{ "best_metric": 0.4838660955429077, "best_model_checkpoint": "V16/checkpoint-368", "epoch": 2.0, "eval_steps": 500, "global_step": 368, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03804347826086957, "grad_norm": 63.9192008972168, "learning_rate": 8.152173913043479e-07, "loss": 2.8905, "step": 7 }, { "epoch": 0.07608695652173914, "grad_norm": 46.61458206176758, "learning_rate": 2.7173913043478263e-06, "loss": 2.4535, "step": 14 }, { "epoch": 0.11413043478260869, "grad_norm": 38.13865280151367, "learning_rate": 4.619565217391304e-06, "loss": 2.3807, "step": 21 }, { "epoch": 0.15217391304347827, "grad_norm": 85.5317153930664, "learning_rate": 6.521739130434783e-06, "loss": 1.6676, "step": 28 }, { "epoch": 0.19021739130434784, "grad_norm": 23.912166595458984, "learning_rate": 8.423913043478262e-06, "loss": 1.4964, "step": 35 }, { "epoch": 0.22826086956521738, "grad_norm": 32.738983154296875, "learning_rate": 1.032608695652174e-05, "loss": 1.4033, "step": 42 }, { "epoch": 0.266304347826087, "grad_norm": 19.695199966430664, "learning_rate": 1.2228260869565218e-05, "loss": 1.0368, "step": 49 }, { "epoch": 0.30434782608695654, "grad_norm": 18.504451751708984, "learning_rate": 1.4130434782608694e-05, "loss": 1.1664, "step": 56 }, { "epoch": 0.3423913043478261, "grad_norm": 24.36909294128418, "learning_rate": 1.6032608695652173e-05, "loss": 1.1207, "step": 63 }, { "epoch": 0.3804347826086957, "grad_norm": 9.182580947875977, "learning_rate": 1.793478260869565e-05, "loss": 0.8311, "step": 70 }, { "epoch": 0.41847826086956524, "grad_norm": 27.781766891479492, "learning_rate": 1.9836956521739132e-05, "loss": 0.8243, "step": 77 }, { "epoch": 0.45652173913043476, "grad_norm": 11.38399887084961, "learning_rate": 2.173913043478261e-05, "loss": 0.6721, "step": 84 }, { "epoch": 0.4945652173913043, "grad_norm": 8.822359085083008, "learning_rate": 2.3641304347826088e-05, "loss": 0.6353, "step": 91 }, { "epoch": 0.532608695652174, "grad_norm": 12.496397018432617, "learning_rate": 2.554347826086957e-05, "loss": 0.6571, "step": 98 }, { "epoch": 0.5706521739130435, "grad_norm": 15.013715744018555, "learning_rate": 2.7445652173913044e-05, "loss": 0.6634, "step": 105 }, { "epoch": 0.6086956521739131, "grad_norm": 19.115114212036133, "learning_rate": 2.9347826086956526e-05, "loss": 0.654, "step": 112 }, { "epoch": 0.6467391304347826, "grad_norm": 8.66184139251709, "learning_rate": 3.125e-05, "loss": 0.6429, "step": 119 }, { "epoch": 0.6847826086956522, "grad_norm": 6.31260347366333, "learning_rate": 3.3152173913043475e-05, "loss": 0.673, "step": 126 }, { "epoch": 0.7228260869565217, "grad_norm": 9.37307071685791, "learning_rate": 3.505434782608696e-05, "loss": 0.5492, "step": 133 }, { "epoch": 0.7608695652173914, "grad_norm": 19.3261661529541, "learning_rate": 3.695652173913043e-05, "loss": 0.6397, "step": 140 }, { "epoch": 0.7989130434782609, "grad_norm": 15.380707740783691, "learning_rate": 3.885869565217391e-05, "loss": 0.6132, "step": 147 }, { "epoch": 0.8369565217391305, "grad_norm": 10.869492530822754, "learning_rate": 4.076086956521739e-05, "loss": 0.7432, "step": 154 }, { "epoch": 0.875, "grad_norm": 29.385499954223633, "learning_rate": 4.266304347826087e-05, "loss": 0.67, "step": 161 }, { "epoch": 0.9130434782608695, "grad_norm": 10.063030242919922, "learning_rate": 4.456521739130435e-05, "loss": 0.6507, "step": 168 }, { "epoch": 0.9510869565217391, "grad_norm": 16.420034408569336, "learning_rate": 4.646739130434783e-05, "loss": 0.629, "step": 175 }, { "epoch": 0.9891304347826086, "grad_norm": 20.766639709472656, "learning_rate": 4.836956521739131e-05, "loss": 0.7708, "step": 182 }, { "epoch": 1.0, "eval_accuracy": 0.6444444444444445, "eval_f1_macro": 0.4861887508626639, "eval_f1_micro": 0.6444444444444445, "eval_f1_weighted": 0.5445304643596568, "eval_loss": 0.7006417512893677, "eval_precision_macro": 0.6824815662344267, "eval_precision_micro": 0.6444444444444445, "eval_precision_weighted": 0.7664683447749809, "eval_recall_macro": 0.6697530864197532, "eval_recall_micro": 0.6444444444444445, "eval_recall_weighted": 0.6444444444444445, "eval_runtime": 0.8522, "eval_samples_per_second": 369.623, "eval_steps_per_second": 23.468, "step": 184 }, { "epoch": 1.0271739130434783, "grad_norm": 5.850246906280518, "learning_rate": 4.9969806763285025e-05, "loss": 0.6568, "step": 189 }, { "epoch": 1.065217391304348, "grad_norm": 13.868361473083496, "learning_rate": 4.9758454106280194e-05, "loss": 0.5564, "step": 196 }, { "epoch": 1.1032608695652173, "grad_norm": 21.7174072265625, "learning_rate": 4.9547101449275364e-05, "loss": 0.6629, "step": 203 }, { "epoch": 1.141304347826087, "grad_norm": 24.38075065612793, "learning_rate": 4.933574879227053e-05, "loss": 0.5487, "step": 210 }, { "epoch": 1.1793478260869565, "grad_norm": 19.306556701660156, "learning_rate": 4.91243961352657e-05, "loss": 0.6809, "step": 217 }, { "epoch": 1.2173913043478262, "grad_norm": 11.268362045288086, "learning_rate": 4.891304347826087e-05, "loss": 0.5441, "step": 224 }, { "epoch": 1.2554347826086958, "grad_norm": 10.799726486206055, "learning_rate": 4.870169082125604e-05, "loss": 0.7062, "step": 231 }, { "epoch": 1.2934782608695652, "grad_norm": 7.134019374847412, "learning_rate": 4.849033816425121e-05, "loss": 0.4888, "step": 238 }, { "epoch": 1.3315217391304348, "grad_norm": 18.638015747070312, "learning_rate": 4.8278985507246374e-05, "loss": 0.5752, "step": 245 }, { "epoch": 1.3695652173913042, "grad_norm": 22.29207992553711, "learning_rate": 4.806763285024155e-05, "loss": 0.4155, "step": 252 }, { "epoch": 1.4076086956521738, "grad_norm": 15.408923149108887, "learning_rate": 4.785628019323672e-05, "loss": 0.5075, "step": 259 }, { "epoch": 1.4456521739130435, "grad_norm": 11.493151664733887, "learning_rate": 4.764492753623189e-05, "loss": 0.4878, "step": 266 }, { "epoch": 1.483695652173913, "grad_norm": 26.615182876586914, "learning_rate": 4.743357487922706e-05, "loss": 0.5762, "step": 273 }, { "epoch": 1.5217391304347827, "grad_norm": 24.798564910888672, "learning_rate": 4.722222222222222e-05, "loss": 0.7204, "step": 280 }, { "epoch": 1.5597826086956523, "grad_norm": 11.025556564331055, "learning_rate": 4.701086956521739e-05, "loss": 0.4635, "step": 287 }, { "epoch": 1.5978260869565217, "grad_norm": 24.296916961669922, "learning_rate": 4.679951690821257e-05, "loss": 0.4529, "step": 294 }, { "epoch": 1.6358695652173914, "grad_norm": 16.77727699279785, "learning_rate": 4.6588164251207736e-05, "loss": 0.5625, "step": 301 }, { "epoch": 1.6739130434782608, "grad_norm": 19.565649032592773, "learning_rate": 4.63768115942029e-05, "loss": 0.6036, "step": 308 }, { "epoch": 1.7119565217391304, "grad_norm": 15.079780578613281, "learning_rate": 4.616545893719807e-05, "loss": 0.3821, "step": 315 }, { "epoch": 1.75, "grad_norm": 11.975324630737305, "learning_rate": 4.595410628019324e-05, "loss": 0.5276, "step": 322 }, { "epoch": 1.7880434782608696, "grad_norm": 8.451642036437988, "learning_rate": 4.574275362318841e-05, "loss": 0.4457, "step": 329 }, { "epoch": 1.8260869565217392, "grad_norm": 10.085076332092285, "learning_rate": 4.553140096618358e-05, "loss": 0.3253, "step": 336 }, { "epoch": 1.8641304347826086, "grad_norm": 4.127739429473877, "learning_rate": 4.5320048309178746e-05, "loss": 0.5676, "step": 343 }, { "epoch": 1.9021739130434783, "grad_norm": 11.594162940979004, "learning_rate": 4.5108695652173916e-05, "loss": 0.6035, "step": 350 }, { "epoch": 1.9402173913043477, "grad_norm": 10.497764587402344, "learning_rate": 4.4897342995169085e-05, "loss": 0.5138, "step": 357 }, { "epoch": 1.9782608695652173, "grad_norm": 8.271937370300293, "learning_rate": 4.4685990338164255e-05, "loss": 0.4024, "step": 364 }, { "epoch": 2.0, "eval_accuracy": 0.7968253968253968, "eval_f1_macro": 0.762273830650919, "eval_f1_micro": 0.7968253968253968, "eval_f1_weighted": 0.7910936557475937, "eval_loss": 0.4838660955429077, "eval_precision_macro": 0.8108958879749956, "eval_precision_micro": 0.7968253968253968, "eval_precision_weighted": 0.79479940517321, "eval_recall_macro": 0.728675645342312, "eval_recall_micro": 0.7968253968253968, "eval_recall_weighted": 0.7968253968253968, "eval_runtime": 0.8585, "eval_samples_per_second": 366.927, "eval_steps_per_second": 23.297, "step": 368 } ], "logging_steps": 7, "max_steps": 1840, "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 500, "total_flos": 193388362030080.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }