| { | |
| "best_metric": 0.8333333333333334, | |
| "best_model_checkpoint": "beit-base-patch16-224-dmae-va-U5-42E/checkpoint-38", | |
| "epoch": 37.935483870967744, | |
| "eval_steps": 500, | |
| "global_step": 294, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9032258064516129, | |
| "eval_accuracy": 0.55, | |
| "eval_loss": 1.2065638303756714, | |
| "eval_runtime": 1.625, | |
| "eval_samples_per_second": 36.924, | |
| "eval_steps_per_second": 1.231, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.2903225806451613, | |
| "grad_norm": 10.190420150756836, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.5038, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.935483870967742, | |
| "eval_accuracy": 0.4166666666666667, | |
| "eval_loss": 1.2166622877120972, | |
| "eval_runtime": 1.5799, | |
| "eval_samples_per_second": 37.978, | |
| "eval_steps_per_second": 1.266, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.5806451612903225, | |
| "grad_norm": 8.352402687072754, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 1.1504, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.967741935483871, | |
| "eval_accuracy": 0.65, | |
| "eval_loss": 0.8820936679840088, | |
| "eval_runtime": 1.6273, | |
| "eval_samples_per_second": 36.871, | |
| "eval_steps_per_second": 1.229, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 3.870967741935484, | |
| "grad_norm": 8.781105041503906, | |
| "learning_rate": 4e-05, | |
| "loss": 0.8299, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.6488708853721619, | |
| "eval_runtime": 1.6663, | |
| "eval_samples_per_second": 36.008, | |
| "eval_steps_per_second": 1.2, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 4.903225806451613, | |
| "eval_accuracy": 0.8333333333333334, | |
| "eval_loss": 0.5796787142753601, | |
| "eval_runtime": 1.6444, | |
| "eval_samples_per_second": 36.488, | |
| "eval_steps_per_second": 1.216, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 5.161290322580645, | |
| "grad_norm": 17.170116424560547, | |
| "learning_rate": 3.848484848484849e-05, | |
| "loss": 0.5887, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.935483870967742, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.6618198752403259, | |
| "eval_runtime": 1.6267, | |
| "eval_samples_per_second": 36.884, | |
| "eval_steps_per_second": 1.229, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 6.451612903225806, | |
| "grad_norm": 8.946195602416992, | |
| "learning_rate": 3.6969696969696974e-05, | |
| "loss": 0.3965, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.967741935483871, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.6530690789222717, | |
| "eval_runtime": 1.6436, | |
| "eval_samples_per_second": 36.505, | |
| "eval_steps_per_second": 1.217, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 7.741935483870968, | |
| "grad_norm": 10.897398948669434, | |
| "learning_rate": 3.545454545454546e-05, | |
| "loss": 0.3089, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7, | |
| "eval_loss": 0.7609087824821472, | |
| "eval_runtime": 1.6283, | |
| "eval_samples_per_second": 36.849, | |
| "eval_steps_per_second": 1.228, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 8.903225806451612, | |
| "eval_accuracy": 0.6833333333333333, | |
| "eval_loss": 0.8608618378639221, | |
| "eval_runtime": 1.8859, | |
| "eval_samples_per_second": 31.815, | |
| "eval_steps_per_second": 1.06, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 9.03225806451613, | |
| "grad_norm": 8.992852210998535, | |
| "learning_rate": 3.3939393939393945e-05, | |
| "loss": 0.2393, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 9.935483870967742, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.6909741759300232, | |
| "eval_runtime": 1.6359, | |
| "eval_samples_per_second": 36.678, | |
| "eval_steps_per_second": 1.223, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 10.32258064516129, | |
| "grad_norm": 8.585020065307617, | |
| "learning_rate": 3.2424242424242423e-05, | |
| "loss": 0.1928, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.967741935483872, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.7773979902267456, | |
| "eval_runtime": 1.636, | |
| "eval_samples_per_second": 36.675, | |
| "eval_steps_per_second": 1.222, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 11.612903225806452, | |
| "grad_norm": 8.469679832458496, | |
| "learning_rate": 3.090909090909091e-05, | |
| "loss": 0.1993, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.8423529863357544, | |
| "eval_runtime": 1.6346, | |
| "eval_samples_per_second": 36.707, | |
| "eval_steps_per_second": 1.224, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 12.903225806451612, | |
| "grad_norm": 5.69382381439209, | |
| "learning_rate": 2.9393939393939398e-05, | |
| "loss": 0.165, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.903225806451612, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.7478328347206116, | |
| "eval_runtime": 1.6366, | |
| "eval_samples_per_second": 36.661, | |
| "eval_steps_per_second": 1.222, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 13.935483870967742, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.757307231426239, | |
| "eval_runtime": 1.6144, | |
| "eval_samples_per_second": 37.166, | |
| "eval_steps_per_second": 1.239, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 14.193548387096774, | |
| "grad_norm": 5.3613386154174805, | |
| "learning_rate": 2.7878787878787883e-05, | |
| "loss": 0.1117, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 14.967741935483872, | |
| "eval_accuracy": 0.8166666666666667, | |
| "eval_loss": 0.8059377074241638, | |
| "eval_runtime": 1.6456, | |
| "eval_samples_per_second": 36.46, | |
| "eval_steps_per_second": 1.215, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 15.483870967741936, | |
| "grad_norm": 6.849861145019531, | |
| "learning_rate": 2.6363636363636365e-05, | |
| "loss": 0.1171, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.8982041478157043, | |
| "eval_runtime": 1.6378, | |
| "eval_samples_per_second": 36.635, | |
| "eval_steps_per_second": 1.221, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 16.774193548387096, | |
| "grad_norm": 4.013183116912842, | |
| "learning_rate": 2.484848484848485e-05, | |
| "loss": 0.0961, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 16.903225806451612, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.9133464097976685, | |
| "eval_runtime": 1.8779, | |
| "eval_samples_per_second": 31.951, | |
| "eval_steps_per_second": 1.065, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 17.93548387096774, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.912105917930603, | |
| "eval_runtime": 1.6239, | |
| "eval_samples_per_second": 36.949, | |
| "eval_steps_per_second": 1.232, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 18.06451612903226, | |
| "grad_norm": 6.759372711181641, | |
| "learning_rate": 2.3333333333333336e-05, | |
| "loss": 0.1359, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 18.967741935483872, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.9297049045562744, | |
| "eval_runtime": 1.7996, | |
| "eval_samples_per_second": 33.341, | |
| "eval_steps_per_second": 1.111, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 19.35483870967742, | |
| "grad_norm": 6.179893970489502, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 0.0981, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.012387990951538, | |
| "eval_runtime": 2.1389, | |
| "eval_samples_per_second": 28.052, | |
| "eval_steps_per_second": 0.935, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 20.64516129032258, | |
| "grad_norm": 5.074409484863281, | |
| "learning_rate": 2.0303030303030303e-05, | |
| "loss": 0.0817, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 20.903225806451612, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.9627708196640015, | |
| "eval_runtime": 1.827, | |
| "eval_samples_per_second": 32.841, | |
| "eval_steps_per_second": 1.095, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 21.93548387096774, | |
| "grad_norm": 5.621190547943115, | |
| "learning_rate": 1.8787878787878792e-05, | |
| "loss": 0.0976, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 21.93548387096774, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.9663652777671814, | |
| "eval_runtime": 1.6503, | |
| "eval_samples_per_second": 36.358, | |
| "eval_steps_per_second": 1.212, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 22.967741935483872, | |
| "eval_accuracy": 0.8166666666666667, | |
| "eval_loss": 0.7979750037193298, | |
| "eval_runtime": 1.6343, | |
| "eval_samples_per_second": 36.714, | |
| "eval_steps_per_second": 1.224, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 23.225806451612904, | |
| "grad_norm": 4.8617706298828125, | |
| "learning_rate": 1.7272727272727274e-05, | |
| "loss": 0.0899, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.8365867733955383, | |
| "eval_runtime": 2.8201, | |
| "eval_samples_per_second": 21.276, | |
| "eval_steps_per_second": 0.709, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 24.516129032258064, | |
| "grad_norm": 5.400606155395508, | |
| "learning_rate": 1.575757575757576e-05, | |
| "loss": 0.1052, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 24.903225806451612, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.9160324335098267, | |
| "eval_runtime": 1.643, | |
| "eval_samples_per_second": 36.519, | |
| "eval_steps_per_second": 1.217, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 25.806451612903224, | |
| "grad_norm": 3.636944055557251, | |
| "learning_rate": 1.4242424242424245e-05, | |
| "loss": 0.0817, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 25.93548387096774, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.970082700252533, | |
| "eval_runtime": 2.149, | |
| "eval_samples_per_second": 27.92, | |
| "eval_steps_per_second": 0.931, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 26.967741935483872, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.9995465874671936, | |
| "eval_runtime": 1.6532, | |
| "eval_samples_per_second": 36.292, | |
| "eval_steps_per_second": 1.21, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 27.096774193548388, | |
| "grad_norm": 10.176094055175781, | |
| "learning_rate": 1.2727272727272728e-05, | |
| "loss": 0.0886, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.8483232855796814, | |
| "eval_runtime": 1.6529, | |
| "eval_samples_per_second": 36.3, | |
| "eval_steps_per_second": 1.21, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 28.387096774193548, | |
| "grad_norm": 4.629071235656738, | |
| "learning_rate": 1.1212121212121212e-05, | |
| "loss": 0.0766, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 28.903225806451612, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.89542156457901, | |
| "eval_runtime": 1.758, | |
| "eval_samples_per_second": 34.129, | |
| "eval_steps_per_second": 1.138, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 29.677419354838708, | |
| "grad_norm": 9.337085723876953, | |
| "learning_rate": 9.696969696969698e-06, | |
| "loss": 0.0923, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 29.93548387096774, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.9605885744094849, | |
| "eval_runtime": 1.6458, | |
| "eval_samples_per_second": 36.456, | |
| "eval_steps_per_second": 1.215, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 30.967741935483872, | |
| "grad_norm": 4.47194242477417, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 0.0579, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 30.967741935483872, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.9958234429359436, | |
| "eval_runtime": 1.8682, | |
| "eval_samples_per_second": 32.116, | |
| "eval_steps_per_second": 1.071, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.9665268063545227, | |
| "eval_runtime": 1.6466, | |
| "eval_samples_per_second": 36.439, | |
| "eval_steps_per_second": 1.215, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 32.25806451612903, | |
| "grad_norm": 3.8999428749084473, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.0707, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 32.903225806451616, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 1.0259429216384888, | |
| "eval_runtime": 1.7687, | |
| "eval_samples_per_second": 33.923, | |
| "eval_steps_per_second": 1.131, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 33.54838709677419, | |
| "grad_norm": 3.87837815284729, | |
| "learning_rate": 5.151515151515152e-06, | |
| "loss": 0.0756, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 33.935483870967744, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.0626757144927979, | |
| "eval_runtime": 2.1538, | |
| "eval_samples_per_second": 27.858, | |
| "eval_steps_per_second": 0.929, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 34.83870967741935, | |
| "grad_norm": 7.991878986358643, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 0.0528, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 34.96774193548387, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 1.050768256187439, | |
| "eval_runtime": 1.6273, | |
| "eval_samples_per_second": 36.87, | |
| "eval_steps_per_second": 1.229, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 1.099774718284607, | |
| "eval_runtime": 1.6638, | |
| "eval_samples_per_second": 36.062, | |
| "eval_steps_per_second": 1.202, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 36.12903225806452, | |
| "grad_norm": 5.349879741668701, | |
| "learning_rate": 2.1212121212121216e-06, | |
| "loss": 0.0706, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 36.903225806451616, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.0693951845169067, | |
| "eval_runtime": 1.675, | |
| "eval_samples_per_second": 35.82, | |
| "eval_steps_per_second": 1.194, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 37.41935483870968, | |
| "grad_norm": 6.116847038269043, | |
| "learning_rate": 6.060606060606061e-07, | |
| "loss": 0.0658, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 37.935483870967744, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 1.0560951232910156, | |
| "eval_runtime": 1.6581, | |
| "eval_samples_per_second": 36.187, | |
| "eval_steps_per_second": 1.206, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 37.935483870967744, | |
| "step": 294, | |
| "total_flos": 2.8633958865108173e+18, | |
| "train_loss": 0.24709520539661653, | |
| "train_runtime": 1707.2205, | |
| "train_samples_per_second": 23.962, | |
| "train_steps_per_second": 0.172 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 294, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 42, | |
| "save_steps": 500, | |
| "total_flos": 2.8633958865108173e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |