| { | |
| "best_metric": 0.8354430379746836, | |
| "best_model_checkpoint": "deit-base-distilled-patch16-224-55-fold4/checkpoint-154", | |
| "epoch": 85.71428571428571, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.8571428571428571, | |
| "eval_accuracy": 0.5063291139240507, | |
| "eval_loss": 0.829655110836029, | |
| "eval_runtime": 1.0986, | |
| "eval_samples_per_second": 71.91, | |
| "eval_steps_per_second": 2.731, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.620253164556962, | |
| "eval_loss": 0.6548452973365784, | |
| "eval_runtime": 1.1343, | |
| "eval_samples_per_second": 69.644, | |
| "eval_steps_per_second": 2.645, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 2.468780040740967, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.6826, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "eval_accuracy": 0.5189873417721519, | |
| "eval_loss": 0.6513058543205261, | |
| "eval_runtime": 1.0973, | |
| "eval_samples_per_second": 71.995, | |
| "eval_steps_per_second": 2.734, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.6023642420768738, | |
| "eval_runtime": 1.1154, | |
| "eval_samples_per_second": 70.826, | |
| "eval_steps_per_second": 2.69, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "eval_accuracy": 0.5822784810126582, | |
| "eval_loss": 0.6472656726837158, | |
| "eval_runtime": 1.1141, | |
| "eval_samples_per_second": 70.912, | |
| "eval_steps_per_second": 2.693, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 2.0615956783294678, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.6198, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6835443037974683, | |
| "eval_loss": 0.5716904997825623, | |
| "eval_runtime": 1.1213, | |
| "eval_samples_per_second": 70.456, | |
| "eval_steps_per_second": 2.676, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "eval_accuracy": 0.5569620253164557, | |
| "eval_loss": 0.7837380766868591, | |
| "eval_runtime": 1.1326, | |
| "eval_samples_per_second": 69.751, | |
| "eval_steps_per_second": 2.649, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.5271920561790466, | |
| "eval_runtime": 1.1388, | |
| "eval_samples_per_second": 69.374, | |
| "eval_steps_per_second": 2.634, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 5.306596279144287, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5409, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.5162191390991211, | |
| "eval_runtime": 1.123, | |
| "eval_samples_per_second": 70.344, | |
| "eval_steps_per_second": 2.671, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6708860759493671, | |
| "eval_loss": 0.6155683994293213, | |
| "eval_runtime": 1.1442, | |
| "eval_samples_per_second": 69.046, | |
| "eval_steps_per_second": 2.622, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 10.857142857142858, | |
| "eval_accuracy": 0.6582278481012658, | |
| "eval_loss": 0.5843802094459534, | |
| "eval_runtime": 1.1353, | |
| "eval_samples_per_second": 69.586, | |
| "eval_steps_per_second": 2.642, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 9.625292778015137, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.5177, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.6708860759493671, | |
| "eval_loss": 0.7012839913368225, | |
| "eval_runtime": 1.1292, | |
| "eval_samples_per_second": 69.962, | |
| "eval_steps_per_second": 2.657, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 12.857142857142858, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.5016757249832153, | |
| "eval_runtime": 1.1469, | |
| "eval_samples_per_second": 68.881, | |
| "eval_steps_per_second": 2.616, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.4722585082054138, | |
| "eval_runtime": 1.1187, | |
| "eval_samples_per_second": 70.62, | |
| "eval_steps_per_second": 2.682, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 2.9092471599578857, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.4995, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 14.857142857142858, | |
| "eval_accuracy": 0.6835443037974683, | |
| "eval_loss": 0.7056770324707031, | |
| "eval_runtime": 1.1276, | |
| "eval_samples_per_second": 70.062, | |
| "eval_steps_per_second": 2.661, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.46686723828315735, | |
| "eval_runtime": 1.1347, | |
| "eval_samples_per_second": 69.622, | |
| "eval_steps_per_second": 2.644, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 16.857142857142858, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.5737015008926392, | |
| "eval_runtime": 1.1389, | |
| "eval_samples_per_second": 69.365, | |
| "eval_steps_per_second": 2.634, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 9.162846565246582, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.3767, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.5034801959991455, | |
| "eval_runtime": 1.1365, | |
| "eval_samples_per_second": 69.513, | |
| "eval_steps_per_second": 2.64, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 18.857142857142858, | |
| "eval_accuracy": 0.620253164556962, | |
| "eval_loss": 0.899321436882019, | |
| "eval_runtime": 1.1354, | |
| "eval_samples_per_second": 69.582, | |
| "eval_steps_per_second": 2.642, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 3.9183833599090576, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.3602, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.5424658060073853, | |
| "eval_runtime": 1.1335, | |
| "eval_samples_per_second": 69.693, | |
| "eval_steps_per_second": 2.647, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 20.857142857142858, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.560451090335846, | |
| "eval_runtime": 1.105, | |
| "eval_samples_per_second": 71.49, | |
| "eval_steps_per_second": 2.715, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.42595717310905457, | |
| "eval_runtime": 1.1303, | |
| "eval_samples_per_second": 69.894, | |
| "eval_steps_per_second": 2.654, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 4.727555274963379, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.3479, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.4116832911968231, | |
| "eval_runtime": 1.13, | |
| "eval_samples_per_second": 69.912, | |
| "eval_steps_per_second": 2.655, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.4017145037651062, | |
| "eval_runtime": 1.1227, | |
| "eval_samples_per_second": 70.364, | |
| "eval_steps_per_second": 2.672, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 24.857142857142858, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.4401136338710785, | |
| "eval_runtime": 1.152, | |
| "eval_samples_per_second": 68.577, | |
| "eval_steps_per_second": 2.604, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 3.6636221408843994, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.2845, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.44896697998046875, | |
| "eval_runtime": 1.1191, | |
| "eval_samples_per_second": 70.591, | |
| "eval_steps_per_second": 2.681, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 26.857142857142858, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.5079873204231262, | |
| "eval_runtime": 1.1016, | |
| "eval_samples_per_second": 71.715, | |
| "eval_steps_per_second": 2.723, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.5014074444770813, | |
| "eval_runtime": 1.1345, | |
| "eval_samples_per_second": 69.632, | |
| "eval_steps_per_second": 2.644, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 5.449157238006592, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.2316, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 28.857142857142858, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.5384952425956726, | |
| "eval_runtime": 1.1315, | |
| "eval_samples_per_second": 69.82, | |
| "eval_steps_per_second": 2.651, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.5642827153205872, | |
| "eval_runtime": 1.1314, | |
| "eval_samples_per_second": 69.826, | |
| "eval_steps_per_second": 2.652, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 30.857142857142858, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.48871567845344543, | |
| "eval_runtime": 1.1405, | |
| "eval_samples_per_second": 69.271, | |
| "eval_steps_per_second": 2.631, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 4.334333896636963, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.2212, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.5253000855445862, | |
| "eval_runtime": 1.1319, | |
| "eval_samples_per_second": 69.795, | |
| "eval_steps_per_second": 2.65, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 32.857142857142854, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.4632531702518463, | |
| "eval_runtime": 1.1344, | |
| "eval_samples_per_second": 69.638, | |
| "eval_steps_per_second": 2.644, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.5078614354133606, | |
| "eval_runtime": 1.1374, | |
| "eval_samples_per_second": 69.459, | |
| "eval_steps_per_second": 2.638, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 2.3969502449035645, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.2093, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 34.857142857142854, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.508246660232544, | |
| "eval_runtime": 1.135, | |
| "eval_samples_per_second": 69.602, | |
| "eval_steps_per_second": 2.643, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.6112725734710693, | |
| "eval_runtime": 1.1521, | |
| "eval_samples_per_second": 68.572, | |
| "eval_steps_per_second": 2.604, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 36.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.5094159841537476, | |
| "eval_runtime": 1.1118, | |
| "eval_samples_per_second": 71.059, | |
| "eval_steps_per_second": 2.698, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 37.142857142857146, | |
| "grad_norm": 4.463730335235596, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.179, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.6253436207771301, | |
| "eval_runtime": 1.1391, | |
| "eval_samples_per_second": 69.35, | |
| "eval_steps_per_second": 2.634, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 38.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.5406025648117065, | |
| "eval_runtime": 1.1575, | |
| "eval_samples_per_second": 68.253, | |
| "eval_steps_per_second": 2.592, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 1.8232030868530273, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.173, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.6686509847640991, | |
| "eval_runtime": 1.1413, | |
| "eval_samples_per_second": 69.221, | |
| "eval_steps_per_second": 2.629, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 40.857142857142854, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.5690881013870239, | |
| "eval_runtime": 1.1296, | |
| "eval_samples_per_second": 69.934, | |
| "eval_steps_per_second": 2.656, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.5776450037956238, | |
| "eval_runtime": 1.1443, | |
| "eval_samples_per_second": 69.035, | |
| "eval_steps_per_second": 2.622, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "grad_norm": 4.304912090301514, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.1881, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.781569242477417, | |
| "eval_runtime": 1.1351, | |
| "eval_samples_per_second": 69.599, | |
| "eval_steps_per_second": 2.643, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.8354430379746836, | |
| "eval_loss": 0.49078744649887085, | |
| "eval_runtime": 1.133, | |
| "eval_samples_per_second": 69.729, | |
| "eval_steps_per_second": 2.648, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 44.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.5329591631889343, | |
| "eval_runtime": 1.1588, | |
| "eval_samples_per_second": 68.172, | |
| "eval_steps_per_second": 2.589, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 45.714285714285715, | |
| "grad_norm": 1.6199402809143066, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.161, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.5350536704063416, | |
| "eval_runtime": 1.1191, | |
| "eval_samples_per_second": 70.592, | |
| "eval_steps_per_second": 2.681, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 46.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.66633540391922, | |
| "eval_runtime": 1.1167, | |
| "eval_samples_per_second": 70.747, | |
| "eval_steps_per_second": 2.687, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.5382458567619324, | |
| "eval_runtime": 1.1327, | |
| "eval_samples_per_second": 69.744, | |
| "eval_steps_per_second": 2.648, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 48.57142857142857, | |
| "grad_norm": 3.656897783279419, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.1446, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 48.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.5876865983009338, | |
| "eval_runtime": 1.1361, | |
| "eval_samples_per_second": 69.536, | |
| "eval_steps_per_second": 2.641, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.6855766773223877, | |
| "eval_runtime": 1.1183, | |
| "eval_samples_per_second": 70.643, | |
| "eval_steps_per_second": 2.683, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 50.857142857142854, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.6043643951416016, | |
| "eval_runtime": 1.1585, | |
| "eval_samples_per_second": 68.189, | |
| "eval_steps_per_second": 2.589, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 51.42857142857143, | |
| "grad_norm": 3.2771971225738525, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.1556, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7954353094100952, | |
| "eval_runtime": 1.1206, | |
| "eval_samples_per_second": 70.5, | |
| "eval_steps_per_second": 2.677, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 52.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7432191967964172, | |
| "eval_runtime": 1.116, | |
| "eval_samples_per_second": 70.79, | |
| "eval_steps_per_second": 2.688, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.7895691394805908, | |
| "eval_runtime": 1.1711, | |
| "eval_samples_per_second": 67.46, | |
| "eval_steps_per_second": 2.562, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 54.285714285714285, | |
| "grad_norm": 2.9898054599761963, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.1602, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 54.857142857142854, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.8399502038955688, | |
| "eval_runtime": 1.1363, | |
| "eval_samples_per_second": 69.523, | |
| "eval_steps_per_second": 2.64, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.8242945671081543, | |
| "eval_runtime": 1.1359, | |
| "eval_samples_per_second": 69.546, | |
| "eval_steps_per_second": 2.641, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 56.857142857142854, | |
| "eval_accuracy": 0.8354430379746836, | |
| "eval_loss": 0.6864063739776611, | |
| "eval_runtime": 1.149, | |
| "eval_samples_per_second": 68.757, | |
| "eval_steps_per_second": 2.611, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 57.142857142857146, | |
| "grad_norm": 2.650709390640259, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.1357, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.7131363749504089, | |
| "eval_runtime": 1.1228, | |
| "eval_samples_per_second": 70.362, | |
| "eval_steps_per_second": 2.672, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 58.857142857142854, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.9191145300865173, | |
| "eval_runtime": 1.1256, | |
| "eval_samples_per_second": 70.186, | |
| "eval_steps_per_second": 2.665, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 2.9468564987182617, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.1262, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.7465390563011169, | |
| "eval_runtime": 1.1233, | |
| "eval_samples_per_second": 70.329, | |
| "eval_steps_per_second": 2.671, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 60.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7127454280853271, | |
| "eval_runtime": 1.131, | |
| "eval_samples_per_second": 69.847, | |
| "eval_steps_per_second": 2.652, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.6973403692245483, | |
| "eval_runtime": 1.1324, | |
| "eval_samples_per_second": 69.763, | |
| "eval_steps_per_second": 2.649, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 62.857142857142854, | |
| "grad_norm": 2.6824991703033447, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.1323, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 62.857142857142854, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7125291228294373, | |
| "eval_runtime": 1.1478, | |
| "eval_samples_per_second": 68.828, | |
| "eval_steps_per_second": 2.614, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7234599590301514, | |
| "eval_runtime": 1.129, | |
| "eval_samples_per_second": 69.973, | |
| "eval_steps_per_second": 2.657, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 64.85714285714286, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.719983696937561, | |
| "eval_runtime": 1.1239, | |
| "eval_samples_per_second": 70.29, | |
| "eval_steps_per_second": 2.669, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 65.71428571428571, | |
| "grad_norm": 3.247549057006836, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.1258, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7616466879844666, | |
| "eval_runtime": 1.1313, | |
| "eval_samples_per_second": 69.831, | |
| "eval_steps_per_second": 2.652, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 66.85714285714286, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.8537306189537048, | |
| "eval_runtime": 1.1248, | |
| "eval_samples_per_second": 70.237, | |
| "eval_steps_per_second": 2.667, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.8222808837890625, | |
| "eval_runtime": 1.1211, | |
| "eval_samples_per_second": 70.469, | |
| "eval_steps_per_second": 2.676, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 68.57142857142857, | |
| "grad_norm": 1.8724284172058105, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1316, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 68.85714285714286, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7751244902610779, | |
| "eval_runtime": 1.1296, | |
| "eval_samples_per_second": 69.935, | |
| "eval_steps_per_second": 2.656, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7689340710639954, | |
| "eval_runtime": 1.132, | |
| "eval_samples_per_second": 69.785, | |
| "eval_steps_per_second": 2.65, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 70.85714285714286, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.7751415371894836, | |
| "eval_runtime": 1.1224, | |
| "eval_samples_per_second": 70.386, | |
| "eval_steps_per_second": 2.673, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 71.42857142857143, | |
| "grad_norm": 1.3768597841262817, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.1104, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.7770127058029175, | |
| "eval_runtime": 1.1362, | |
| "eval_samples_per_second": 69.53, | |
| "eval_steps_per_second": 2.64, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 72.85714285714286, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.7916089296340942, | |
| "eval_runtime": 1.1361, | |
| "eval_samples_per_second": 69.534, | |
| "eval_steps_per_second": 2.641, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7994998693466187, | |
| "eval_runtime": 1.1345, | |
| "eval_samples_per_second": 69.633, | |
| "eval_steps_per_second": 2.644, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 74.28571428571429, | |
| "grad_norm": 1.7055776119232178, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.1239, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 74.85714285714286, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7980771064758301, | |
| "eval_runtime": 1.1415, | |
| "eval_samples_per_second": 69.209, | |
| "eval_steps_per_second": 2.628, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7889901995658875, | |
| "eval_runtime": 1.1265, | |
| "eval_samples_per_second": 70.129, | |
| "eval_steps_per_second": 2.663, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 76.85714285714286, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7888072729110718, | |
| "eval_runtime": 1.1364, | |
| "eval_samples_per_second": 69.519, | |
| "eval_steps_per_second": 2.64, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 77.14285714285714, | |
| "grad_norm": 1.5552798509597778, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.1057, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7956684231758118, | |
| "eval_runtime": 1.1326, | |
| "eval_samples_per_second": 69.75, | |
| "eval_steps_per_second": 2.649, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 78.85714285714286, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7978534698486328, | |
| "eval_runtime": 1.1385, | |
| "eval_samples_per_second": 69.387, | |
| "eval_steps_per_second": 2.635, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 1.6866562366485596, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.1015, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.7958642244338989, | |
| "eval_runtime": 1.1663, | |
| "eval_samples_per_second": 67.738, | |
| "eval_steps_per_second": 2.572, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 80.85714285714286, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.7934866547584534, | |
| "eval_runtime": 1.1341, | |
| "eval_samples_per_second": 69.659, | |
| "eval_steps_per_second": 2.645, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "eval_accuracy": 0.8227848101265823, | |
| "eval_loss": 0.7973299026489258, | |
| "eval_runtime": 1.1543, | |
| "eval_samples_per_second": 68.44, | |
| "eval_steps_per_second": 2.599, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 82.85714285714286, | |
| "grad_norm": 2.308314085006714, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.101, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 82.85714285714286, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.8032561540603638, | |
| "eval_runtime": 1.1376, | |
| "eval_samples_per_second": 69.442, | |
| "eval_steps_per_second": 2.637, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.8071873784065247, | |
| "eval_runtime": 1.1388, | |
| "eval_samples_per_second": 69.372, | |
| "eval_steps_per_second": 2.634, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 84.85714285714286, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.8049139976501465, | |
| "eval_runtime": 1.1314, | |
| "eval_samples_per_second": 69.822, | |
| "eval_steps_per_second": 2.651, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "grad_norm": 2.8343937397003174, | |
| "learning_rate": 0.0, | |
| "loss": 0.1116, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.8041517734527588, | |
| "eval_runtime": 1.1245, | |
| "eval_samples_per_second": 70.256, | |
| "eval_steps_per_second": 2.668, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "step": 300, | |
| "total_flos": 2.9362240500074496e+18, | |
| "train_loss": 0.24529865741729737, | |
| "train_runtime": 1655.9157, | |
| "train_samples_per_second": 26.692, | |
| "train_steps_per_second": 0.181 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "eval_accuracy": 0.8354430379746836, | |
| "eval_loss": 0.49078744649887085, | |
| "eval_runtime": 1.1234, | |
| "eval_samples_per_second": 70.325, | |
| "eval_steps_per_second": 2.671, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "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": 2.9362240500074496e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |