| { | |
| "best_metric": 0.6918547749519348, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-2_60-samples_config-1_auto/checkpoint-40", | |
| "epoch": 13.91304347826087, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.6925010681152344, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9774, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.8720422983169556, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.0716, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.7599945068359375, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.9547, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 1.0679796934127808, | |
| "eval_runtime": 5.4001, | |
| "eval_samples_per_second": 2.222, | |
| "eval_steps_per_second": 2.222, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.7053505778312683, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.9544, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.6913086175918579, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.9348, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.686968982219696, | |
| "learning_rate": 4e-05, | |
| "loss": 0.9627, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "eval_loss": 0.9489891529083252, | |
| "eval_runtime": 5.1633, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.5743587613105774, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.8689, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.5342193245887756, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.7842, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.5124989151954651, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.7983, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "eval_loss": 0.8335420489311218, | |
| "eval_runtime": 5.1672, | |
| "eval_samples_per_second": 2.322, | |
| "eval_steps_per_second": 2.322, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.4865790605545044, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.7147, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.47428378462791443, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6653, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 0.451888769865036, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.6484, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.7680513262748718, | |
| "eval_runtime": 5.1611, | |
| "eval_samples_per_second": 2.325, | |
| "eval_steps_per_second": 2.325, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.4443926215171814, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.6834, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.4753190279006958, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.6431, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.49259427189826965, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.5716, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.725757896900177, | |
| "eval_runtime": 5.1607, | |
| "eval_samples_per_second": 2.325, | |
| "eval_steps_per_second": 2.325, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.44385942816734314, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.5186, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.4831661283969879, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.5655, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 0.47351720929145813, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.5093, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.6976701617240906, | |
| "eval_runtime": 5.1647, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.4175882339477539, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.4491, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.461471825838089, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.4795, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.4657042920589447, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.4159, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.6918547749519348, | |
| "eval_runtime": 5.161, | |
| "eval_samples_per_second": 2.325, | |
| "eval_steps_per_second": 2.325, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.304347826086957, | |
| "grad_norm": 0.4652663767337799, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.4287, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.4810912311077118, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.3742, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.5502631068229675, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.358, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.720707893371582, | |
| "eval_runtime": 5.1647, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.521979808807373, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.3205, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.5816531181335449, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.304, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.869565217391305, | |
| "eval_loss": 0.7752299904823303, | |
| "eval_runtime": 5.1707, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.7092567682266235, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.2826, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.391304347826088, | |
| "grad_norm": 0.5844910740852356, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.2004, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.7829347848892212, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.2095, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 9.91304347826087, | |
| "eval_loss": 0.9146395325660706, | |
| "eval_runtime": 5.1605, | |
| "eval_samples_per_second": 2.325, | |
| "eval_steps_per_second": 2.325, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 10.08695652173913, | |
| "grad_norm": 0.8053140640258789, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.2218, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.7154636979103088, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.1292, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "grad_norm": 1.2132505178451538, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.1363, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 10.956521739130435, | |
| "eval_loss": 1.0753557682037354, | |
| "eval_runtime": 5.1664, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.7205609679222107, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.1054, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 11.478260869565217, | |
| "grad_norm": 0.8497998714447021, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.0847, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.8081735968589783, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.0658, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.2560852766036987, | |
| "eval_runtime": 5.1595, | |
| "eval_samples_per_second": 2.326, | |
| "eval_steps_per_second": 2.326, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 0.522749125957489, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0461, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.7770083546638489, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.0328, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "grad_norm": 0.9416211247444153, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.0395, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.4314936399459839, | |
| "eval_runtime": 5.1673, | |
| "eval_samples_per_second": 2.322, | |
| "eval_steps_per_second": 2.322, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.4895729422569275, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.0258, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 13.565217391304348, | |
| "grad_norm": 0.5921884179115295, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.0167, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.5972926616668701, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0209, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.531657099723816, | |
| "eval_runtime": 5.1639, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "step": 80, | |
| "total_flos": 2.2012891167195136e+16, | |
| "train_loss": 0.43889961135573685, | |
| "train_runtime": 953.8956, | |
| "train_samples_per_second": 2.411, | |
| "train_steps_per_second": 0.262 | |
| } | |
| ], | |
| "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": 2.2012891167195136e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |