| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 830, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.060350030175015085, | |
| "grad_norm": 0.2168818861246109, | |
| "learning_rate": 8.18181818181818e-05, | |
| "loss": 1.8651, | |
| "mean_token_accuracy": 0.614892452955246, | |
| "num_tokens": 157222.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12070006035003017, | |
| "grad_norm": 0.3058246970176697, | |
| "learning_rate": 0.00016704545454545452, | |
| "loss": 1.075, | |
| "mean_token_accuracy": 0.7393987023830414, | |
| "num_tokens": 285335.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18105009052504525, | |
| "grad_norm": 0.21806401014328003, | |
| "learning_rate": 0.0002522727272727273, | |
| "loss": 0.6678, | |
| "mean_token_accuracy": 0.8195570611953735, | |
| "num_tokens": 443748.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.24140012070006034, | |
| "grad_norm": 0.42833107709884644, | |
| "learning_rate": 0.0002999887132933212, | |
| "loss": 0.5362, | |
| "mean_token_accuracy": 0.8508000493049621, | |
| "num_tokens": 570813.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.30175015087507545, | |
| "grad_norm": 0.2923378050327301, | |
| "learning_rate": 0.0002998791256978121, | |
| "loss": 0.4351, | |
| "mean_token_accuracy": 0.8748408752679825, | |
| "num_tokens": 727818.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3621001810500905, | |
| "grad_norm": 0.5519189238548279, | |
| "learning_rate": 0.0002996530399366737, | |
| "loss": 0.353, | |
| "mean_token_accuracy": 0.9003111439943313, | |
| "num_tokens": 852739.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4224502112251056, | |
| "grad_norm": 0.2830982804298401, | |
| "learning_rate": 0.00029931063174202567, | |
| "loss": 0.2802, | |
| "mean_token_accuracy": 0.9183641839027404, | |
| "num_tokens": 1009253.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4828002414001207, | |
| "grad_norm": 0.3710537254810333, | |
| "learning_rate": 0.00029885216726118104, | |
| "loss": 0.2359, | |
| "mean_token_accuracy": 0.9340034741163253, | |
| "num_tokens": 1135711.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5431502715751357, | |
| "grad_norm": 0.3129195272922516, | |
| "learning_rate": 0.00029827800284977474, | |
| "loss": 0.1936, | |
| "mean_token_accuracy": 0.9437149178981781, | |
| "num_tokens": 1294974.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6035003017501509, | |
| "grad_norm": 0.3972657322883606, | |
| "learning_rate": 0.00029758858479477575, | |
| "loss": 0.1992, | |
| "mean_token_accuracy": 0.9438802194595337, | |
| "num_tokens": 1422454.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.663850331925166, | |
| "grad_norm": 0.3286677300930023, | |
| "learning_rate": 0.0002967844489675963, | |
| "loss": 0.1467, | |
| "mean_token_accuracy": 0.9580748552083969, | |
| "num_tokens": 1578961.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.724200362100181, | |
| "grad_norm": 0.3204401135444641, | |
| "learning_rate": 0.00029586622040756957, | |
| "loss": 0.1492, | |
| "mean_token_accuracy": 0.9577940207719803, | |
| "num_tokens": 1704389.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7845503922751962, | |
| "grad_norm": 0.18775713443756104, | |
| "learning_rate": 0.0002948346128361186, | |
| "loss": 0.1163, | |
| "mean_token_accuracy": 0.9667265516519546, | |
| "num_tokens": 1863333.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8449004224502112, | |
| "grad_norm": 0.2705429494380951, | |
| "learning_rate": 0.00029369042810199416, | |
| "loss": 0.1007, | |
| "mean_token_accuracy": 0.9721928060054779, | |
| "num_tokens": 1990936.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9052504526252263, | |
| "grad_norm": 0.19949601590633392, | |
| "learning_rate": 0.0002924345555580135, | |
| "loss": 0.0936, | |
| "mean_token_accuracy": 0.9747320550680161, | |
| "num_tokens": 2147408.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9656004828002414, | |
| "grad_norm": 0.30361101031303406, | |
| "learning_rate": 0.000291067971369783, | |
| "loss": 0.0975, | |
| "mean_token_accuracy": 0.972884624004364, | |
| "num_tokens": 2273014.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.08620278537273407, | |
| "eval_mean_token_accuracy": 0.9761469618694202, | |
| "eval_num_tokens": 2354180.0, | |
| "eval_runtime": 61.6733, | |
| "eval_samples_per_second": 5.983, | |
| "eval_steps_per_second": 3.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.024140012070006, | |
| "grad_norm": 0.09305181354284286, | |
| "learning_rate": 0.0002895917377569438, | |
| "loss": 0.0907, | |
| "mean_token_accuracy": 0.974765161878055, | |
| "num_tokens": 2423410.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.0844900422450212, | |
| "grad_norm": 0.3788456916809082, | |
| "learning_rate": 0.00028800700216752875, | |
| "loss": 0.0593, | |
| "mean_token_accuracy": 0.9832174813747406, | |
| "num_tokens": 2566972.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1448400724200363, | |
| "grad_norm": 0.14668315649032593, | |
| "learning_rate": 0.00028631499638607285, | |
| "loss": 0.0762, | |
| "mean_token_accuracy": 0.9784989726543426, | |
| "num_tokens": 2708394.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.2051901025950513, | |
| "grad_norm": 0.10731452703475952, | |
| "learning_rate": 0.0002845170355761712, | |
| "loss": 0.0532, | |
| "mean_token_accuracy": 0.9849696922302246, | |
| "num_tokens": 2850550.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2655401327700664, | |
| "grad_norm": 0.12523867189884186, | |
| "learning_rate": 0.0002826145172582274, | |
| "loss": 0.0664, | |
| "mean_token_accuracy": 0.9817911142110824, | |
| "num_tokens": 2991599.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.3258901629450814, | |
| "grad_norm": 0.12591515481472015, | |
| "learning_rate": 0.00028060892022318764, | |
| "loss": 0.0483, | |
| "mean_token_accuracy": 0.9862601357698441, | |
| "num_tokens": 3132335.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3862401931200965, | |
| "grad_norm": 0.1305922567844391, | |
| "learning_rate": 0.0002785018033831051, | |
| "loss": 0.0681, | |
| "mean_token_accuracy": 0.9813536697626114, | |
| "num_tokens": 3272427.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.4465902232951118, | |
| "grad_norm": 0.15813934803009033, | |
| "learning_rate": 0.0002762948045594276, | |
| "loss": 0.0493, | |
| "mean_token_accuracy": 0.9858457553386688, | |
| "num_tokens": 3413641.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5069402534701268, | |
| "grad_norm": 0.13504762947559357, | |
| "learning_rate": 0.0002739896392099502, | |
| "loss": 0.061, | |
| "mean_token_accuracy": 0.98312191426754, | |
| "num_tokens": 3553593.0, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.567290283645142, | |
| "grad_norm": 0.11668980866670609, | |
| "learning_rate": 0.00027158809909542307, | |
| "loss": 0.0452, | |
| "mean_token_accuracy": 0.9869613242149353, | |
| "num_tokens": 3696166.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.627640313820157, | |
| "grad_norm": 0.1464385837316513, | |
| "learning_rate": 0.00026909205088685, | |
| "loss": 0.0613, | |
| "mean_token_accuracy": 0.9831733548641205, | |
| "num_tokens": 3839563.0, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.687990343995172, | |
| "grad_norm": 0.1125669926404953, | |
| "learning_rate": 0.0002665034347145612, | |
| "loss": 0.0435, | |
| "mean_token_accuracy": 0.9872170311212539, | |
| "num_tokens": 3984309.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.748340374170187, | |
| "grad_norm": 0.10721296817064285, | |
| "learning_rate": 0.000263824262660187, | |
| "loss": 0.0622, | |
| "mean_token_accuracy": 0.9823342782258987, | |
| "num_tokens": 4127163.0, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.8086904043452021, | |
| "grad_norm": 0.09685028344392776, | |
| "learning_rate": 0.0002610566171927056, | |
| "loss": 0.043, | |
| "mean_token_accuracy": 0.9875605070590973, | |
| "num_tokens": 4270459.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.8690404345202172, | |
| "grad_norm": 0.09882082790136337, | |
| "learning_rate": 0.00025820264954977976, | |
| "loss": 0.056, | |
| "mean_token_accuracy": 0.984576758146286, | |
| "num_tokens": 4412568.0, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.9293904646952322, | |
| "grad_norm": 0.07612240314483643, | |
| "learning_rate": 0.00025526457806564136, | |
| "loss": 0.041, | |
| "mean_token_accuracy": 0.9879619979858398, | |
| "num_tokens": 4553943.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9897404948702473, | |
| "grad_norm": 0.10256911814212799, | |
| "learning_rate": 0.00025224468644682245, | |
| "loss": 0.0496, | |
| "mean_token_accuracy": 0.9865266931056976, | |
| "num_tokens": 4687925.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.052460573613643646, | |
| "eval_mean_token_accuracy": 0.9856111246186334, | |
| "eval_num_tokens": 4708360.0, | |
| "eval_runtime": 61.6647, | |
| "eval_samples_per_second": 5.984, | |
| "eval_steps_per_second": 3.0, | |
| "step": 830 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2905, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0227038191145677e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |