| { | |
| "best_metric": 0.9329698524695318, | |
| "best_model_checkpoint": "results/facebook/hubert-base-ls960/42/_retain/checkpoint-2800", | |
| "epoch": 4.017216642754663, | |
| "eval_steps": 400, | |
| "global_step": 2800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14347202295552366, | |
| "grad_norm": 7.395946025848389, | |
| "learning_rate": 0.00017857142857142857, | |
| "loss": 3.276, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28694404591104733, | |
| "grad_norm": 64.37727355957031, | |
| "learning_rate": 0.00035714285714285714, | |
| "loss": 2.5895, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.430416068866571, | |
| "grad_norm": 37.723262786865234, | |
| "learning_rate": 0.000496031746031746, | |
| "loss": 2.0702, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5738880918220947, | |
| "grad_norm": 34.24871826171875, | |
| "learning_rate": 0.0004761904761904762, | |
| "loss": 1.9753, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5738880918220947, | |
| "eval_accuracy": 0.39191789608723543, | |
| "eval_f1_macro": 0.2435662437168276, | |
| "eval_loss": 1.6742650270462036, | |
| "eval_runtime": 12.5972, | |
| "eval_samples_per_second": 247.516, | |
| "eval_steps_per_second": 30.959, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7173601147776184, | |
| "grad_norm": 54.40435791015625, | |
| "learning_rate": 0.0004563492063492063, | |
| "loss": 1.8347, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.860832137733142, | |
| "grad_norm": 95.11190795898438, | |
| "learning_rate": 0.0004365079365079365, | |
| "loss": 1.6797, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0043041606886658, | |
| "grad_norm": 57.426300048828125, | |
| "learning_rate": 0.0004166666666666667, | |
| "loss": 1.5546, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1477761836441893, | |
| "grad_norm": 26.83301544189453, | |
| "learning_rate": 0.0003968253968253968, | |
| "loss": 1.4852, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1477761836441893, | |
| "eval_accuracy": 0.43585631815266196, | |
| "eval_f1_macro": 0.35693788277668864, | |
| "eval_loss": 1.400166392326355, | |
| "eval_runtime": 12.5264, | |
| "eval_samples_per_second": 248.914, | |
| "eval_steps_per_second": 31.134, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.291248206599713, | |
| "grad_norm": 110.81356048583984, | |
| "learning_rate": 0.000376984126984127, | |
| "loss": 1.3436, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4347202295552366, | |
| "grad_norm": 32.543121337890625, | |
| "learning_rate": 0.00035714285714285714, | |
| "loss": 1.2294, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5781922525107603, | |
| "grad_norm": 56.1235237121582, | |
| "learning_rate": 0.00033730158730158733, | |
| "loss": 1.2193, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.721664275466284, | |
| "grad_norm": 45.8999137878418, | |
| "learning_rate": 0.00031746031746031746, | |
| "loss": 1.1012, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.721664275466284, | |
| "eval_accuracy": 0.5, | |
| "eval_f1_macro": 0.40820204518718034, | |
| "eval_loss": 1.1420663595199585, | |
| "eval_runtime": 12.5482, | |
| "eval_samples_per_second": 248.481, | |
| "eval_steps_per_second": 31.08, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8651362984218078, | |
| "grad_norm": 61.26133728027344, | |
| "learning_rate": 0.00029761904761904765, | |
| "loss": 1.0592, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0086083213773316, | |
| "grad_norm": 14.982177734375, | |
| "learning_rate": 0.0002777777777777778, | |
| "loss": 0.9699, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.152080344332855, | |
| "grad_norm": 81.27915954589844, | |
| "learning_rate": 0.00025793650793650796, | |
| "loss": 0.8729, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.2955523672883786, | |
| "grad_norm": 20.86075210571289, | |
| "learning_rate": 0.0002380952380952381, | |
| "loss": 0.8175, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.2955523672883786, | |
| "eval_accuracy": 0.7113534316869788, | |
| "eval_f1_macro": 0.7266787228808548, | |
| "eval_loss": 0.9044921398162842, | |
| "eval_runtime": 12.5282, | |
| "eval_samples_per_second": 248.878, | |
| "eval_steps_per_second": 31.13, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4390243902439024, | |
| "grad_norm": 25.487239837646484, | |
| "learning_rate": 0.00021825396825396825, | |
| "loss": 0.7922, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.582496413199426, | |
| "grad_norm": 63.39326477050781, | |
| "learning_rate": 0.0001984126984126984, | |
| "loss": 0.6825, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.72596843615495, | |
| "grad_norm": 28.30609893798828, | |
| "learning_rate": 0.00017857142857142857, | |
| "loss": 0.6263, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.869440459110473, | |
| "grad_norm": 20.287073135375977, | |
| "learning_rate": 0.00015873015873015873, | |
| "loss": 0.5372, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.869440459110473, | |
| "eval_accuracy": 0.8787684413085312, | |
| "eval_f1_macro": 0.8885767692084414, | |
| "eval_loss": 0.48355233669281006, | |
| "eval_runtime": 12.5225, | |
| "eval_samples_per_second": 248.991, | |
| "eval_steps_per_second": 31.144, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.012912482065997, | |
| "grad_norm": 23.214784622192383, | |
| "learning_rate": 0.0001388888888888889, | |
| "loss": 0.4395, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.1563845050215207, | |
| "grad_norm": 60.17699432373047, | |
| "learning_rate": 0.00011904761904761905, | |
| "loss": 0.3497, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.2998565279770444, | |
| "grad_norm": 1.4547343254089355, | |
| "learning_rate": 9.92063492063492e-05, | |
| "loss": 0.3281, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.443328550932568, | |
| "grad_norm": 8.235297203063965, | |
| "learning_rate": 7.936507936507937e-05, | |
| "loss": 0.3182, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.443328550932568, | |
| "eval_accuracy": 0.9153303399615138, | |
| "eval_f1_macro": 0.916711729246869, | |
| "eval_loss": 0.41809144616127014, | |
| "eval_runtime": 12.7913, | |
| "eval_samples_per_second": 243.759, | |
| "eval_steps_per_second": 30.489, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.586800573888092, | |
| "grad_norm": 24.18486213684082, | |
| "learning_rate": 5.9523809523809524e-05, | |
| "loss": 0.2798, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.7302725968436157, | |
| "grad_norm": 12.47843074798584, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 0.2553, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.8737446197991394, | |
| "grad_norm": 24.14739990234375, | |
| "learning_rate": 1.984126984126984e-05, | |
| "loss": 0.231, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.017216642754663, | |
| "grad_norm": 4.063531398773193, | |
| "learning_rate": 0.0, | |
| "loss": 0.2035, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.017216642754663, | |
| "eval_accuracy": 0.9329698524695318, | |
| "eval_f1_macro": 0.9320770657825714, | |
| "eval_loss": 0.33222800493240356, | |
| "eval_runtime": 12.7416, | |
| "eval_samples_per_second": 244.711, | |
| "eval_steps_per_second": 30.608, | |
| "step": 2800 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 2800, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.2540383690752e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |