| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9998339008388006, | |
| "eval_steps": 500, | |
| "global_step": 12040, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08304958059961798, | |
| "grad_norm": 0.3119010627269745, | |
| "learning_rate": 4.7940199335548175e-05, | |
| "loss": 0.3439, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16609916119923596, | |
| "grad_norm": 0.9808832406997681, | |
| "learning_rate": 4.5863787375415286e-05, | |
| "loss": 0.2923, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2491487417988539, | |
| "grad_norm": 0.4935096800327301, | |
| "learning_rate": 4.379568106312292e-05, | |
| "loss": 0.4114, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3321983223984719, | |
| "grad_norm": 0.8764172196388245, | |
| "learning_rate": 4.171926910299003e-05, | |
| "loss": 0.309, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.41524790299808984, | |
| "grad_norm": 0.9836592674255371, | |
| "learning_rate": 3.964285714285714e-05, | |
| "loss": 0.323, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4982974835977078, | |
| "grad_norm": 0.8834519386291504, | |
| "learning_rate": 3.7566445182724254e-05, | |
| "loss": 0.2963, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5813470641973258, | |
| "grad_norm": 0.0723366066813469, | |
| "learning_rate": 3.5490033222591366e-05, | |
| "loss": 0.2738, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6643966447969438, | |
| "grad_norm": 0.799544095993042, | |
| "learning_rate": 3.341362126245847e-05, | |
| "loss": 0.2729, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7474462253965617, | |
| "grad_norm": 0.26322752237319946, | |
| "learning_rate": 3.133720930232558e-05, | |
| "loss": 0.2726, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8304958059961797, | |
| "grad_norm": 1.7207447290420532, | |
| "learning_rate": 2.9264950166112958e-05, | |
| "loss": 0.2854, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9135453865957976, | |
| "grad_norm": 0.27596575021743774, | |
| "learning_rate": 2.718853820598007e-05, | |
| "loss": 0.2919, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.9965949671954156, | |
| "grad_norm": 0.3219141364097595, | |
| "learning_rate": 2.5112126245847177e-05, | |
| "loss": 0.2948, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.0796445477950336, | |
| "grad_norm": 1.064741849899292, | |
| "learning_rate": 2.3035714285714285e-05, | |
| "loss": 0.2182, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1626941283946517, | |
| "grad_norm": 1.5456783771514893, | |
| "learning_rate": 2.0959302325581396e-05, | |
| "loss": 0.1974, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2457437089942696, | |
| "grad_norm": 1.0485806465148926, | |
| "learning_rate": 1.8882890365448504e-05, | |
| "loss": 0.2068, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3287932895938876, | |
| "grad_norm": 0.9721829891204834, | |
| "learning_rate": 1.6806478405315615e-05, | |
| "loss": 0.211, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4118428701935055, | |
| "grad_norm": 0.15295402705669403, | |
| "learning_rate": 1.4730066445182724e-05, | |
| "loss": 0.2258, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.4948924507931234, | |
| "grad_norm": 1.4515725374221802, | |
| "learning_rate": 1.2657807308970101e-05, | |
| "loss": 0.2239, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.5779420313927415, | |
| "grad_norm": 1.0659480094909668, | |
| "learning_rate": 1.058139534883721e-05, | |
| "loss": 0.2222, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.6609916119923596, | |
| "grad_norm": 2.279104232788086, | |
| "learning_rate": 8.50498338870432e-06, | |
| "loss": 0.232, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.7440411925919774, | |
| "grad_norm": 0.7458949089050293, | |
| "learning_rate": 6.428571428571429e-06, | |
| "loss": 0.2378, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.8270907731915953, | |
| "grad_norm": 1.7988941669464111, | |
| "learning_rate": 4.352159468438539e-06, | |
| "loss": 0.223, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.9101403537912134, | |
| "grad_norm": 0.9544153213500977, | |
| "learning_rate": 2.275747508305648e-06, | |
| "loss": 0.2448, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.9931899343908315, | |
| "grad_norm": 2.081665277481079, | |
| "learning_rate": 1.9933554817275746e-07, | |
| "loss": 0.2509, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.9998339008388006, | |
| "step": 12040, | |
| "total_flos": 1.2491757048924733e+18, | |
| "train_loss": 0.2651168168185161, | |
| "train_runtime": 32469.1467, | |
| "train_samples_per_second": 2.967, | |
| "train_steps_per_second": 0.371 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 12040, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 1.2491757048924733e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |