| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.011010184420589045, | |
| "eval_steps": 9, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00011010184420589045, | |
| "eval_loss": 1.441235899925232, | |
| "eval_runtime": 1176.9455, | |
| "eval_samples_per_second": 12.997, | |
| "eval_steps_per_second": 1.625, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00033030553261767135, | |
| "grad_norm": 47.746360778808594, | |
| "learning_rate": 1.5e-05, | |
| "loss": 5.1907, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0006606110652353427, | |
| "grad_norm": 45.7549934387207, | |
| "learning_rate": 3e-05, | |
| "loss": 4.5615, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.000990916597853014, | |
| "grad_norm": 54.48334503173828, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.5264, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.000990916597853014, | |
| "eval_loss": 0.35898175835609436, | |
| "eval_runtime": 1183.3259, | |
| "eval_samples_per_second": 12.927, | |
| "eval_steps_per_second": 1.617, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0013212221304706854, | |
| "grad_norm": 23.894742965698242, | |
| "learning_rate": 4.993910125649561e-05, | |
| "loss": 1.0379, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0016515276630883566, | |
| "grad_norm": 19.3171443939209, | |
| "learning_rate": 4.962019382530521e-05, | |
| "loss": 0.5869, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.001981833195706028, | |
| "grad_norm": 13.86229419708252, | |
| "learning_rate": 4.9031542398457974e-05, | |
| "loss": 0.4021, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.001981833195706028, | |
| "eval_loss": 0.061614975333213806, | |
| "eval_runtime": 1183.6671, | |
| "eval_samples_per_second": 12.923, | |
| "eval_steps_per_second": 1.616, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0023121387283236996, | |
| "grad_norm": 8.881821632385254, | |
| "learning_rate": 4.817959636416969e-05, | |
| "loss": 0.1436, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.002642444260941371, | |
| "grad_norm": 5.539639472961426, | |
| "learning_rate": 4.707368982147318e-05, | |
| "loss": 0.0862, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.002972749793559042, | |
| "grad_norm": 12.02042293548584, | |
| "learning_rate": 4.572593931387604e-05, | |
| "loss": 0.0929, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.002972749793559042, | |
| "eval_loss": 0.014203645288944244, | |
| "eval_runtime": 1185.2725, | |
| "eval_samples_per_second": 12.906, | |
| "eval_steps_per_second": 1.614, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0033030553261767133, | |
| "grad_norm": 1.4277693033218384, | |
| "learning_rate": 4.415111107797445e-05, | |
| "loss": 0.0259, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003633360858794385, | |
| "grad_norm": 11.9326753616333, | |
| "learning_rate": 4.2366459261474933e-05, | |
| "loss": 0.1501, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.003963666391412056, | |
| "grad_norm": 13.255081176757812, | |
| "learning_rate": 4.039153688314145e-05, | |
| "loss": 0.0348, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.003963666391412056, | |
| "eval_loss": 0.010344818234443665, | |
| "eval_runtime": 1185.2904, | |
| "eval_samples_per_second": 12.906, | |
| "eval_steps_per_second": 1.614, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0042939719240297275, | |
| "grad_norm": 0.062791608273983, | |
| "learning_rate": 3.824798160583012e-05, | |
| "loss": 0.0094, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.004624277456647399, | |
| "grad_norm": 1.287218689918518, | |
| "learning_rate": 3.5959278669726935e-05, | |
| "loss": 0.0155, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.00495458298926507, | |
| "grad_norm": 0.8373615741729736, | |
| "learning_rate": 3.355050358314172e-05, | |
| "loss": 0.0018, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.00495458298926507, | |
| "eval_loss": 0.009160873480141163, | |
| "eval_runtime": 1185.2739, | |
| "eval_samples_per_second": 12.906, | |
| "eval_steps_per_second": 1.614, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.005284888521882742, | |
| "grad_norm": 0.1402716338634491, | |
| "learning_rate": 3.104804738999169e-05, | |
| "loss": 0.0432, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.005615194054500413, | |
| "grad_norm": 0.2631000578403473, | |
| "learning_rate": 2.8479327524001636e-05, | |
| "loss": 0.0085, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.005945499587118084, | |
| "grad_norm": 3.119053840637207, | |
| "learning_rate": 2.587248741756253e-05, | |
| "loss": 0.055, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.005945499587118084, | |
| "eval_loss": 0.007310016546398401, | |
| "eval_runtime": 1184.9738, | |
| "eval_samples_per_second": 12.909, | |
| "eval_steps_per_second": 1.614, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.006275805119735756, | |
| "grad_norm": 1.840917944908142, | |
| "learning_rate": 2.3256088156396868e-05, | |
| "loss": 0.0032, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.006606110652353427, | |
| "grad_norm": 4.891698360443115, | |
| "learning_rate": 2.0658795558326743e-05, | |
| "loss": 0.0493, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.006936416184971098, | |
| "grad_norm": 7.879451751708984, | |
| "learning_rate": 1.8109066104575023e-05, | |
| "loss": 0.0499, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.006936416184971098, | |
| "eval_loss": 0.0055978428572416306, | |
| "eval_runtime": 1185.2837, | |
| "eval_samples_per_second": 12.906, | |
| "eval_steps_per_second": 1.614, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.00726672171758877, | |
| "grad_norm": 1.2567558288574219, | |
| "learning_rate": 1.56348351646022e-05, | |
| "loss": 0.0473, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.007597027250206441, | |
| "grad_norm": 0.13767965137958527, | |
| "learning_rate": 1.3263210930352737e-05, | |
| "loss": 0.0037, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.007927332782824112, | |
| "grad_norm": 0.10938674211502075, | |
| "learning_rate": 1.1020177413231334e-05, | |
| "loss": 0.0007, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.007927332782824112, | |
| "eval_loss": 0.004890562035143375, | |
| "eval_runtime": 1184.4337, | |
| "eval_samples_per_second": 12.915, | |
| "eval_steps_per_second": 1.615, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.008257638315441783, | |
| "grad_norm": 0.06881111860275269, | |
| "learning_rate": 8.930309757836517e-06, | |
| "loss": 0.005, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.008587943848059455, | |
| "grad_norm": 1.8729628324508667, | |
| "learning_rate": 7.016504991533726e-06, | |
| "loss": 0.0321, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.008918249380677127, | |
| "grad_norm": 0.050177041441202164, | |
| "learning_rate": 5.299731159831953e-06, | |
| "loss": 0.0043, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.008918249380677127, | |
| "eval_loss": 0.004884020891040564, | |
| "eval_runtime": 1184.5953, | |
| "eval_samples_per_second": 12.913, | |
| "eval_steps_per_second": 1.615, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.009248554913294798, | |
| "grad_norm": 0.036499105393886566, | |
| "learning_rate": 3.798797596089351e-06, | |
| "loss": 0.002, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.009578860445912468, | |
| "grad_norm": 0.19988898932933807, | |
| "learning_rate": 2.5301488425208296e-06, | |
| "loss": 0.0006, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.00990916597853014, | |
| "grad_norm": 0.11661279201507568, | |
| "learning_rate": 1.5076844803522922e-06, | |
| "loss": 0.0005, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.00990916597853014, | |
| "eval_loss": 0.004694274626672268, | |
| "eval_runtime": 1184.8106, | |
| "eval_samples_per_second": 12.911, | |
| "eval_steps_per_second": 1.615, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.010239471511147812, | |
| "grad_norm": 0.025197148323059082, | |
| "learning_rate": 7.426068431000882e-07, | |
| "loss": 0.0007, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.010569777043765483, | |
| "grad_norm": 10.560782432556152, | |
| "learning_rate": 2.4329828146074095e-07, | |
| "loss": 0.0412, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.010900082576383155, | |
| "grad_norm": 0.04284638911485672, | |
| "learning_rate": 1.522932452260595e-08, | |
| "loss": 0.0004, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.010900082576383155, | |
| "eval_loss": 0.004690830130130053, | |
| "eval_runtime": 1184.8269, | |
| "eval_samples_per_second": 12.911, | |
| "eval_steps_per_second": 1.615, | |
| "step": 99 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 9, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5082145020536422e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |