17.41025641025641, | |
| "grad_norm": 12.726486206054688, | |
| "learning_rate": 1.2948717948717952e-06, | |
| "loss": 0.2506, | |
| "step": 2037 | |
| }, | |
| { | |
| "epoch": 17.418803418803417, | |
| "grad_norm": 2.249285936355591, | |
| "learning_rate": 1.2905982905982908e-06, | |
| "loss": 0.0532, | |
| "step": 2038 | |
| }, | |
| { | |
| "epoch": 17.427350427350426, | |
| "grad_norm": 0.7129601836204529, | |
| "learning_rate": 1.2863247863247863e-06, | |
| "loss": 0.0207, | |
| "step": 2039 | |
| }, | |
| { | |
| "epoch": 17.435897435897434, | |
| "grad_norm": 1.9362183809280396, | |
| "learning_rate": 1.282051282051282e-06, | |
| "loss": 0.0311, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 17.444444444444443, | |
| "grad_norm": 2.253690242767334, | |
| "learning_rate": 1.2777777777777779e-06, | |
| "loss": 0.1203, | |
| "step": 2041 | |
| }, | |
| { | |
| "epoch": 17.45299145299145, | |
| "grad_norm": 3.835174798965454, | |
| "learning_rate": 1.2735042735042736e-06, | |
| "loss": 0.0928, | |
| "step": 2042 | |
| }, | |
| { | |
| "epoch": 17.46153846153846, | |
| "grad_norm": 143.36563110351562, | |
| "learning_rate": 1.2692307692307692e-06, | |
| "loss": 0.2984, | |
| "step": 2043 | |
| }, | |
| { | |
| "epoch": 17.47008547008547, | |
| "grad_norm": 0.6122754216194153, | |
| "learning_rate": 1.264957264957265e-06, | |
| "loss": 0.0171, | |
| "step": 2044 | |
| }, | |
| { | |
| "epoch": 17.478632478632477, | |
| "grad_norm": 3.0697991847991943, | |
| "learning_rate": 1.2606837606837608e-06, | |
| "loss": 0.1412, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 17.487179487179485, | |
| "grad_norm": 1.0684096813201904, | |
| "learning_rate": 1.2564102564102565e-06, | |
| "loss": 0.0278, | |
| "step": 2046 | |
| }, | |
| { | |
| "epoch": 17.495726495726494, | |
| "grad_norm": 5.379480838775635, | |
| "learning_rate": 1.2521367521367522e-06, | |
| "loss": 0.1114, | |
| "step": 2047 | |
| }, | |
| { | |
| "epoch": 17.504273504273506, | |
| "grad_norm": 3.893343448638916, | |
| "learning_rate": 1.247863247863248e-06, | |
| "loss": 0.1499, | |
| "step": 2048 | |
| }, | |
| { | |
| "epoch": 17.51282051282051, | |
| "grad_norm": 1.0436211824417114, | |
| "learning_rate": 1.2435897435897437e-06, | |
| "loss": 0.0259, | |
| "step": 2049 | |
| }, | |
| { | |
| "epoch": 17.521367521367523, | |
| "grad_norm": 2.8706037998199463, | |
| "learning_rate": 1.2393162393162394e-06, | |
| "loss": 0.1071, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 17.52991452991453, | |
| "grad_norm": 1.5661158561706543, | |
| "learning_rate": 1.2350427350427353e-06, | |
| "loss": 0.0392, | |
| "step": 2051 | |
| }, | |
| { | |
| "epoch": 17.53846153846154, | |
| "grad_norm": 3.7152199745178223, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.0698, | |
| "step": 2052 | |
| }, | |
| { | |
| "epoch": 17.54700854700855, | |
| "grad_norm": 2.6527271270751953, | |
| "learning_rate": 1.2264957264957264e-06, | |
| "loss": 0.1276, | |
| "step": 2053 | |
| }, | |
| { | |
| "epoch": 17.555555555555557, | |
| "grad_norm": 0.9018534421920776, | |
| "learning_rate": 1.2222222222222223e-06, | |
| "loss": 0.066, | |
| "step": 2054 | |
| }, | |
| { | |
| "epoch": 17.564102564102566, | |
| "grad_norm": 7.11035680770874, | |
| "learning_rate": 1.217948717948718e-06, | |
| "loss": 0.0836, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 17.572649572649574, | |
| "grad_norm": 2.5168066024780273, | |
| "learning_rate": 1.2136752136752137e-06, | |
| "loss": 0.0662, | |
| "step": 2056 | |
| }, | |
| { | |
| "epoch": 17.581196581196583, | |
| "grad_norm": 0.7215616703033447, | |
| "learning_rate": 1.2094017094017096e-06, | |
| "loss": 0.0186, | |
| "step": 2057 | |
| }, | |
| { | |
| "epoch": 17.58974358974359, | |
| "grad_norm": 7.076876640319824, | |
| "learning_rate": 1.2051282051282053e-06, | |
| "loss": 0.1493, | |
| "step": 2058 | |
| }, | |
| { | |
| "epoch": 17.5982905982906, | |
| "grad_norm": 1.1687662601470947, | |
| "learning_rate": 1.200854700854701e-06, | |
| "loss": 0.0368, | |
| "step": 2059 | |
| }, | |
| { | |
| "epoch": 17.60683760683761, | |
| "grad_norm": 2.5085737705230713, | |
| "learning_rate": 1.1965811965811968e-06, | |
| "loss": 0.1567, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 17.615384615384617, | |
| "grad_norm": 0.43566644191741943, | |
| "learning_rate": 1.1923076923076925e-06, | |
| "loss": 0.0097, | |
| "step": 2061 | |
| }, | |
| { | |
| "epoch": 17.623931623931625, | |
| "grad_norm": 0.7698078155517578, | |
| "learning_rate": 1.188034188034188e-06, | |
| "loss": 0.0231, | |
| "step": 2062 | |
| }, | |
| { | |
| "epoch": 17.632478632478634, | |
| "grad_norm": 1.8352185487747192, | |
| "learning_rate": 1.1837606837606839e-06, | |
| "loss": 0.0324, | |
| "step": 2063 | |
| }, | |
| { | |
| "epoch": 17.641025641025642, | |
| "grad_norm": 12.11907958984375, | |
| "learning_rate": 1.1794871794871795e-06, | |
| "loss": 0.6052, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 17.64957264957265, | |
| "grad_norm": 0.49942728877067566, | |
| "learning_rate": 1.1752136752136752e-06, | |
| "loss": 0.0111, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 17.65811965811966, | |
| "grad_norm": 3.579129457473755, | |
| "learning_rate": 1.1709401709401711e-06, | |
| "loss": 0.1706, | |
| "step": 2066 | |
| }, | |
| { | |
| "epoch": 17.666666666666668, | |
| "grad_norm": 2.112550973892212, | |
| "learning_rate": 1.1666666666666668e-06, | |
| "loss": 0.0438, | |
| "step": 2067 | |
| }, | |
| { | |
| "epoch": 17.675213675213676, | |
| "grad_norm": 2.4429895877838135, | |
| "learning_rate": 1.1623931623931625e-06, | |
| "loss": 0.0498, | |
| "step": 2068 | |
| }, | |
| { | |
| "epoch": 17.683760683760685, | |
| "grad_norm": 1.8436684608459473, | |
| "learning_rate": 1.1581196581196584e-06, | |
| "loss": 0.1228, | |
| "step": 2069 | |
| }, | |
| { | |
| "epoch": 17.692307692307693, | |
| "grad_norm": 4.679569244384766, | |
| "learning_rate": 1.153846153846154e-06, | |
| "loss": 0.1505, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 17.700854700854702, | |
| "grad_norm": 2.4409713745117188, | |
| "learning_rate": 1.1495726495726495e-06, | |
| "loss": 0.0603, | |
| "step": 2071 | |
| }, | |
| { | |
| "epoch": 17.70940170940171, | |
| "grad_norm": 3.577721118927002, | |
| "learning_rate": 1.1452991452991454e-06, | |
| "loss": 0.1078, | |
| "step": 2072 | |
| }, | |
| { | |
| "epoch": 17.71794871794872, | |
| "grad_norm": 3.774958372116089, | |
| "learning_rate": 1.141025641025641e-06, | |
| "loss": 0.3782, | |
| "step": 2073 | |
| }, | |
| { | |
| "epoch": 17.726495726495727, | |
| "grad_norm": 2.9011383056640625, | |
| "learning_rate": 1.1367521367521368e-06, | |
| "loss": 0.0714, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.735042735042736, | |
| "grad_norm": 1.7296162843704224, | |
| "learning_rate": 1.1324786324786326e-06, | |
| "loss": 0.0463, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 17.743589743589745, | |
| "grad_norm": 1.8955838680267334, | |
| "learning_rate": 1.1282051282051283e-06, | |
| "loss": 0.0641, | |
| "step": 2076 | |
| }, | |
| { | |
| "epoch": 17.752136752136753, | |
| "grad_norm": 3.0198490619659424, | |
| "learning_rate": 1.123931623931624e-06, | |
| "loss": 0.1516, | |
| "step": 2077 | |
| }, | |
| { | |
| "epoch": 17.76068376068376, | |
| "grad_norm": 1.5012823343276978, | |
| "learning_rate": 1.1196581196581199e-06, | |
| "loss": 0.0206, | |
| "step": 2078 | |
| }, | |
| { | |
| "epoch": 17.76923076923077, | |
| "grad_norm": 2.4390790462493896, | |
| "learning_rate": 1.1153846153846156e-06, | |
| "loss": 0.0458, | |
| "step": 2079 | |
| }, | |
| { | |
| "epoch": 17.77777777777778, | |
| "grad_norm": 5.728135585784912, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 0.0443, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 17.786324786324787, | |
| "grad_norm": 1.423771858215332, | |
| "learning_rate": 1.106837606837607e-06, | |
| "loss": 0.0223, | |
| "step": 2081 | |
| }, | |
| { | |
| "epoch": 17.794871794871796, | |
| "grad_norm": 2.524941921234131, | |
| "learning_rate": 1.1025641025641026e-06, | |
| "loss": 0.0587, | |
| "step": 2082 | |
| }, | |
| { | |
| "epoch": 17.803418803418804, | |
| "grad_norm": 0.9632331132888794, | |
| "learning_rate": 1.0982905982905983e-06, | |
| "loss": 0.0324, | |
| "step": 2083 | |
| }, | |
| { | |
| "epoch": 17.811965811965813, | |
| "grad_norm": 1.8369181156158447, | |
| "learning_rate": 1.0940170940170942e-06, | |
| "loss": 0.0182, | |
| "step": 2084 | |
| }, | |
| { | |
| "epoch": 17.82051282051282, | |
| "grad_norm": 2.547654867172241, | |
| "learning_rate": 1.0897435897435899e-06, | |
| "loss": 0.1395, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 17.82905982905983, | |
| "grad_norm": 3.516977310180664, | |
| "learning_rate": 1.0854700854700855e-06, | |
| "loss": 0.1044, | |
| "step": 2086 | |
| }, | |
| { | |
| "epoch": 17.837606837606838, | |
| "grad_norm": 1.7064217329025269, | |
| "learning_rate": 1.0811965811965814e-06, | |
| "loss": 0.0302, | |
| "step": 2087 | |
| }, | |
| { | |
| "epoch": 17.846153846153847, | |
| "grad_norm": 1.7427505254745483, | |
| "learning_rate": 1.076923076923077e-06, | |
| "loss": 0.0298, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 17.854700854700855, | |
| "grad_norm": 1.3395370244979858, | |
| "learning_rate": 1.0726495726495726e-06, | |
| "loss": 0.0302, | |
| "step": 2089 | |
| }, | |
| { | |
| "epoch": 17.863247863247864, | |
| "grad_norm": 7.244344711303711, | |
| "learning_rate": 1.0683760683760685e-06, | |
| "loss": 0.1925, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 17.871794871794872, | |
| "grad_norm": 5.942878723144531, | |
| "learning_rate": 1.0641025641025641e-06, | |
| "loss": 0.489, | |
| "step": 2091 | |
| }, | |
| { | |
| "epoch": 17.88034188034188, | |
| "grad_norm": 3.244260787963867, | |
| "learning_rate": 1.0598290598290598e-06, | |
| "loss": 0.2538, | |
| "step": 2092 | |
| }, | |
| { | |
| "epoch": 17.88888888888889, | |
| "grad_norm": 0.9833334684371948, | |
| "learning_rate": 1.0555555555555557e-06, | |
| "loss": 0.0215, | |
| "step": 2093 | |
| }, | |
| { | |
| "epoch": 17.897435897435898, | |
| "grad_norm": 3.0194849967956543, | |
| "learning_rate": 1.0512820512820514e-06, | |
| "loss": 0.07, | |
| "step": 2094 | |
| }, | |
| { | |
| "epoch": 17.905982905982906, | |
| "grad_norm": 0.48535388708114624, | |
| "learning_rate": 1.047008547008547e-06, | |
| "loss": 0.0113, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 17.914529914529915, | |
| "grad_norm": 4.334452152252197, | |
| "learning_rate": 1.042735042735043e-06, | |
| "loss": 0.127, | |
| "step": 2096 | |
| }, | |
| { | |
| "epoch": 17.923076923076923, | |
| "grad_norm": 3.54429030418396, | |
| "learning_rate": 1.0384615384615386e-06, | |
| "loss": 0.0704, | |
| "step": 2097 | |
| }, | |
| { | |
| "epoch": 17.931623931623932, | |
| "grad_norm": 1.1745219230651855, | |
| "learning_rate": 1.034188034188034e-06, | |
| "loss": 0.0418, | |
| "step": 2098 | |
| }, | |
| { | |
| "epoch": 17.94017094017094, | |
| "grad_norm": 5.157544136047363, | |
| "learning_rate": 1.02991452991453e-06, | |
| "loss": 0.2562, | |
| "step": 2099 | |
| }, | |
| { | |
| "epoch": 17.94871794871795, | |
| "grad_norm": 4.454767227172852, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.1141, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 17.957264957264957, | |
| "grad_norm": 12.859573364257812, | |
| "learning_rate": 1.0213675213675213e-06, | |
| "loss": 0.3516, | |
| "step": 2101 | |
| }, | |
| { | |
| "epoch": 17.965811965811966, | |
| "grad_norm": 5.780513763427734, | |
| "learning_rate": 1.0170940170940172e-06, | |
| "loss": 0.1663, | |
| "step": 2102 | |
| }, | |
| { | |
| "epoch": 17.974358974358974, | |
| "grad_norm": 2.762153387069702, | |
| "learning_rate": 1.012820512820513e-06, | |
| "loss": 0.19, | |
| "step": 2103 | |
| }, | |
| { | |
| "epoch": 17.982905982905983, | |
| "grad_norm": 5.649252891540527, | |
| "learning_rate": 1.0085470085470086e-06, | |
| "loss": 0.1736, | |
| "step": 2104 | |
| }, | |
| { | |
| "epoch": 17.99145299145299, | |
| "grad_norm": 5.10836124420166, | |
| "learning_rate": 1.0042735042735045e-06, | |
| "loss": 0.1739, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 6.474237442016602, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.3239, | |
| "step": 2106 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 0.052614517509937286, | |
| "eval_runtime": 9.28, | |
| "eval_samples_per_second": 50.216, | |
| "eval_steps_per_second": 6.358, | |
| "step": 2106 | |
| }, | |
| { | |
| "epoch": 18.00854700854701, | |
| "grad_norm": 0.8820164203643799, | |
| "learning_rate": 9.957264957264958e-07, | |
| "loss": 0.0237, | |
| "step": 2107 | |
| }, | |
| { | |
| "epoch": 18.017094017094017, | |
| "grad_norm": 2.692166566848755, | |
| "learning_rate": 9.914529914529915e-07, | |
| "loss": 0.0962, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 18.025641025641026, | |
| "grad_norm": 0.8048399090766907, | |
| "learning_rate": 9.871794871794872e-07, | |
| "loss": 0.0232, | |
| "step": 2109 | |
| }, | |
| { | |
| "epoch": 18.034188034188034, | |
| "grad_norm": 4.4439826011657715, | |
| "learning_rate": 9.829059829059829e-07, | |
| "loss": 0.064, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 18.042735042735043, | |
| "grad_norm": 1.62433660030365, | |
| "learning_rate": 9.786324786324788e-07, | |
| "loss": 0.1263, | |
| "step": 2111 | |
| }, | |
| { | |
| "epoch": 18.05128205128205, | |
| "grad_norm": 4.766104221343994, | |
| "learning_rate": 9.743589743589745e-07, | |
| "loss": 0.2108, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 18.05982905982906, | |
| "grad_norm": 139.34445190429688, | |
| "learning_rate": 9.700854700854701e-07, | |
| "loss": 0.237, | |
| "step": 2113 | |
| }, | |
| { | |
| "epoch": 18.068376068376068, | |
| "grad_norm": 0.6069220900535583, | |
| "learning_rate": 9.65811965811966e-07, | |
| "loss": 0.0135, | |
| "step": 2114 | |
| }, | |
| { | |
| "epoch": 18.076923076923077, | |
| "grad_norm": 2.7833995819091797, | |
| "learning_rate": 9.615384615384617e-07, | |
| "loss": 0.1677, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 18.085470085470085, | |
| "grad_norm": 4.570268630981445, | |
| "learning_rate": 9.572649572649574e-07, | |
| "loss": 0.2304, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 18.094017094017094, | |
| "grad_norm": 4.7644805908203125, | |
| "learning_rate": 9.529914529914531e-07, | |
| "loss": 0.138, | |
| "step": 2117 | |
| }, | |
| { | |
| "epoch": 18.102564102564102, | |
| "grad_norm": 1.9438762664794922, | |
| "learning_rate": 9.487179487179487e-07, | |
| "loss": 0.0488, | |
| "step": 2118 | |
| }, | |
| { | |
| "epoch": 18.11111111111111, | |
| "grad_norm": 1.4188040494918823, | |
| "learning_rate": 9.444444444444445e-07, | |
| "loss": 0.0545, | |
| "step": 2119 | |
| }, | |
| { | |
| "epoch": 18.11965811965812, | |
| "grad_norm": 0.357928603887558, | |
| "learning_rate": 9.401709401709402e-07, | |
| "loss": 0.0092, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 18.128205128205128, | |
| "grad_norm": 1.8646256923675537, | |
| "learning_rate": 9.35897435897436e-07, | |
| "loss": 0.086, | |
| "step": 2121 | |
| }, | |
| { | |
| "epoch": 18.136752136752136, | |
| "grad_norm": 2.111544609069824, | |
| "learning_rate": 9.316239316239318e-07, | |
| "loss": 0.0319, | |
| "step": 2122 | |
| }, | |
| { | |
| "epoch": 18.145299145299145, | |
| "grad_norm": 3.0686893463134766, | |
| "learning_rate": 9.273504273504274e-07, | |
| "loss": 0.0689, | |
| "step": 2123 | |
| }, | |
| { | |
| "epoch": 18.153846153846153, | |
| "grad_norm": 4.028079509735107, | |
| "learning_rate": 9.230769230769232e-07, | |
| "loss": 0.125, | |
| "step": 2124 | |
| }, | |
| { | |
| "epoch": 18.162393162393162, | |
| "grad_norm": 1.0433181524276733, | |
| "learning_rate": 9.188034188034189e-07, | |
| "loss": 0.0174, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 18.17094017094017, | |
| "grad_norm": 3.4533402919769287, | |
| "learning_rate": 9.145299145299146e-07, | |
| "loss": 0.1556, | |
| "step": 2126 | |
| }, | |
| { | |
| "epoch": 18.17948717948718, | |
| "grad_norm": 11.187241554260254, | |
| "learning_rate": 9.102564102564103e-07, | |
| "loss": 0.2578, | |
| "step": 2127 | |
| }, | |
| { | |
| "epoch": 18.188034188034187, | |
| "grad_norm": 2.544975757598877, | |
| "learning_rate": 9.05982905982906e-07, | |
| "loss": 0.0868, | |
| "step": 2128 | |
| }, | |
| { | |
| "epoch": 18.196581196581196, | |
| "grad_norm": 2.490493059158325, | |
| "learning_rate": 9.017094017094017e-07, | |
| "loss": 0.1575, | |
| "step": 2129 | |
| }, | |
| { | |
| "epoch": 18.205128205128204, | |
| "grad_norm": 4.665895938873291, | |
| "learning_rate": 8.974358974358975e-07, | |
| "loss": 0.1644, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 18.213675213675213, | |
| "grad_norm": 3.135772943496704, | |
| "learning_rate": 8.931623931623933e-07, | |
| "loss": 0.205, | |
| "step": 2131 | |
| }, | |
| { | |
| "epoch": 18.22222222222222, | |
| "grad_norm": 1.5636606216430664, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 0.0541, | |
| "step": 2132 | |
| }, | |
| { | |
| "epoch": 18.23076923076923, | |
| "grad_norm": 3.603691816329956, | |
| "learning_rate": 8.846153846153848e-07, | |
| "loss": 0.0478, | |
| "step": 2133 | |
| }, | |
| { | |
| "epoch": 18.23931623931624, | |
| "grad_norm": 2.6537222862243652, | |
| "learning_rate": 8.803418803418804e-07, | |
| "loss": 0.1206, | |
| "step": 2134 | |
| }, | |
| { | |
| "epoch": 18.247863247863247, | |
| "grad_norm": 5.086421966552734, | |
| "learning_rate": 8.760683760683761e-07, | |
| "loss": 0.1212, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 18.256410256410255, | |
| "grad_norm": 4.673394203186035, | |
| "learning_rate": 8.717948717948718e-07, | |
| "loss": 0.0588, | |
| "step": 2136 | |
| }, | |
| { | |
| "epoch": 18.264957264957264, | |
| "grad_norm": 2.1376845836639404, | |
| "learning_rate": 8.675213675213676e-07, | |
| "loss": 0.0492, | |
| "step": 2137 | |
| }, | |
| { | |
| "epoch": 18.273504273504273, | |
| "grad_norm": 2.8616504669189453, | |
| "learning_rate": 8.632478632478633e-07, | |
| "loss": 0.1834, | |
| "step": 2138 | |
| }, | |
| { | |
| "epoch": 18.28205128205128, | |
| "grad_norm": 2.7179784774780273, | |
| "learning_rate": 8.58974358974359e-07, | |
| "loss": 0.1508, | |
| "step": 2139 | |
| }, | |
| { | |
| "epoch": 18.29059829059829, | |
| "grad_norm": 1.1909416913986206, | |
| "learning_rate": 8.547008547008548e-07, | |
| "loss": 0.0721, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 18.299145299145298, | |
| "grad_norm": 1.8272216320037842, | |
| "learning_rate": 8.504273504273505e-07, | |
| "loss": 0.0797, | |
| "step": 2141 | |
| }, | |
| { | |
| "epoch": 18.307692307692307, | |
| "grad_norm": 4.394528388977051, | |
| "learning_rate": 8.461538461538463e-07, | |
| "loss": 0.2762, | |
| "step": 2142 | |
| }, | |
| { | |
| "epoch": 18.316239316239315, | |
| "grad_norm": 4.276169776916504, | |
| "learning_rate": 8.41880341880342e-07, | |
| "loss": 0.0969, | |
| "step": 2143 | |
| }, | |
| { | |
| "epoch": 18.324786324786324, | |
| "grad_norm": 2.0932376384735107, | |
| "learning_rate": 8.376068376068377e-07, | |
| "loss": 0.0595, | |
| "step": 2144 | |
| }, | |
| { | |
| "epoch": 18.333333333333332, | |
| "grad_norm": 5.714378833770752, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 0.1176, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 18.34188034188034, | |
| "grad_norm": 1.1050394773483276, | |
| "learning_rate": 8.290598290598291e-07, | |
| "loss": 0.0284, | |
| "step": 2146 | |
| }, | |
| { | |
| "epoch": 18.35042735042735, | |
| "grad_norm": 3.2809271812438965, | |
| "learning_rate": 8.247863247863248e-07, | |
| "loss": 0.0737, | |
| "step": 2147 | |
| }, | |
| { | |
| "epoch": 18.358974358974358, | |
| "grad_norm": 2.102889060974121, | |
| "learning_rate": 8.205128205128206e-07, | |
| "loss": 0.0477, | |
| "step": 2148 | |
| }, | |
| { | |
| "epoch": 18.367521367521366, | |
| "grad_norm": 1.5728402137756348, | |
| "learning_rate": 8.162393162393164e-07, | |
| "loss": 0.0476, | |
| "step": 2149 | |
| }, | |
| { | |
| "epoch": 18.376068376068375, | |
| "grad_norm": 2.0337905883789062, | |
| "learning_rate": 8.11965811965812e-07, | |
| "loss": 0.019, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 18.384615384615383, | |
| "grad_norm": 5.475340843200684, | |
| "learning_rate": 8.076923076923078e-07, | |
| "loss": 0.1625, | |
| "step": 2151 | |
| }, | |
| { | |
| "epoch": 18.39316239316239, | |
| "grad_norm": 0.4993753135204315, | |
| "learning_rate": 8.034188034188035e-07, | |
| "loss": 0.0132, | |
| "step": 2152 | |
| }, | |
| { | |
| "epoch": 18.4017094017094, | |
| "grad_norm": 4.052933216094971, | |
| "learning_rate": 7.991452991452992e-07, | |
| "loss": 0.1603, | |
| "step": 2153 | |
| }, | |
| { | |
| "epoch": 18.41025641025641, | |
| "grad_norm": 3.005293607711792, | |
| "learning_rate": 7.948717948717949e-07, | |
| "loss": 0.0399, | |
| "step": 2154 | |
| }, | |
| { | |
| "epoch": 18.418803418803417, | |
| "grad_norm": 3.0186731815338135, | |
| "learning_rate": 7.905982905982906e-07, | |
| "loss": 0.0564, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 18.427350427350426, | |
| "grad_norm": 5.522226333618164, | |
| "learning_rate": 7.863247863247863e-07, | |
| "loss": 0.1138, | |
| "step": 2156 | |
| }, | |
| { | |
| "epoch": 18.435897435897434, | |
| "grad_norm": 5.463916301727295, | |
| "learning_rate": 7.820512820512821e-07, | |
| "loss": 0.4811, | |
| "step": 2157 | |
| }, | |
| { | |
| "epoch": 18.444444444444443, | |
| "grad_norm": 0.41404595971107483, | |
| "learning_rate": 7.777777777777779e-07, | |
| "loss": 0.0114, | |
| "step": 2158 | |
| }, | |
| { | |
| "epoch": 18.45299145299145, | |
| "grad_norm": 0.9279537200927734, | |
| "learning_rate": 7.735042735042736e-07, | |
| "loss": 0.0268, | |
| "step": 2159 | |
| }, | |
| { | |
| "epoch": 18.46153846153846, | |
| "grad_norm": 0.5745738744735718, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 0.0155, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 18.47008547008547, | |
| "grad_norm": 2.329507827758789, | |
| "learning_rate": 7.64957264957265e-07, | |
| "loss": 0.0421, | |
| "step": 2161 | |
| }, | |
| { | |
| "epoch": 18.478632478632477, | |
| "grad_norm": 2.934424638748169, | |
| "learning_rate": 7.606837606837607e-07, | |
| "loss": 0.0925, | |
| "step": 2162 | |
| }, | |
| { | |
| "epoch": 18.487179487179485, | |
| "grad_norm": 3.226261854171753, | |
| "learning_rate": 7.564102564102564e-07, | |
| "loss": 0.1914, | |
| "step": 2163 | |
| }, | |
| { | |
| "epoch": 18.495726495726494, | |
| "grad_norm": 1.2033684253692627, | |
| "learning_rate": 7.521367521367522e-07, | |
| "loss": 0.0218, | |
| "step": 2164 | |
| }, | |
| { | |
| "epoch": 18.504273504273506, | |
| "grad_norm": 1.092015266418457, | |
| "learning_rate": 7.478632478632479e-07, | |
| "loss": 0.0165, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 18.51282051282051, | |
| "grad_norm": 1.2283809185028076, | |
| "learning_rate": 7.435897435897436e-07, | |
| "loss": 0.025, | |
| "step": 2166 | |
| }, | |
| { | |
| "epoch": 18.521367521367523, | |
| "grad_norm": 6.3457722663879395, | |
| "learning_rate": 7.393162393162394e-07, | |
| "loss": 0.2224, | |
| "step": 2167 | |
| }, | |
| { | |
| "epoch": 18.52991452991453, | |
| "grad_norm": 4.920536518096924, | |
| "learning_rate": 7.350427350427351e-07, | |
| "loss": 0.1381, | |
| "step": 2168 | |
| }, | |
| { | |
| "epoch": 18.53846153846154, | |
| "grad_norm": 4.16088342666626, | |
| "learning_rate": 7.307692307692309e-07, | |
| "loss": 0.2725, | |
| "step": 2169 | |
| }, | |
| { | |
| "epoch": 18.54700854700855, | |
| "grad_norm": 1.4776932001113892, | |
| "learning_rate": 7.264957264957266e-07, | |
| "loss": 0.0236, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 18.555555555555557, | |
| "grad_norm": 5.517492294311523, | |
| "learning_rate": 7.222222222222222e-07, | |
| "loss": 0.3427, | |
| "step": 2171 | |
| }, | |
| { | |
| "epoch": 18.564102564102566, | |
| "grad_norm": 0.7798398733139038, | |
| "learning_rate": 7.179487179487179e-07, | |
| "loss": 0.0139, | |
| "step": 2172 | |
| }, | |
| { | |
| "epoch": 18.572649572649574, | |
| "grad_norm": 0.7174245119094849, | |
| "learning_rate": 7.136752136752137e-07, | |
| "loss": 0.0144, | |
| "step": 2173 | |
| }, | |
| { | |
| "epoch": 18.581196581196583, | |
| "grad_norm": 5.118779182434082, | |
| "learning_rate": 7.094017094017094e-07, | |
| "loss": 0.1899, | |
| "step": 2174 | |
| }, | |
| { | |
| "epoch": 18.58974358974359, | |
| "grad_norm": 2.8726353645324707, | |
| "learning_rate": 7.051282051282052e-07, | |
| "loss": 0.1177, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 18.5982905982906, | |
| "grad_norm": 2.3775036334991455, | |
| "learning_rate": 7.00854700854701e-07, | |
| "loss": 0.1183, | |
| "step": 2176 | |
| }, | |
| { | |
| "epoch": 18.60683760683761, | |
| "grad_norm": 19.23975944519043, | |
| "learning_rate": 6.965811965811966e-07, | |
| "loss": 0.4534, | |
| "step": 2177 | |
| }, | |
| { | |
| "epoch": 18.615384615384617, | |
| "grad_norm": 1.3832803964614868, | |
| "learning_rate": 6.923076923076924e-07, | |
| "loss": 0.0309, | |
| "step": 2178 | |
| }, | |
| { | |
| "epoch": 18.623931623931625, | |
| "grad_norm": 1.6752214431762695, | |
| "learning_rate": 6.880341880341881e-07, | |
| "loss": 0.0201, | |
| "step": 2179 | |
| }, | |
| { | |
| "epoch": 18.632478632478634, | |
| "grad_norm": 3.1885950565338135, | |
| "learning_rate": 6.837606837606839e-07, | |
| "loss": 0.1242, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 18.641025641025642, | |
| "grad_norm": 0.9290790557861328, | |
| "learning_rate": 6.794871794871795e-07, | |
| "loss": 0.0189, | |
| "step": 2181 | |
| }, | |
| { | |
| "epoch": 18.64957264957265, | |
| "grad_norm": 0.25725051760673523, | |
| "learning_rate": 6.752136752136752e-07, | |
| "loss": 0.0065, | |
| "step": 2182 | |
| }, | |
| { | |
| "epoch": 18.65811965811966, | |
| "grad_norm": 1.9815839529037476, | |
| "learning_rate": 6.709401709401709e-07, | |
| "loss": 0.0576, | |
| "step": 2183 | |
| }, | |
| { | |
| "epoch": 18.666666666666668, | |
| "grad_norm": 1.924490213394165, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.0671, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 18.675213675213676, | |
| "grad_norm": 2.9947164058685303, | |
| "learning_rate": 6.623931623931625e-07, | |
| "loss": 0.1859, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 18.683760683760685, | |
| "grad_norm": 1.8680211305618286, | |
| "learning_rate": 6.581196581196582e-07, | |
| "loss": 0.1028, | |
| "step": 2186 | |
| }, | |
| { | |
| "epoch": 18.692307692307693, | |
| "grad_norm": 0.823103666305542, | |
| "learning_rate": 6.53846153846154e-07, | |
| "loss": 0.0198, | |
| "step": 2187 | |
| }, | |
| { | |
| "epoch": 18.700854700854702, | |
| "grad_norm": 2.3616061210632324, | |
| "learning_rate": 6.495726495726496e-07, | |
| "loss": 0.1025, | |
| "step": 2188 | |
| }, | |
| { | |
| "epoch": 18.70940170940171, | |
| "grad_norm": 3.1370067596435547, | |
| "learning_rate": 6.452991452991454e-07, | |
| "loss": 0.0438, | |
| "step": 2189 | |
| }, | |
| { | |
| "epoch": 18.71794871794872, | |
| "grad_norm": 8.058025360107422, | |
| "learning_rate": 6.41025641025641e-07, | |
| "loss": 0.082, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 18.726495726495727, | |
| "grad_norm": 2.1969916820526123, | |
| "learning_rate": 6.367521367521368e-07, | |
| "loss": 0.1074, | |
| "step": 2191 | |
| }, | |
| { | |
| "epoch": 18.735042735042736, | |
| "grad_norm": 2.5845255851745605, | |
| "learning_rate": 6.324786324786325e-07, | |
| "loss": 0.0795, | |
| "step": 2192 | |
| }, | |
| { | |
| "epoch": 18.743589743589745, | |
| "grad_norm": 3.578331708908081, | |
| "learning_rate": 6.282051282051282e-07, | |
| "loss": 0.1111, | |
| "step": 2193 | |
| }, | |
| { | |
| "epoch": 18.752136752136753, | |
| "grad_norm": 1.5390626192092896, | |
| "learning_rate": 6.23931623931624e-07, | |
| "loss": 0.064, | |
| "step": 2194 | |
| }, | |
| { | |
| "epoch": 18.76068376068376, | |
| "grad_norm": 3.1742804050445557, | |
| "learning_rate": 6.196581196581197e-07, | |
| "loss": 0.0971, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 18.76923076923077, | |
| "grad_norm": 1.7017542123794556, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 0.0424, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.77777777777778, | |
| "grad_norm": 2.642102003097534, | |
| "learning_rate": 6.111111111111112e-07, | |
| "loss": 0.1243, | |
| "step": 2197 | |
| }, | |
| { | |
| "epoch": 18.786324786324787, | |
| "grad_norm": 1.2010291814804077, | |
| "learning_rate": 6.068376068376068e-07, | |
| "loss": 0.0375, | |
| "step": 2198 | |
| }, | |
| { | |
| "epoch": 18.794871794871796, | |
| "grad_norm": 3.1580190658569336, | |
| "learning_rate": 6.025641025641026e-07, | |
| "loss": 0.0565, | |
| "step": 2199 | |
| }, | |
| { | |
| "epoch": 18.803418803418804, | |
| "grad_norm": 2.7660391330718994, | |
| "learning_rate": 5.982905982905984e-07, | |
| "loss": 0.0385, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 18.811965811965813, | |
| "grad_norm": 0.7716617584228516, | |
| "learning_rate": 5.94017094017094e-07, | |
| "loss": 0.0159, | |
| "step": 2201 | |
| }, | |
| { | |
| "epoch": 18.82051282051282, | |
| "grad_norm": 3.190251588821411, | |
| "learning_rate": 5.897435897435898e-07, | |
| "loss": 0.241, | |
| "step": 2202 | |
| }, | |
| { | |
| "epoch": 18.82905982905983, | |
| "grad_norm": 7.115220069885254, | |
| "learning_rate": 5.854700854700856e-07, | |
| "loss": 0.1777, | |
| "step": 2203 | |
| }, | |
| { | |
| "epoch": 18.837606837606838, | |
| "grad_norm": 5.071573257446289, | |
| "learning_rate": 5.811965811965812e-07, | |
| "loss": 0.5421, | |
| "step": 2204 | |
| }, | |
| { | |
| "epoch": 18.846153846153847, | |
| "grad_norm": 3.8419785499572754, | |
| "learning_rate": 5.76923076923077e-07, | |
| "loss": 0.0784, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 18.854700854700855, | |
| "grad_norm": 2.8234896659851074, | |
| "learning_rate": 5.726495726495727e-07, | |
| "loss": 0.1071, | |
| "step": 2206 | |
| }, | |
| { | |
| "epoch": 18.863247863247864, | |
| "grad_norm": 1.4067480564117432, | |
| "learning_rate": 5.683760683760684e-07, | |
| "loss": 0.0375, | |
| "step": 2207 | |
| }, | |
| { | |
| "epoch": 18.871794871794872, | |
| "grad_norm": 2.508589029312134, | |
| "learning_rate": 5.641025641025642e-07, | |
| "loss": 0.0921, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 18.88034188034188, | |
| "grad_norm": 7.314038276672363, | |
| "learning_rate": 5.598290598290599e-07, | |
| "loss": 0.3581, | |
| "step": 2209 | |
| }, | |
| { | |
| "epoch": 18.88888888888889, | |
| "grad_norm": 4.375041961669922, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 0.1115, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 18.897435897435898, | |
| "grad_norm": 4.789741516113281, | |
| "learning_rate": 5.512820512820513e-07, | |
| "loss": 0.1813, | |
| "step": 2211 | |
| }, | |
| { | |
| "epoch": 18.905982905982906, | |
| "grad_norm": 3.008720874786377, | |
| "learning_rate": 5.470085470085471e-07, | |
| "loss": 0.104, | |
| "step": 2212 | |
| }, | |
| { | |
| "epoch": 18.914529914529915, | |
| "grad_norm": 0.6364433765411377, | |
| "learning_rate": 5.427350427350428e-07, | |
| "loss": 0.0153, | |
| "step": 2213 | |
| }, | |
| { | |
| "epoch": 18.923076923076923, | |
| "grad_norm": 1.4009958505630493, | |
| "learning_rate": 5.384615384615386e-07, | |
| "loss": 0.0499, | |
| "step": 2214 | |
| }, | |
| { | |
| "epoch": 18.931623931623932, | |
| "grad_norm": 4.53135347366333, | |
| "learning_rate": 5.341880341880342e-07, | |
| "loss": 0.1021, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 18.94017094017094, | |
| "grad_norm": 0.7855163216590881, | |
| "learning_rate": 5.299145299145299e-07, | |
| "loss": 0.0297, | |
| "step": 2216 | |
| }, | |
| { | |
| "epoch": 18.94871794871795, | |
| "grad_norm": 1.5316343307495117, | |
| "learning_rate": 5.256410256410257e-07, | |
| "loss": 0.0438, | |
| "step": 2217 | |
| }, | |
| { | |
| "epoch": 18.957264957264957, | |
| "grad_norm": 1.2713849544525146, | |
| "learning_rate": 5.213675213675215e-07, | |
| "loss": 0.0311, | |
| "step": 2218 | |
| }, | |
| { | |
| "epoch": 18.965811965811966, | |
| "grad_norm": 1.612418293952942, | |
| "learning_rate": 5.17094017094017e-07, | |
| "loss": 0.0796, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 18.974358974358974, | |
| "grad_norm": 6.046596527099609, | |
| "learning_rate": 5.128205128205128e-07, | |
| "loss": 0.0835, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 18.982905982905983, | |
| "grad_norm": 2.527993679046631, | |
| "learning_rate": 5.085470085470086e-07, | |
| "loss": 0.0448, | |
| "step": 2221 | |
| }, | |
| { | |
| "epoch": 18.99145299145299, | |
| "grad_norm": 0.9519897699356079, | |
| "learning_rate": 5.042735042735043e-07, | |
| "loss": 0.0223, | |
| "step": 2222 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 14.08708667755127, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.6753, | |
| "step": 2223 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 0.05170569196343422, | |
| "eval_runtime": 9.3972, | |
| "eval_samples_per_second": 49.589, | |
| "eval_steps_per_second": 6.278, | |
| "step": 2223 | |
| }, | |
| { | |
| "epoch": 19.00854700854701, | |
| "grad_norm": 5.215019702911377, | |
| "learning_rate": 4.957264957264958e-07, | |
| "loss": 0.1614, | |
| "step": 2224 | |
| }, | |
| { | |
| "epoch": 19.017094017094017, | |
| "grad_norm": 2.855567216873169, | |
| "learning_rate": 4.914529914529914e-07, | |
| "loss": 0.1051, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 19.025641025641026, | |
| "grad_norm": 4.078762054443359, | |
| "learning_rate": 4.871794871794872e-07, | |
| "loss": 0.2859, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 19.034188034188034, | |
| "grad_norm": 0.9259152412414551, | |
| "learning_rate": 4.82905982905983e-07, | |
| "loss": 0.0257, | |
| "step": 2227 | |
| }, | |
| { | |
| "epoch": 19.042735042735043, | |
| "grad_norm": 3.629925012588501, | |
| "learning_rate": 4.786324786324787e-07, | |
| "loss": 0.1283, | |
| "step": 2228 | |
| }, | |
| { | |
| "epoch": 19.05128205128205, | |
| "grad_norm": 3.104196310043335, | |
| "learning_rate": 4.7435897435897437e-07, | |
| "loss": 0.0701, | |
| "step": 2229 | |
| }, | |
| { | |
| "epoch": 19.05982905982906, | |
| "grad_norm": 8.760592460632324, | |
| "learning_rate": 4.700854700854701e-07, | |
| "loss": 0.5793, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 19.068376068376068, | |
| "grad_norm": 1.2966917753219604, | |
| "learning_rate": 4.658119658119659e-07, | |
| "loss": 0.0573, | |
| "step": 2231 | |
| }, | |
| { | |
| "epoch": 19.076923076923077, | |
| "grad_norm": 1.7045038938522339, | |
| "learning_rate": 4.615384615384616e-07, | |
| "loss": 0.0497, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 19.085470085470085, | |
| "grad_norm": 7.805142402648926, | |
| "learning_rate": 4.572649572649573e-07, | |
| "loss": 0.2898, | |
| "step": 2233 | |
| }, | |
| { | |
| "epoch": 19.094017094017094, | |
| "grad_norm": 0.5019100308418274, | |
| "learning_rate": 4.52991452991453e-07, | |
| "loss": 0.0132, | |
| "step": 2234 | |
| }, | |
| { | |
| "epoch": 19.102564102564102, | |
| "grad_norm": 3.1100540161132812, | |
| "learning_rate": 4.4871794871794876e-07, | |
| "loss": 0.0874, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 19.11111111111111, | |
| "grad_norm": 0.40422680974006653, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 0.012, | |
| "step": 2236 | |
| }, | |
| { | |
| "epoch": 19.11965811965812, | |
| "grad_norm": 1.2845938205718994, | |
| "learning_rate": 4.401709401709402e-07, | |
| "loss": 0.0259, | |
| "step": 2237 | |
| }, | |
| { | |
| "epoch": 19.128205128205128, | |
| "grad_norm": 4.621537208557129, | |
| "learning_rate": 4.358974358974359e-07, | |
| "loss": 0.246, | |
| "step": 2238 | |
| }, | |
| { | |
| "epoch": 19.136752136752136, | |
| "grad_norm": 1.1688278913497925, | |
| "learning_rate": 4.3162393162393163e-07, | |
| "loss": 0.0804, | |
| "step": 2239 | |
| }, | |
| { | |
| "epoch": 19.145299145299145, | |
| "grad_norm": 10.896872520446777, | |
| "learning_rate": 4.273504273504274e-07, | |
| "loss": 0.2695, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 19.153846153846153, | |
| "grad_norm": 2.7485415935516357, | |
| "learning_rate": 4.2307692307692315e-07, | |
| "loss": 0.0474, | |
| "step": 2241 | |
| }, | |
| { | |
| "epoch": 19.162393162393162, | |
| "grad_norm": 1.1686739921569824, | |
| "learning_rate": 4.188034188034188e-07, | |
| "loss": 0.0257, | |
| "step": 2242 | |
| }, | |
| { | |
| "epoch": 19.17094017094017, | |
| "grad_norm": 3.5579254627227783, | |
| "learning_rate": 4.1452991452991456e-07, | |
| "loss": 0.0419, | |
| "step": 2243 | |
| }, | |
| { | |
| "epoch": 19.17948717948718, | |
| "grad_norm": 3.088649034500122, | |
| "learning_rate": 4.102564102564103e-07, | |
| "loss": 0.1229, | |
| "step": 2244 | |
| }, | |
| { | |
| "epoch": 19.188034188034187, | |
| "grad_norm": 1.4894665479660034, | |
| "learning_rate": 4.05982905982906e-07, | |
| "loss": 0.0414, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 19.196581196581196, | |
| "grad_norm": 5.022091865539551, | |
| "learning_rate": 4.0170940170940175e-07, | |
| "loss": 0.1423, | |
| "step": 2246 | |
| }, | |
| { | |
| "epoch": 19.205128205128204, | |
| "grad_norm": 1.6117054224014282, | |
| "learning_rate": 3.9743589743589743e-07, | |
| "loss": 0.0244, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 19.213675213675213, | |
| "grad_norm": 0.5429085493087769, | |
| "learning_rate": 3.9316239316239316e-07, | |
| "loss": 0.0122, | |
| "step": 2248 | |
| }, | |
| { | |
| "epoch": 19.22222222222222, | |
| "grad_norm": 7.429282188415527, | |
| "learning_rate": 3.8888888888888895e-07, | |
| "loss": 0.122, | |
| "step": 2249 | |
| }, | |
| { | |
| "epoch": 19.23076923076923, | |
| "grad_norm": 4.492022514343262, | |
| "learning_rate": 3.846153846153847e-07, | |
| "loss": 0.3181, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 19.23931623931624, | |
| "grad_norm": 5.219499588012695, | |
| "learning_rate": 3.8034188034188036e-07, | |
| "loss": 0.1374, | |
| "step": 2251 | |
| }, | |
| { | |
| "epoch": 19.247863247863247, | |
| "grad_norm": 3.454345941543579, | |
| "learning_rate": 3.760683760683761e-07, | |
| "loss": 0.147, | |
| "step": 2252 | |
| }, | |
| { | |
| "epoch": 19.256410256410255, | |
| "grad_norm": 0.6370477080345154, | |
| "learning_rate": 3.717948717948718e-07, | |
| "loss": 0.0154, | |
| "step": 2253 | |
| }, | |
| { | |
| "epoch": 19.264957264957264, | |
| "grad_norm": 1.7189971208572388, | |
| "learning_rate": 3.6752136752136755e-07, | |
| "loss": 0.0635, | |
| "step": 2254 | |
| }, | |
| { | |
| "epoch": 19.273504273504273, | |
| "grad_norm": 2.716744899749756, | |
| "learning_rate": 3.632478632478633e-07, | |
| "loss": 0.0966, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 19.28205128205128, | |
| "grad_norm": 2.4959864616394043, | |
| "learning_rate": 3.5897435897435896e-07, | |
| "loss": 0.0779, | |
| "step": 2256 | |
| }, | |
| { | |
| "epoch": 19.29059829059829, | |
| "grad_norm": 3.625793218612671, | |
| "learning_rate": 3.547008547008547e-07, | |
| "loss": 0.3238, | |
| "step": 2257 | |
| }, | |
| { | |
| "epoch": 19.299145299145298, | |
| "grad_norm": 1.8783844709396362, | |
| "learning_rate": 3.504273504273505e-07, | |
| "loss": 0.0319, | |
| "step": 2258 | |
| }, | |
| { | |
| "epoch": 19.307692307692307, | |
| "grad_norm": 1.6740922927856445, | |
| "learning_rate": 3.461538461538462e-07, | |
| "loss": 0.0844, | |
| "step": 2259 | |
| }, | |
| { | |
| "epoch": 19.316239316239315, | |
| "grad_norm": 2.8891098499298096, | |
| "learning_rate": 3.4188034188034194e-07, | |
| "loss": 0.1916, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 19.324786324786324, | |
| "grad_norm": 0.9975456595420837, | |
| "learning_rate": 3.376068376068376e-07, | |
| "loss": 0.0266, | |
| "step": 2261 | |
| }, | |
| { | |
| "epoch": 19.333333333333332, | |
| "grad_norm": 2.576789379119873, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 0.0722, | |
| "step": 2262 | |
| }, | |
| { | |
| "epoch": 19.34188034188034, | |
| "grad_norm": 9.070858001708984, | |
| "learning_rate": 3.290598290598291e-07, | |
| "loss": 0.2998, | |
| "step": 2263 | |
| }, | |
| { | |
| "epoch": 19.35042735042735, | |
| "grad_norm": 3.052319049835205, | |
| "learning_rate": 3.247863247863248e-07, | |
| "loss": 0.0435, | |
| "step": 2264 | |
| }, | |
| { | |
| "epoch": 19.358974358974358, | |
| "grad_norm": 0.8035821318626404, | |
| "learning_rate": 3.205128205128205e-07, | |
| "loss": 0.0233, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 19.367521367521366, | |
| "grad_norm": 3.7658371925354004, | |
| "learning_rate": 3.1623931623931623e-07, | |
| "loss": 0.3007, | |
| "step": 2266 | |
| }, | |
| { | |
| "epoch": 19.376068376068375, | |
| "grad_norm": 1.210494875907898, | |
| "learning_rate": 3.11965811965812e-07, | |
| "loss": 0.0344, | |
| "step": 2267 | |
| }, | |
| { | |
| "epoch": 19.384615384615383, | |
| "grad_norm": 1.1121772527694702, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.054, | |
| "step": 2268 | |
| }, | |
| { | |
| "epoch": 19.39316239316239, | |
| "grad_norm": 2.842228412628174, | |
| "learning_rate": 3.034188034188034e-07, | |
| "loss": 0.0814, | |
| "step": 2269 | |
| }, | |
| { | |
| "epoch": 19.4017094017094, | |
| "grad_norm": 1.9269556999206543, | |
| "learning_rate": 2.991452991452992e-07, | |
| "loss": 0.0354, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 19.41025641025641, | |
| "grad_norm": 7.359715938568115, | |
| "learning_rate": 2.948717948717949e-07, | |
| "loss": 0.3288, | |
| "step": 2271 | |
| }, | |
| { | |
| "epoch": 19.418803418803417, | |
| "grad_norm": 1.7621564865112305, | |
| "learning_rate": 2.905982905982906e-07, | |
| "loss": 0.0313, | |
| "step": 2272 | |
| }, | |
| { | |
| "epoch": 19.427350427350426, | |
| "grad_norm": 2.5410284996032715, | |
| "learning_rate": 2.8632478632478635e-07, | |
| "loss": 0.076, | |
| "step": 2273 | |
| }, | |
| { | |
| "epoch": 19.435897435897434, | |
| "grad_norm": 5.633874416351318, | |
| "learning_rate": 2.820512820512821e-07, | |
| "loss": 0.1903, | |
| "step": 2274 | |
| }, | |
| { | |
| "epoch": 19.444444444444443, | |
| "grad_norm": 1.935703158378601, | |
| "learning_rate": 2.7777777777777776e-07, | |
| "loss": 0.3778, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 19.45299145299145, | |
| "grad_norm": 7.559366703033447, | |
| "learning_rate": 2.7350427350427354e-07, | |
| "loss": 0.2684, | |
| "step": 2276 | |
| }, | |
| { | |
| "epoch": 19.46153846153846, | |
| "grad_norm": 9.240869522094727, | |
| "learning_rate": 2.692307692307693e-07, | |
| "loss": 0.2982, | |
| "step": 2277 | |
| }, | |
| { | |
| "epoch": 19.47008547008547, | |
| "grad_norm": 6.940350532531738, | |
| "learning_rate": 2.6495726495726495e-07, | |
| "loss": 0.3131, | |
| "step": 2278 | |
| }, | |
| { | |
| "epoch": 19.478632478632477, | |
| "grad_norm": 1.3201594352722168, | |
| "learning_rate": 2.6068376068376074e-07, | |
| "loss": 0.0191, | |
| "step": 2279 | |
| }, | |
| { | |
| "epoch": 19.487179487179485, | |
| "grad_norm": 1.626806616783142, | |
| "learning_rate": 2.564102564102564e-07, | |
| "loss": 0.0361, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 19.495726495726494, | |
| "grad_norm": 8.687582969665527, | |
| "learning_rate": 2.5213675213675215e-07, | |
| "loss": 0.1942, | |
| "step": 2281 | |
| }, | |
| { | |
| "epoch": 19.504273504273506, | |
| "grad_norm": 5.104561805725098, | |
| "learning_rate": 2.478632478632479e-07, | |
| "loss": 0.1906, | |
| "step": 2282 | |
| }, | |
| { | |
| "epoch": 19.51282051282051, | |
| "grad_norm": 2.8611207008361816, | |
| "learning_rate": 2.435897435897436e-07, | |
| "loss": 0.1258, | |
| "step": 2283 | |
| }, | |
| { | |
| "epoch": 19.521367521367523, | |
| "grad_norm": 1.2258422374725342, | |
| "learning_rate": 2.3931623931623934e-07, | |
| "loss": 0.0186, | |
| "step": 2284 | |
| }, | |
| { | |
| "epoch": 19.52991452991453, | |
| "grad_norm": 5.307450294494629, | |
| "learning_rate": 2.3504273504273505e-07, | |
| "loss": 0.1356, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 19.53846153846154, | |
| "grad_norm": 2.0854647159576416, | |
| "learning_rate": 2.307692307692308e-07, | |
| "loss": 0.0533, | |
| "step": 2286 | |
| }, | |
| { | |
| "epoch": 19.54700854700855, | |
| "grad_norm": 1.8560184240341187, | |
| "learning_rate": 2.264957264957265e-07, | |
| "loss": 0.048, | |
| "step": 2287 | |
| }, | |
| { | |
| "epoch": 19.555555555555557, | |
| "grad_norm": 5.781933307647705, | |
| "learning_rate": 2.2222222222222224e-07, | |
| "loss": 0.2769, | |
| "step": 2288 | |
| }, | |
| { | |
| "epoch": 19.564102564102566, | |
| "grad_norm": 4.858759880065918, | |
| "learning_rate": 2.1794871794871795e-07, | |
| "loss": 0.4217, | |
| "step": 2289 | |
| }, | |
| { | |
| "epoch": 19.572649572649574, | |
| "grad_norm": 3.7598235607147217, | |
| "learning_rate": 2.136752136752137e-07, | |
| "loss": 0.162, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 19.581196581196583, | |
| "grad_norm": 0.5706556439399719, | |
| "learning_rate": 2.094017094017094e-07, | |
| "loss": 0.0151, | |
| "step": 2291 | |
| }, | |
| { | |
| "epoch": 19.58974358974359, | |
| "grad_norm": 5.697900295257568, | |
| "learning_rate": 2.0512820512820514e-07, | |
| "loss": 0.1015, | |
| "step": 2292 | |
| }, | |
| { | |
| "epoch": 19.5982905982906, | |
| "grad_norm": 4.635442733764648, | |
| "learning_rate": 2.0085470085470088e-07, | |
| "loss": 0.1827, | |
| "step": 2293 | |
| }, | |
| { | |
| "epoch": 19.60683760683761, | |
| "grad_norm": 3.070131778717041, | |
| "learning_rate": 1.9658119658119658e-07, | |
| "loss": 0.0802, | |
| "step": 2294 | |
| }, | |
| { | |
| "epoch": 19.615384615384617, | |
| "grad_norm": 0.979217529296875, | |
| "learning_rate": 1.9230769230769234e-07, | |
| "loss": 0.0237, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 19.623931623931625, | |
| "grad_norm": 5.640648365020752, | |
| "learning_rate": 1.8803418803418804e-07, | |
| "loss": 0.0588, | |
| "step": 2296 | |
| }, | |
| { | |
| "epoch": 19.632478632478634, | |
| "grad_norm": 7.1512861251831055, | |
| "learning_rate": 1.8376068376068378e-07, | |
| "loss": 0.1942, | |
| "step": 2297 | |
| }, | |
| { | |
| "epoch": 19.641025641025642, | |
| "grad_norm": 12.868803024291992, | |
| "learning_rate": 1.7948717948717948e-07, | |
| "loss": 0.2771, | |
| "step": 2298 | |
| }, | |
| { | |
| "epoch": 19.64957264957265, | |
| "grad_norm": 2.954000234603882, | |
| "learning_rate": 1.7521367521367524e-07, | |
| "loss": 0.1124, | |
| "step": 2299 | |
| }, | |
| { | |
| "epoch": 19.65811965811966, | |
| "grad_norm": 0.47206825017929077, | |
| "learning_rate": 1.7094017094017097e-07, | |
| "loss": 0.0104, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 19.666666666666668, | |
| "grad_norm": 0.6243001818656921, | |
| "learning_rate": 1.6666666666666668e-07, | |
| "loss": 0.0145, | |
| "step": 2301 | |
| }, | |
| { | |
| "epoch": 19.675213675213676, | |
| "grad_norm": 1.6680350303649902, | |
| "learning_rate": 1.623931623931624e-07, | |
| "loss": 0.0634, | |
| "step": 2302 | |
| }, | |
| { | |
| "epoch": 19.683760683760685, | |
| "grad_norm": 6.298573017120361, | |
| "learning_rate": 1.5811965811965811e-07, | |
| "loss": 0.2083, | |
| "step": 2303 | |
| }, | |
| { | |
| "epoch": 19.692307692307693, | |
| "grad_norm": 0.622466504573822, | |
| "learning_rate": 1.5384615384615387e-07, | |
| "loss": 0.0155, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 19.700854700854702, | |
| "grad_norm": 2.289080858230591, | |
| "learning_rate": 1.495726495726496e-07, | |
| "loss": 0.0698, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 19.70940170940171, | |
| "grad_norm": 13.065472602844238, | |
| "learning_rate": 1.452991452991453e-07, | |
| "loss": 0.2587, | |
| "step": 2306 | |
| }, | |
| { | |
| "epoch": 19.71794871794872, | |
| "grad_norm": 0.903513491153717, | |
| "learning_rate": 1.4102564102564104e-07, | |
| "loss": 0.0222, | |
| "step": 2307 | |
| }, | |
| { | |
| "epoch": 19.726495726495727, | |
| "grad_norm": 1.3763283491134644, | |
| "learning_rate": 1.3675213675213677e-07, | |
| "loss": 0.042, | |
| "step": 2308 | |
| }, | |
| { | |
| "epoch": 19.735042735042736, | |
| "grad_norm": 3.3493802547454834, | |
| "learning_rate": 1.3247863247863248e-07, | |
| "loss": 0.1042, | |
| "step": 2309 | |
| }, | |
| { | |
| "epoch": 19.743589743589745, | |
| "grad_norm": 12.862226486206055, | |
| "learning_rate": 1.282051282051282e-07, | |
| "loss": 0.359, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 19.752136752136753, | |
| "grad_norm": 5.56069278717041, | |
| "learning_rate": 1.2393162393162394e-07, | |
| "loss": 0.1645, | |
| "step": 2311 | |
| }, | |
| { | |
| "epoch": 19.76068376068376, | |
| "grad_norm": 2.900381326675415, | |
| "learning_rate": 1.1965811965811967e-07, | |
| "loss": 0.1641, | |
| "step": 2312 | |
| }, | |
| { | |
| "epoch": 19.76923076923077, | |
| "grad_norm": 1.3674333095550537, | |
| "learning_rate": 1.153846153846154e-07, | |
| "loss": 0.0428, | |
| "step": 2313 | |
| }, | |
| { | |
| "epoch": 19.77777777777778, | |
| "grad_norm": 2.06278657913208, | |
| "learning_rate": 1.1111111111111112e-07, | |
| "loss": 0.0404, | |
| "step": 2314 | |
| }, | |
| { | |
| "epoch": 19.786324786324787, | |
| "grad_norm": 5.760499954223633, | |
| "learning_rate": 1.0683760683760685e-07, | |
| "loss": 0.1298, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 19.794871794871796, | |
| "grad_norm": 3.2554516792297363, | |
| "learning_rate": 1.0256410256410257e-07, | |
| "loss": 0.0432, | |
| "step": 2316 | |
| }, | |
| { | |
| "epoch": 19.803418803418804, | |
| "grad_norm": 1.7984355688095093, | |
| "learning_rate": 9.829059829059829e-08, | |
| "loss": 0.0461, | |
| "step": 2317 | |
| }, | |
| { | |
| "epoch": 19.811965811965813, | |
| "grad_norm": 1.633736491203308, | |
| "learning_rate": 9.401709401709402e-08, | |
| "loss": 0.0746, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.82051282051282, | |
| "grad_norm": 2.6958866119384766, | |
| "learning_rate": 8.974358974358974e-08, | |
| "loss": 0.0852, | |
| "step": 2319 | |
| }, | |
| { | |
| "epoch": 19.82905982905983, | |
| "grad_norm": 0.9744161367416382, | |
| "learning_rate": 8.547008547008549e-08, | |
| "loss": 0.0368, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 19.837606837606838, | |
| "grad_norm": 1.2404037714004517, | |
| "learning_rate": 8.11965811965812e-08, | |
| "loss": 0.0547, | |
| "step": 2321 | |
| }, | |
| { | |
| "epoch": 19.846153846153847, | |
| "grad_norm": 1.6044564247131348, | |
| "learning_rate": 7.692307692307694e-08, | |
| "loss": 0.0441, | |
| "step": 2322 | |
| }, | |
| { | |
| "epoch": 19.854700854700855, | |
| "grad_norm": 0.47167596220970154, | |
| "learning_rate": 7.264957264957265e-08, | |
| "loss": 0.0099, | |
| "step": 2323 | |
| }, | |
| { | |
| "epoch": 19.863247863247864, | |
| "grad_norm": 1.6729376316070557, | |
| "learning_rate": 6.837606837606839e-08, | |
| "loss": 0.0258, | |
| "step": 2324 | |
| }, | |
| { | |
| "epoch": 19.871794871794872, | |
| "grad_norm": 0.5823857188224792, | |
| "learning_rate": 6.41025641025641e-08, | |
| "loss": 0.0131, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 19.88034188034188, | |
| "grad_norm": 4.055545806884766, | |
| "learning_rate": 5.982905982905984e-08, | |
| "loss": 0.073, | |
| "step": 2326 | |
| }, | |
| { | |
| "epoch": 19.88888888888889, | |
| "grad_norm": 2.693838596343994, | |
| "learning_rate": 5.555555555555556e-08, | |
| "loss": 0.0845, | |
| "step": 2327 | |
| }, | |
| { | |
| "epoch": 19.897435897435898, | |
| "grad_norm": 0.9895898103713989, | |
| "learning_rate": 5.1282051282051286e-08, | |
| "loss": 0.0205, | |
| "step": 2328 | |
| }, | |
| { | |
| "epoch": 19.905982905982906, | |
| "grad_norm": 3.560816526412964, | |
| "learning_rate": 4.700854700854701e-08, | |
| "loss": 0.0989, | |
| "step": 2329 | |
| }, | |
| { | |
| "epoch": 19.914529914529915, | |
| "grad_norm": 5.152528762817383, | |
| "learning_rate": 4.273504273504274e-08, | |
| "loss": 0.0133, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 19.923076923076923, | |
| "grad_norm": 1.709021806716919, | |
| "learning_rate": 3.846153846153847e-08, | |
| "loss": 0.068, | |
| "step": 2331 | |
| }, | |
| { | |
| "epoch": 19.931623931623932, | |
| "grad_norm": 0.4786951541900635, | |
| "learning_rate": 3.418803418803419e-08, | |
| "loss": 0.0141, | |
| "step": 2332 | |
| }, | |
| { | |
| "epoch": 19.94017094017094, | |
| "grad_norm": 1.5413727760314941, | |
| "learning_rate": 2.991452991452992e-08, | |
| "loss": 0.0246, | |
| "step": 2333 | |
| }, | |
| { | |
| "epoch": 19.94871794871795, | |
| "grad_norm": 1.019601583480835, | |
| "learning_rate": 2.5641025641025643e-08, | |
| "loss": 0.0199, | |
| "step": 2334 | |
| }, | |
| { | |
| "epoch": 19.957264957264957, | |
| "grad_norm": 1.6115524768829346, | |
| "learning_rate": 2.136752136752137e-08, | |
| "loss": 0.0752, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 19.965811965811966, | |
| "grad_norm": 2.381624698638916, | |
| "learning_rate": 1.7094017094017096e-08, | |
| "loss": 0.0609, | |
| "step": 2336 | |
| }, | |
| { | |
| "epoch": 19.974358974358974, | |
| "grad_norm": 1.688704013824463, | |
| "learning_rate": 1.2820512820512822e-08, | |
| "loss": 0.0419, | |
| "step": 2337 | |
| }, | |
| { | |
| "epoch": 19.982905982905983, | |
| "grad_norm": 1.643002986907959, | |
| "learning_rate": 8.547008547008548e-09, | |
| "loss": 0.0456, | |
| "step": 2338 | |
| }, | |
| { | |
| "epoch": 19.99145299145299, | |
| "grad_norm": 3.5371882915496826, | |
| "learning_rate": 4.273504273504274e-09, | |
| "loss": 0.0392, | |
| "step": 2339 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 4.692568302154541, | |
| "learning_rate": 0.0, | |
| "loss": 0.1751, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.051427390426397324, | |
| "eval_runtime": 9.301, | |
| "eval_samples_per_second": 50.102, | |
| "eval_steps_per_second": 6.343, | |
| "step": 2340 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 2340, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 560912565657600.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |