| { | |
| "best_metric": 4.681754112243652, | |
| "best_model_checkpoint": "car_brands_image_detection/checkpoint-21765", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 21765, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.245316505432129, | |
| "learning_rate": 4.8963849873359435e-06, | |
| "loss": 5.7646, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.3855398893356323, | |
| "learning_rate": 4.781257195486991e-06, | |
| "loss": 5.7198, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.5914552211761475, | |
| "learning_rate": 4.666129403638039e-06, | |
| "loss": 5.6722, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.574438214302063, | |
| "learning_rate": 4.551001611789087e-06, | |
| "loss": 5.6256, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.6384121179580688, | |
| "learning_rate": 4.435873819940134e-06, | |
| "loss": 5.5753, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.455289363861084, | |
| "learning_rate": 4.320746028091182e-06, | |
| "loss": 5.5277, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.637087345123291, | |
| "learning_rate": 4.205618236242229e-06, | |
| "loss": 5.4828, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.887794017791748, | |
| "learning_rate": 4.090490444393277e-06, | |
| "loss": 5.4315, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.16143021252153936, | |
| "eval_loss": 5.4130473136901855, | |
| "eval_runtime": 373.3699, | |
| "eval_samples_per_second": 93.259, | |
| "eval_steps_per_second": 11.659, | |
| "step": 4353 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.1020355224609375, | |
| "learning_rate": 3.975362652544324e-06, | |
| "loss": 5.3915, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.868577241897583, | |
| "learning_rate": 3.860234860695372e-06, | |
| "loss": 5.34, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.8365103006362915, | |
| "learning_rate": 3.74510706884642e-06, | |
| "loss": 5.2992, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.9328190088272095, | |
| "learning_rate": 3.6299792769974675e-06, | |
| "loss": 5.2557, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.9776891469955444, | |
| "learning_rate": 3.514851485148515e-06, | |
| "loss": 5.2221, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 1.8611873388290405, | |
| "learning_rate": 3.3997236932995626e-06, | |
| "loss": 5.1845, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 2.0713303089141846, | |
| "learning_rate": 3.2845959014506106e-06, | |
| "loss": 5.1466, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 2.1336920261383057, | |
| "learning_rate": 3.1694681096016582e-06, | |
| "loss": 5.1209, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 2.0790388584136963, | |
| "learning_rate": 3.054340317752706e-06, | |
| "loss": 5.0821, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.2534175761056864, | |
| "eval_loss": 5.101416110992432, | |
| "eval_runtime": 379.9834, | |
| "eval_samples_per_second": 91.636, | |
| "eval_steps_per_second": 11.456, | |
| "step": 8706 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 2.2780089378356934, | |
| "learning_rate": 2.9392125259037534e-06, | |
| "loss": 5.0448, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 2.555509090423584, | |
| "learning_rate": 2.824084734054801e-06, | |
| "loss": 5.005, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 2.188175916671753, | |
| "learning_rate": 2.7089569422058486e-06, | |
| "loss": 4.9819, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 2.2957634925842285, | |
| "learning_rate": 2.593829150356896e-06, | |
| "loss": 4.9499, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.6697535514831543, | |
| "learning_rate": 2.4787013585079438e-06, | |
| "loss": 4.9236, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.6424810886383057, | |
| "learning_rate": 2.363573566658992e-06, | |
| "loss": 4.8961, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.5476348400115967, | |
| "learning_rate": 2.2484457748100394e-06, | |
| "loss": 4.8669, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 2.3084092140197754, | |
| "learning_rate": 2.133317982961087e-06, | |
| "loss": 4.8464, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 2.4545116424560547, | |
| "learning_rate": 2.0181901911121346e-06, | |
| "loss": 4.825, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.3051120045950603, | |
| "eval_loss": 4.873249530792236, | |
| "eval_runtime": 383.4043, | |
| "eval_samples_per_second": 90.818, | |
| "eval_steps_per_second": 11.354, | |
| "step": 13059 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 2.7930803298950195, | |
| "learning_rate": 1.9030623992631822e-06, | |
| "loss": 4.7866, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 3.1879501342773438, | |
| "learning_rate": 1.7879346074142297e-06, | |
| "loss": 4.7743, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 2.784158945083618, | |
| "learning_rate": 1.6728068155652778e-06, | |
| "loss": 4.7578, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 2.792235851287842, | |
| "learning_rate": 1.5576790237163253e-06, | |
| "loss": 4.7372, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 2.644819974899292, | |
| "learning_rate": 1.442551231867373e-06, | |
| "loss": 4.7102, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 2.5567691326141357, | |
| "learning_rate": 1.3274234400184205e-06, | |
| "loss": 4.705, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 2.979523181915283, | |
| "learning_rate": 1.2122956481694683e-06, | |
| "loss": 4.6817, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 2.797534227371216, | |
| "learning_rate": 1.097167856320516e-06, | |
| "loss": 4.6712, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.3327110855829983, | |
| "eval_loss": 4.730710029602051, | |
| "eval_runtime": 371.2204, | |
| "eval_samples_per_second": 93.799, | |
| "eval_steps_per_second": 11.726, | |
| "step": 17412 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.7830564975738525, | |
| "learning_rate": 9.820400644715635e-07, | |
| "loss": 4.6613, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 2.9348843097686768, | |
| "learning_rate": 8.669122726226112e-07, | |
| "loss": 4.6467, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 2.592219829559326, | |
| "learning_rate": 7.517844807736588e-07, | |
| "loss": 4.6333, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 4.5050129890441895, | |
| "learning_rate": 6.366566889247065e-07, | |
| "loss": 4.6238, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 3.7189841270446777, | |
| "learning_rate": 5.215288970757541e-07, | |
| "loss": 4.6177, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 2.521855115890503, | |
| "learning_rate": 4.064011052268018e-07, | |
| "loss": 4.6131, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 2.892822027206421, | |
| "learning_rate": 2.912733133778494e-07, | |
| "loss": 4.6138, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 3.660858631134033, | |
| "learning_rate": 1.761455215288971e-07, | |
| "loss": 4.5955, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 2.4211411476135254, | |
| "learning_rate": 6.101772967994475e-08, | |
| "loss": 4.592, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.3435956346927053, | |
| "eval_loss": 4.681754112243652, | |
| "eval_runtime": 375.04, | |
| "eval_samples_per_second": 92.843, | |
| "eval_steps_per_second": 11.607, | |
| "step": 21765 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 21765, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.411951551988482e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |