| { | |
| "best_metric": 0.4903908967971802, | |
| "best_model_checkpoint": "./beans_outputs/checkpoint-621", | |
| "epoch": 50.0, | |
| "eval_steps": 500, | |
| "global_step": 1150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 3.492309093475342, | |
| "learning_rate": 1.9826086956521742e-05, | |
| "loss": 0.7305, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 4.003854751586914, | |
| "learning_rate": 1.965217391304348e-05, | |
| "loss": 0.6794, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6062992125984252, | |
| "eval_loss": 0.6559741497039795, | |
| "eval_runtime": 4.6669, | |
| "eval_samples_per_second": 54.425, | |
| "eval_steps_per_second": 0.857, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 4.675185680389404, | |
| "learning_rate": 1.947826086956522e-05, | |
| "loss": 0.6399, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 4.350035667419434, | |
| "learning_rate": 1.9304347826086957e-05, | |
| "loss": 0.6215, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7362204724409449, | |
| "eval_loss": 0.5833372473716736, | |
| "eval_runtime": 3.7107, | |
| "eval_samples_per_second": 68.451, | |
| "eval_steps_per_second": 1.078, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 2.1739130434782608, | |
| "grad_norm": 3.180147886276245, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.5964, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 3.246190309524536, | |
| "learning_rate": 1.8956521739130434e-05, | |
| "loss": 0.5784, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7598425196850394, | |
| "eval_loss": 0.5489528179168701, | |
| "eval_runtime": 4.3517, | |
| "eval_samples_per_second": 58.367, | |
| "eval_steps_per_second": 0.919, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 3.0434782608695654, | |
| "grad_norm": 2.3400914669036865, | |
| "learning_rate": 1.8782608695652175e-05, | |
| "loss": 0.5412, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 3.5264837741851807, | |
| "learning_rate": 1.8608695652173912e-05, | |
| "loss": 0.5659, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.9130434782608696, | |
| "grad_norm": 4.993140697479248, | |
| "learning_rate": 1.8434782608695653e-05, | |
| "loss": 0.5347, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7637795275590551, | |
| "eval_loss": 0.5305963754653931, | |
| "eval_runtime": 3.9321, | |
| "eval_samples_per_second": 64.596, | |
| "eval_steps_per_second": 1.017, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 2.20806622505188, | |
| "learning_rate": 1.8260869565217393e-05, | |
| "loss": 0.5086, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.782608695652174, | |
| "grad_norm": 4.256261825561523, | |
| "learning_rate": 1.808695652173913e-05, | |
| "loss": 0.5307, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7637795275590551, | |
| "eval_loss": 0.5235078930854797, | |
| "eval_runtime": 3.7141, | |
| "eval_samples_per_second": 68.389, | |
| "eval_steps_per_second": 1.077, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 2.6543545722961426, | |
| "learning_rate": 1.791304347826087e-05, | |
| "loss": 0.5085, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 5.6521739130434785, | |
| "grad_norm": 4.274487495422363, | |
| "learning_rate": 1.773913043478261e-05, | |
| "loss": 0.5391, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.5090441703796387, | |
| "eval_runtime": 4.3438, | |
| "eval_samples_per_second": 58.475, | |
| "eval_steps_per_second": 0.921, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 6.086956521739131, | |
| "grad_norm": 3.147414445877075, | |
| "learning_rate": 1.756521739130435e-05, | |
| "loss": 0.4977, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 6.521739130434782, | |
| "grad_norm": 4.254673004150391, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.5297, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 2.083784818649292, | |
| "learning_rate": 1.721739130434783e-05, | |
| "loss": 0.48, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5108471512794495, | |
| "eval_runtime": 4.1877, | |
| "eval_samples_per_second": 60.653, | |
| "eval_steps_per_second": 0.955, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 7.391304347826087, | |
| "grad_norm": 4.193545818328857, | |
| "learning_rate": 1.7043478260869566e-05, | |
| "loss": 0.4826, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 7.826086956521739, | |
| "grad_norm": 2.05076003074646, | |
| "learning_rate": 1.6869565217391307e-05, | |
| "loss": 0.473, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.5028324127197266, | |
| "eval_runtime": 3.7021, | |
| "eval_samples_per_second": 68.61, | |
| "eval_steps_per_second": 1.08, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 8.26086956521739, | |
| "grad_norm": 3.007233142852783, | |
| "learning_rate": 1.6695652173913044e-05, | |
| "loss": 0.5255, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 2.196945905685425, | |
| "learning_rate": 1.6521739130434785e-05, | |
| "loss": 0.5014, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5054498314857483, | |
| "eval_runtime": 4.225, | |
| "eval_samples_per_second": 60.119, | |
| "eval_steps_per_second": 0.947, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 9.130434782608695, | |
| "grad_norm": 2.184353828430176, | |
| "learning_rate": 1.6347826086956525e-05, | |
| "loss": 0.5044, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 9.565217391304348, | |
| "grad_norm": 4.106619358062744, | |
| "learning_rate": 1.6173913043478262e-05, | |
| "loss": 0.4822, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 4.000082969665527, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.496, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5039955973625183, | |
| "eval_runtime": 5.3498, | |
| "eval_samples_per_second": 47.478, | |
| "eval_steps_per_second": 0.748, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 5.726933002471924, | |
| "learning_rate": 1.582608695652174e-05, | |
| "loss": 0.5101, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 10.869565217391305, | |
| "grad_norm": 4.100568771362305, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.4688, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.4972316324710846, | |
| "eval_runtime": 3.7607, | |
| "eval_samples_per_second": 67.54, | |
| "eval_steps_per_second": 1.064, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 11.304347826086957, | |
| "grad_norm": 2.6119587421417236, | |
| "learning_rate": 1.5478260869565217e-05, | |
| "loss": 0.485, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 11.73913043478261, | |
| "grad_norm": 3.003861427307129, | |
| "learning_rate": 1.5304347826086958e-05, | |
| "loss": 0.4943, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7637795275590551, | |
| "eval_loss": 0.49771231412887573, | |
| "eval_runtime": 4.9203, | |
| "eval_samples_per_second": 51.622, | |
| "eval_steps_per_second": 0.813, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 2.9490270614624023, | |
| "learning_rate": 1.5130434782608697e-05, | |
| "loss": 0.4505, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 12.608695652173914, | |
| "grad_norm": 2.8131847381591797, | |
| "learning_rate": 1.4956521739130436e-05, | |
| "loss": 0.5012, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5057242512702942, | |
| "eval_runtime": 3.7024, | |
| "eval_samples_per_second": 68.605, | |
| "eval_steps_per_second": 1.08, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 13.043478260869565, | |
| "grad_norm": 17.65978240966797, | |
| "learning_rate": 1.4782608695652174e-05, | |
| "loss": 0.4768, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 13.478260869565217, | |
| "grad_norm": 2.085587978363037, | |
| "learning_rate": 1.4608695652173915e-05, | |
| "loss": 0.4729, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 4.59744119644165, | |
| "learning_rate": 1.4434782608695654e-05, | |
| "loss": 0.4639, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5010089874267578, | |
| "eval_runtime": 3.7018, | |
| "eval_samples_per_second": 68.616, | |
| "eval_steps_per_second": 1.081, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 14.347826086956522, | |
| "grad_norm": 2.4057395458221436, | |
| "learning_rate": 1.4260869565217392e-05, | |
| "loss": 0.4751, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 14.782608695652174, | |
| "grad_norm": 3.549567222595215, | |
| "learning_rate": 1.4086956521739133e-05, | |
| "loss": 0.4709, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.4948899447917938, | |
| "eval_runtime": 4.9714, | |
| "eval_samples_per_second": 51.092, | |
| "eval_steps_per_second": 0.805, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 15.217391304347826, | |
| "grad_norm": 6.705427646636963, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.4379, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 15.652173913043478, | |
| "grad_norm": 2.444533348083496, | |
| "learning_rate": 1.373913043478261e-05, | |
| "loss": 0.4888, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7834645669291339, | |
| "eval_loss": 0.49550917744636536, | |
| "eval_runtime": 3.6768, | |
| "eval_samples_per_second": 69.081, | |
| "eval_steps_per_second": 1.088, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 16.08695652173913, | |
| "grad_norm": 5.470461845397949, | |
| "learning_rate": 1.3565217391304348e-05, | |
| "loss": 0.4952, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 16.52173913043478, | |
| "grad_norm": 2.0678608417510986, | |
| "learning_rate": 1.3391304347826088e-05, | |
| "loss": 0.4784, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 16.956521739130434, | |
| "grad_norm": 6.63480806350708, | |
| "learning_rate": 1.3217391304347827e-05, | |
| "loss": 0.4594, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.49856194853782654, | |
| "eval_runtime": 3.7219, | |
| "eval_samples_per_second": 68.245, | |
| "eval_steps_per_second": 1.075, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 4.448991298675537, | |
| "learning_rate": 1.3043478260869566e-05, | |
| "loss": 0.4607, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 17.82608695652174, | |
| "grad_norm": 2.716780424118042, | |
| "learning_rate": 1.2869565217391305e-05, | |
| "loss": 0.4745, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.501070499420166, | |
| "eval_runtime": 4.5054, | |
| "eval_samples_per_second": 56.377, | |
| "eval_steps_per_second": 0.888, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 18.26086956521739, | |
| "grad_norm": 2.406355857849121, | |
| "learning_rate": 1.2695652173913045e-05, | |
| "loss": 0.4639, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 18.695652173913043, | |
| "grad_norm": 5.627669811248779, | |
| "learning_rate": 1.2521739130434784e-05, | |
| "loss": 0.4667, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.4928434491157532, | |
| "eval_runtime": 4.0475, | |
| "eval_samples_per_second": 62.756, | |
| "eval_steps_per_second": 0.988, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 19.130434782608695, | |
| "grad_norm": 4.074652671813965, | |
| "learning_rate": 1.2347826086956523e-05, | |
| "loss": 0.4671, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 19.565217391304348, | |
| "grad_norm": 5.88148832321167, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.4442, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 3.00347900390625, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.4551, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.5055357217788696, | |
| "eval_runtime": 3.6885, | |
| "eval_samples_per_second": 68.862, | |
| "eval_steps_per_second": 1.084, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 20.434782608695652, | |
| "grad_norm": 10.164237976074219, | |
| "learning_rate": 1.182608695652174e-05, | |
| "loss": 0.4657, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 2.1962711811065674, | |
| "learning_rate": 1.1652173913043478e-05, | |
| "loss": 0.4657, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.4928124248981476, | |
| "eval_runtime": 4.4478, | |
| "eval_samples_per_second": 57.107, | |
| "eval_steps_per_second": 0.899, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 21.304347826086957, | |
| "grad_norm": 5.0302228927612305, | |
| "learning_rate": 1.1478260869565218e-05, | |
| "loss": 0.4564, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 21.73913043478261, | |
| "grad_norm": 3.5275819301605225, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.4818, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.5001721978187561, | |
| "eval_runtime": 4.0355, | |
| "eval_samples_per_second": 62.942, | |
| "eval_steps_per_second": 0.991, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 22.17391304347826, | |
| "grad_norm": 6.920666694641113, | |
| "learning_rate": 1.1130434782608696e-05, | |
| "loss": 0.4608, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 22.608695652173914, | |
| "grad_norm": 2.2840707302093506, | |
| "learning_rate": 1.0956521739130435e-05, | |
| "loss": 0.4633, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.7834645669291339, | |
| "eval_loss": 0.49459317326545715, | |
| "eval_runtime": 3.7179, | |
| "eval_samples_per_second": 68.319, | |
| "eval_steps_per_second": 1.076, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 23.043478260869566, | |
| "grad_norm": 6.509201526641846, | |
| "learning_rate": 1.0782608695652175e-05, | |
| "loss": 0.4694, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 23.47826086956522, | |
| "grad_norm": 2.403275489807129, | |
| "learning_rate": 1.0608695652173914e-05, | |
| "loss": 0.4874, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 23.91304347826087, | |
| "grad_norm": 2.1320598125457764, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.4779, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.49417200684547424, | |
| "eval_runtime": 4.3215, | |
| "eval_samples_per_second": 58.776, | |
| "eval_steps_per_second": 0.926, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 24.347826086956523, | |
| "grad_norm": 3.7421488761901855, | |
| "learning_rate": 1.0260869565217393e-05, | |
| "loss": 0.4579, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 24.782608695652176, | |
| "grad_norm": 3.07060170173645, | |
| "learning_rate": 1.008695652173913e-05, | |
| "loss": 0.4718, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.7834645669291339, | |
| "eval_loss": 0.49625155329704285, | |
| "eval_runtime": 5.5612, | |
| "eval_samples_per_second": 45.674, | |
| "eval_steps_per_second": 0.719, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 25.217391304347824, | |
| "grad_norm": 4.446998596191406, | |
| "learning_rate": 9.913043478260871e-06, | |
| "loss": 0.443, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 25.652173913043477, | |
| "grad_norm": 2.4786624908447266, | |
| "learning_rate": 9.73913043478261e-06, | |
| "loss": 0.4511, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5011107325553894, | |
| "eval_runtime": 3.7637, | |
| "eval_samples_per_second": 67.487, | |
| "eval_steps_per_second": 1.063, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 26.08695652173913, | |
| "grad_norm": 5.552999019622803, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.4631, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 26.52173913043478, | |
| "grad_norm": 5.050811290740967, | |
| "learning_rate": 9.391304347826087e-06, | |
| "loss": 0.4564, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 26.956521739130434, | |
| "grad_norm": 2.0711512565612793, | |
| "learning_rate": 9.217391304347826e-06, | |
| "loss": 0.4798, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.7874015748031497, | |
| "eval_loss": 0.4903908967971802, | |
| "eval_runtime": 4.9056, | |
| "eval_samples_per_second": 51.777, | |
| "eval_steps_per_second": 0.815, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 27.391304347826086, | |
| "grad_norm": 4.117509365081787, | |
| "learning_rate": 9.043478260869565e-06, | |
| "loss": 0.4411, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 27.82608695652174, | |
| "grad_norm": 4.448685646057129, | |
| "learning_rate": 8.869565217391306e-06, | |
| "loss": 0.4868, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.7834645669291339, | |
| "eval_loss": 0.4982087016105652, | |
| "eval_runtime": 3.7322, | |
| "eval_samples_per_second": 68.057, | |
| "eval_steps_per_second": 1.072, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 28.26086956521739, | |
| "grad_norm": 3.0993807315826416, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.4414, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 28.695652173913043, | |
| "grad_norm": 4.982347011566162, | |
| "learning_rate": 8.521739130434783e-06, | |
| "loss": 0.4653, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.7874015748031497, | |
| "eval_loss": 0.498798668384552, | |
| "eval_runtime": 3.7347, | |
| "eval_samples_per_second": 68.012, | |
| "eval_steps_per_second": 1.071, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 29.130434782608695, | |
| "grad_norm": 3.081833600997925, | |
| "learning_rate": 8.347826086956522e-06, | |
| "loss": 0.4503, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 29.565217391304348, | |
| "grad_norm": 4.352429389953613, | |
| "learning_rate": 8.173913043478263e-06, | |
| "loss": 0.4674, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 5.281393051147461, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4613, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.49851593375205994, | |
| "eval_runtime": 4.8766, | |
| "eval_samples_per_second": 52.085, | |
| "eval_steps_per_second": 0.82, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 30.434782608695652, | |
| "grad_norm": 2.2079997062683105, | |
| "learning_rate": 7.82608695652174e-06, | |
| "loss": 0.4574, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 30.869565217391305, | |
| "grad_norm": 4.6935858726501465, | |
| "learning_rate": 7.652173913043479e-06, | |
| "loss": 0.4675, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5060083270072937, | |
| "eval_runtime": 3.7305, | |
| "eval_samples_per_second": 68.087, | |
| "eval_steps_per_second": 1.072, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 31.304347826086957, | |
| "grad_norm": 4.8790602684021, | |
| "learning_rate": 7.478260869565218e-06, | |
| "loss": 0.4802, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 31.73913043478261, | |
| "grad_norm": 5.6365485191345215, | |
| "learning_rate": 7.304347826086957e-06, | |
| "loss": 0.4587, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5059147477149963, | |
| "eval_runtime": 3.7699, | |
| "eval_samples_per_second": 67.376, | |
| "eval_steps_per_second": 1.061, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 32.17391304347826, | |
| "grad_norm": 5.480165004730225, | |
| "learning_rate": 7.130434782608696e-06, | |
| "loss": 0.4541, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 32.608695652173914, | |
| "grad_norm": 2.053098440170288, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.464, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.5041583180427551, | |
| "eval_runtime": 4.9981, | |
| "eval_samples_per_second": 50.82, | |
| "eval_steps_per_second": 0.8, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 33.04347826086956, | |
| "grad_norm": 3.6429481506347656, | |
| "learning_rate": 6.782608695652174e-06, | |
| "loss": 0.454, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 33.47826086956522, | |
| "grad_norm": 2.436143636703491, | |
| "learning_rate": 6.6086956521739135e-06, | |
| "loss": 0.4612, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 33.91304347826087, | |
| "grad_norm": 2.5793776512145996, | |
| "learning_rate": 6.434782608695652e-06, | |
| "loss": 0.4374, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.5063456296920776, | |
| "eval_runtime": 3.7117, | |
| "eval_samples_per_second": 68.432, | |
| "eval_steps_per_second": 1.078, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 34.34782608695652, | |
| "grad_norm": 3.71374773979187, | |
| "learning_rate": 6.260869565217392e-06, | |
| "loss": 0.4667, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 34.78260869565217, | |
| "grad_norm": 4.282368183135986, | |
| "learning_rate": 6.086956521739132e-06, | |
| "loss": 0.4864, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.5039507150650024, | |
| "eval_runtime": 3.6837, | |
| "eval_samples_per_second": 68.952, | |
| "eval_steps_per_second": 1.086, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 35.21739130434783, | |
| "grad_norm": 2.896638870239258, | |
| "learning_rate": 5.91304347826087e-06, | |
| "loss": 0.4922, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 35.65217391304348, | |
| "grad_norm": 2.2342097759246826, | |
| "learning_rate": 5.739130434782609e-06, | |
| "loss": 0.4354, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5108994841575623, | |
| "eval_runtime": 4.9899, | |
| "eval_samples_per_second": 50.902, | |
| "eval_steps_per_second": 0.802, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 36.08695652173913, | |
| "grad_norm": 8.385408401489258, | |
| "learning_rate": 5.565217391304348e-06, | |
| "loss": 0.4585, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 36.52173913043478, | |
| "grad_norm": 2.839411497116089, | |
| "learning_rate": 5.391304347826088e-06, | |
| "loss": 0.4497, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 36.95652173913044, | |
| "grad_norm": 2.479076623916626, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.4655, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.510716438293457, | |
| "eval_runtime": 3.6997, | |
| "eval_samples_per_second": 68.653, | |
| "eval_steps_per_second": 1.081, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 37.391304347826086, | |
| "grad_norm": 2.271686553955078, | |
| "learning_rate": 5.043478260869565e-06, | |
| "loss": 0.4462, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 37.82608695652174, | |
| "grad_norm": 3.4210402965545654, | |
| "learning_rate": 4.869565217391305e-06, | |
| "loss": 0.4691, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.5093376636505127, | |
| "eval_runtime": 3.7287, | |
| "eval_samples_per_second": 68.119, | |
| "eval_steps_per_second": 1.073, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 38.26086956521739, | |
| "grad_norm": 5.694761276245117, | |
| "learning_rate": 4.695652173913044e-06, | |
| "loss": 0.4592, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 38.69565217391305, | |
| "grad_norm": 2.2949883937835693, | |
| "learning_rate": 4.5217391304347826e-06, | |
| "loss": 0.4826, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5044277906417847, | |
| "eval_runtime": 4.9781, | |
| "eval_samples_per_second": 51.024, | |
| "eval_steps_per_second": 0.804, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 39.130434782608695, | |
| "grad_norm": 3.4144210815429688, | |
| "learning_rate": 4.347826086956522e-06, | |
| "loss": 0.4407, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 39.56521739130435, | |
| "grad_norm": 2.22868013381958, | |
| "learning_rate": 4.173913043478261e-06, | |
| "loss": 0.4482, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 3.2193689346313477, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.4577, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.4999626874923706, | |
| "eval_runtime": 3.6952, | |
| "eval_samples_per_second": 68.738, | |
| "eval_steps_per_second": 1.082, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 40.43478260869565, | |
| "grad_norm": 4.500718593597412, | |
| "learning_rate": 3.8260869565217395e-06, | |
| "loss": 0.4585, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 40.869565217391305, | |
| "grad_norm": 1.9281222820281982, | |
| "learning_rate": 3.6521739130434787e-06, | |
| "loss": 0.4636, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.4962589144706726, | |
| "eval_runtime": 3.6977, | |
| "eval_samples_per_second": 68.69, | |
| "eval_steps_per_second": 1.082, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 41.30434782608695, | |
| "grad_norm": 2.193452835083008, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 0.4306, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 41.73913043478261, | |
| "grad_norm": 2.2370336055755615, | |
| "learning_rate": 3.3043478260869567e-06, | |
| "loss": 0.4361, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.4958040118217468, | |
| "eval_runtime": 4.9548, | |
| "eval_samples_per_second": 51.264, | |
| "eval_steps_per_second": 0.807, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 42.17391304347826, | |
| "grad_norm": 3.6354355812072754, | |
| "learning_rate": 3.130434782608696e-06, | |
| "loss": 0.4514, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 42.608695652173914, | |
| "grad_norm": 1.8955118656158447, | |
| "learning_rate": 2.956521739130435e-06, | |
| "loss": 0.4534, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.5007808208465576, | |
| "eval_runtime": 3.7121, | |
| "eval_samples_per_second": 68.424, | |
| "eval_steps_per_second": 1.078, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 43.04347826086956, | |
| "grad_norm": 2.2034902572631836, | |
| "learning_rate": 2.782608695652174e-06, | |
| "loss": 0.4176, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 43.47826086956522, | |
| "grad_norm": 4.387076377868652, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 0.4748, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 43.91304347826087, | |
| "grad_norm": 5.444644927978516, | |
| "learning_rate": 2.4347826086956525e-06, | |
| "loss": 0.4559, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.5025174021720886, | |
| "eval_runtime": 3.7093, | |
| "eval_samples_per_second": 68.476, | |
| "eval_steps_per_second": 1.078, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 44.34782608695652, | |
| "grad_norm": 2.2067017555236816, | |
| "learning_rate": 2.2608695652173913e-06, | |
| "loss": 0.4882, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 44.78260869565217, | |
| "grad_norm": 3.562736988067627, | |
| "learning_rate": 2.0869565217391305e-06, | |
| "loss": 0.4189, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.5014046430587769, | |
| "eval_runtime": 4.9992, | |
| "eval_samples_per_second": 50.808, | |
| "eval_steps_per_second": 0.8, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 45.21739130434783, | |
| "grad_norm": 10.402663230895996, | |
| "learning_rate": 1.9130434782608697e-06, | |
| "loss": 0.4432, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 45.65217391304348, | |
| "grad_norm": 4.949878215789795, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 0.4861, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_accuracy": 0.7677165354330708, | |
| "eval_loss": 0.5003762245178223, | |
| "eval_runtime": 3.7019, | |
| "eval_samples_per_second": 68.614, | |
| "eval_steps_per_second": 1.081, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 46.08695652173913, | |
| "grad_norm": 1.938593864440918, | |
| "learning_rate": 1.565217391304348e-06, | |
| "loss": 0.4326, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 46.52173913043478, | |
| "grad_norm": 3.236699342727661, | |
| "learning_rate": 1.391304347826087e-06, | |
| "loss": 0.4726, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 46.95652173913044, | |
| "grad_norm": 3.047184944152832, | |
| "learning_rate": 1.2173913043478262e-06, | |
| "loss": 0.4709, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "eval_accuracy": 0.7795275590551181, | |
| "eval_loss": 0.5004997849464417, | |
| "eval_runtime": 3.7143, | |
| "eval_samples_per_second": 68.384, | |
| "eval_steps_per_second": 1.077, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 47.391304347826086, | |
| "grad_norm": 2.8639461994171143, | |
| "learning_rate": 1.0434782608695653e-06, | |
| "loss": 0.4649, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 47.82608695652174, | |
| "grad_norm": 3.7704715728759766, | |
| "learning_rate": 8.695652173913044e-07, | |
| "loss": 0.4726, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_accuracy": 0.7716535433070866, | |
| "eval_loss": 0.5007592439651489, | |
| "eval_runtime": 4.8498, | |
| "eval_samples_per_second": 52.373, | |
| "eval_steps_per_second": 0.825, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 48.26086956521739, | |
| "grad_norm": 4.941337585449219, | |
| "learning_rate": 6.956521739130435e-07, | |
| "loss": 0.4314, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 48.69565217391305, | |
| "grad_norm": 3.2265655994415283, | |
| "learning_rate": 5.217391304347826e-07, | |
| "loss": 0.4441, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.4987953305244446, | |
| "eval_runtime": 3.6681, | |
| "eval_samples_per_second": 69.246, | |
| "eval_steps_per_second": 1.09, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 49.130434782608695, | |
| "grad_norm": 3.7678611278533936, | |
| "learning_rate": 3.4782608695652175e-07, | |
| "loss": 0.4571, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 49.56521739130435, | |
| "grad_norm": 3.657460927963257, | |
| "learning_rate": 1.7391304347826088e-07, | |
| "loss": 0.4558, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 3.096832513809204, | |
| "learning_rate": 0.0, | |
| "loss": 0.4579, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_accuracy": 0.7755905511811023, | |
| "eval_loss": 0.499985009431839, | |
| "eval_runtime": 3.8189, | |
| "eval_samples_per_second": 66.512, | |
| "eval_steps_per_second": 1.047, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 1150, | |
| "total_flos": 2.72467378584576e+17, | |
| "train_loss": 0.4791690407628598, | |
| "train_runtime": 1616.3187, | |
| "train_samples_per_second": 44.391, | |
| "train_steps_per_second": 0.711 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.72467378584576e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |