| { | |
| "best_metric": 0.8254, | |
| "best_model_checkpoint": "imagenet-tiger/checkpoint-3750", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 3750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 11.370553970336914, | |
| "learning_rate": 9.982464296247522e-05, | |
| "loss": 1.1275, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 11.340897560119629, | |
| "learning_rate": 9.929980185352526e-05, | |
| "loss": 1.1066, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 9.576211929321289, | |
| "learning_rate": 9.842915805643155e-05, | |
| "loss": 1.1377, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 10.26335334777832, | |
| "learning_rate": 9.721881851187406e-05, | |
| "loss": 1.1071, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 9.477747917175293, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.1081, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 7.301119804382324, | |
| "learning_rate": 9.381533400219318e-05, | |
| "loss": 1.0873, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 11.125399589538574, | |
| "learning_rate": 9.164606203550497e-05, | |
| "loss": 1.1031, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 13.228848457336426, | |
| "learning_rate": 8.9184672866292e-05, | |
| "loss": 1.0457, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 7.878283500671387, | |
| "learning_rate": 8.644843137107059e-05, | |
| "loss": 1.1162, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.097326755523682, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.1242, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 12.475173950195312, | |
| "learning_rate": 8.022995574311876e-05, | |
| "loss": 1.1094, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 12.513256072998047, | |
| "learning_rate": 7.679133974894983e-05, | |
| "loss": 1.0771, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7971, | |
| "eval_loss": 0.7663370370864868, | |
| "eval_runtime": 202.227, | |
| "eval_samples_per_second": 49.449, | |
| "eval_steps_per_second": 1.548, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 10.96892261505127, | |
| "learning_rate": 7.316480175599309e-05, | |
| "loss": 1.1045, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 12.19508171081543, | |
| "learning_rate": 6.937577932260515e-05, | |
| "loss": 0.9819, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 16.93886375427246, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.967, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 12.414902687072754, | |
| "learning_rate": 6.141754350553279e-05, | |
| "loss": 0.9287, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 16.11931610107422, | |
| "learning_rate": 5.730415142812059e-05, | |
| "loss": 0.8632, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 10.42251968383789, | |
| "learning_rate": 5.313952597646568e-05, | |
| "loss": 0.9125, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 11.504471778869629, | |
| "learning_rate": 4.895287900583216e-05, | |
| "loss": 0.9045, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 9.115998268127441, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.8667, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 10.520715713500977, | |
| "learning_rate": 4.063093427071376e-05, | |
| "loss": 0.8806, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 12.112250328063965, | |
| "learning_rate": 3.655400896923672e-05, | |
| "loss": 0.884, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 12.806370735168457, | |
| "learning_rate": 3.257139763390925e-05, | |
| "loss": 0.807, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 12.071837425231934, | |
| "learning_rate": 2.8711035421746367e-05, | |
| "loss": 0.8634, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 9.2099609375, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.8206, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8207, | |
| "eval_loss": 0.6771770119667053, | |
| "eval_runtime": 200.5917, | |
| "eval_samples_per_second": 49.853, | |
| "eval_steps_per_second": 1.56, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 9.109099388122559, | |
| "learning_rate": 2.1464321615778422e-05, | |
| "loss": 0.7565, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 7.973142623901367, | |
| "learning_rate": 1.8128800512565513e-05, | |
| "loss": 0.7475, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 13.218464851379395, | |
| "learning_rate": 1.5016832974331724e-05, | |
| "loss": 0.7201, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 7.9871134757995605, | |
| "learning_rate": 1.2150247217412186e-05, | |
| "loss": 0.7449, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 7.4544291496276855, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.7013, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 8.663176536560059, | |
| "learning_rate": 7.2317869919746705e-06, | |
| "loss": 0.7347, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 12.630393028259277, | |
| "learning_rate": 5.214411988029355e-06, | |
| "loss": 0.7542, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 8.864751815795898, | |
| "learning_rate": 3.511175705587433e-06, | |
| "loss": 0.6819, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 6.5524725914001465, | |
| "learning_rate": 2.134025123396638e-06, | |
| "loss": 0.7919, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 7.316872596740723, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.6864, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 9.99939250946045, | |
| "learning_rate": 3.9426493427611177e-07, | |
| "loss": 0.7364, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 12.940382957458496, | |
| "learning_rate": 4.385849505708084e-08, | |
| "loss": 0.7212, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8254, | |
| "eval_loss": 0.6595176458358765, | |
| "eval_runtime": 198.5571, | |
| "eval_samples_per_second": 50.363, | |
| "eval_steps_per_second": 1.576, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 3750, | |
| "total_flos": 9.37824189087744e+18, | |
| "train_loss": 0.9114908182779948, | |
| "train_runtime": 5966.5057, | |
| "train_samples_per_second": 20.112, | |
| "train_steps_per_second": 0.629 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 3750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 9.37824189087744e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |