| { | |
| "best_metric": 0.8246700763702393, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-1_60-samples_config-1_full_auto/checkpoint-46", | |
| "epoch": 14.956521739130435, | |
| "eval_steps": 500, | |
| "global_step": 86, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 1.8825360536575317, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.134, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 1.797432780265808, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.2794, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 1.3481346368789673, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.1899, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 2.0615599155426025, | |
| "eval_runtime": 5.0274, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 1.4031779766082764, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.1487, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 1.6137045621871948, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.0567, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 1.0302890539169312, | |
| "learning_rate": 4e-05, | |
| "loss": 1.9473, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "eval_loss": 1.7736454010009766, | |
| "eval_runtime": 5.0277, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 1.3873261213302612, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.9107, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.8510915637016296, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.7363, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.6782408356666565, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.6049, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "eval_loss": 1.4598970413208008, | |
| "eval_runtime": 5.0449, | |
| "eval_samples_per_second": 2.379, | |
| "eval_steps_per_second": 2.379, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.7629269361495972, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.5614, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.9661106467247009, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3718, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 0.9902026653289795, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.1934, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.0349758863449097, | |
| "eval_runtime": 5.0242, | |
| "eval_samples_per_second": 2.388, | |
| "eval_steps_per_second": 2.388, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.8432597517967224, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.0786, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.5363503694534302, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.9666, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.5397949814796448, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.8711, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.906165361404419, | |
| "eval_runtime": 5.0276, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.4831950068473816, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.8453, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.45016416907310486, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.8762, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 0.41508305072784424, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.8035, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.8637191653251648, | |
| "eval_runtime": 5.0295, | |
| "eval_samples_per_second": 2.386, | |
| "eval_steps_per_second": 2.386, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.4248453378677368, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.7614, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.4545738995075226, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.7764, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.4439631700515747, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.7366, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.8405813574790955, | |
| "eval_runtime": 5.0307, | |
| "eval_samples_per_second": 2.385, | |
| "eval_steps_per_second": 2.385, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.304347826086957, | |
| "grad_norm": 0.5056247115135193, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.7489, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.44373199343681335, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.6803, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.5359618067741394, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.6995, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.8246700763702393, | |
| "eval_runtime": 5.0242, | |
| "eval_samples_per_second": 2.388, | |
| "eval_steps_per_second": 2.388, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.5545627474784851, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.6248, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.6266648173332214, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.6613, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.869565217391305, | |
| "eval_loss": 0.8258533477783203, | |
| "eval_runtime": 5.0277, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.7139117121696472, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.6253, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.391304347826088, | |
| "grad_norm": 0.7528420686721802, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.5488, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.7976906895637512, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.5531, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 9.91304347826087, | |
| "eval_loss": 0.8317675590515137, | |
| "eval_runtime": 5.0285, | |
| "eval_samples_per_second": 2.386, | |
| "eval_steps_per_second": 2.386, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 10.08695652173913, | |
| "grad_norm": 0.7176584005355835, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.5712, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.6586626768112183, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.4544, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "grad_norm": 0.8539354205131531, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.5061, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 10.956521739130435, | |
| "eval_loss": 0.8597629070281982, | |
| "eval_runtime": 5.029, | |
| "eval_samples_per_second": 2.386, | |
| "eval_steps_per_second": 2.386, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.6799076199531555, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.4456, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 11.478260869565217, | |
| "grad_norm": 0.6904007792472839, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.3692, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.926986813545227, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.3776, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9167086482048035, | |
| "eval_runtime": 5.0287, | |
| "eval_samples_per_second": 2.386, | |
| "eval_steps_per_second": 2.386, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 0.8241488337516785, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.3262, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.8996691703796387, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.3029, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "grad_norm": 1.148189663887024, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.264, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.0157562494277954, | |
| "eval_runtime": 5.0275, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.9193633794784546, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.2267, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 13.565217391304348, | |
| "grad_norm": 1.1552317142486572, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.1833, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.9492861032485962, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.1878, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.1250226497650146, | |
| "eval_runtime": 5.0348, | |
| "eval_samples_per_second": 2.383, | |
| "eval_steps_per_second": 2.383, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.26086956521739, | |
| "grad_norm": 1.2470932006835938, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.1453, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 1.0260671377182007, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.126, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "grad_norm": 1.2034170627593994, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.1417, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.203776240348816, | |
| "eval_runtime": 5.0321, | |
| "eval_samples_per_second": 2.385, | |
| "eval_steps_per_second": 2.385, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "step": 86, | |
| "total_flos": 1.1330036965572608e+16, | |
| "train_loss": 0.8706800944150187, | |
| "train_runtime": 834.1495, | |
| "train_samples_per_second": 2.757, | |
| "train_steps_per_second": 0.3 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1330036965572608e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |