| { | |
| "best_metric": 0.32041043043136597, | |
| "best_model_checkpoint": "/mnt/beegfs/farid/mlora/outputs/xnli/aya-101/en/rank4_lr5e-5/checkpoint-5500", | |
| "epoch": 0.24445893089960888, | |
| "eval_steps": 500, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.020371577574967405, | |
| "grad_norm": 11.233623504638672, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 1.0562, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.020371577574967405, | |
| "eval_accuracy": 0.7180722891566265, | |
| "eval_f1": 0.6924640340883744, | |
| "eval_loss": 0.6811789870262146, | |
| "eval_runtime": 413.6821, | |
| "eval_samples_per_second": 6.019, | |
| "eval_steps_per_second": 0.377, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04074315514993481, | |
| "grad_norm": 9.590261459350586, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.5179, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.04074315514993481, | |
| "eval_accuracy": 0.8461847389558232, | |
| "eval_f1": 0.8450597019988969, | |
| "eval_loss": 0.45314186811447144, | |
| "eval_runtime": 412.2609, | |
| "eval_samples_per_second": 6.04, | |
| "eval_steps_per_second": 0.378, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06111473272490222, | |
| "grad_norm": 22.115205764770508, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.4515, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.06111473272490222, | |
| "eval_accuracy": 0.8654618473895582, | |
| "eval_f1": 0.8652148804287579, | |
| "eval_loss": 0.3968792259693146, | |
| "eval_runtime": 431.7684, | |
| "eval_samples_per_second": 5.767, | |
| "eval_steps_per_second": 0.361, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.08148631029986962, | |
| "grad_norm": 7.419126033782959, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.4078, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.08148631029986962, | |
| "eval_accuracy": 0.8710843373493976, | |
| "eval_f1": 0.8706568816583564, | |
| "eval_loss": 0.38109150528907776, | |
| "eval_runtime": 411.4289, | |
| "eval_samples_per_second": 6.052, | |
| "eval_steps_per_second": 0.379, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.10185788787483703, | |
| "grad_norm": 4.023831844329834, | |
| "learning_rate": 3.240740740740741e-05, | |
| "loss": 0.4042, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.10185788787483703, | |
| "eval_accuracy": 0.876706827309237, | |
| "eval_f1": 0.8758798869119646, | |
| "eval_loss": 0.37034499645233154, | |
| "eval_runtime": 411.4246, | |
| "eval_samples_per_second": 6.052, | |
| "eval_steps_per_second": 0.379, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.12222946544980444, | |
| "grad_norm": 10.224037170410156, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.3805, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.12222946544980444, | |
| "eval_accuracy": 0.8863453815261044, | |
| "eval_f1": 0.8862348263748947, | |
| "eval_loss": 0.3436357080936432, | |
| "eval_runtime": 411.4342, | |
| "eval_samples_per_second": 6.052, | |
| "eval_steps_per_second": 0.379, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.14260104302477183, | |
| "grad_norm": 7.620177745819092, | |
| "learning_rate": 2.314814814814815e-05, | |
| "loss": 0.3747, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.14260104302477183, | |
| "eval_accuracy": 0.8887550200803213, | |
| "eval_f1": 0.8885562445706028, | |
| "eval_loss": 0.32690730690956116, | |
| "eval_runtime": 411.9499, | |
| "eval_samples_per_second": 6.044, | |
| "eval_steps_per_second": 0.379, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.16297262059973924, | |
| "grad_norm": 4.451201915740967, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.3802, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.16297262059973924, | |
| "eval_accuracy": 0.8947791164658635, | |
| "eval_f1": 0.8943233881142542, | |
| "eval_loss": 0.3284657895565033, | |
| "eval_runtime": 411.4408, | |
| "eval_samples_per_second": 6.052, | |
| "eval_steps_per_second": 0.379, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.18334419817470665, | |
| "grad_norm": 8.444794654846191, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.3636, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.18334419817470665, | |
| "eval_accuracy": 0.8947791164658635, | |
| "eval_f1": 0.8945672170700908, | |
| "eval_loss": 0.3261275589466095, | |
| "eval_runtime": 411.7339, | |
| "eval_samples_per_second": 6.048, | |
| "eval_steps_per_second": 0.379, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.20371577574967406, | |
| "grad_norm": 10.426863670349121, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.374, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.20371577574967406, | |
| "eval_accuracy": 0.893574297188755, | |
| "eval_f1": 0.8932728411563985, | |
| "eval_loss": 0.3269090950489044, | |
| "eval_runtime": 411.5544, | |
| "eval_samples_per_second": 6.05, | |
| "eval_steps_per_second": 0.379, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.22408735332464147, | |
| "grad_norm": 4.699596405029297, | |
| "learning_rate": 4.6296296296296296e-06, | |
| "loss": 0.3692, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.22408735332464147, | |
| "eval_accuracy": 0.8931726907630522, | |
| "eval_f1": 0.8929461976278104, | |
| "eval_loss": 0.32041043043136597, | |
| "eval_runtime": 411.5803, | |
| "eval_samples_per_second": 6.05, | |
| "eval_steps_per_second": 0.379, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.24445893089960888, | |
| "grad_norm": 7.466755390167236, | |
| "learning_rate": 0.0, | |
| "loss": 0.3548, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.24445893089960888, | |
| "eval_accuracy": 0.8939759036144578, | |
| "eval_f1": 0.8935742825082755, | |
| "eval_loss": 0.3236287534236908, | |
| "eval_runtime": 411.5661, | |
| "eval_samples_per_second": 6.05, | |
| "eval_steps_per_second": 0.379, | |
| "step": 6000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 8.03166870528e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |