| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9788235294117648, | |
| "eval_steps": 500, | |
| "global_step": 26, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.103447222745244, | |
| "learning_rate": 0.0001, | |
| "loss": 1.5918, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.930916372129095, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 1.6299, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.398070948517042, | |
| "learning_rate": 9.842915805643155e-05, | |
| "loss": 1.6372, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.17279784653199, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 1.5483, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.251687596185326, | |
| "learning_rate": 9.381533400219318e-05, | |
| "loss": 1.6069, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.1869658645951886, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.4985, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.7267817051984844, | |
| "learning_rate": 8.644843137107059e-05, | |
| "loss": 1.4424, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.9101891237074138, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 1.3213, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.6330633181587038, | |
| "learning_rate": 7.679133974894983e-05, | |
| "loss": 1.4426, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.6305240585237435, | |
| "learning_rate": 7.128896457825364e-05, | |
| "loss": 1.3071, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.3944693110209463, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.3665, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.4126280206859627, | |
| "learning_rate": 5.9369065729286245e-05, | |
| "loss": 1.2878, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.367322681350601, | |
| "learning_rate": 5.313952597646568e-05, | |
| "loss": 1.3589, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.2053182377155975, | |
| "learning_rate": 4.6860474023534335e-05, | |
| "loss": 1.3247, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.109279281261198, | |
| "learning_rate": 4.063093427071376e-05, | |
| "loss": 1.2959, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.3045532630263748, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.3459, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.197638462603245, | |
| "learning_rate": 2.8711035421746367e-05, | |
| "loss": 1.4053, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.2235614357026008, | |
| "learning_rate": 2.3208660251050158e-05, | |
| "loss": 1.301, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.4513710687985, | |
| "learning_rate": 1.8128800512565513e-05, | |
| "loss": 1.3291, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.0341249332438227, | |
| "learning_rate": 1.3551568628929434e-05, | |
| "loss": 1.394, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.251919439255797, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.3943, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.1326570760452561, | |
| "learning_rate": 6.184665997806832e-06, | |
| "loss": 1.311, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.163110395778291, | |
| "learning_rate": 3.511175705587433e-06, | |
| "loss": 1.4033, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.1416927151375318, | |
| "learning_rate": 1.5708419435684462e-06, | |
| "loss": 1.271, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.0311072964787094, | |
| "learning_rate": 3.9426493427611177e-07, | |
| "loss": 1.2318, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.9861863614033847, | |
| "learning_rate": 0.0, | |
| "loss": 1.2832, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "step": 26, | |
| "total_flos": 1.1771536628462387e+17, | |
| "train_loss": 1.3973060021033654, | |
| "train_runtime": 832.4023, | |
| "train_samples_per_second": 4.085, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 26, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "total_flos": 1.1771536628462387e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |