| { | |
| "best_metric": 0.49125435948371887, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_t5/t5_base_twitter/checkpoint-250", | |
| "epoch": 0.9191176470588235, | |
| "eval_steps": 50, | |
| "global_step": 250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.5012608766555786, | |
| "learning_rate": 0.0004938725490196079, | |
| "loss": 0.6085, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0885683298110962, | |
| "learning_rate": 0.0004877450980392157, | |
| "loss": 0.5402, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.7890007495880127, | |
| "learning_rate": 0.00048161764705882356, | |
| "loss": 0.5591, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2080425024032593, | |
| "learning_rate": 0.00047549019607843134, | |
| "loss": 0.5687, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9302119612693787, | |
| "learning_rate": 0.0004693627450980392, | |
| "loss": 0.4808, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.7444852941176471, | |
| "eval_f1_macro": 0.6739663773514942, | |
| "eval_f1_micro": 0.7444852941176471, | |
| "eval_loss": 0.5169704556465149, | |
| "eval_runtime": 2.6255, | |
| "eval_samples_per_second": 414.395, | |
| "eval_steps_per_second": 12.95, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.9698123931884766, | |
| "learning_rate": 0.0004632352941176471, | |
| "loss": 0.4972, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.0578800439834595, | |
| "learning_rate": 0.0004571078431372549, | |
| "loss": 0.4967, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.512136697769165, | |
| "learning_rate": 0.0004509803921568628, | |
| "loss": 0.4987, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.629062533378601, | |
| "learning_rate": 0.00044485294117647056, | |
| "loss": 0.4347, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.581597089767456, | |
| "learning_rate": 0.00043872549019607844, | |
| "loss": 0.5169, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.7555147058823529, | |
| "eval_f1_macro": 0.7269323671497585, | |
| "eval_f1_micro": 0.7555147058823529, | |
| "eval_loss": 0.5100224018096924, | |
| "eval_runtime": 2.6367, | |
| "eval_samples_per_second": 412.631, | |
| "eval_steps_per_second": 12.895, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.149016261100769, | |
| "learning_rate": 0.0004325980392156863, | |
| "loss": 0.4898, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.1472299098968506, | |
| "learning_rate": 0.0004264705882352941, | |
| "loss": 0.497, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.7374740839004517, | |
| "learning_rate": 0.000420343137254902, | |
| "loss": 0.4811, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.1296411752700806, | |
| "learning_rate": 0.0004142156862745098, | |
| "loss": 0.5066, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.9138337969779968, | |
| "learning_rate": 0.00040808823529411766, | |
| "loss": 0.4548, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_accuracy": 0.7647058823529411, | |
| "eval_f1_macro": 0.7017083911650301, | |
| "eval_f1_micro": 0.7647058823529411, | |
| "eval_loss": 0.49222531914711, | |
| "eval_runtime": 2.6971, | |
| "eval_samples_per_second": 403.396, | |
| "eval_steps_per_second": 12.606, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1084121465682983, | |
| "learning_rate": 0.0004019607843137255, | |
| "loss": 0.4683, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.0122885704040527, | |
| "learning_rate": 0.0003958333333333333, | |
| "loss": 0.5217, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.2070980072021484, | |
| "learning_rate": 0.0003897058823529412, | |
| "loss": 0.524, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.27259361743927, | |
| "learning_rate": 0.00038357843137254904, | |
| "loss": 0.4955, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.8282164931297302, | |
| "learning_rate": 0.0003774509803921569, | |
| "loss": 0.498, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.7518382352941176, | |
| "eval_f1_macro": 0.6776331672629108, | |
| "eval_f1_micro": 0.7518382352941176, | |
| "eval_loss": 0.5056816935539246, | |
| "eval_runtime": 2.6992, | |
| "eval_samples_per_second": 403.083, | |
| "eval_steps_per_second": 12.596, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.8028568029403687, | |
| "learning_rate": 0.0003713235294117647, | |
| "loss": 0.4926, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 2.6962406635284424, | |
| "learning_rate": 0.00036519607843137254, | |
| "loss": 0.5134, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.7592838406562805, | |
| "learning_rate": 0.0003590686274509804, | |
| "loss": 0.469, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.637353777885437, | |
| "learning_rate": 0.00035294117647058826, | |
| "loss": 0.4381, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.9144633412361145, | |
| "learning_rate": 0.0003468137254901961, | |
| "loss": 0.4844, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.765625, | |
| "eval_f1_macro": 0.7266124240384777, | |
| "eval_f1_micro": 0.765625, | |
| "eval_loss": 0.49125435948371887, | |
| "eval_runtime": 2.6614, | |
| "eval_samples_per_second": 408.809, | |
| "eval_steps_per_second": 12.775, | |
| "step": 250 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 816, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 1221553815552000.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |