| { | |
| "best_metric": 0.6120218634605408, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/microsoft/phi_2_ledgar/checkpoint-2800", | |
| "epoch": 3.0, | |
| "eval_steps": 100, | |
| "global_step": 2814, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 278.5504455566406, | |
| "learning_rate": 4.9555792466240235e-06, | |
| "loss": 4.8556, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 332.1685485839844, | |
| "learning_rate": 4.911158493248046e-06, | |
| "loss": 4.2381, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 238.07369995117188, | |
| "learning_rate": 4.866737739872069e-06, | |
| "loss": 4.0038, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 152.84237670898438, | |
| "learning_rate": 4.822316986496091e-06, | |
| "loss": 3.6034, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.337, | |
| "eval_f1_macro": 0.12358690595802334, | |
| "eval_f1_micro": 0.337, | |
| "eval_loss": 3.211400032043457, | |
| "eval_runtime": 47.3513, | |
| "eval_samples_per_second": 211.188, | |
| "eval_steps_per_second": 3.316, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 143.5316619873047, | |
| "learning_rate": 4.777896233120114e-06, | |
| "loss": 3.1831, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 152.6884765625, | |
| "learning_rate": 4.733475479744136e-06, | |
| "loss": 2.9609, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 116.26598358154297, | |
| "learning_rate": 4.6890547263681595e-06, | |
| "loss": 2.4773, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 102.64595031738281, | |
| "learning_rate": 4.644633972992183e-06, | |
| "loss": 2.2678, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_accuracy": 0.5623, | |
| "eval_f1_macro": 0.3331251034728065, | |
| "eval_f1_micro": 0.5623, | |
| "eval_loss": 1.9837249517440796, | |
| "eval_runtime": 47.4491, | |
| "eval_samples_per_second": 210.752, | |
| "eval_steps_per_second": 3.309, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 107.59564208984375, | |
| "learning_rate": 4.600213219616206e-06, | |
| "loss": 2.0387, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 129.82203674316406, | |
| "learning_rate": 4.555792466240228e-06, | |
| "loss": 1.7772, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 106.58737182617188, | |
| "learning_rate": 4.51137171286425e-06, | |
| "loss": 1.5959, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 219.3987274169922, | |
| "learning_rate": 4.466950959488273e-06, | |
| "loss": 1.4927, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.6712, | |
| "eval_f1_macro": 0.48839865479998884, | |
| "eval_f1_micro": 0.6712, | |
| "eval_loss": 1.3369125127792358, | |
| "eval_runtime": 47.4603, | |
| "eval_samples_per_second": 210.702, | |
| "eval_steps_per_second": 3.308, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 62.788822174072266, | |
| "learning_rate": 4.422530206112296e-06, | |
| "loss": 1.4203, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 63.6900634765625, | |
| "learning_rate": 4.378109452736319e-06, | |
| "loss": 1.2841, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 75.33383178710938, | |
| "learning_rate": 4.333688699360342e-06, | |
| "loss": 1.1642, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 89.42117309570312, | |
| "learning_rate": 4.289267945984365e-06, | |
| "loss": 1.1518, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_accuracy": 0.7243, | |
| "eval_f1_macro": 0.5613056174383468, | |
| "eval_f1_micro": 0.7243, | |
| "eval_loss": 1.052606225013733, | |
| "eval_runtime": 47.3838, | |
| "eval_samples_per_second": 211.043, | |
| "eval_steps_per_second": 3.313, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 74.13993072509766, | |
| "learning_rate": 4.244847192608387e-06, | |
| "loss": 1.135, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 80.19258117675781, | |
| "learning_rate": 4.200426439232409e-06, | |
| "loss": 1.176, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 70.83063507080078, | |
| "learning_rate": 4.156005685856432e-06, | |
| "loss": 1.1386, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 109.83588409423828, | |
| "learning_rate": 4.1115849324804554e-06, | |
| "loss": 1.1041, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.7521, | |
| "eval_f1_macro": 0.6206030644902993, | |
| "eval_f1_micro": 0.7521, | |
| "eval_loss": 0.9304812550544739, | |
| "eval_runtime": 47.4284, | |
| "eval_samples_per_second": 210.844, | |
| "eval_steps_per_second": 3.31, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 51.40995788574219, | |
| "learning_rate": 4.067164179104478e-06, | |
| "loss": 0.9872, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 75.6723861694336, | |
| "learning_rate": 4.022743425728501e-06, | |
| "loss": 0.9832, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 86.0229721069336, | |
| "learning_rate": 3.978322672352524e-06, | |
| "loss": 0.9485, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 113.96233367919922, | |
| "learning_rate": 3.933901918976546e-06, | |
| "loss": 1.0144, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.7574, | |
| "eval_f1_macro": 0.6294250374822845, | |
| "eval_f1_micro": 0.7574, | |
| "eval_loss": 0.9067593812942505, | |
| "eval_runtime": 47.4873, | |
| "eval_samples_per_second": 210.583, | |
| "eval_steps_per_second": 3.306, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 77.80159759521484, | |
| "learning_rate": 3.889481165600569e-06, | |
| "loss": 1.0435, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 77.97108459472656, | |
| "learning_rate": 3.8450604122245914e-06, | |
| "loss": 0.9593, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 68.20707702636719, | |
| "learning_rate": 3.8006396588486145e-06, | |
| "loss": 0.9818, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 103.11054229736328, | |
| "learning_rate": 3.756218905472637e-06, | |
| "loss": 0.9892, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_accuracy": 0.7669, | |
| "eval_f1_macro": 0.642987522646621, | |
| "eval_f1_micro": 0.7669, | |
| "eval_loss": 0.8712124824523926, | |
| "eval_runtime": 47.4988, | |
| "eval_samples_per_second": 210.531, | |
| "eval_steps_per_second": 3.305, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 88.3270263671875, | |
| "learning_rate": 3.71179815209666e-06, | |
| "loss": 0.8937, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 67.907958984375, | |
| "learning_rate": 3.667377398720683e-06, | |
| "loss": 0.9788, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 68.60499572753906, | |
| "learning_rate": 3.622956645344705e-06, | |
| "loss": 0.9459, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 71.7964859008789, | |
| "learning_rate": 3.578535891968728e-06, | |
| "loss": 0.9972, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_accuracy": 0.7675, | |
| "eval_f1_macro": 0.6369038817239384, | |
| "eval_f1_micro": 0.7675, | |
| "eval_loss": 0.8591374754905701, | |
| "eval_runtime": 47.4288, | |
| "eval_samples_per_second": 210.842, | |
| "eval_steps_per_second": 3.31, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 99.17341613769531, | |
| "learning_rate": 3.534115138592751e-06, | |
| "loss": 0.9663, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 198.6553955078125, | |
| "learning_rate": 3.4896943852167736e-06, | |
| "loss": 0.8617, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 82.58099365234375, | |
| "learning_rate": 3.4452736318407963e-06, | |
| "loss": 0.8735, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 63.692623138427734, | |
| "learning_rate": 3.4008528784648194e-06, | |
| "loss": 0.8439, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.7848, | |
| "eval_f1_macro": 0.6834829058415216, | |
| "eval_f1_micro": 0.7848, | |
| "eval_loss": 0.7894874811172485, | |
| "eval_runtime": 47.4489, | |
| "eval_samples_per_second": 210.753, | |
| "eval_steps_per_second": 3.309, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 78.88030242919922, | |
| "learning_rate": 3.3564321250888416e-06, | |
| "loss": 0.8582, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 45.49142837524414, | |
| "learning_rate": 3.3120113717128643e-06, | |
| "loss": 0.8621, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 37.98048400878906, | |
| "learning_rate": 3.2675906183368874e-06, | |
| "loss": 0.8396, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 37.72966384887695, | |
| "learning_rate": 3.22316986496091e-06, | |
| "loss": 0.7409, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.7944, | |
| "eval_f1_macro": 0.6808979493082343, | |
| "eval_f1_micro": 0.7944, | |
| "eval_loss": 0.7614499926567078, | |
| "eval_runtime": 47.4823, | |
| "eval_samples_per_second": 210.605, | |
| "eval_steps_per_second": 3.306, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 57.926578521728516, | |
| "learning_rate": 3.1787491115849327e-06, | |
| "loss": 0.7475, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 86.9485855102539, | |
| "learning_rate": 3.1343283582089558e-06, | |
| "loss": 0.7197, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 54.83045959472656, | |
| "learning_rate": 3.0899076048329785e-06, | |
| "loss": 0.8314, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 68.1777114868164, | |
| "learning_rate": 3.0454868514570007e-06, | |
| "loss": 0.7627, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.7946, | |
| "eval_f1_macro": 0.6809735037021439, | |
| "eval_f1_micro": 0.7946, | |
| "eval_loss": 0.7538655996322632, | |
| "eval_runtime": 47.4181, | |
| "eval_samples_per_second": 210.89, | |
| "eval_steps_per_second": 3.311, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 31.217445373535156, | |
| "learning_rate": 3.0010660980810234e-06, | |
| "loss": 0.7501, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 38.76744842529297, | |
| "learning_rate": 2.9566453447050464e-06, | |
| "loss": 0.6884, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 50.481388092041016, | |
| "learning_rate": 2.912224591329069e-06, | |
| "loss": 0.6911, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 54.91836929321289, | |
| "learning_rate": 2.867803837953092e-06, | |
| "loss": 0.8065, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.8008, | |
| "eval_f1_macro": 0.6945223141031294, | |
| "eval_f1_micro": 0.8008, | |
| "eval_loss": 0.7289281487464905, | |
| "eval_runtime": 47.4458, | |
| "eval_samples_per_second": 210.767, | |
| "eval_steps_per_second": 3.309, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 72.01395416259766, | |
| "learning_rate": 2.823383084577115e-06, | |
| "loss": 0.8075, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 83.60379791259766, | |
| "learning_rate": 2.7789623312011375e-06, | |
| "loss": 0.8068, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 51.42044448852539, | |
| "learning_rate": 2.7345415778251598e-06, | |
| "loss": 0.7549, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 44.68650436401367, | |
| "learning_rate": 2.690120824449183e-06, | |
| "loss": 0.7359, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.8034, | |
| "eval_f1_macro": 0.6975565105528009, | |
| "eval_f1_micro": 0.8034, | |
| "eval_loss": 0.7253593802452087, | |
| "eval_runtime": 47.5217, | |
| "eval_samples_per_second": 210.43, | |
| "eval_steps_per_second": 3.304, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 56.39244842529297, | |
| "learning_rate": 2.6457000710732055e-06, | |
| "loss": 0.7201, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 70.84113311767578, | |
| "learning_rate": 2.601279317697228e-06, | |
| "loss": 0.7346, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 69.2415542602539, | |
| "learning_rate": 2.5568585643212513e-06, | |
| "loss": 0.6693, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 50.06897735595703, | |
| "learning_rate": 2.512437810945274e-06, | |
| "loss": 0.6525, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_accuracy": 0.8065, | |
| "eval_f1_macro": 0.7050317065636842, | |
| "eval_f1_micro": 0.8065, | |
| "eval_loss": 0.7072968482971191, | |
| "eval_runtime": 47.4893, | |
| "eval_samples_per_second": 210.574, | |
| "eval_steps_per_second": 3.306, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 45.797054290771484, | |
| "learning_rate": 2.4680170575692966e-06, | |
| "loss": 0.7271, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 46.304473876953125, | |
| "learning_rate": 2.4235963041933193e-06, | |
| "loss": 0.8033, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 80.502685546875, | |
| "learning_rate": 2.379175550817342e-06, | |
| "loss": 0.7464, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 89.19364166259766, | |
| "learning_rate": 2.3347547974413646e-06, | |
| "loss": 0.7359, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.8033, | |
| "eval_f1_macro": 0.6948589139841357, | |
| "eval_f1_micro": 0.8033, | |
| "eval_loss": 0.720634400844574, | |
| "eval_runtime": 47.454, | |
| "eval_samples_per_second": 210.73, | |
| "eval_steps_per_second": 3.308, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 36.966548919677734, | |
| "learning_rate": 2.2903340440653877e-06, | |
| "loss": 0.697, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 37.08382797241211, | |
| "learning_rate": 2.24591329068941e-06, | |
| "loss": 0.6736, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 37.009403228759766, | |
| "learning_rate": 2.201492537313433e-06, | |
| "loss": 0.6441, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 29.34271240234375, | |
| "learning_rate": 2.1570717839374557e-06, | |
| "loss": 0.7291, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_accuracy": 0.8089, | |
| "eval_f1_macro": 0.7065968196696667, | |
| "eval_f1_micro": 0.8089, | |
| "eval_loss": 0.6923750042915344, | |
| "eval_runtime": 47.4715, | |
| "eval_samples_per_second": 210.653, | |
| "eval_steps_per_second": 3.307, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 43.28564453125, | |
| "learning_rate": 2.112651030561479e-06, | |
| "loss": 0.7153, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 27.64020347595215, | |
| "learning_rate": 2.068230277185501e-06, | |
| "loss": 0.6721, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 46.75600051879883, | |
| "learning_rate": 2.023809523809524e-06, | |
| "loss": 0.6566, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 34.07355499267578, | |
| "learning_rate": 1.979388770433547e-06, | |
| "loss": 0.7072, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "eval_accuracy": 0.8102, | |
| "eval_f1_macro": 0.7070176051219198, | |
| "eval_f1_micro": 0.8102, | |
| "eval_loss": 0.6763781309127808, | |
| "eval_runtime": 47.526, | |
| "eval_samples_per_second": 210.411, | |
| "eval_steps_per_second": 3.303, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 43.36367416381836, | |
| "learning_rate": 1.9349680170575695e-06, | |
| "loss": 0.7289, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 34.61235046386719, | |
| "learning_rate": 1.8905472636815921e-06, | |
| "loss": 0.6787, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 44.31460952758789, | |
| "learning_rate": 1.846126510305615e-06, | |
| "loss": 0.7188, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 53.3677864074707, | |
| "learning_rate": 1.8017057569296375e-06, | |
| "loss": 0.6688, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.814, | |
| "eval_f1_macro": 0.7128286594201773, | |
| "eval_f1_micro": 0.814, | |
| "eval_loss": 0.6546062231063843, | |
| "eval_runtime": 47.5163, | |
| "eval_samples_per_second": 210.454, | |
| "eval_steps_per_second": 3.304, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 37.4492073059082, | |
| "learning_rate": 1.7572850035536603e-06, | |
| "loss": 0.7817, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 41.068538665771484, | |
| "learning_rate": 1.7128642501776832e-06, | |
| "loss": 0.7044, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 67.96875762939453, | |
| "learning_rate": 1.668443496801706e-06, | |
| "loss": 0.5948, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 51.58526611328125, | |
| "learning_rate": 1.6240227434257286e-06, | |
| "loss": 0.6253, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.8158, | |
| "eval_f1_macro": 0.7058775462036815, | |
| "eval_f1_micro": 0.8158, | |
| "eval_loss": 0.6505562663078308, | |
| "eval_runtime": 47.3282, | |
| "eval_samples_per_second": 211.29, | |
| "eval_steps_per_second": 3.317, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 45.662109375, | |
| "learning_rate": 1.5796019900497514e-06, | |
| "loss": 0.645, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 66.76183319091797, | |
| "learning_rate": 1.5351812366737743e-06, | |
| "loss": 0.6412, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 56.84629440307617, | |
| "learning_rate": 1.4907604832977968e-06, | |
| "loss": 0.6511, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 53.2021484375, | |
| "learning_rate": 1.4463397299218196e-06, | |
| "loss": 0.6044, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_accuracy": 0.8155, | |
| "eval_f1_macro": 0.716530877903978, | |
| "eval_f1_micro": 0.8155, | |
| "eval_loss": 0.6602968573570251, | |
| "eval_runtime": 47.473, | |
| "eval_samples_per_second": 210.646, | |
| "eval_steps_per_second": 3.307, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 71.7922134399414, | |
| "learning_rate": 1.4019189765458423e-06, | |
| "loss": 0.6381, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 32.50026321411133, | |
| "learning_rate": 1.357498223169865e-06, | |
| "loss": 0.6622, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 50.483219146728516, | |
| "learning_rate": 1.3130774697938879e-06, | |
| "loss": 0.6165, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 40.9278450012207, | |
| "learning_rate": 1.2686567164179105e-06, | |
| "loss": 0.6414, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_accuracy": 0.8138, | |
| "eval_f1_macro": 0.7185019905510756, | |
| "eval_f1_micro": 0.8138, | |
| "eval_loss": 0.6434906125068665, | |
| "eval_runtime": 47.5169, | |
| "eval_samples_per_second": 210.452, | |
| "eval_steps_per_second": 3.304, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 40.769439697265625, | |
| "learning_rate": 1.2242359630419332e-06, | |
| "loss": 0.619, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 23.210596084594727, | |
| "learning_rate": 1.179815209665956e-06, | |
| "loss": 0.6168, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 43.848052978515625, | |
| "learning_rate": 1.1353944562899787e-06, | |
| "loss": 0.5244, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 40.81973648071289, | |
| "learning_rate": 1.0909737029140014e-06, | |
| "loss": 0.6115, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_accuracy": 0.8216, | |
| "eval_f1_macro": 0.727953382739938, | |
| "eval_f1_micro": 0.8216, | |
| "eval_loss": 0.63681560754776, | |
| "eval_runtime": 47.4723, | |
| "eval_samples_per_second": 210.649, | |
| "eval_steps_per_second": 3.307, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 28.034944534301758, | |
| "learning_rate": 1.0465529495380243e-06, | |
| "loss": 0.5613, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 68.5925521850586, | |
| "learning_rate": 1.002132196162047e-06, | |
| "loss": 0.5987, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 30.3986759185791, | |
| "learning_rate": 9.577114427860696e-07, | |
| "loss": 0.6119, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 38.481117248535156, | |
| "learning_rate": 9.132906894100925e-07, | |
| "loss": 0.6331, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "eval_accuracy": 0.8208, | |
| "eval_f1_macro": 0.7251441527998863, | |
| "eval_f1_micro": 0.8208, | |
| "eval_loss": 0.6273249983787537, | |
| "eval_runtime": 47.475, | |
| "eval_samples_per_second": 210.637, | |
| "eval_steps_per_second": 3.307, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 52.91291427612305, | |
| "learning_rate": 8.688699360341152e-07, | |
| "loss": 0.6261, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 43.07246398925781, | |
| "learning_rate": 8.24449182658138e-07, | |
| "loss": 0.6191, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 40.49489974975586, | |
| "learning_rate": 7.800284292821607e-07, | |
| "loss": 0.5856, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 30.555904388427734, | |
| "learning_rate": 7.356076759061834e-07, | |
| "loss": 0.608, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.8232, | |
| "eval_f1_macro": 0.7285501243647127, | |
| "eval_f1_micro": 0.8232, | |
| "eval_loss": 0.6251906156539917, | |
| "eval_runtime": 47.5564, | |
| "eval_samples_per_second": 210.277, | |
| "eval_steps_per_second": 3.301, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 175.31057739257812, | |
| "learning_rate": 6.911869225302062e-07, | |
| "loss": 0.6802, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 30.15215301513672, | |
| "learning_rate": 6.467661691542289e-07, | |
| "loss": 0.6307, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 34.21811294555664, | |
| "learning_rate": 6.023454157782517e-07, | |
| "loss": 0.6029, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 56.559364318847656, | |
| "learning_rate": 5.579246624022743e-07, | |
| "loss": 0.5879, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_accuracy": 0.8241, | |
| "eval_f1_macro": 0.7307624039691639, | |
| "eval_f1_micro": 0.8241, | |
| "eval_loss": 0.6171656250953674, | |
| "eval_runtime": 47.4737, | |
| "eval_samples_per_second": 210.643, | |
| "eval_steps_per_second": 3.307, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 43.20689392089844, | |
| "learning_rate": 5.135039090262971e-07, | |
| "loss": 0.5858, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 46.331817626953125, | |
| "learning_rate": 4.690831556503199e-07, | |
| "loss": 0.5959, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 36.399810791015625, | |
| "learning_rate": 4.2466240227434256e-07, | |
| "loss": 0.6069, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 79.88939666748047, | |
| "learning_rate": 3.8024164889836533e-07, | |
| "loss": 0.6056, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.8257, | |
| "eval_f1_macro": 0.7346364945364393, | |
| "eval_f1_micro": 0.8257, | |
| "eval_loss": 0.6157156229019165, | |
| "eval_runtime": 47.4846, | |
| "eval_samples_per_second": 210.595, | |
| "eval_steps_per_second": 3.306, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 49.164161682128906, | |
| "learning_rate": 3.358208955223881e-07, | |
| "loss": 0.5675, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 46.98076248168945, | |
| "learning_rate": 2.914001421464108e-07, | |
| "loss": 0.6607, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 23.488380432128906, | |
| "learning_rate": 2.4697938877043354e-07, | |
| "loss": 0.5866, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 31.710861206054688, | |
| "learning_rate": 2.0255863539445632e-07, | |
| "loss": 0.5711, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_accuracy": 0.8253, | |
| "eval_f1_macro": 0.7341228101010588, | |
| "eval_f1_micro": 0.8253, | |
| "eval_loss": 0.6129437685012817, | |
| "eval_runtime": 47.5391, | |
| "eval_samples_per_second": 210.353, | |
| "eval_steps_per_second": 3.303, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 33.91932678222656, | |
| "learning_rate": 1.5813788201847903e-07, | |
| "loss": 0.5565, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 35.29643630981445, | |
| "learning_rate": 1.1371712864250178e-07, | |
| "loss": 0.5793, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 33.85700988769531, | |
| "learning_rate": 6.929637526652453e-08, | |
| "loss": 0.5425, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 40.65436935424805, | |
| "learning_rate": 2.4875621890547265e-08, | |
| "loss": 0.5802, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_accuracy": 0.826, | |
| "eval_f1_macro": 0.7355084015446216, | |
| "eval_f1_micro": 0.826, | |
| "eval_loss": 0.6120218634605408, | |
| "eval_runtime": 47.528, | |
| "eval_samples_per_second": 210.402, | |
| "eval_steps_per_second": 3.303, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2814, | |
| "total_flos": 3.482386379487314e+17, | |
| "train_loss": 0.9835220694796108, | |
| "train_runtime": 6489.4436, | |
| "train_samples_per_second": 27.737, | |
| "train_steps_per_second": 0.434 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2814, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "total_flos": 3.482386379487314e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |