{ "best_metric": 2.097651720046997, "best_model_checkpoint": "/content/models/openai-community-gpt2-large/checkpoint-1460", "epoch": 5.0, "eval_steps": 365, "global_step": 7290, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.25, "grad_norm": 4.67961311340332, "learning_rate": 4.7496570644718796e-05, "loss": 2.3306, "step": 365 }, { "epoch": 0.25, "eval_loss": 2.2394158840179443, "eval_runtime": 3.8004, "eval_samples_per_second": 53.942, "eval_steps_per_second": 6.841, "step": 365 }, { "epoch": 0.5, "grad_norm": 3.010347366333008, "learning_rate": 4.499314128943759e-05, "loss": 2.1698, "step": 730 }, { "epoch": 0.5, "eval_loss": 2.153510332107544, "eval_runtime": 3.8111, "eval_samples_per_second": 53.79, "eval_steps_per_second": 6.822, "step": 730 }, { "epoch": 0.75, "grad_norm": 2.377396583557129, "learning_rate": 4.248971193415638e-05, "loss": 2.1126, "step": 1095 }, { "epoch": 0.75, "eval_loss": 2.1220626831054688, "eval_runtime": 3.8077, "eval_samples_per_second": 53.839, "eval_steps_per_second": 6.828, "step": 1095 }, { "epoch": 1.0, "grad_norm": 2.0250906944274902, "learning_rate": 3.9986282578875175e-05, "loss": 2.0649, "step": 1460 }, { "epoch": 1.0, "eval_loss": 2.097651720046997, "eval_runtime": 3.8144, "eval_samples_per_second": 53.744, "eval_steps_per_second": 6.816, "step": 1460 }, { "epoch": 1.25, "grad_norm": 3.449071168899536, "learning_rate": 3.748285322359397e-05, "loss": 1.6887, "step": 1825 }, { "epoch": 1.25, "eval_loss": 2.125108003616333, "eval_runtime": 3.8167, "eval_samples_per_second": 53.712, "eval_steps_per_second": 6.812, "step": 1825 }, { "epoch": 1.5, "grad_norm": 2.925269365310669, "learning_rate": 3.497942386831276e-05, "loss": 1.7126, "step": 2190 }, { "epoch": 1.5, "eval_loss": 2.1161348819732666, "eval_runtime": 3.828, "eval_samples_per_second": 53.553, "eval_steps_per_second": 6.792, "step": 2190 }, { "epoch": 1.75, "grad_norm": 3.348761796951294, "learning_rate": 3.247599451303155e-05, "loss": 1.7253, "step": 2555 }, { "epoch": 1.75, "eval_loss": 2.1163294315338135, "eval_runtime": 3.8247, "eval_samples_per_second": 53.599, "eval_steps_per_second": 6.798, "step": 2555 }, { "epoch": 2.0, "grad_norm": 2.501504898071289, "learning_rate": 2.9972565157750345e-05, "loss": 1.6857, "step": 2920 }, { "epoch": 2.0, "eval_loss": 2.1025500297546387, "eval_runtime": 3.8312, "eval_samples_per_second": 53.508, "eval_steps_per_second": 6.786, "step": 2920 }, { "epoch": 2.25, "grad_norm": 4.346751689910889, "learning_rate": 2.7469135802469138e-05, "loss": 1.1799, "step": 3285 }, { "epoch": 2.25, "eval_loss": 2.34588360786438, "eval_runtime": 3.8313, "eval_samples_per_second": 53.507, "eval_steps_per_second": 6.786, "step": 3285 }, { "epoch": 2.5, "grad_norm": 5.11303186416626, "learning_rate": 2.496570644718793e-05, "loss": 1.1877, "step": 3650 }, { "epoch": 2.5, "eval_loss": 2.343195676803589, "eval_runtime": 3.8294, "eval_samples_per_second": 53.534, "eval_steps_per_second": 6.79, "step": 3650 }, { "epoch": 2.75, "grad_norm": 4.008133411407471, "learning_rate": 2.246227709190672e-05, "loss": 1.1877, "step": 4015 }, { "epoch": 2.75, "eval_loss": 2.378342390060425, "eval_runtime": 3.8264, "eval_samples_per_second": 53.574, "eval_steps_per_second": 6.795, "step": 4015 }, { "epoch": 3.0, "grad_norm": 3.2181403636932373, "learning_rate": 1.9958847736625517e-05, "loss": 1.1932, "step": 4380 }, { "epoch": 3.0, "eval_loss": 2.3641903400421143, "eval_runtime": 3.8349, "eval_samples_per_second": 53.457, "eval_steps_per_second": 6.78, "step": 4380 }, { "epoch": 3.25, "grad_norm": 5.110302448272705, "learning_rate": 1.7455418381344307e-05, "loss": 0.7276, "step": 4745 }, { "epoch": 3.25, "eval_loss": 2.6530215740203857, "eval_runtime": 3.834, "eval_samples_per_second": 53.469, "eval_steps_per_second": 6.781, "step": 4745 }, { "epoch": 3.5, "grad_norm": 4.757246971130371, "learning_rate": 1.49519890260631e-05, "loss": 0.7138, "step": 5110 }, { "epoch": 3.5, "eval_loss": 2.721984386444092, "eval_runtime": 3.8333, "eval_samples_per_second": 53.479, "eval_steps_per_second": 6.783, "step": 5110 }, { "epoch": 3.76, "grad_norm": 3.532853603363037, "learning_rate": 1.2448559670781894e-05, "loss": 0.7135, "step": 5475 }, { "epoch": 3.76, "eval_loss": 2.6952061653137207, "eval_runtime": 3.8325, "eval_samples_per_second": 53.489, "eval_steps_per_second": 6.784, "step": 5475 }, { "epoch": 4.01, "grad_norm": 3.3618454933166504, "learning_rate": 9.945130315500687e-06, "loss": 0.6983, "step": 5840 }, { "epoch": 4.01, "eval_loss": 2.7304930686950684, "eval_runtime": 3.8317, "eval_samples_per_second": 53.502, "eval_steps_per_second": 6.786, "step": 5840 }, { "epoch": 4.26, "grad_norm": 4.5132622718811035, "learning_rate": 7.441700960219479e-06, "loss": 0.4195, "step": 6205 }, { "epoch": 4.26, "eval_loss": 2.9191341400146484, "eval_runtime": 3.8313, "eval_samples_per_second": 53.506, "eval_steps_per_second": 6.786, "step": 6205 }, { "epoch": 4.51, "grad_norm": 5.9186201095581055, "learning_rate": 4.938271604938272e-06, "loss": 0.4193, "step": 6570 }, { "epoch": 4.51, "eval_loss": 2.9266867637634277, "eval_runtime": 3.8312, "eval_samples_per_second": 53.509, "eval_steps_per_second": 6.786, "step": 6570 }, { "epoch": 4.76, "grad_norm": 5.357140064239502, "learning_rate": 2.4348422496570644e-06, "loss": 0.4239, "step": 6935 }, { "epoch": 4.76, "eval_loss": 2.922820568084717, "eval_runtime": 3.8273, "eval_samples_per_second": 53.563, "eval_steps_per_second": 6.793, "step": 6935 }, { "epoch": 5.0, "step": 7290, "total_flos": 3966079500288000.0, "train_loss": 1.2395105746547872, "train_runtime": 1985.2268, "train_samples_per_second": 3.672, "train_steps_per_second": 3.672 } ], "logging_steps": 365, "max_steps": 7290, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 365, "total_flos": 3966079500288000.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }