| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9989594172736732, | |
| "eval_steps": 20, | |
| "global_step": 480, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04162330905306972, | |
| "grad_norm": 2.6779890060424805, | |
| "learning_rate": 0.00019828326180257511, | |
| "loss": 5.368, | |
| "mean_token_accuracy": 0.8072595901787281, | |
| "num_tokens": 161548.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04162330905306972, | |
| "eval_loss": 0.12305498868227005, | |
| "eval_mean_token_accuracy": 0.9847841203212738, | |
| "eval_num_tokens": 161548.0, | |
| "eval_runtime": 68.9772, | |
| "eval_samples_per_second": 2.9, | |
| "eval_steps_per_second": 0.725, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08324661810613944, | |
| "grad_norm": 0.5366652011871338, | |
| "learning_rate": 0.00018969957081545064, | |
| "loss": 0.1516, | |
| "mean_token_accuracy": 0.9917096085846424, | |
| "num_tokens": 323168.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08324661810613944, | |
| "eval_loss": 0.013321125879883766, | |
| "eval_mean_token_accuracy": 0.9938336944580078, | |
| "eval_num_tokens": 323168.0, | |
| "eval_runtime": 68.8811, | |
| "eval_samples_per_second": 2.904, | |
| "eval_steps_per_second": 0.726, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12486992715920915, | |
| "grad_norm": 0.4724605977535248, | |
| "learning_rate": 0.0001811158798283262, | |
| "loss": 0.0462, | |
| "mean_token_accuracy": 0.9939016968011856, | |
| "num_tokens": 484780.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12486992715920915, | |
| "eval_loss": 0.010577572509646416, | |
| "eval_mean_token_accuracy": 0.99473925948143, | |
| "eval_num_tokens": 484780.0, | |
| "eval_runtime": 68.7899, | |
| "eval_samples_per_second": 2.907, | |
| "eval_steps_per_second": 0.727, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16649323621227888, | |
| "grad_norm": 0.33190304040908813, | |
| "learning_rate": 0.00017253218884120172, | |
| "loss": 0.0388, | |
| "mean_token_accuracy": 0.9948750860989094, | |
| "num_tokens": 646431.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.16649323621227888, | |
| "eval_loss": 0.00940256379544735, | |
| "eval_mean_token_accuracy": 0.9951227140426636, | |
| "eval_num_tokens": 646431.0, | |
| "eval_runtime": 69.1687, | |
| "eval_samples_per_second": 2.891, | |
| "eval_steps_per_second": 0.723, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2081165452653486, | |
| "grad_norm": 0.20264093577861786, | |
| "learning_rate": 0.00016394849785407727, | |
| "loss": 0.0367, | |
| "mean_token_accuracy": 0.9950500458478928, | |
| "num_tokens": 808054.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2081165452653486, | |
| "eval_loss": 0.008246215991675854, | |
| "eval_mean_token_accuracy": 0.9951834440231323, | |
| "eval_num_tokens": 808054.0, | |
| "eval_runtime": 68.7059, | |
| "eval_samples_per_second": 2.911, | |
| "eval_steps_per_second": 0.728, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2497398543184183, | |
| "grad_norm": 0.30675262212753296, | |
| "learning_rate": 0.0001553648068669528, | |
| "loss": 0.0341, | |
| "mean_token_accuracy": 0.994870014488697, | |
| "num_tokens": 969623.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2497398543184183, | |
| "eval_loss": 0.00862042885273695, | |
| "eval_mean_token_accuracy": 0.994941633939743, | |
| "eval_num_tokens": 969623.0, | |
| "eval_runtime": 68.8073, | |
| "eval_samples_per_second": 2.907, | |
| "eval_steps_per_second": 0.727, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.29136316337148804, | |
| "grad_norm": 0.1850423365831375, | |
| "learning_rate": 0.00014678111587982832, | |
| "loss": 0.034, | |
| "mean_token_accuracy": 0.9949115067720413, | |
| "num_tokens": 1131222.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.29136316337148804, | |
| "eval_loss": 0.007929541170597076, | |
| "eval_mean_token_accuracy": 0.9951228404045105, | |
| "eval_num_tokens": 1131222.0, | |
| "eval_runtime": 68.7004, | |
| "eval_samples_per_second": 2.911, | |
| "eval_steps_per_second": 0.728, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.33298647242455776, | |
| "grad_norm": 0.3047815263271332, | |
| "learning_rate": 0.00013819742489270387, | |
| "loss": 0.0328, | |
| "mean_token_accuracy": 0.9950883395969867, | |
| "num_tokens": 1292839.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.33298647242455776, | |
| "eval_loss": 0.007993862964212894, | |
| "eval_mean_token_accuracy": 0.9951430022716522, | |
| "eval_num_tokens": 1292839.0, | |
| "eval_runtime": 68.1668, | |
| "eval_samples_per_second": 2.934, | |
| "eval_steps_per_second": 0.733, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.37460978147762747, | |
| "grad_norm": 0.13929370045661926, | |
| "learning_rate": 0.0001296137339055794, | |
| "loss": 0.0331, | |
| "mean_token_accuracy": 0.9951471641659737, | |
| "num_tokens": 1454417.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.37460978147762747, | |
| "eval_loss": 0.007801079656928778, | |
| "eval_mean_token_accuracy": 0.9954658925533295, | |
| "eval_num_tokens": 1454417.0, | |
| "eval_runtime": 68.0889, | |
| "eval_samples_per_second": 2.937, | |
| "eval_steps_per_second": 0.734, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4162330905306972, | |
| "grad_norm": 0.29847878217697144, | |
| "learning_rate": 0.00012103004291845495, | |
| "loss": 0.0308, | |
| "mean_token_accuracy": 0.9953480623662472, | |
| "num_tokens": 1615982.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4162330905306972, | |
| "eval_loss": 0.007313677109777927, | |
| "eval_mean_token_accuracy": 0.9954455089569092, | |
| "eval_num_tokens": 1615982.0, | |
| "eval_runtime": 68.4814, | |
| "eval_samples_per_second": 2.921, | |
| "eval_steps_per_second": 0.73, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4578563995837669, | |
| "grad_norm": 0.14924530684947968, | |
| "learning_rate": 0.00011244635193133049, | |
| "loss": 0.0298, | |
| "mean_token_accuracy": 0.9952255949378014, | |
| "num_tokens": 1777586.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4578563995837669, | |
| "eval_loss": 0.0077740405686199665, | |
| "eval_mean_token_accuracy": 0.9953651452064514, | |
| "eval_num_tokens": 1777586.0, | |
| "eval_runtime": 68.0902, | |
| "eval_samples_per_second": 2.937, | |
| "eval_steps_per_second": 0.734, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4994797086368366, | |
| "grad_norm": 0.31947973370552063, | |
| "learning_rate": 0.000103862660944206, | |
| "loss": 0.0299, | |
| "mean_token_accuracy": 0.9952647179365158, | |
| "num_tokens": 1939204.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4994797086368366, | |
| "eval_loss": 0.007334359921514988, | |
| "eval_mean_token_accuracy": 0.9954258131980896, | |
| "eval_num_tokens": 1939204.0, | |
| "eval_runtime": 68.026, | |
| "eval_samples_per_second": 2.94, | |
| "eval_steps_per_second": 0.735, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5411030176899063, | |
| "grad_norm": 0.15298807621002197, | |
| "learning_rate": 9.527896995708155e-05, | |
| "loss": 0.0305, | |
| "mean_token_accuracy": 0.9951752610504627, | |
| "num_tokens": 2100821.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5411030176899063, | |
| "eval_loss": 0.007329077925533056, | |
| "eval_mean_token_accuracy": 0.9955062341690063, | |
| "eval_num_tokens": 2100821.0, | |
| "eval_runtime": 68.3167, | |
| "eval_samples_per_second": 2.928, | |
| "eval_steps_per_second": 0.732, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5827263267429761, | |
| "grad_norm": 0.1927741914987564, | |
| "learning_rate": 8.669527896995709e-05, | |
| "loss": 0.0301, | |
| "mean_token_accuracy": 0.9952726632356643, | |
| "num_tokens": 2262372.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5827263267429761, | |
| "eval_loss": 0.007559146732091904, | |
| "eval_mean_token_accuracy": 0.9953450250625611, | |
| "eval_num_tokens": 2262372.0, | |
| "eval_runtime": 68.1437, | |
| "eval_samples_per_second": 2.935, | |
| "eval_steps_per_second": 0.734, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6243496357960457, | |
| "grad_norm": 0.15931323170661926, | |
| "learning_rate": 7.811158798283263e-05, | |
| "loss": 0.028, | |
| "mean_token_accuracy": 0.9956002615392208, | |
| "num_tokens": 2423931.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6243496357960457, | |
| "eval_loss": 0.007200200576335192, | |
| "eval_mean_token_accuracy": 0.9955260717868805, | |
| "eval_num_tokens": 2423931.0, | |
| "eval_runtime": 68.2059, | |
| "eval_samples_per_second": 2.932, | |
| "eval_steps_per_second": 0.733, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6659729448491155, | |
| "grad_norm": 0.2115289717912674, | |
| "learning_rate": 6.952789699570815e-05, | |
| "loss": 0.0274, | |
| "mean_token_accuracy": 0.9953135840594769, | |
| "num_tokens": 2585532.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6659729448491155, | |
| "eval_loss": 0.00743951927870512, | |
| "eval_mean_token_accuracy": 0.9954253661632538, | |
| "eval_num_tokens": 2585532.0, | |
| "eval_runtime": 67.978, | |
| "eval_samples_per_second": 2.942, | |
| "eval_steps_per_second": 0.736, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7075962539021852, | |
| "grad_norm": 0.19469623267650604, | |
| "learning_rate": 6.094420600858369e-05, | |
| "loss": 0.0291, | |
| "mean_token_accuracy": 0.9953210979700089, | |
| "num_tokens": 2747068.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7075962539021852, | |
| "eval_loss": 0.0072942511178553104, | |
| "eval_mean_token_accuracy": 0.9953449034690857, | |
| "eval_num_tokens": 2747068.0, | |
| "eval_runtime": 68.6035, | |
| "eval_samples_per_second": 2.915, | |
| "eval_steps_per_second": 0.729, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7492195629552549, | |
| "grad_norm": 0.14993388950824738, | |
| "learning_rate": 5.2360515021459236e-05, | |
| "loss": 0.029, | |
| "mean_token_accuracy": 0.9953904427587986, | |
| "num_tokens": 2908697.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7492195629552549, | |
| "eval_loss": 0.0072441427037119865, | |
| "eval_mean_token_accuracy": 0.9954857683181763, | |
| "eval_num_tokens": 2908697.0, | |
| "eval_runtime": 68.0534, | |
| "eval_samples_per_second": 2.939, | |
| "eval_steps_per_second": 0.735, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7908428720083247, | |
| "grad_norm": 0.19807656109333038, | |
| "learning_rate": 4.377682403433477e-05, | |
| "loss": 0.0289, | |
| "mean_token_accuracy": 0.9953253343701363, | |
| "num_tokens": 3070284.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7908428720083247, | |
| "eval_loss": 0.00713132182136178, | |
| "eval_mean_token_accuracy": 0.9954454469680786, | |
| "eval_num_tokens": 3070284.0, | |
| "eval_runtime": 68.0742, | |
| "eval_samples_per_second": 2.938, | |
| "eval_steps_per_second": 0.734, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8324661810613944, | |
| "grad_norm": 0.124539814889431, | |
| "learning_rate": 3.51931330472103e-05, | |
| "loss": 0.0282, | |
| "mean_token_accuracy": 0.9952500656247139, | |
| "num_tokens": 3231892.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8324661810613944, | |
| "eval_loss": 0.007116459775716066, | |
| "eval_mean_token_accuracy": 0.995526316165924, | |
| "eval_num_tokens": 3231892.0, | |
| "eval_runtime": 68.1093, | |
| "eval_samples_per_second": 2.936, | |
| "eval_steps_per_second": 0.734, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8740894901144641, | |
| "grad_norm": 0.16263636946678162, | |
| "learning_rate": 2.660944206008584e-05, | |
| "loss": 0.0287, | |
| "mean_token_accuracy": 0.9955162301659584, | |
| "num_tokens": 3393525.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8740894901144641, | |
| "eval_loss": 0.007074729073792696, | |
| "eval_mean_token_accuracy": 0.9955260527133941, | |
| "eval_num_tokens": 3393525.0, | |
| "eval_runtime": 68.1728, | |
| "eval_samples_per_second": 2.934, | |
| "eval_steps_per_second": 0.733, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9157127991675338, | |
| "grad_norm": 0.14117111265659332, | |
| "learning_rate": 1.8025751072961374e-05, | |
| "loss": 0.0277, | |
| "mean_token_accuracy": 0.9953380830585956, | |
| "num_tokens": 3555132.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9157127991675338, | |
| "eval_loss": 0.006947609595954418, | |
| "eval_mean_token_accuracy": 0.9955261528491974, | |
| "eval_num_tokens": 3555132.0, | |
| "eval_runtime": 67.9755, | |
| "eval_samples_per_second": 2.942, | |
| "eval_steps_per_second": 0.736, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9573361082206036, | |
| "grad_norm": 0.1456322818994522, | |
| "learning_rate": 9.44206008583691e-06, | |
| "loss": 0.028, | |
| "mean_token_accuracy": 0.9955740444362163, | |
| "num_tokens": 3716673.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9573361082206036, | |
| "eval_loss": 0.006936792749911547, | |
| "eval_mean_token_accuracy": 0.9955462729930877, | |
| "eval_num_tokens": 3716673.0, | |
| "eval_runtime": 68.1211, | |
| "eval_samples_per_second": 2.936, | |
| "eval_steps_per_second": 0.734, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9989594172736732, | |
| "grad_norm": 0.17603395879268646, | |
| "learning_rate": 8.583690987124464e-07, | |
| "loss": 0.0277, | |
| "mean_token_accuracy": 0.9955169551074505, | |
| "num_tokens": 3878323.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9989594172736732, | |
| "eval_loss": 0.006938955280929804, | |
| "eval_mean_token_accuracy": 0.9956067180633545, | |
| "eval_num_tokens": 3878323.0, | |
| "eval_runtime": 67.9722, | |
| "eval_samples_per_second": 2.942, | |
| "eval_steps_per_second": 0.736, | |
| "step": 480 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 481, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 40, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.012785327107474e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |