| { | |
| "best_metric": 0.16805602610111237, | |
| "best_model_checkpoint": "./vit-handwritten-characters/checkpoint-8400", | |
| "epoch": 2.952548330404218, | |
| "global_step": 8400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9.882835383714118e-05, | |
| "loss": 2.6668, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "eval_accuracy": 0.7636050763605077, | |
| "eval_loss": 1.8537834882736206, | |
| "eval_runtime": 53.1177, | |
| "eval_samples_per_second": 87.523, | |
| "eval_steps_per_second": 2.749, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.765670767428236e-05, | |
| "loss": 1.3385, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "eval_accuracy": 0.8698644869864487, | |
| "eval_loss": 1.0306403636932373, | |
| "eval_runtime": 50.6679, | |
| "eval_samples_per_second": 91.754, | |
| "eval_steps_per_second": 2.882, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.648506151142355e-05, | |
| "loss": 0.8372, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.8657775865777586, | |
| "eval_loss": 0.7356651425361633, | |
| "eval_runtime": 51.2063, | |
| "eval_samples_per_second": 90.79, | |
| "eval_steps_per_second": 2.851, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.531341534856473e-05, | |
| "loss": 0.5838, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "eval_accuracy": 0.899118089911809, | |
| "eval_loss": 0.5292066335678101, | |
| "eval_runtime": 51.0683, | |
| "eval_samples_per_second": 91.035, | |
| "eval_steps_per_second": 2.859, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.414176918570592e-05, | |
| "loss": 0.4572, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.901699290169929, | |
| "eval_loss": 0.4536021947860718, | |
| "eval_runtime": 51.5623, | |
| "eval_samples_per_second": 90.163, | |
| "eval_steps_per_second": 2.832, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 9.29701230228471e-05, | |
| "loss": 0.3808, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_accuracy": 0.9141750914175092, | |
| "eval_loss": 0.37226173281669617, | |
| "eval_runtime": 51.6002, | |
| "eval_samples_per_second": 90.097, | |
| "eval_steps_per_second": 2.829, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.17984768599883e-05, | |
| "loss": 0.393, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_accuracy": 0.899763389976339, | |
| "eval_loss": 0.42711377143859863, | |
| "eval_runtime": 53.3282, | |
| "eval_samples_per_second": 87.177, | |
| "eval_steps_per_second": 2.738, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.062683069712947e-05, | |
| "loss": 0.3611, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_accuracy": 0.8793288879328888, | |
| "eval_loss": 0.41316691040992737, | |
| "eval_runtime": 52.5044, | |
| "eval_samples_per_second": 88.545, | |
| "eval_steps_per_second": 2.781, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 8.945518453427065e-05, | |
| "loss": 0.3436, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.9068616906861691, | |
| "eval_loss": 0.34593629837036133, | |
| "eval_runtime": 52.4712, | |
| "eval_samples_per_second": 88.601, | |
| "eval_steps_per_second": 2.782, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 8.828353837141183e-05, | |
| "loss": 0.3147, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "eval_accuracy": 0.9135297913529792, | |
| "eval_loss": 0.32545194029808044, | |
| "eval_runtime": 55.2872, | |
| "eval_samples_per_second": 84.088, | |
| "eval_steps_per_second": 2.641, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 8.711189220855302e-05, | |
| "loss": 0.2887, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_accuracy": 0.9137448913744891, | |
| "eval_loss": 0.3017228841781616, | |
| "eval_runtime": 52.7469, | |
| "eval_samples_per_second": 88.138, | |
| "eval_steps_per_second": 2.768, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 8.59402460456942e-05, | |
| "loss": 0.2768, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_accuracy": 0.9130995913099591, | |
| "eval_loss": 0.3049895465373993, | |
| "eval_runtime": 52.9848, | |
| "eval_samples_per_second": 87.742, | |
| "eval_steps_per_second": 2.756, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 8.476859988283538e-05, | |
| "loss": 0.3113, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_accuracy": 0.9167562916756292, | |
| "eval_loss": 0.2875942885875702, | |
| "eval_runtime": 54.7094, | |
| "eval_samples_per_second": 84.976, | |
| "eval_steps_per_second": 2.669, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 8.359695371997657e-05, | |
| "loss": 0.2607, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "eval_accuracy": 0.9098730909873091, | |
| "eval_loss": 0.33047187328338623, | |
| "eval_runtime": 55.3261, | |
| "eval_samples_per_second": 84.029, | |
| "eval_steps_per_second": 2.639, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 8.242530755711776e-05, | |
| "loss": 0.288, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.9098730909873091, | |
| "eval_loss": 0.3238196074962616, | |
| "eval_runtime": 53.2815, | |
| "eval_samples_per_second": 87.254, | |
| "eval_steps_per_second": 2.74, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 8.125366139425894e-05, | |
| "loss": 0.311, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_accuracy": 0.9128844912884492, | |
| "eval_loss": 0.2932077944278717, | |
| "eval_runtime": 54.1056, | |
| "eval_samples_per_second": 85.925, | |
| "eval_steps_per_second": 2.698, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 8.008201523140012e-05, | |
| "loss": 0.2702, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_accuracy": 0.9214884921488492, | |
| "eval_loss": 0.281431645154953, | |
| "eval_runtime": 55.0433, | |
| "eval_samples_per_second": 84.461, | |
| "eval_steps_per_second": 2.652, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 7.89103690685413e-05, | |
| "loss": 0.2476, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_accuracy": 0.899118089911809, | |
| "eval_loss": 0.29802632331848145, | |
| "eval_runtime": 53.6155, | |
| "eval_samples_per_second": 86.71, | |
| "eval_steps_per_second": 2.723, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 7.773872290568249e-05, | |
| "loss": 0.2833, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_accuracy": 0.9283716928371692, | |
| "eval_loss": 0.2524644136428833, | |
| "eval_runtime": 53.6383, | |
| "eval_samples_per_second": 86.673, | |
| "eval_steps_per_second": 2.722, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 7.656707674282367e-05, | |
| "loss": 0.2542, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_accuracy": 0.9247149924714992, | |
| "eval_loss": 0.2760772705078125, | |
| "eval_runtime": 53.5886, | |
| "eval_samples_per_second": 86.753, | |
| "eval_steps_per_second": 2.724, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 7.539543057996485e-05, | |
| "loss": 0.2405, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.9229941922994193, | |
| "eval_loss": 0.2586613595485687, | |
| "eval_runtime": 55.7532, | |
| "eval_samples_per_second": 83.385, | |
| "eval_steps_per_second": 2.619, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.422378441710603e-05, | |
| "loss": 0.2404, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_accuracy": 0.9184770918477092, | |
| "eval_loss": 0.25114524364471436, | |
| "eval_runtime": 54.8709, | |
| "eval_samples_per_second": 84.726, | |
| "eval_steps_per_second": 2.661, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 7.305213825424722e-05, | |
| "loss": 0.2277, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "eval_accuracy": 0.9191223919122392, | |
| "eval_loss": 0.2526193857192993, | |
| "eval_runtime": 54.7258, | |
| "eval_samples_per_second": 84.951, | |
| "eval_steps_per_second": 2.668, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 7.188049209138841e-05, | |
| "loss": 0.2399, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_accuracy": 0.9214884921488492, | |
| "eval_loss": 0.2550754249095917, | |
| "eval_runtime": 55.744, | |
| "eval_samples_per_second": 83.399, | |
| "eval_steps_per_second": 2.619, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 7.070884592852959e-05, | |
| "loss": 0.2336, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_accuracy": 0.9219186921918692, | |
| "eval_loss": 0.23437733948230743, | |
| "eval_runtime": 53.9586, | |
| "eval_samples_per_second": 86.159, | |
| "eval_steps_per_second": 2.706, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 6.953719976567077e-05, | |
| "loss": 0.2246, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "eval_accuracy": 0.9197676919767692, | |
| "eval_loss": 0.24808667600154877, | |
| "eval_runtime": 53.4909, | |
| "eval_samples_per_second": 86.912, | |
| "eval_steps_per_second": 2.729, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 6.836555360281195e-05, | |
| "loss": 0.2284, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_accuracy": 0.9322434932243493, | |
| "eval_loss": 0.21941034495830536, | |
| "eval_runtime": 53.0587, | |
| "eval_samples_per_second": 87.62, | |
| "eval_steps_per_second": 2.752, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 6.719390743995314e-05, | |
| "loss": 0.2335, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.9331038933103893, | |
| "eval_loss": 0.2168220579624176, | |
| "eval_runtime": 56.1589, | |
| "eval_samples_per_second": 82.783, | |
| "eval_steps_per_second": 2.6, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 6.602226127709432e-05, | |
| "loss": 0.1883, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.9229941922994193, | |
| "eval_loss": 0.24746185541152954, | |
| "eval_runtime": 54.8887, | |
| "eval_samples_per_second": 84.699, | |
| "eval_steps_per_second": 2.66, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 6.48506151142355e-05, | |
| "loss": 0.194, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_accuracy": 0.9328887932888793, | |
| "eval_loss": 0.2247306853532791, | |
| "eval_runtime": 54.695, | |
| "eval_samples_per_second": 84.999, | |
| "eval_steps_per_second": 2.669, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 6.367896895137668e-05, | |
| "loss": 0.2097, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "eval_accuracy": 0.9272961927296193, | |
| "eval_loss": 0.21885570883750916, | |
| "eval_runtime": 55.7057, | |
| "eval_samples_per_second": 83.456, | |
| "eval_steps_per_second": 2.621, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 6.250732278851787e-05, | |
| "loss": 0.1663, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_accuracy": 0.9380511938051194, | |
| "eval_loss": 0.2038552463054657, | |
| "eval_runtime": 57.0481, | |
| "eval_samples_per_second": 81.493, | |
| "eval_steps_per_second": 2.559, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 6.133567662565906e-05, | |
| "loss": 0.1847, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "eval_accuracy": 0.9266508926650893, | |
| "eval_loss": 0.2347552329301834, | |
| "eval_runtime": 56.4891, | |
| "eval_samples_per_second": 82.299, | |
| "eval_steps_per_second": 2.585, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 6.016403046280023e-05, | |
| "loss": 0.1809, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_accuracy": 0.9236394923639493, | |
| "eval_loss": 0.24315237998962402, | |
| "eval_runtime": 54.6687, | |
| "eval_samples_per_second": 85.04, | |
| "eval_steps_per_second": 2.671, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 5.899238429994142e-05, | |
| "loss": 0.1787, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_accuracy": 0.9339642933964294, | |
| "eval_loss": 0.21847452223300934, | |
| "eval_runtime": 54.0398, | |
| "eval_samples_per_second": 86.029, | |
| "eval_steps_per_second": 2.702, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 5.782073813708261e-05, | |
| "loss": 0.183, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "eval_accuracy": 0.9363303936330394, | |
| "eval_loss": 0.20778249204158783, | |
| "eval_runtime": 54.441, | |
| "eval_samples_per_second": 85.395, | |
| "eval_steps_per_second": 2.682, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 5.664909197422379e-05, | |
| "loss": 0.1704, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_accuracy": 0.9152505915250592, | |
| "eval_loss": 0.23343944549560547, | |
| "eval_runtime": 56.1215, | |
| "eval_samples_per_second": 82.838, | |
| "eval_steps_per_second": 2.601, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 5.547744581136497e-05, | |
| "loss": 0.2002, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "eval_accuracy": 0.9212733921273392, | |
| "eval_loss": 0.23881135880947113, | |
| "eval_runtime": 54.8777, | |
| "eval_samples_per_second": 84.716, | |
| "eval_steps_per_second": 2.66, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 5.430579964850615e-05, | |
| "loss": 0.1787, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "eval_accuracy": 0.9260055926005593, | |
| "eval_loss": 0.22407278418540955, | |
| "eval_runtime": 54.5169, | |
| "eval_samples_per_second": 85.276, | |
| "eval_steps_per_second": 2.678, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 5.3134153485647345e-05, | |
| "loss": 0.1757, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "eval_accuracy": 0.9324585932458593, | |
| "eval_loss": 0.2097584456205368, | |
| "eval_runtime": 54.3298, | |
| "eval_samples_per_second": 85.57, | |
| "eval_steps_per_second": 2.687, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 5.1962507322788524e-05, | |
| "loss": 0.1667, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_accuracy": 0.9348246934824693, | |
| "eval_loss": 0.21198517084121704, | |
| "eval_runtime": 53.7704, | |
| "eval_samples_per_second": 86.46, | |
| "eval_steps_per_second": 2.715, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 5.07908611599297e-05, | |
| "loss": 0.2078, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_accuracy": 0.9348246934824693, | |
| "eval_loss": 0.2187454104423523, | |
| "eval_runtime": 54.7943, | |
| "eval_samples_per_second": 84.845, | |
| "eval_steps_per_second": 2.665, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 4.961921499707089e-05, | |
| "loss": 0.1775, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_accuracy": 0.9294471929447193, | |
| "eval_loss": 0.226194366812706, | |
| "eval_runtime": 54.2742, | |
| "eval_samples_per_second": 85.658, | |
| "eval_steps_per_second": 2.69, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 4.844756883421207e-05, | |
| "loss": 0.1771, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "eval_accuracy": 0.9311679931167993, | |
| "eval_loss": 0.2125757783651352, | |
| "eval_runtime": 53.3871, | |
| "eval_samples_per_second": 87.081, | |
| "eval_steps_per_second": 2.735, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 4.727592267135325e-05, | |
| "loss": 0.158, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "eval_accuracy": 0.9343944934394494, | |
| "eval_loss": 0.20714816451072693, | |
| "eval_runtime": 53.7213, | |
| "eval_samples_per_second": 86.539, | |
| "eval_steps_per_second": 2.718, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.610427650849444e-05, | |
| "loss": 0.1792, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "eval_accuracy": 0.9275112927511293, | |
| "eval_loss": 0.227621927857399, | |
| "eval_runtime": 54.9746, | |
| "eval_samples_per_second": 84.566, | |
| "eval_steps_per_second": 2.656, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 4.4932630345635615e-05, | |
| "loss": 0.1523, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "eval_accuracy": 0.9268659926865993, | |
| "eval_loss": 0.21835899353027344, | |
| "eval_runtime": 55.0826, | |
| "eval_samples_per_second": 84.4, | |
| "eval_steps_per_second": 2.651, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 4.37609841827768e-05, | |
| "loss": 0.1739, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "eval_accuracy": 0.9322434932243493, | |
| "eval_loss": 0.22359773516654968, | |
| "eval_runtime": 53.2692, | |
| "eval_samples_per_second": 87.274, | |
| "eval_steps_per_second": 2.741, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 4.2589338019917986e-05, | |
| "loss": 0.1638, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "eval_accuracy": 0.9402021940202194, | |
| "eval_loss": 0.18819227814674377, | |
| "eval_runtime": 52.562, | |
| "eval_samples_per_second": 88.448, | |
| "eval_steps_per_second": 2.778, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 4.141769185705917e-05, | |
| "loss": 0.1653, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "eval_accuracy": 0.9309528930952893, | |
| "eval_loss": 0.21739357709884644, | |
| "eval_runtime": 54.6403, | |
| "eval_samples_per_second": 85.084, | |
| "eval_steps_per_second": 2.672, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.024604569420035e-05, | |
| "loss": 0.1584, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 0.9384813938481393, | |
| "eval_loss": 0.197902649641037, | |
| "eval_runtime": 54.1444, | |
| "eval_samples_per_second": 85.863, | |
| "eval_steps_per_second": 2.696, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 3.9074399531341535e-05, | |
| "loss": 0.1621, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_accuracy": 0.9371907937190793, | |
| "eval_loss": 0.20193301141262054, | |
| "eval_runtime": 52.6121, | |
| "eval_samples_per_second": 88.364, | |
| "eval_steps_per_second": 2.775, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 3.790275336848272e-05, | |
| "loss": 0.156, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "eval_accuracy": 0.9322434932243493, | |
| "eval_loss": 0.20995856821537018, | |
| "eval_runtime": 52.8452, | |
| "eval_samples_per_second": 87.974, | |
| "eval_steps_per_second": 2.763, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 3.6731107205623906e-05, | |
| "loss": 0.1819, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_accuracy": 0.9384813938481393, | |
| "eval_loss": 0.18913036584854126, | |
| "eval_runtime": 53.1008, | |
| "eval_samples_per_second": 87.55, | |
| "eval_steps_per_second": 2.749, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 3.5559461042765085e-05, | |
| "loss": 0.1676, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "eval_accuracy": 0.9397719939771993, | |
| "eval_loss": 0.18966327607631683, | |
| "eval_runtime": 54.5409, | |
| "eval_samples_per_second": 85.239, | |
| "eval_steps_per_second": 2.677, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 3.438781487990627e-05, | |
| "loss": 0.1675, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_accuracy": 0.9393417939341794, | |
| "eval_loss": 0.1928705871105194, | |
| "eval_runtime": 54.2748, | |
| "eval_samples_per_second": 85.657, | |
| "eval_steps_per_second": 2.69, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.3216168717047456e-05, | |
| "loss": 0.1389, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9389115938911594, | |
| "eval_loss": 0.1925574094057083, | |
| "eval_runtime": 53.5821, | |
| "eval_samples_per_second": 86.764, | |
| "eval_steps_per_second": 2.725, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3.204452255418864e-05, | |
| "loss": 0.1217, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_accuracy": 0.9324585932458593, | |
| "eval_loss": 0.1982860118150711, | |
| "eval_runtime": 53.4582, | |
| "eval_samples_per_second": 86.965, | |
| "eval_steps_per_second": 2.731, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 3.087287639132982e-05, | |
| "loss": 0.1158, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "eval_accuracy": 0.9391266939126693, | |
| "eval_loss": 0.1853022575378418, | |
| "eval_runtime": 52.3558, | |
| "eval_samples_per_second": 88.796, | |
| "eval_steps_per_second": 2.789, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 2.9701230228471005e-05, | |
| "loss": 0.1375, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "eval_accuracy": 0.9371907937190793, | |
| "eval_loss": 0.19718600809574127, | |
| "eval_runtime": 53.8918, | |
| "eval_samples_per_second": 86.265, | |
| "eval_steps_per_second": 2.709, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 2.8541300527240778e-05, | |
| "loss": 0.1195, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_accuracy": 0.9384813938481393, | |
| "eval_loss": 0.19445277750492096, | |
| "eval_runtime": 53.8229, | |
| "eval_samples_per_second": 86.376, | |
| "eval_steps_per_second": 2.713, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 2.7369654364381957e-05, | |
| "loss": 0.1249, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "eval_accuracy": 0.9350397935039794, | |
| "eval_loss": 0.19001764059066772, | |
| "eval_runtime": 52.3182, | |
| "eval_samples_per_second": 88.86, | |
| "eval_steps_per_second": 2.791, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 2.619800820152314e-05, | |
| "loss": 0.1204, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "eval_accuracy": 0.9324585932458593, | |
| "eval_loss": 0.20320840179920197, | |
| "eval_runtime": 52.8746, | |
| "eval_samples_per_second": 87.925, | |
| "eval_steps_per_second": 2.761, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 2.5026362038664324e-05, | |
| "loss": 0.122, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "eval_accuracy": 0.9376209937620994, | |
| "eval_loss": 0.17533761262893677, | |
| "eval_runtime": 52.222, | |
| "eval_samples_per_second": 89.024, | |
| "eval_steps_per_second": 2.796, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2.385471587580551e-05, | |
| "loss": 0.1136, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_accuracy": 0.9406323940632394, | |
| "eval_loss": 0.18234318494796753, | |
| "eval_runtime": 53.2472, | |
| "eval_samples_per_second": 87.31, | |
| "eval_steps_per_second": 2.742, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 2.268306971294669e-05, | |
| "loss": 0.125, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "eval_accuracy": 0.9391266939126693, | |
| "eval_loss": 0.1814049780368805, | |
| "eval_runtime": 55.3063, | |
| "eval_samples_per_second": 84.059, | |
| "eval_steps_per_second": 2.64, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 2.1511423550087877e-05, | |
| "loss": 0.1142, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "eval_accuracy": 0.9402021940202194, | |
| "eval_loss": 0.1924043744802475, | |
| "eval_runtime": 53.3253, | |
| "eval_samples_per_second": 87.182, | |
| "eval_steps_per_second": 2.738, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 2.033977738722906e-05, | |
| "loss": 0.1258, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "eval_accuracy": 0.9421380942138095, | |
| "eval_loss": 0.1804322451353073, | |
| "eval_runtime": 52.0681, | |
| "eval_samples_per_second": 89.287, | |
| "eval_steps_per_second": 2.804, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 1.916813122437024e-05, | |
| "loss": 0.1178, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "eval_accuracy": 0.9374058937405894, | |
| "eval_loss": 0.18380628526210785, | |
| "eval_runtime": 52.0076, | |
| "eval_samples_per_second": 89.391, | |
| "eval_steps_per_second": 2.807, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 1.7996485061511426e-05, | |
| "loss": 0.1093, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "eval_accuracy": 0.9391266939126693, | |
| "eval_loss": 0.17910005152225494, | |
| "eval_runtime": 51.873, | |
| "eval_samples_per_second": 89.623, | |
| "eval_steps_per_second": 2.815, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 1.6824838898652608e-05, | |
| "loss": 0.1069, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_accuracy": 0.9391266939126693, | |
| "eval_loss": 0.18617549538612366, | |
| "eval_runtime": 54.5914, | |
| "eval_samples_per_second": 85.16, | |
| "eval_steps_per_second": 2.674, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 1.565319273579379e-05, | |
| "loss": 0.1361, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_accuracy": 0.9417078941707894, | |
| "eval_loss": 0.17183491587638855, | |
| "eval_runtime": 53.9807, | |
| "eval_samples_per_second": 86.123, | |
| "eval_steps_per_second": 2.705, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 1.4481546572934974e-05, | |
| "loss": 0.1167, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "eval_accuracy": 0.9380511938051194, | |
| "eval_loss": 0.18089379370212555, | |
| "eval_runtime": 52.2028, | |
| "eval_samples_per_second": 89.056, | |
| "eval_steps_per_second": 2.797, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 1.3309900410076156e-05, | |
| "loss": 0.127, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "eval_accuracy": 0.9445041944504194, | |
| "eval_loss": 0.17267315089702606, | |
| "eval_runtime": 51.7533, | |
| "eval_samples_per_second": 89.83, | |
| "eval_steps_per_second": 2.821, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 1.2138254247217341e-05, | |
| "loss": 0.1215, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "eval_accuracy": 0.9402021940202194, | |
| "eval_loss": 0.18060338497161865, | |
| "eval_runtime": 53.2058, | |
| "eval_samples_per_second": 87.378, | |
| "eval_steps_per_second": 2.744, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 1.0966608084358525e-05, | |
| "loss": 0.1162, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_accuracy": 0.9404172940417294, | |
| "eval_loss": 0.18343867361545563, | |
| "eval_runtime": 56.0804, | |
| "eval_samples_per_second": 82.899, | |
| "eval_steps_per_second": 2.603, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 9.794961921499708e-06, | |
| "loss": 0.1182, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "eval_accuracy": 0.9406323940632394, | |
| "eval_loss": 0.17546848952770233, | |
| "eval_runtime": 53.4141, | |
| "eval_samples_per_second": 87.037, | |
| "eval_steps_per_second": 2.733, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 8.62331575864089e-06, | |
| "loss": 0.1179, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "eval_accuracy": 0.9410625941062594, | |
| "eval_loss": 0.179013192653656, | |
| "eval_runtime": 54.9709, | |
| "eval_samples_per_second": 84.572, | |
| "eval_steps_per_second": 2.656, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 7.451669595782074e-06, | |
| "loss": 0.1012, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "eval_accuracy": 0.9427833942783395, | |
| "eval_loss": 0.17139938473701477, | |
| "eval_runtime": 53.4602, | |
| "eval_samples_per_second": 86.962, | |
| "eval_steps_per_second": 2.731, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 6.280023432923257e-06, | |
| "loss": 0.1186, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "eval_accuracy": 0.9393417939341794, | |
| "eval_loss": 0.17887155711650848, | |
| "eval_runtime": 53.0073, | |
| "eval_samples_per_second": 87.705, | |
| "eval_steps_per_second": 2.754, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 5.1083772700644406e-06, | |
| "loss": 0.1305, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "eval_accuracy": 0.9408474940847494, | |
| "eval_loss": 0.17335866391658783, | |
| "eval_runtime": 52.8785, | |
| "eval_samples_per_second": 87.919, | |
| "eval_steps_per_second": 2.761, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 3.936731107205624e-06, | |
| "loss": 0.1126, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_accuracy": 0.9397719939771993, | |
| "eval_loss": 0.1753973364830017, | |
| "eval_runtime": 51.7413, | |
| "eval_samples_per_second": 89.851, | |
| "eval_steps_per_second": 2.822, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 2.776801405975396e-06, | |
| "loss": 0.1128, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "eval_accuracy": 0.9445041944504194, | |
| "eval_loss": 0.1688772737979889, | |
| "eval_runtime": 51.7644, | |
| "eval_samples_per_second": 89.811, | |
| "eval_steps_per_second": 2.82, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 1.605155243116579e-06, | |
| "loss": 0.1108, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "eval_accuracy": 0.9442890944289094, | |
| "eval_loss": 0.16805602610111237, | |
| "eval_runtime": 53.4934, | |
| "eval_samples_per_second": 86.908, | |
| "eval_steps_per_second": 2.729, | |
| "step": 8400 | |
| } | |
| ], | |
| "max_steps": 8535, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.083300191455087e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |