| { | |
| "best_metric": 0.4994313716888428, | |
| "best_model_checkpoint": "./vit-weight-decay-1e-2/checkpoint-5457", | |
| "epoch": 27.0, | |
| "eval_steps": 500, | |
| "global_step": 8667, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 27.05959129333496, | |
| "learning_rate": 2.6004922067268256e-05, | |
| "loss": 1.7124, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6924410540915396, | |
| "eval_f1": 0.603022144115533, | |
| "eval_loss": 0.8697461485862732, | |
| "eval_precision": 0.6656411815509768, | |
| "eval_recall": 0.6924410540915396, | |
| "eval_runtime": 24.2005, | |
| "eval_samples_per_second": 119.171, | |
| "eval_steps_per_second": 14.917, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 31.123794555664062, | |
| "learning_rate": 5.233798195242002e-05, | |
| "loss": 1.1476, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6990291262135923, | |
| "eval_f1": 0.714853056600681, | |
| "eval_loss": 0.7271208763122559, | |
| "eval_precision": 0.7684437503161472, | |
| "eval_recall": 0.6990291262135923, | |
| "eval_runtime": 23.1919, | |
| "eval_samples_per_second": 124.354, | |
| "eval_steps_per_second": 15.566, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.652539253234863, | |
| "learning_rate": 7.867104183757178e-05, | |
| "loss": 1.0734, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7687239944521498, | |
| "eval_f1": 0.7417221561170182, | |
| "eval_loss": 0.6440889239311218, | |
| "eval_precision": 0.7568176957187778, | |
| "eval_recall": 0.7687239944521498, | |
| "eval_runtime": 23.9098, | |
| "eval_samples_per_second": 120.62, | |
| "eval_steps_per_second": 15.098, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 26.467485427856445, | |
| "learning_rate": 9.99845966779335e-05, | |
| "loss": 1.0271, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7773925104022191, | |
| "eval_f1": 0.7814315937772335, | |
| "eval_loss": 0.5854852199554443, | |
| "eval_precision": 0.78834249113171, | |
| "eval_recall": 0.7773925104022191, | |
| "eval_runtime": 23.2363, | |
| "eval_samples_per_second": 124.116, | |
| "eval_steps_per_second": 15.536, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 9.265125274658203, | |
| "learning_rate": 9.93971225198763e-05, | |
| "loss": 0.9158, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7635228848821082, | |
| "eval_f1": 0.7661961115296833, | |
| "eval_loss": 0.700226366519928, | |
| "eval_precision": 0.7929662231425049, | |
| "eval_recall": 0.7635228848821082, | |
| "eval_runtime": 23.9812, | |
| "eval_samples_per_second": 120.261, | |
| "eval_steps_per_second": 15.053, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.993426322937012, | |
| "learning_rate": 9.796799913911281e-05, | |
| "loss": 0.9167, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7812066574202496, | |
| "eval_f1": 0.7900429974573581, | |
| "eval_loss": 0.5867117047309875, | |
| "eval_precision": 0.806466933291998, | |
| "eval_recall": 0.7812066574202496, | |
| "eval_runtime": 23.5885, | |
| "eval_samples_per_second": 122.263, | |
| "eval_steps_per_second": 15.304, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 8.043280601501465, | |
| "learning_rate": 9.572157654878572e-05, | |
| "loss": 0.786, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7340499306518724, | |
| "eval_f1": 0.751486114744647, | |
| "eval_loss": 0.6516677737236023, | |
| "eval_precision": 0.8047120902571854, | |
| "eval_recall": 0.7340499306518724, | |
| "eval_runtime": 24.7074, | |
| "eval_samples_per_second": 116.726, | |
| "eval_steps_per_second": 14.611, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.0447545051574707, | |
| "learning_rate": 9.26961302542397e-05, | |
| "loss": 0.7406, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7066574202496533, | |
| "eval_f1": 0.7330088250984578, | |
| "eval_loss": 0.6647158265113831, | |
| "eval_precision": 0.8133714017605812, | |
| "eval_recall": 0.7066574202496533, | |
| "eval_runtime": 24.3728, | |
| "eval_samples_per_second": 118.329, | |
| "eval_steps_per_second": 14.812, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 3.4238717555999756, | |
| "learning_rate": 8.89432090986511e-05, | |
| "loss": 0.682, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8228155339805825, | |
| "eval_f1": 0.8207044848679155, | |
| "eval_loss": 0.510608434677124, | |
| "eval_precision": 0.8230905088203688, | |
| "eval_recall": 0.8228155339805825, | |
| "eval_runtime": 24.4807, | |
| "eval_samples_per_second": 117.807, | |
| "eval_steps_per_second": 14.746, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 11.989164352416992, | |
| "learning_rate": 8.45267569518721e-05, | |
| "loss": 0.6427, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8165742024965326, | |
| "eval_f1": 0.8221884365803362, | |
| "eval_loss": 0.5032415390014648, | |
| "eval_precision": 0.835376334433007, | |
| "eval_recall": 0.8165742024965326, | |
| "eval_runtime": 23.2354, | |
| "eval_samples_per_second": 124.121, | |
| "eval_steps_per_second": 15.537, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 34.21263885498047, | |
| "learning_rate": 7.952202320752798e-05, | |
| "loss": 0.5663, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8151872399445215, | |
| "eval_f1": 0.8216396746544562, | |
| "eval_loss": 0.5357819199562073, | |
| "eval_precision": 0.8325502393616235, | |
| "eval_recall": 0.8151872399445215, | |
| "eval_runtime": 24.4554, | |
| "eval_samples_per_second": 117.929, | |
| "eval_steps_per_second": 14.762, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 44.03895568847656, | |
| "learning_rate": 7.401428065178325e-05, | |
| "loss": 0.5395, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8248959778085991, | |
| "eval_f1": 0.8298561156988427, | |
| "eval_loss": 0.5487632155418396, | |
| "eval_precision": 0.8391637957530821, | |
| "eval_recall": 0.8248959778085991, | |
| "eval_runtime": 23.3681, | |
| "eval_samples_per_second": 123.416, | |
| "eval_steps_per_second": 15.448, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 27.093250274658203, | |
| "learning_rate": 6.80973725492743e-05, | |
| "loss": 0.4468, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8231622746185853, | |
| "eval_f1": 0.8260339295474823, | |
| "eval_loss": 0.578988254070282, | |
| "eval_precision": 0.8397229815779756, | |
| "eval_recall": 0.8231622746185853, | |
| "eval_runtime": 24.8132, | |
| "eval_samples_per_second": 116.229, | |
| "eval_steps_per_second": 14.549, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 5.494964122772217, | |
| "learning_rate": 6.187211370157784e-05, | |
| "loss": 0.4247, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8415395284327323, | |
| "eval_f1": 0.8448790763195109, | |
| "eval_loss": 0.5437958240509033, | |
| "eval_precision": 0.8570099067237934, | |
| "eval_recall": 0.8415395284327323, | |
| "eval_runtime": 23.4933, | |
| "eval_samples_per_second": 122.758, | |
| "eval_steps_per_second": 15.366, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 5.022720813751221, | |
| "learning_rate": 5.544457272166217e-05, | |
| "loss": 0.3495, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8453536754507628, | |
| "eval_f1": 0.846725089255697, | |
| "eval_loss": 0.5135474801063538, | |
| "eval_precision": 0.8518606648557052, | |
| "eval_recall": 0.8453536754507628, | |
| "eval_runtime": 22.9349, | |
| "eval_samples_per_second": 125.747, | |
| "eval_steps_per_second": 15.74, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 21.08028221130371, | |
| "learning_rate": 4.894460661440583e-05, | |
| "loss": 0.3039, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8408460471567267, | |
| "eval_f1": 0.8447631505343497, | |
| "eval_loss": 0.5631198287010193, | |
| "eval_precision": 0.8520329480153485, | |
| "eval_recall": 0.8408460471567267, | |
| "eval_runtime": 22.9052, | |
| "eval_samples_per_second": 125.91, | |
| "eval_steps_per_second": 15.761, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.3130456805229187, | |
| "learning_rate": 4.244239774409037e-05, | |
| "loss": 0.2602, | |
| "step": 5457 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8602635228848821, | |
| "eval_f1": 0.8599959330449201, | |
| "eval_loss": 0.4994313716888428, | |
| "eval_precision": 0.8617536928422816, | |
| "eval_recall": 0.8602635228848821, | |
| "eval_runtime": 24.0474, | |
| "eval_samples_per_second": 119.93, | |
| "eval_steps_per_second": 15.012, | |
| "step": 5457 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 13.50483226776123, | |
| "learning_rate": 3.606895852147351e-05, | |
| "loss": 0.2616, | |
| "step": 5778 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8564493758668515, | |
| "eval_f1": 0.8585397395861224, | |
| "eval_loss": 0.5405685901641846, | |
| "eval_precision": 0.8621724878350435, | |
| "eval_recall": 0.8564493758668515, | |
| "eval_runtime": 23.9163, | |
| "eval_samples_per_second": 120.587, | |
| "eval_steps_per_second": 15.094, | |
| "step": 5778 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.11348175257444382, | |
| "learning_rate": 2.9932882319894417e-05, | |
| "loss": 0.1876, | |
| "step": 6099 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.848127600554785, | |
| "eval_f1": 0.8525273952669322, | |
| "eval_loss": 0.5612274408340454, | |
| "eval_precision": 0.8629290477513907, | |
| "eval_recall": 0.848127600554785, | |
| "eval_runtime": 24.2059, | |
| "eval_samples_per_second": 119.145, | |
| "eval_steps_per_second": 14.914, | |
| "step": 6099 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.7572256326675415, | |
| "learning_rate": 2.4138718220394167e-05, | |
| "loss": 0.2052, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8429264909847434, | |
| "eval_f1": 0.8427565344137952, | |
| "eval_loss": 0.6802518367767334, | |
| "eval_precision": 0.8502279276035353, | |
| "eval_recall": 0.8429264909847434, | |
| "eval_runtime": 23.6557, | |
| "eval_samples_per_second": 121.916, | |
| "eval_steps_per_second": 15.261, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 11.348849296569824, | |
| "learning_rate": 1.8785189659922232e-05, | |
| "loss": 0.1533, | |
| "step": 6741 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.8734396671289875, | |
| "eval_f1": 0.870883732217841, | |
| "eval_loss": 0.546351432800293, | |
| "eval_precision": 0.8698420453848273, | |
| "eval_recall": 0.8734396671289875, | |
| "eval_runtime": 23.8505, | |
| "eval_samples_per_second": 120.92, | |
| "eval_steps_per_second": 15.136, | |
| "step": 6741 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 54.87451171875, | |
| "learning_rate": 1.396351233934956e-05, | |
| "loss": 0.1175, | |
| "step": 7062 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.8685852981969486, | |
| "eval_f1": 0.8672514399266257, | |
| "eval_loss": 0.5572792291641235, | |
| "eval_precision": 0.8667361486336195, | |
| "eval_recall": 0.8685852981969486, | |
| "eval_runtime": 23.4628, | |
| "eval_samples_per_second": 122.918, | |
| "eval_steps_per_second": 15.386, | |
| "step": 7062 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.37277576327323914, | |
| "learning_rate": 9.755840051487997e-06, | |
| "loss": 0.1218, | |
| "step": 7383 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.8703190013869625, | |
| "eval_f1": 0.8669113876377326, | |
| "eval_loss": 0.6043308973312378, | |
| "eval_precision": 0.8680623901339135, | |
| "eval_recall": 0.8703190013869625, | |
| "eval_runtime": 23.6979, | |
| "eval_samples_per_second": 121.698, | |
| "eval_steps_per_second": 15.233, | |
| "step": 7383 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 1.1696586608886719, | |
| "learning_rate": 6.233864909760889e-06, | |
| "loss": 0.114, | |
| "step": 7704 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.871012482662968, | |
| "eval_f1": 0.8692580136075093, | |
| "eval_loss": 0.5944604277610779, | |
| "eval_precision": 0.8705845000186183, | |
| "eval_recall": 0.871012482662968, | |
| "eval_runtime": 22.9273, | |
| "eval_samples_per_second": 125.789, | |
| "eval_steps_per_second": 15.745, | |
| "step": 7704 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.24613961577415466, | |
| "learning_rate": 3.457595827424931e-06, | |
| "loss": 0.104, | |
| "step": 8025 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.8765603328710125, | |
| "eval_f1": 0.8751858009450587, | |
| "eval_loss": 0.5850355625152588, | |
| "eval_precision": 0.8753430843268125, | |
| "eval_recall": 0.8765603328710125, | |
| "eval_runtime": 23.185, | |
| "eval_samples_per_second": 124.391, | |
| "eval_steps_per_second": 15.57, | |
| "step": 8025 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 15.901623725891113, | |
| "learning_rate": 1.4743360601349622e-06, | |
| "loss": 0.0752, | |
| "step": 8346 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.8782940360610264, | |
| "eval_f1": 0.8757212451946566, | |
| "eval_loss": 0.5867504477500916, | |
| "eval_precision": 0.8747328684004348, | |
| "eval_recall": 0.8782940360610264, | |
| "eval_runtime": 24.8645, | |
| "eval_samples_per_second": 115.989, | |
| "eval_steps_per_second": 14.519, | |
| "step": 8346 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.1058092936873436, | |
| "learning_rate": 3.1787723291717977e-07, | |
| "loss": 0.1309, | |
| "step": 8667 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.8786407766990292, | |
| "eval_f1": 0.876119808222944, | |
| "eval_loss": 0.5839141011238098, | |
| "eval_precision": 0.875311708483151, | |
| "eval_recall": 0.8786407766990292, | |
| "eval_runtime": 24.8375, | |
| "eval_samples_per_second": 116.115, | |
| "eval_steps_per_second": 14.534, | |
| "step": 8667 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "step": 8667, | |
| "total_flos": 1.0729711727592948e+19, | |
| "train_loss": 0.5187536703752126, | |
| "train_runtime": 3224.0652, | |
| "train_samples_per_second": 159.054, | |
| "train_steps_per_second": 9.956 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 1.0729711727592948e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |