| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.006246769713578172, | |
| "eval_steps": 2000, | |
| "global_step": 8400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.451894760131836, | |
| "learning_rate": 1.9999999959757473e-05, | |
| "loss": 1.835, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.7373883724212646, | |
| "learning_rate": 1.9999999832252933e-05, | |
| "loss": 1.6278, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.7490854263305664, | |
| "learning_rate": 1.9999999617416517e-05, | |
| "loss": 1.6314, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 10.143038749694824, | |
| "learning_rate": 1.999999931524823e-05, | |
| "loss": 1.5416, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.783194065093994, | |
| "learning_rate": 1.999999892574807e-05, | |
| "loss": 1.5775, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.1446919441223145, | |
| "learning_rate": 1.9999998448916044e-05, | |
| "loss": 1.6922, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.6168997287750244, | |
| "learning_rate": 1.9999997884752155e-05, | |
| "loss": 1.6211, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.068266868591309, | |
| "learning_rate": 1.9999997233256404e-05, | |
| "loss": 1.6001, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.046320676803589, | |
| "learning_rate": 1.9999996494428805e-05, | |
| "loss": 1.5682, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.574249267578125, | |
| "learning_rate": 1.9999995668269356e-05, | |
| "loss": 1.5658, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.401742935180664, | |
| "learning_rate": 1.999999475956276e-05, | |
| "loss": 1.6152, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.141517162322998, | |
| "learning_rate": 1.9999993759176304e-05, | |
| "loss": 1.564, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.8213422298431396, | |
| "learning_rate": 1.9999992671458023e-05, | |
| "loss": 1.5586, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.3063032627105713, | |
| "learning_rate": 1.999999149640793e-05, | |
| "loss": 1.6118, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.5887880325317383, | |
| "learning_rate": 1.9999990234026036e-05, | |
| "loss": 1.586, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.8140385150909424, | |
| "learning_rate": 1.9999988884312347e-05, | |
| "loss": 1.6221, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.5657193660736084, | |
| "learning_rate": 1.9999987447266877e-05, | |
| "loss": 1.5533, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.193918466567993, | |
| "learning_rate": 1.9999985922889644e-05, | |
| "loss": 1.5725, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.9052414894104004, | |
| "learning_rate": 1.9999984311180655e-05, | |
| "loss": 1.5804, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.269617557525635, | |
| "learning_rate": 1.999998261213993e-05, | |
| "loss": 1.6025, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.5482230186462402, | |
| "learning_rate": 1.9999980825767474e-05, | |
| "loss": 1.5963, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.360860824584961, | |
| "learning_rate": 1.999997896164907e-05, | |
| "loss": 1.5837, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.9968528747558594, | |
| "learning_rate": 1.9999977001049872e-05, | |
| "loss": 1.5586, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.3270204067230225, | |
| "learning_rate": 1.9999974953119e-05, | |
| "loss": 1.597, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.4163918495178223, | |
| "learning_rate": 1.999997281785647e-05, | |
| "loss": 1.5405, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.7667906284332275, | |
| "learning_rate": 1.9999970595262297e-05, | |
| "loss": 1.5714, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.6416239738464355, | |
| "learning_rate": 1.9999968297103373e-05, | |
| "loss": 1.5909, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.743027210235596, | |
| "learning_rate": 1.999996590028264e-05, | |
| "loss": 1.5651, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.262922763824463, | |
| "learning_rate": 1.9999963416130326e-05, | |
| "loss": 1.6067, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.953801393508911, | |
| "learning_rate": 1.999996084464646e-05, | |
| "loss": 1.6252, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.8496161699295044, | |
| "learning_rate": 1.9999958185831053e-05, | |
| "loss": 1.5803, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.5667736530303955, | |
| "learning_rate": 1.999995543968414e-05, | |
| "loss": 1.6274, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.9622883796691895, | |
| "learning_rate": 1.9999952606205736e-05, | |
| "loss": 1.6222, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.0667927265167236, | |
| "learning_rate": 1.999994968539587e-05, | |
| "loss": 1.5443, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.8644936084747314, | |
| "learning_rate": 1.9999946677254565e-05, | |
| "loss": 1.5163, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.813724994659424, | |
| "learning_rate": 1.999994358178185e-05, | |
| "loss": 1.5569, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.9615787267684937, | |
| "learning_rate": 1.999994039897775e-05, | |
| "loss": 1.5864, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.2379679679870605, | |
| "learning_rate": 1.9999937128842296e-05, | |
| "loss": 1.6411, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.499472141265869, | |
| "learning_rate": 1.9999933771375512e-05, | |
| "loss": 1.5566, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.320197582244873, | |
| "learning_rate": 1.9999930326577432e-05, | |
| "loss": 1.5593, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.441413640975952, | |
| "learning_rate": 1.999992679444808e-05, | |
| "loss": 1.5118, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.383536338806152, | |
| "learning_rate": 1.9999923174987494e-05, | |
| "loss": 1.5533, | |
| "step": 8400 | |
| } | |
| ], | |
| "logging_steps": 200, | |
| "max_steps": 6723475, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 200, | |
| "total_flos": 1.0986977552390554e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |