| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.0, |
| "eval_steps": 2000, |
| "global_step": 30786, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 6.496459429610862e-05, |
| "grad_norm": 8.543986320495605, |
| "learning_rate": 0.0, |
| "loss": 2.8055, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.12992918859221725, |
| "grad_norm": 4.228544235229492, |
| "learning_rate": 3.896179833679834e-06, |
| "loss": 2.2736, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.12992918859221725, |
| "eval_cosine_accuracy": 0.9490216970443726, |
| "eval_loss": 0.9151816368103027, |
| "eval_runtime": 24.6065, |
| "eval_samples_per_second": 384.249, |
| "eval_steps_per_second": 1.504, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.2598583771844345, |
| "grad_norm": 4.337920665740967, |
| "learning_rate": 7.794308731808732e-06, |
| "loss": 1.6559, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.2598583771844345, |
| "eval_cosine_accuracy": 0.9477525353431702, |
| "eval_loss": 0.8329138159751892, |
| "eval_runtime": 24.8331, |
| "eval_samples_per_second": 380.742, |
| "eval_steps_per_second": 1.49, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.38978756577665175, |
| "grad_norm": 4.304419994354248, |
| "learning_rate": 1.1690488565488565e-05, |
| "loss": 1.3098, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.38978756577665175, |
| "eval_cosine_accuracy": 0.9482813477516174, |
| "eval_loss": 0.8458878397941589, |
| "eval_runtime": 24.8269, |
| "eval_samples_per_second": 380.836, |
| "eval_steps_per_second": 1.49, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.519716754368869, |
| "grad_norm": 3.9155142307281494, |
| "learning_rate": 1.5588617463617465e-05, |
| "loss": 1.0964, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.519716754368869, |
| "eval_cosine_accuracy": 0.9482813477516174, |
| "eval_loss": 0.8095585107803345, |
| "eval_runtime": 24.7407, |
| "eval_samples_per_second": 382.164, |
| "eval_steps_per_second": 1.496, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.6496459429610862, |
| "grad_norm": 3.452397108078003, |
| "learning_rate": 1.9484797297297298e-05, |
| "loss": 0.9505, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.6496459429610862, |
| "eval_cosine_accuracy": 0.9518772959709167, |
| "eval_loss": 0.8028092980384827, |
| "eval_runtime": 24.6342, |
| "eval_samples_per_second": 383.816, |
| "eval_steps_per_second": 1.502, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.7795751315533035, |
| "grad_norm": 3.464122772216797, |
| "learning_rate": 2.33829261954262e-05, |
| "loss": 0.8452, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.7795751315533035, |
| "eval_cosine_accuracy": 0.9518772959709167, |
| "eval_loss": 0.7998226881027222, |
| "eval_runtime": 24.6439, |
| "eval_samples_per_second": 383.665, |
| "eval_steps_per_second": 1.501, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.9095043201455207, |
| "grad_norm": 3.4652106761932373, |
| "learning_rate": 2.7277156964656967e-05, |
| "loss": 0.7666, |
| "step": 14000 |
| }, |
| { |
| "epoch": 0.9095043201455207, |
| "eval_cosine_accuracy": 0.9551560282707214, |
| "eval_loss": 0.7942920923233032, |
| "eval_runtime": 24.635, |
| "eval_samples_per_second": 383.803, |
| "eval_steps_per_second": 1.502, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.039433508737738, |
| "grad_norm": 3.053743362426758, |
| "learning_rate": 2.9987422757534067e-05, |
| "loss": 1.0153, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.039433508737738, |
| "eval_cosine_accuracy": 0.9563193917274475, |
| "eval_loss": 0.8067317008972168, |
| "eval_runtime": 24.8609, |
| "eval_samples_per_second": 380.316, |
| "eval_steps_per_second": 1.488, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.1693626973299551, |
| "grad_norm": 2.9687039852142334, |
| "learning_rate": 2.9765792692770182e-05, |
| "loss": 0.6566, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.1693626973299551, |
| "eval_cosine_accuracy": 0.9573770761489868, |
| "eval_loss": 0.7977977991104126, |
| "eval_runtime": 24.5047, |
| "eval_samples_per_second": 385.844, |
| "eval_steps_per_second": 1.51, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.2992918859221725, |
| "grad_norm": 2.7921295166015625, |
| "learning_rate": 2.927119886582247e-05, |
| "loss": 0.6184, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.2992918859221725, |
| "eval_cosine_accuracy": 0.9610788226127625, |
| "eval_loss": 0.7389510273933411, |
| "eval_runtime": 24.7006, |
| "eval_samples_per_second": 382.785, |
| "eval_steps_per_second": 1.498, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.4292210745143896, |
| "grad_norm": 2.440551996231079, |
| "learning_rate": 2.8512824795403465e-05, |
| "loss": 0.589, |
| "step": 22000 |
| }, |
| { |
| "epoch": 1.4292210745143896, |
| "eval_cosine_accuracy": 0.9605499505996704, |
| "eval_loss": 0.752509593963623, |
| "eval_runtime": 24.6945, |
| "eval_samples_per_second": 382.879, |
| "eval_steps_per_second": 1.498, |
| "step": 22000 |
| }, |
| { |
| "epoch": 1.559150263106607, |
| "grad_norm": 2.729365825653076, |
| "learning_rate": 2.750525137009602e-05, |
| "loss": 0.5661, |
| "step": 24000 |
| }, |
| { |
| "epoch": 1.559150263106607, |
| "eval_cosine_accuracy": 0.9626652598381042, |
| "eval_loss": 0.7153900861740112, |
| "eval_runtime": 24.7624, |
| "eval_samples_per_second": 381.829, |
| "eval_steps_per_second": 1.494, |
| "step": 24000 |
| }, |
| { |
| "epoch": 1.6890794516988241, |
| "grad_norm": 2.3202743530273438, |
| "learning_rate": 2.6266095449088885e-05, |
| "loss": 0.5497, |
| "step": 26000 |
| }, |
| { |
| "epoch": 1.6890794516988241, |
| "eval_cosine_accuracy": 0.9619249105453491, |
| "eval_loss": 0.704952597618103, |
| "eval_runtime": 24.6441, |
| "eval_samples_per_second": 383.661, |
| "eval_steps_per_second": 1.501, |
| "step": 26000 |
| }, |
| { |
| "epoch": 1.8190086402910413, |
| "grad_norm": 2.6536314487457275, |
| "learning_rate": 2.4819475227824236e-05, |
| "loss": 0.5298, |
| "step": 28000 |
| }, |
| { |
| "epoch": 1.8190086402910413, |
| "eval_cosine_accuracy": 0.9615018367767334, |
| "eval_loss": 0.7386171817779541, |
| "eval_runtime": 24.7742, |
| "eval_samples_per_second": 381.646, |
| "eval_steps_per_second": 1.493, |
| "step": 28000 |
| }, |
| { |
| "epoch": 1.9489378288832586, |
| "grad_norm": 2.2802300453186035, |
| "learning_rate": 2.319068405376193e-05, |
| "loss": 0.5168, |
| "step": 30000 |
| }, |
| { |
| "epoch": 1.9489378288832586, |
| "eval_cosine_accuracy": 0.96308833360672, |
| "eval_loss": 0.7362696528434753, |
| "eval_runtime": 24.795, |
| "eval_samples_per_second": 381.328, |
| "eval_steps_per_second": 1.492, |
| "step": 30000 |
| } |
| ], |
| "logging_steps": 2000, |
| "max_steps": 61572, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "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": 0.0, |
| "train_batch_size": 256, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
