| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.41379310344827586, | |
| "eval_steps": 1, | |
| "global_step": 30, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013793103448275862, | |
| "grad_norm": 6.989287853240967, | |
| "learning_rate": 2.2727272727272728e-06, | |
| "loss": 1.8926, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013793103448275862, | |
| "eval_loss": 2.144650459289551, | |
| "eval_runtime": 17.2592, | |
| "eval_samples_per_second": 1.159, | |
| "eval_steps_per_second": 0.579, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.027586206896551724, | |
| "grad_norm": 6.281332492828369, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 1.986, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.027586206896551724, | |
| "eval_loss": 2.0905685424804688, | |
| "eval_runtime": 17.6184, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 0.568, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.041379310344827586, | |
| "grad_norm": 4.347537040710449, | |
| "learning_rate": 6.818181818181818e-06, | |
| "loss": 1.9355, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.041379310344827586, | |
| "eval_loss": 1.9983774423599243, | |
| "eval_runtime": 17.5928, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 0.568, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05517241379310345, | |
| "grad_norm": 4.809764385223389, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 1.7509, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05517241379310345, | |
| "eval_loss": 1.8737837076187134, | |
| "eval_runtime": 17.6375, | |
| "eval_samples_per_second": 1.134, | |
| "eval_steps_per_second": 0.567, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 3.4548702239990234, | |
| "learning_rate": 1.1363636363636365e-05, | |
| "loss": 1.8838, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "eval_loss": 1.7746165990829468, | |
| "eval_runtime": 18.0257, | |
| "eval_samples_per_second": 1.11, | |
| "eval_steps_per_second": 0.555, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08275862068965517, | |
| "grad_norm": 3.1943702697753906, | |
| "learning_rate": 1.3636363636363637e-05, | |
| "loss": 1.7707, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.08275862068965517, | |
| "eval_loss": 1.6792665719985962, | |
| "eval_runtime": 17.7498, | |
| "eval_samples_per_second": 1.127, | |
| "eval_steps_per_second": 0.563, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09655172413793103, | |
| "grad_norm": 3.318288564682007, | |
| "learning_rate": 1.590909090909091e-05, | |
| "loss": 1.7171, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.09655172413793103, | |
| "eval_loss": 1.5874873399734497, | |
| "eval_runtime": 17.6295, | |
| "eval_samples_per_second": 1.134, | |
| "eval_steps_per_second": 0.567, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1103448275862069, | |
| "grad_norm": 3.210330009460449, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 1.5734, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1103448275862069, | |
| "eval_loss": 1.535287618637085, | |
| "eval_runtime": 17.6232, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 0.567, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.12413793103448276, | |
| "grad_norm": 3.2319107055664062, | |
| "learning_rate": 2.0454545454545457e-05, | |
| "loss": 1.7986, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.12413793103448276, | |
| "eval_loss": 1.467301607131958, | |
| "eval_runtime": 17.5824, | |
| "eval_samples_per_second": 1.138, | |
| "eval_steps_per_second": 0.569, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 2.79286789894104, | |
| "learning_rate": 2.272727272727273e-05, | |
| "loss": 1.5025, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "eval_loss": 1.3961191177368164, | |
| "eval_runtime": 18.3446, | |
| "eval_samples_per_second": 1.09, | |
| "eval_steps_per_second": 0.545, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15172413793103448, | |
| "grad_norm": 2.885422706604004, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.5477, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15172413793103448, | |
| "eval_loss": 1.3420469760894775, | |
| "eval_runtime": 17.7683, | |
| "eval_samples_per_second": 1.126, | |
| "eval_steps_per_second": 0.563, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.16551724137931034, | |
| "grad_norm": 2.7671327590942383, | |
| "learning_rate": 2.7272727272727273e-05, | |
| "loss": 1.6921, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.16551724137931034, | |
| "eval_loss": 1.3071445226669312, | |
| "eval_runtime": 17.652, | |
| "eval_samples_per_second": 1.133, | |
| "eval_steps_per_second": 0.567, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1793103448275862, | |
| "grad_norm": 2.9047963619232178, | |
| "learning_rate": 2.954545454545455e-05, | |
| "loss": 1.5365, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1793103448275862, | |
| "eval_loss": 1.2601890563964844, | |
| "eval_runtime": 17.5232, | |
| "eval_samples_per_second": 1.141, | |
| "eval_steps_per_second": 0.571, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.19310344827586207, | |
| "grad_norm": 2.6054675579071045, | |
| "learning_rate": 3.181818181818182e-05, | |
| "loss": 1.6621, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19310344827586207, | |
| "eval_loss": 1.2506535053253174, | |
| "eval_runtime": 17.6295, | |
| "eval_samples_per_second": 1.134, | |
| "eval_steps_per_second": 0.567, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 2.538036823272705, | |
| "learning_rate": 3.409090909090909e-05, | |
| "loss": 1.6763, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "eval_loss": 1.2366451025009155, | |
| "eval_runtime": 18.6172, | |
| "eval_samples_per_second": 1.074, | |
| "eval_steps_per_second": 0.537, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2206896551724138, | |
| "grad_norm": 2.5125789642333984, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.668, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2206896551724138, | |
| "eval_loss": 1.2205184698104858, | |
| "eval_runtime": 17.7529, | |
| "eval_samples_per_second": 1.127, | |
| "eval_steps_per_second": 0.563, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.23448275862068965, | |
| "grad_norm": 5.055665969848633, | |
| "learning_rate": 3.8636363636363636e-05, | |
| "loss": 1.5703, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23448275862068965, | |
| "eval_loss": 1.167407751083374, | |
| "eval_runtime": 17.5902, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 0.568, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2482758620689655, | |
| "grad_norm": 2.567411422729492, | |
| "learning_rate": 4.0909090909090915e-05, | |
| "loss": 1.4859, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2482758620689655, | |
| "eval_loss": 1.1367636919021606, | |
| "eval_runtime": 17.4832, | |
| "eval_samples_per_second": 1.144, | |
| "eval_steps_per_second": 0.572, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2620689655172414, | |
| "grad_norm": 2.3214948177337646, | |
| "learning_rate": 4.318181818181819e-05, | |
| "loss": 1.4511, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2620689655172414, | |
| "eval_loss": 1.1296402215957642, | |
| "eval_runtime": 17.6655, | |
| "eval_samples_per_second": 1.132, | |
| "eval_steps_per_second": 0.566, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 2.390448570251465, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 1.7181, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "eval_loss": 1.126497507095337, | |
| "eval_runtime": 17.9736, | |
| "eval_samples_per_second": 1.113, | |
| "eval_steps_per_second": 0.556, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2896551724137931, | |
| "grad_norm": 2.3728342056274414, | |
| "learning_rate": 4.772727272727273e-05, | |
| "loss": 1.4155, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2896551724137931, | |
| "eval_loss": 1.09345281124115, | |
| "eval_runtime": 17.8283, | |
| "eval_samples_per_second": 1.122, | |
| "eval_steps_per_second": 0.561, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.30344827586206896, | |
| "grad_norm": 2.4872097969055176, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3752, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.30344827586206896, | |
| "eval_loss": 1.0705276727676392, | |
| "eval_runtime": 17.6481, | |
| "eval_samples_per_second": 1.133, | |
| "eval_steps_per_second": 0.567, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.31724137931034485, | |
| "grad_norm": 2.953234910964966, | |
| "learning_rate": 4.999672209164081e-05, | |
| "loss": 1.4449, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.31724137931034485, | |
| "eval_loss": 1.0468412637710571, | |
| "eval_runtime": 17.6053, | |
| "eval_samples_per_second": 1.136, | |
| "eval_steps_per_second": 0.568, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3310344827586207, | |
| "grad_norm": 2.47603702545166, | |
| "learning_rate": 4.998688922613788e-05, | |
| "loss": 1.4286, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3310344827586207, | |
| "eval_loss": 1.0292497873306274, | |
| "eval_runtime": 17.5777, | |
| "eval_samples_per_second": 1.138, | |
| "eval_steps_per_second": 0.569, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 2.2879106998443604, | |
| "learning_rate": 4.997050398198977e-05, | |
| "loss": 1.5164, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "eval_loss": 1.0196115970611572, | |
| "eval_runtime": 17.9439, | |
| "eval_samples_per_second": 1.115, | |
| "eval_steps_per_second": 0.557, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3586206896551724, | |
| "grad_norm": 2.319134473800659, | |
| "learning_rate": 4.9947570655942796e-05, | |
| "loss": 1.5282, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3586206896551724, | |
| "eval_loss": 1.013381004333496, | |
| "eval_runtime": 17.7628, | |
| "eval_samples_per_second": 1.126, | |
| "eval_steps_per_second": 0.563, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3724137931034483, | |
| "grad_norm": 2.259608745574951, | |
| "learning_rate": 4.991809526186424e-05, | |
| "loss": 1.4901, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3724137931034483, | |
| "eval_loss": 1.0091207027435303, | |
| "eval_runtime": 17.619, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 0.568, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.38620689655172413, | |
| "grad_norm": 2.2252631187438965, | |
| "learning_rate": 4.988208552916535e-05, | |
| "loss": 1.5518, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.38620689655172413, | |
| "eval_loss": 1.0063353776931763, | |
| "eval_runtime": 17.4778, | |
| "eval_samples_per_second": 1.144, | |
| "eval_steps_per_second": 0.572, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.2154901027679443, | |
| "learning_rate": 4.983955090077444e-05, | |
| "loss": 1.4682, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.99261075258255, | |
| "eval_runtime": 17.4894, | |
| "eval_samples_per_second": 1.144, | |
| "eval_steps_per_second": 0.572, | |
| "step": 29 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 216, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.786199396619059e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |