| { | |
| "best_metric": 0.6583548784255981, | |
| "best_model_checkpoint": "facial_emotions_image_detection_rafdb_microsoft_vit/checkpoint-12540", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 12540, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 14.677521705627441, | |
| "learning_rate": 2.7281514297221103e-06, | |
| "loss": 1.7706, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.45075587486903157, | |
| "eval_loss": 1.4894078969955444, | |
| "eval_runtime": 69.9167, | |
| "eval_samples_per_second": 191.113, | |
| "eval_steps_per_second": 11.957, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 17.35078239440918, | |
| "learning_rate": 2.4260974627466776e-06, | |
| "loss": 1.4285, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5424337674000899, | |
| "eval_loss": 1.2471543550491333, | |
| "eval_runtime": 65.1157, | |
| "eval_samples_per_second": 205.204, | |
| "eval_steps_per_second": 12.839, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 15.841980934143066, | |
| "learning_rate": 2.1240434957712445e-06, | |
| "loss": 1.2563, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5873372249663225, | |
| "eval_loss": 1.1151762008666992, | |
| "eval_runtime": 65.1132, | |
| "eval_samples_per_second": 205.212, | |
| "eval_steps_per_second": 12.839, | |
| "step": 1881 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 24.84258270263672, | |
| "learning_rate": 1.8219895287958114e-06, | |
| "loss": 1.1575, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 14.571057319641113, | |
| "learning_rate": 1.5199355618203787e-06, | |
| "loss": 1.0741, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6200419098937285, | |
| "eval_loss": 1.0384080410003662, | |
| "eval_runtime": 65.4313, | |
| "eval_samples_per_second": 204.214, | |
| "eval_steps_per_second": 12.777, | |
| "step": 2508 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 19.834932327270508, | |
| "learning_rate": 1.2178815948449456e-06, | |
| "loss": 1.0238, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6430923514443946, | |
| "eval_loss": 0.981325626373291, | |
| "eval_runtime": 66.0473, | |
| "eval_samples_per_second": 202.31, | |
| "eval_steps_per_second": 12.658, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "grad_norm": 17.73134994506836, | |
| "learning_rate": 9.158276278695127e-07, | |
| "loss": 0.9837, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6525969166292471, | |
| "eval_loss": 0.9464946985244751, | |
| "eval_runtime": 66.383, | |
| "eval_samples_per_second": 201.286, | |
| "eval_steps_per_second": 12.594, | |
| "step": 3762 | |
| }, | |
| { | |
| "epoch": 6.38, | |
| "grad_norm": 14.271998405456543, | |
| "learning_rate": 6.137736608940798e-07, | |
| "loss": 0.9694, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6600059871276756, | |
| "eval_loss": 0.9257609248161316, | |
| "eval_runtime": 66.6168, | |
| "eval_samples_per_second": 200.58, | |
| "eval_steps_per_second": 12.549, | |
| "step": 4389 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "grad_norm": 26.607189178466797, | |
| "learning_rate": 3.117196939186468e-07, | |
| "loss": 0.9391, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.97, | |
| "grad_norm": 20.43568992614746, | |
| "learning_rate": 9.665726943213855e-09, | |
| "loss": 0.9235, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.663897620116749, | |
| "eval_loss": 0.9192785620689392, | |
| "eval_runtime": 64.6255, | |
| "eval_samples_per_second": 206.761, | |
| "eval_steps_per_second": 12.936, | |
| "step": 5016 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "grad_norm": 20.14082908630371, | |
| "learning_rate": 1.6909527622097679e-06, | |
| "loss": 0.9239, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6873970962430774, | |
| "eval_loss": 0.8600204586982727, | |
| "eval_runtime": 66.5155, | |
| "eval_samples_per_second": 200.886, | |
| "eval_steps_per_second": 12.569, | |
| "step": 5643 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "grad_norm": 20.530790328979492, | |
| "learning_rate": 1.5708566853482785e-06, | |
| "loss": 0.8971, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6999700643616225, | |
| "eval_loss": 0.81741863489151, | |
| "eval_runtime": 66.4357, | |
| "eval_samples_per_second": 201.127, | |
| "eval_steps_per_second": 12.584, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 10.37, | |
| "grad_norm": 17.738765716552734, | |
| "learning_rate": 1.4507606084867894e-06, | |
| "loss": 0.8592, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7122436760963927, | |
| "eval_loss": 0.7865827083587646, | |
| "eval_runtime": 66.2499, | |
| "eval_samples_per_second": 201.691, | |
| "eval_steps_per_second": 12.619, | |
| "step": 6897 | |
| }, | |
| { | |
| "epoch": 11.16, | |
| "grad_norm": 19.051393508911133, | |
| "learning_rate": 1.3306645316253003e-06, | |
| "loss": 0.8255, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 11.96, | |
| "grad_norm": 25.116697311401367, | |
| "learning_rate": 1.2105684547638111e-06, | |
| "loss": 0.8041, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.727211495285137, | |
| "eval_loss": 0.7529184222221375, | |
| "eval_runtime": 66.27, | |
| "eval_samples_per_second": 201.63, | |
| "eval_steps_per_second": 12.615, | |
| "step": 7524 | |
| }, | |
| { | |
| "epoch": 12.76, | |
| "grad_norm": 18.908737182617188, | |
| "learning_rate": 1.0904723779023218e-06, | |
| "loss": 0.7815, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7366412213740458, | |
| "eval_loss": 0.7270856499671936, | |
| "eval_runtime": 64.8125, | |
| "eval_samples_per_second": 206.164, | |
| "eval_steps_per_second": 12.899, | |
| "step": 8151 | |
| }, | |
| { | |
| "epoch": 13.56, | |
| "grad_norm": 21.439659118652344, | |
| "learning_rate": 9.703763010408327e-07, | |
| "loss": 0.7533, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7433767400089807, | |
| "eval_loss": 0.7089582085609436, | |
| "eval_runtime": 65.4385, | |
| "eval_samples_per_second": 204.192, | |
| "eval_steps_per_second": 12.775, | |
| "step": 8778 | |
| }, | |
| { | |
| "epoch": 14.35, | |
| "grad_norm": 19.495594024658203, | |
| "learning_rate": 8.502802241793434e-07, | |
| "loss": 0.7425, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7480916030534351, | |
| "eval_loss": 0.6946737170219421, | |
| "eval_runtime": 65.0525, | |
| "eval_samples_per_second": 205.403, | |
| "eval_steps_per_second": 12.851, | |
| "step": 9405 | |
| }, | |
| { | |
| "epoch": 15.15, | |
| "grad_norm": 17.079021453857422, | |
| "learning_rate": 7.301841473178543e-07, | |
| "loss": 0.721, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 15.95, | |
| "grad_norm": 16.454561233520508, | |
| "learning_rate": 6.100880704563651e-07, | |
| "loss": 0.7184, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7560993863194133, | |
| "eval_loss": 0.6799051761627197, | |
| "eval_runtime": 65.3787, | |
| "eval_samples_per_second": 204.378, | |
| "eval_steps_per_second": 12.787, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 16.75, | |
| "grad_norm": 16.228721618652344, | |
| "learning_rate": 4.899919935948759e-07, | |
| "loss": 0.6994, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.758569076485556, | |
| "eval_loss": 0.6724166870117188, | |
| "eval_runtime": 65.0284, | |
| "eval_samples_per_second": 205.479, | |
| "eval_steps_per_second": 12.856, | |
| "step": 10659 | |
| }, | |
| { | |
| "epoch": 17.54, | |
| "grad_norm": 18.473840713500977, | |
| "learning_rate": 3.698959167333867e-07, | |
| "loss": 0.7012, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7650052387367161, | |
| "eval_loss": 0.6634852290153503, | |
| "eval_runtime": 66.3039, | |
| "eval_samples_per_second": 201.527, | |
| "eval_steps_per_second": 12.609, | |
| "step": 11286 | |
| }, | |
| { | |
| "epoch": 18.34, | |
| "grad_norm": 18.456228256225586, | |
| "learning_rate": 2.497998398718975e-07, | |
| "loss": 0.6899, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7647058823529411, | |
| "eval_loss": 0.6589637994766235, | |
| "eval_runtime": 67.4002, | |
| "eval_samples_per_second": 198.249, | |
| "eval_steps_per_second": 12.404, | |
| "step": 11913 | |
| }, | |
| { | |
| "epoch": 19.14, | |
| "grad_norm": 21.776016235351562, | |
| "learning_rate": 1.2970376301040834e-07, | |
| "loss": 0.6772, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 19.94, | |
| "grad_norm": 20.911270141601562, | |
| "learning_rate": 9.607686148919135e-09, | |
| "loss": 0.6831, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7655291124083221, | |
| "eval_loss": 0.6583548784255981, | |
| "eval_runtime": 66.3383, | |
| "eval_samples_per_second": 201.422, | |
| "eval_steps_per_second": 12.602, | |
| "step": 12540 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 12540, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 9.964666614526525e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |