| { | |
| "best_metric": 0.986404299736023, | |
| "best_model_checkpoint": "./outputs/dinov2-small-types-of-film-shots-vN/checkpoint-221", | |
| "epoch": 11.695431472081218, | |
| "eval_steps": 500, | |
| "global_step": 288, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 2.3969, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 2.0526, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8415, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.9633699633699634e-05, | |
| "loss": 1.6177, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_accuracy": 0.41007194244604317, | |
| "eval_loss": 1.5501453876495361, | |
| "eval_runtime": 0.6217, | |
| "eval_samples_per_second": 223.574, | |
| "eval_steps_per_second": 56.296, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.926739926739927e-05, | |
| "loss": 1.5922, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.8901098901098903e-05, | |
| "loss": 1.3521, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 1.8534798534798536e-05, | |
| "loss": 1.4178, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 1.816849816849817e-05, | |
| "loss": 1.2448, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.78021978021978e-05, | |
| "loss": 1.3029, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_accuracy": 0.5107913669064749, | |
| "eval_loss": 1.2448219060897827, | |
| "eval_runtime": 0.633, | |
| "eval_samples_per_second": 219.573, | |
| "eval_steps_per_second": 55.288, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 1.1814, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 1.706959706959707e-05, | |
| "loss": 1.0996, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.6703296703296707e-05, | |
| "loss": 1.098, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 1.633699633699634e-05, | |
| "loss": 1.041, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 1.5970695970695972e-05, | |
| "loss": 1.1785, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_accuracy": 0.5251798561151079, | |
| "eval_loss": 1.0555555820465088, | |
| "eval_runtime": 0.6319, | |
| "eval_samples_per_second": 219.987, | |
| "eval_steps_per_second": 55.392, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 1.5604395604395605e-05, | |
| "loss": 1.116, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 1.523809523809524e-05, | |
| "loss": 1.0266, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "learning_rate": 1.4871794871794874e-05, | |
| "loss": 0.9728, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 1.4505494505494506e-05, | |
| "loss": 0.9575, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 1.4139194139194141e-05, | |
| "loss": 1.2146, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "eval_accuracy": 0.539568345323741, | |
| "eval_loss": 1.2316133975982666, | |
| "eval_runtime": 0.6178, | |
| "eval_samples_per_second": 225.001, | |
| "eval_steps_per_second": 56.655, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "learning_rate": 1.3772893772893774e-05, | |
| "loss": 0.8369, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "learning_rate": 1.3406593406593406e-05, | |
| "loss": 0.9136, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 1.3040293040293041e-05, | |
| "loss": 0.9703, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 1.2673992673992674e-05, | |
| "loss": 0.9575, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8389, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "eval_accuracy": 0.5971223021582733, | |
| "eval_loss": 1.0235121250152588, | |
| "eval_runtime": 0.6239, | |
| "eval_samples_per_second": 222.778, | |
| "eval_steps_per_second": 56.095, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "learning_rate": 1.1941391941391942e-05, | |
| "loss": 0.8346, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "learning_rate": 1.1575091575091577e-05, | |
| "loss": 0.8608, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "learning_rate": 1.120879120879121e-05, | |
| "loss": 0.855, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "learning_rate": 1.0842490842490842e-05, | |
| "loss": 0.8467, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "learning_rate": 1.0476190476190477e-05, | |
| "loss": 0.7883, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "eval_accuracy": 0.6258992805755396, | |
| "eval_loss": 0.9959595799446106, | |
| "eval_runtime": 0.6274, | |
| "eval_samples_per_second": 221.561, | |
| "eval_steps_per_second": 55.789, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 6.09, | |
| "learning_rate": 1.010989010989011e-05, | |
| "loss": 0.7238, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 9.743589743589744e-06, | |
| "loss": 0.688, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 9.377289377289379e-06, | |
| "loss": 0.6678, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "learning_rate": 9.010989010989011e-06, | |
| "loss": 0.7223, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "learning_rate": 8.644688644688646e-06, | |
| "loss": 0.7899, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.98, | |
| "eval_accuracy": 0.5539568345323741, | |
| "eval_loss": 1.1354069709777832, | |
| "eval_runtime": 0.6338, | |
| "eval_samples_per_second": 219.304, | |
| "eval_steps_per_second": 55.22, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "learning_rate": 8.278388278388278e-06, | |
| "loss": 0.7066, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "learning_rate": 7.912087912087913e-06, | |
| "loss": 0.6458, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 7.5457875457875465e-06, | |
| "loss": 0.6245, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.6187, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "learning_rate": 6.813186813186814e-06, | |
| "loss": 0.663, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5827338129496403, | |
| "eval_loss": 1.097096562385559, | |
| "eval_runtime": 0.6204, | |
| "eval_samples_per_second": 224.06, | |
| "eval_steps_per_second": 56.418, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "learning_rate": 6.446886446886448e-06, | |
| "loss": 0.6473, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 8.32, | |
| "learning_rate": 6.080586080586081e-06, | |
| "loss": 0.5863, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.5137, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.73, | |
| "learning_rate": 5.347985347985348e-06, | |
| "loss": 0.5704, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "learning_rate": 4.9816849816849826e-06, | |
| "loss": 0.6013, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_accuracy": 0.6258992805755396, | |
| "eval_loss": 0.986404299736023, | |
| "eval_runtime": 0.6376, | |
| "eval_samples_per_second": 217.99, | |
| "eval_steps_per_second": 54.89, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.4918, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 9.34, | |
| "learning_rate": 4.249084249084249e-06, | |
| "loss": 0.5953, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 9.54, | |
| "learning_rate": 3.882783882783883e-06, | |
| "loss": 0.5561, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "learning_rate": 3.516483516483517e-06, | |
| "loss": 0.4991, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "learning_rate": 3.1501831501831505e-06, | |
| "loss": 0.6276, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 9.99, | |
| "eval_accuracy": 0.6115107913669064, | |
| "eval_loss": 1.018210768699646, | |
| "eval_runtime": 0.6196, | |
| "eval_samples_per_second": 224.345, | |
| "eval_steps_per_second": 56.49, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 10.15, | |
| "learning_rate": 2.783882783882784e-06, | |
| "loss": 0.393, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "learning_rate": 2.4175824175824177e-06, | |
| "loss": 0.5476, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 10.56, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.457, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.76, | |
| "learning_rate": 1.6849816849816852e-06, | |
| "loss": 0.4764, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 10.96, | |
| "learning_rate": 1.3186813186813187e-06, | |
| "loss": 0.5196, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 10.96, | |
| "eval_accuracy": 0.6546762589928058, | |
| "eval_loss": 1.0074387788772583, | |
| "eval_runtime": 0.6224, | |
| "eval_samples_per_second": 223.337, | |
| "eval_steps_per_second": 56.236, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 11.17, | |
| "learning_rate": 9.523809523809525e-07, | |
| "loss": 0.4737, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 11.37, | |
| "learning_rate": 5.860805860805862e-07, | |
| "loss": 0.547, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 11.57, | |
| "learning_rate": 2.197802197802198e-07, | |
| "loss": 0.4761, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "eval_accuracy": 0.6762589928057554, | |
| "eval_loss": 0.9956228137016296, | |
| "eval_runtime": 0.6444, | |
| "eval_samples_per_second": 215.696, | |
| "eval_steps_per_second": 54.312, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "step": 288, | |
| "total_flos": 2.3928536971188634e+17, | |
| "train_loss": 0.8871624300049411, | |
| "train_runtime": 139.3419, | |
| "train_samples_per_second": 67.69, | |
| "train_steps_per_second": 2.067 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 288, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 500, | |
| "total_flos": 2.3928536971188634e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |