|
{ |
|
"best_metric": 0.7663551401869159, |
|
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-435", |
|
"epoch": 93.33333333333333, |
|
"eval_steps": 500, |
|
"global_step": 700, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.9333333333333333, |
|
"eval_accuracy": 0.08411214953271028, |
|
"eval_loss": 3.889423370361328, |
|
"eval_runtime": 0.6153, |
|
"eval_samples_per_second": 173.891, |
|
"eval_steps_per_second": 6.501, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 1.3333333333333333, |
|
"grad_norm": 4.184154033660889, |
|
"learning_rate": 7.142857142857143e-06, |
|
"loss": 3.897, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.08411214953271028, |
|
"eval_loss": 3.8185415267944336, |
|
"eval_runtime": 0.6194, |
|
"eval_samples_per_second": 172.745, |
|
"eval_steps_per_second": 6.458, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 2.6666666666666665, |
|
"grad_norm": 4.875367164611816, |
|
"learning_rate": 1.4285714285714285e-05, |
|
"loss": 3.8553, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 2.9333333333333336, |
|
"eval_accuracy": 0.07476635514018691, |
|
"eval_loss": 3.7401788234710693, |
|
"eval_runtime": 0.6021, |
|
"eval_samples_per_second": 177.698, |
|
"eval_steps_per_second": 6.643, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 5.443772315979004, |
|
"learning_rate": 2.1428571428571428e-05, |
|
"loss": 3.7568, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.07476635514018691, |
|
"eval_loss": 3.637192964553833, |
|
"eval_runtime": 0.7262, |
|
"eval_samples_per_second": 147.335, |
|
"eval_steps_per_second": 5.508, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 4.933333333333334, |
|
"eval_accuracy": 0.08411214953271028, |
|
"eval_loss": 3.5481796264648438, |
|
"eval_runtime": 0.6382, |
|
"eval_samples_per_second": 167.648, |
|
"eval_steps_per_second": 6.267, |
|
"step": 37 |
|
}, |
|
{ |
|
"epoch": 5.333333333333333, |
|
"grad_norm": 5.656874656677246, |
|
"learning_rate": 2.857142857142857e-05, |
|
"loss": 3.5912, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.11214953271028037, |
|
"eval_loss": 3.406933069229126, |
|
"eval_runtime": 0.6307, |
|
"eval_samples_per_second": 169.654, |
|
"eval_steps_per_second": 6.342, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 6.666666666666667, |
|
"grad_norm": 6.326891899108887, |
|
"learning_rate": 3.571428571428572e-05, |
|
"loss": 3.4342, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 6.933333333333334, |
|
"eval_accuracy": 0.1308411214953271, |
|
"eval_loss": 3.293895721435547, |
|
"eval_runtime": 0.6244, |
|
"eval_samples_per_second": 171.371, |
|
"eval_steps_per_second": 6.406, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"grad_norm": 12.78520393371582, |
|
"learning_rate": 4.2857142857142856e-05, |
|
"loss": 3.2601, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.21495327102803738, |
|
"eval_loss": 3.178621768951416, |
|
"eval_runtime": 0.7022, |
|
"eval_samples_per_second": 152.374, |
|
"eval_steps_per_second": 5.696, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 8.933333333333334, |
|
"eval_accuracy": 0.2336448598130841, |
|
"eval_loss": 3.032252311706543, |
|
"eval_runtime": 0.6285, |
|
"eval_samples_per_second": 170.255, |
|
"eval_steps_per_second": 6.365, |
|
"step": 67 |
|
}, |
|
{ |
|
"epoch": 9.333333333333334, |
|
"grad_norm": 9.798720359802246, |
|
"learning_rate": 5e-05, |
|
"loss": 3.0498, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.2616822429906542, |
|
"eval_loss": 2.869462728500366, |
|
"eval_runtime": 0.6405, |
|
"eval_samples_per_second": 167.062, |
|
"eval_steps_per_second": 6.245, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 10.666666666666666, |
|
"grad_norm": 9.29806137084961, |
|
"learning_rate": 4.9206349206349204e-05, |
|
"loss": 2.849, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 10.933333333333334, |
|
"eval_accuracy": 0.2523364485981308, |
|
"eval_loss": 2.8504700660705566, |
|
"eval_runtime": 0.6315, |
|
"eval_samples_per_second": 169.425, |
|
"eval_steps_per_second": 6.334, |
|
"step": 82 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"grad_norm": 15.086236000061035, |
|
"learning_rate": 4.841269841269841e-05, |
|
"loss": 2.6452, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.2803738317757009, |
|
"eval_loss": 2.63193416595459, |
|
"eval_runtime": 0.7011, |
|
"eval_samples_per_second": 152.622, |
|
"eval_steps_per_second": 5.706, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 12.933333333333334, |
|
"eval_accuracy": 0.32710280373831774, |
|
"eval_loss": 2.465355157852173, |
|
"eval_runtime": 0.6287, |
|
"eval_samples_per_second": 170.194, |
|
"eval_steps_per_second": 6.362, |
|
"step": 97 |
|
}, |
|
{ |
|
"epoch": 13.333333333333334, |
|
"grad_norm": 10.405584335327148, |
|
"learning_rate": 4.761904761904762e-05, |
|
"loss": 2.4123, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.3364485981308411, |
|
"eval_loss": 2.399456262588501, |
|
"eval_runtime": 0.6357, |
|
"eval_samples_per_second": 168.322, |
|
"eval_steps_per_second": 6.292, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 14.666666666666666, |
|
"grad_norm": 12.884756088256836, |
|
"learning_rate": 4.682539682539683e-05, |
|
"loss": 2.2561, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 14.933333333333334, |
|
"eval_accuracy": 0.40186915887850466, |
|
"eval_loss": 2.258385419845581, |
|
"eval_runtime": 0.6927, |
|
"eval_samples_per_second": 154.461, |
|
"eval_steps_per_second": 5.774, |
|
"step": 112 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"grad_norm": 11.66384506225586, |
|
"learning_rate": 4.603174603174603e-05, |
|
"loss": 2.0447, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.42990654205607476, |
|
"eval_loss": 2.1999597549438477, |
|
"eval_runtime": 0.6407, |
|
"eval_samples_per_second": 167.013, |
|
"eval_steps_per_second": 6.243, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 16.933333333333334, |
|
"eval_accuracy": 0.4392523364485981, |
|
"eval_loss": 2.080615282058716, |
|
"eval_runtime": 0.6391, |
|
"eval_samples_per_second": 167.427, |
|
"eval_steps_per_second": 6.259, |
|
"step": 127 |
|
}, |
|
{ |
|
"epoch": 17.333333333333332, |
|
"grad_norm": 9.476760864257812, |
|
"learning_rate": 4.523809523809524e-05, |
|
"loss": 1.8569, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.4392523364485981, |
|
"eval_loss": 2.0593273639678955, |
|
"eval_runtime": 0.6302, |
|
"eval_samples_per_second": 169.788, |
|
"eval_steps_per_second": 6.347, |
|
"step": 135 |
|
}, |
|
{ |
|
"epoch": 18.666666666666668, |
|
"grad_norm": 11.518675804138184, |
|
"learning_rate": 4.4444444444444447e-05, |
|
"loss": 1.7447, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 18.933333333333334, |
|
"eval_accuracy": 0.4672897196261682, |
|
"eval_loss": 1.8832261562347412, |
|
"eval_runtime": 0.7003, |
|
"eval_samples_per_second": 152.783, |
|
"eval_steps_per_second": 5.712, |
|
"step": 142 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 10.125329971313477, |
|
"learning_rate": 4.3650793650793655e-05, |
|
"loss": 1.5821, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.5046728971962616, |
|
"eval_loss": 1.8217570781707764, |
|
"eval_runtime": 0.6316, |
|
"eval_samples_per_second": 169.416, |
|
"eval_steps_per_second": 6.333, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 20.933333333333334, |
|
"eval_accuracy": 0.5420560747663551, |
|
"eval_loss": 1.7333636283874512, |
|
"eval_runtime": 0.6313, |
|
"eval_samples_per_second": 169.494, |
|
"eval_steps_per_second": 6.336, |
|
"step": 157 |
|
}, |
|
{ |
|
"epoch": 21.333333333333332, |
|
"grad_norm": 11.70013427734375, |
|
"learning_rate": 4.2857142857142856e-05, |
|
"loss": 1.3999, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_accuracy": 0.5514018691588785, |
|
"eval_loss": 1.6213181018829346, |
|
"eval_runtime": 0.6355, |
|
"eval_samples_per_second": 168.373, |
|
"eval_steps_per_second": 6.294, |
|
"step": 165 |
|
}, |
|
{ |
|
"epoch": 22.666666666666668, |
|
"grad_norm": 9.099847793579102, |
|
"learning_rate": 4.2063492063492065e-05, |
|
"loss": 1.2901, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 22.933333333333334, |
|
"eval_accuracy": 0.5233644859813084, |
|
"eval_loss": 1.593188762664795, |
|
"eval_runtime": 0.7026, |
|
"eval_samples_per_second": 152.289, |
|
"eval_steps_per_second": 5.693, |
|
"step": 172 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"grad_norm": 10.077893257141113, |
|
"learning_rate": 4.126984126984127e-05, |
|
"loss": 1.1569, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.5700934579439252, |
|
"eval_loss": 1.5255681276321411, |
|
"eval_runtime": 0.6286, |
|
"eval_samples_per_second": 170.213, |
|
"eval_steps_per_second": 6.363, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 24.933333333333334, |
|
"eval_accuracy": 0.5887850467289719, |
|
"eval_loss": 1.428060531616211, |
|
"eval_runtime": 0.6322, |
|
"eval_samples_per_second": 169.258, |
|
"eval_steps_per_second": 6.327, |
|
"step": 187 |
|
}, |
|
{ |
|
"epoch": 25.333333333333332, |
|
"grad_norm": 10.525726318359375, |
|
"learning_rate": 4.047619047619048e-05, |
|
"loss": 1.0903, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 26.0, |
|
"eval_accuracy": 0.5794392523364486, |
|
"eval_loss": 1.3997470140457153, |
|
"eval_runtime": 0.6928, |
|
"eval_samples_per_second": 154.457, |
|
"eval_steps_per_second": 5.774, |
|
"step": 195 |
|
}, |
|
{ |
|
"epoch": 26.666666666666668, |
|
"grad_norm": 8.83031177520752, |
|
"learning_rate": 3.968253968253968e-05, |
|
"loss": 0.9674, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 26.933333333333334, |
|
"eval_accuracy": 0.5887850467289719, |
|
"eval_loss": 1.4017095565795898, |
|
"eval_runtime": 0.634, |
|
"eval_samples_per_second": 168.758, |
|
"eval_steps_per_second": 6.309, |
|
"step": 202 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"grad_norm": 9.685677528381348, |
|
"learning_rate": 3.888888888888889e-05, |
|
"loss": 0.98, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_accuracy": 0.5981308411214953, |
|
"eval_loss": 1.2915815114974976, |
|
"eval_runtime": 0.6453, |
|
"eval_samples_per_second": 165.816, |
|
"eval_steps_per_second": 6.199, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 28.933333333333334, |
|
"eval_accuracy": 0.5981308411214953, |
|
"eval_loss": 1.301841378211975, |
|
"eval_runtime": 0.6256, |
|
"eval_samples_per_second": 171.047, |
|
"eval_steps_per_second": 6.394, |
|
"step": 217 |
|
}, |
|
{ |
|
"epoch": 29.333333333333332, |
|
"grad_norm": 11.74543285369873, |
|
"learning_rate": 3.809523809523809e-05, |
|
"loss": 0.8772, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 30.0, |
|
"eval_accuracy": 0.6355140186915887, |
|
"eval_loss": 1.2552070617675781, |
|
"eval_runtime": 0.6964, |
|
"eval_samples_per_second": 153.648, |
|
"eval_steps_per_second": 5.744, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 30.666666666666668, |
|
"grad_norm": 11.836760520935059, |
|
"learning_rate": 3.730158730158731e-05, |
|
"loss": 0.7842, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 30.933333333333334, |
|
"eval_accuracy": 0.6074766355140186, |
|
"eval_loss": 1.2371814250946045, |
|
"eval_runtime": 0.6411, |
|
"eval_samples_per_second": 166.89, |
|
"eval_steps_per_second": 6.239, |
|
"step": 232 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"grad_norm": 8.279513359069824, |
|
"learning_rate": 3.650793650793651e-05, |
|
"loss": 0.7438, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_accuracy": 0.616822429906542, |
|
"eval_loss": 1.1908384561538696, |
|
"eval_runtime": 0.636, |
|
"eval_samples_per_second": 168.249, |
|
"eval_steps_per_second": 6.29, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 32.93333333333333, |
|
"eval_accuracy": 0.6635514018691588, |
|
"eval_loss": 1.1566777229309082, |
|
"eval_runtime": 0.6312, |
|
"eval_samples_per_second": 169.507, |
|
"eval_steps_per_second": 6.337, |
|
"step": 247 |
|
}, |
|
{ |
|
"epoch": 33.333333333333336, |
|
"grad_norm": 8.077630996704102, |
|
"learning_rate": 3.571428571428572e-05, |
|
"loss": 0.725, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 34.0, |
|
"eval_accuracy": 0.6261682242990654, |
|
"eval_loss": 1.1541680097579956, |
|
"eval_runtime": 0.6806, |
|
"eval_samples_per_second": 157.204, |
|
"eval_steps_per_second": 5.877, |
|
"step": 255 |
|
}, |
|
{ |
|
"epoch": 34.666666666666664, |
|
"grad_norm": 10.355536460876465, |
|
"learning_rate": 3.492063492063492e-05, |
|
"loss": 0.6709, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 34.93333333333333, |
|
"eval_accuracy": 0.6261682242990654, |
|
"eval_loss": 1.137677550315857, |
|
"eval_runtime": 0.6265, |
|
"eval_samples_per_second": 170.787, |
|
"eval_steps_per_second": 6.385, |
|
"step": 262 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"grad_norm": 7.819301128387451, |
|
"learning_rate": 3.412698412698413e-05, |
|
"loss": 0.6898, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_accuracy": 0.6635514018691588, |
|
"eval_loss": 1.0523799657821655, |
|
"eval_runtime": 0.6376, |
|
"eval_samples_per_second": 167.806, |
|
"eval_steps_per_second": 6.273, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 36.93333333333333, |
|
"eval_accuracy": 0.6728971962616822, |
|
"eval_loss": 1.027221441268921, |
|
"eval_runtime": 0.6345, |
|
"eval_samples_per_second": 168.642, |
|
"eval_steps_per_second": 6.304, |
|
"step": 277 |
|
}, |
|
{ |
|
"epoch": 37.333333333333336, |
|
"grad_norm": 12.395262718200684, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.6125, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 38.0, |
|
"eval_accuracy": 0.6355140186915887, |
|
"eval_loss": 1.0398948192596436, |
|
"eval_runtime": 0.6985, |
|
"eval_samples_per_second": 153.19, |
|
"eval_steps_per_second": 5.727, |
|
"step": 285 |
|
}, |
|
{ |
|
"epoch": 38.666666666666664, |
|
"grad_norm": 8.95246410369873, |
|
"learning_rate": 3.253968253968254e-05, |
|
"loss": 0.6153, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 38.93333333333333, |
|
"eval_accuracy": 0.6822429906542056, |
|
"eval_loss": 1.0307663679122925, |
|
"eval_runtime": 0.6229, |
|
"eval_samples_per_second": 171.769, |
|
"eval_steps_per_second": 6.421, |
|
"step": 292 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"grad_norm": 10.705389022827148, |
|
"learning_rate": 3.1746031746031745e-05, |
|
"loss": 0.5898, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"eval_accuracy": 0.7009345794392523, |
|
"eval_loss": 1.0150656700134277, |
|
"eval_runtime": 0.6404, |
|
"eval_samples_per_second": 167.072, |
|
"eval_steps_per_second": 6.246, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 40.93333333333333, |
|
"eval_accuracy": 0.6542056074766355, |
|
"eval_loss": 1.0482978820800781, |
|
"eval_runtime": 0.696, |
|
"eval_samples_per_second": 153.738, |
|
"eval_steps_per_second": 5.747, |
|
"step": 307 |
|
}, |
|
{ |
|
"epoch": 41.333333333333336, |
|
"grad_norm": 9.051219940185547, |
|
"learning_rate": 3.095238095238095e-05, |
|
"loss": 0.5881, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 42.0, |
|
"eval_accuracy": 0.7009345794392523, |
|
"eval_loss": 0.992605984210968, |
|
"eval_runtime": 0.6887, |
|
"eval_samples_per_second": 155.368, |
|
"eval_steps_per_second": 5.808, |
|
"step": 315 |
|
}, |
|
{ |
|
"epoch": 42.666666666666664, |
|
"grad_norm": 8.400652885437012, |
|
"learning_rate": 3.0158730158730158e-05, |
|
"loss": 0.54, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 42.93333333333333, |
|
"eval_accuracy": 0.6915887850467289, |
|
"eval_loss": 1.0300043821334839, |
|
"eval_runtime": 0.6255, |
|
"eval_samples_per_second": 171.052, |
|
"eval_steps_per_second": 6.394, |
|
"step": 322 |
|
}, |
|
{ |
|
"epoch": 44.0, |
|
"grad_norm": 10.61039924621582, |
|
"learning_rate": 2.9365079365079366e-05, |
|
"loss": 0.4515, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 44.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.926239013671875, |
|
"eval_runtime": 0.689, |
|
"eval_samples_per_second": 155.308, |
|
"eval_steps_per_second": 5.806, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 44.93333333333333, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.9486252069473267, |
|
"eval_runtime": 0.6286, |
|
"eval_samples_per_second": 170.217, |
|
"eval_steps_per_second": 6.363, |
|
"step": 337 |
|
}, |
|
{ |
|
"epoch": 45.333333333333336, |
|
"grad_norm": 11.857452392578125, |
|
"learning_rate": 2.857142857142857e-05, |
|
"loss": 0.5057, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 46.0, |
|
"eval_accuracy": 0.7102803738317757, |
|
"eval_loss": 0.9219488501548767, |
|
"eval_runtime": 0.6365, |
|
"eval_samples_per_second": 168.094, |
|
"eval_steps_per_second": 6.284, |
|
"step": 345 |
|
}, |
|
{ |
|
"epoch": 46.666666666666664, |
|
"grad_norm": 8.10464096069336, |
|
"learning_rate": 2.777777777777778e-05, |
|
"loss": 0.4905, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 46.93333333333333, |
|
"eval_accuracy": 0.6822429906542056, |
|
"eval_loss": 1.0184197425842285, |
|
"eval_runtime": 0.6292, |
|
"eval_samples_per_second": 170.066, |
|
"eval_steps_per_second": 6.358, |
|
"step": 352 |
|
}, |
|
{ |
|
"epoch": 48.0, |
|
"grad_norm": 9.08785629272461, |
|
"learning_rate": 2.6984126984126984e-05, |
|
"loss": 0.4669, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 48.0, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.9337471127510071, |
|
"eval_runtime": 0.6967, |
|
"eval_samples_per_second": 153.592, |
|
"eval_steps_per_second": 5.742, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 48.93333333333333, |
|
"eval_accuracy": 0.7102803738317757, |
|
"eval_loss": 0.9431414604187012, |
|
"eval_runtime": 0.6378, |
|
"eval_samples_per_second": 167.775, |
|
"eval_steps_per_second": 6.272, |
|
"step": 367 |
|
}, |
|
{ |
|
"epoch": 49.333333333333336, |
|
"grad_norm": 8.805204391479492, |
|
"learning_rate": 2.6190476190476192e-05, |
|
"loss": 0.4437, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 50.0, |
|
"eval_accuracy": 0.7009345794392523, |
|
"eval_loss": 0.9311835169792175, |
|
"eval_runtime": 0.6277, |
|
"eval_samples_per_second": 170.465, |
|
"eval_steps_per_second": 6.373, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 50.666666666666664, |
|
"grad_norm": 7.111200332641602, |
|
"learning_rate": 2.5396825396825397e-05, |
|
"loss": 0.4754, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 50.93333333333333, |
|
"eval_accuracy": 0.719626168224299, |
|
"eval_loss": 0.9244596362113953, |
|
"eval_runtime": 0.6252, |
|
"eval_samples_per_second": 171.138, |
|
"eval_steps_per_second": 6.398, |
|
"step": 382 |
|
}, |
|
{ |
|
"epoch": 52.0, |
|
"grad_norm": 7.307917594909668, |
|
"learning_rate": 2.4603174603174602e-05, |
|
"loss": 0.4119, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 52.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8826178908348083, |
|
"eval_runtime": 0.6908, |
|
"eval_samples_per_second": 154.896, |
|
"eval_steps_per_second": 5.79, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 52.93333333333333, |
|
"eval_accuracy": 0.719626168224299, |
|
"eval_loss": 0.9261904358863831, |
|
"eval_runtime": 0.6228, |
|
"eval_samples_per_second": 171.799, |
|
"eval_steps_per_second": 6.422, |
|
"step": 397 |
|
}, |
|
{ |
|
"epoch": 53.333333333333336, |
|
"grad_norm": 8.19266128540039, |
|
"learning_rate": 2.380952380952381e-05, |
|
"loss": 0.4087, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 54.0, |
|
"eval_accuracy": 0.7476635514018691, |
|
"eval_loss": 0.888160765171051, |
|
"eval_runtime": 0.6231, |
|
"eval_samples_per_second": 171.734, |
|
"eval_steps_per_second": 6.42, |
|
"step": 405 |
|
}, |
|
{ |
|
"epoch": 54.666666666666664, |
|
"grad_norm": 7.831826210021973, |
|
"learning_rate": 2.3015873015873015e-05, |
|
"loss": 0.3987, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 54.93333333333333, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.9281949400901794, |
|
"eval_runtime": 0.63, |
|
"eval_samples_per_second": 169.85, |
|
"eval_steps_per_second": 6.35, |
|
"step": 412 |
|
}, |
|
{ |
|
"epoch": 56.0, |
|
"grad_norm": 8.277617454528809, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.4253, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 56.0, |
|
"eval_accuracy": 0.7476635514018691, |
|
"eval_loss": 0.9003600478172302, |
|
"eval_runtime": 0.738, |
|
"eval_samples_per_second": 144.99, |
|
"eval_steps_per_second": 5.42, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 56.93333333333333, |
|
"eval_accuracy": 0.7476635514018691, |
|
"eval_loss": 0.8783094882965088, |
|
"eval_runtime": 0.6305, |
|
"eval_samples_per_second": 169.701, |
|
"eval_steps_per_second": 6.344, |
|
"step": 427 |
|
}, |
|
{ |
|
"epoch": 57.333333333333336, |
|
"grad_norm": 7.786144256591797, |
|
"learning_rate": 2.1428571428571428e-05, |
|
"loss": 0.4134, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 58.0, |
|
"eval_accuracy": 0.7663551401869159, |
|
"eval_loss": 0.835954487323761, |
|
"eval_runtime": 0.6311, |
|
"eval_samples_per_second": 169.557, |
|
"eval_steps_per_second": 6.339, |
|
"step": 435 |
|
}, |
|
{ |
|
"epoch": 58.666666666666664, |
|
"grad_norm": 8.696057319641113, |
|
"learning_rate": 2.0634920634920636e-05, |
|
"loss": 0.4024, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 58.93333333333333, |
|
"eval_accuracy": 0.719626168224299, |
|
"eval_loss": 0.901554524898529, |
|
"eval_runtime": 0.6321, |
|
"eval_samples_per_second": 169.266, |
|
"eval_steps_per_second": 6.328, |
|
"step": 442 |
|
}, |
|
{ |
|
"epoch": 60.0, |
|
"grad_norm": 7.4623284339904785, |
|
"learning_rate": 1.984126984126984e-05, |
|
"loss": 0.3688, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 60.0, |
|
"eval_accuracy": 0.6822429906542056, |
|
"eval_loss": 0.9250590205192566, |
|
"eval_runtime": 0.7012, |
|
"eval_samples_per_second": 152.597, |
|
"eval_steps_per_second": 5.705, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 60.93333333333333, |
|
"eval_accuracy": 0.7102803738317757, |
|
"eval_loss": 0.9085938930511475, |
|
"eval_runtime": 0.6226, |
|
"eval_samples_per_second": 171.866, |
|
"eval_steps_per_second": 6.425, |
|
"step": 457 |
|
}, |
|
{ |
|
"epoch": 61.333333333333336, |
|
"grad_norm": 8.650228500366211, |
|
"learning_rate": 1.9047619047619046e-05, |
|
"loss": 0.3833, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 62.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8493680953979492, |
|
"eval_runtime": 0.6222, |
|
"eval_samples_per_second": 171.977, |
|
"eval_steps_per_second": 6.429, |
|
"step": 465 |
|
}, |
|
{ |
|
"epoch": 62.666666666666664, |
|
"grad_norm": 9.089720726013184, |
|
"learning_rate": 1.8253968253968254e-05, |
|
"loss": 0.3614, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 62.93333333333333, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8298574686050415, |
|
"eval_runtime": 0.6823, |
|
"eval_samples_per_second": 156.816, |
|
"eval_steps_per_second": 5.862, |
|
"step": 472 |
|
}, |
|
{ |
|
"epoch": 64.0, |
|
"grad_norm": 9.999822616577148, |
|
"learning_rate": 1.746031746031746e-05, |
|
"loss": 0.3792, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 64.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.9015009999275208, |
|
"eval_runtime": 0.647, |
|
"eval_samples_per_second": 165.386, |
|
"eval_steps_per_second": 6.183, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 64.93333333333334, |
|
"eval_accuracy": 0.719626168224299, |
|
"eval_loss": 0.8801712989807129, |
|
"eval_runtime": 0.6283, |
|
"eval_samples_per_second": 170.303, |
|
"eval_steps_per_second": 6.366, |
|
"step": 487 |
|
}, |
|
{ |
|
"epoch": 65.33333333333333, |
|
"grad_norm": 7.207292556762695, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.3632, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 66.0, |
|
"eval_accuracy": 0.7009345794392523, |
|
"eval_loss": 0.8881424069404602, |
|
"eval_runtime": 0.6298, |
|
"eval_samples_per_second": 169.883, |
|
"eval_steps_per_second": 6.351, |
|
"step": 495 |
|
}, |
|
{ |
|
"epoch": 66.66666666666667, |
|
"grad_norm": 9.152532577514648, |
|
"learning_rate": 1.5873015873015872e-05, |
|
"loss": 0.3405, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 66.93333333333334, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.857825517654419, |
|
"eval_runtime": 0.7094, |
|
"eval_samples_per_second": 150.84, |
|
"eval_steps_per_second": 5.639, |
|
"step": 502 |
|
}, |
|
{ |
|
"epoch": 68.0, |
|
"grad_norm": 6.808733940124512, |
|
"learning_rate": 1.5079365079365079e-05, |
|
"loss": 0.3673, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 68.0, |
|
"eval_accuracy": 0.7570093457943925, |
|
"eval_loss": 0.8540030717849731, |
|
"eval_runtime": 0.6324, |
|
"eval_samples_per_second": 169.206, |
|
"eval_steps_per_second": 6.325, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 68.93333333333334, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.834481954574585, |
|
"eval_runtime": 0.6329, |
|
"eval_samples_per_second": 169.057, |
|
"eval_steps_per_second": 6.32, |
|
"step": 517 |
|
}, |
|
{ |
|
"epoch": 69.33333333333333, |
|
"grad_norm": 7.036710262298584, |
|
"learning_rate": 1.4285714285714285e-05, |
|
"loss": 0.3379, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 70.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.7918941974639893, |
|
"eval_runtime": 0.6863, |
|
"eval_samples_per_second": 155.912, |
|
"eval_steps_per_second": 5.828, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 70.66666666666667, |
|
"grad_norm": 8.039285659790039, |
|
"learning_rate": 1.3492063492063492e-05, |
|
"loss": 0.3389, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 70.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8383651375770569, |
|
"eval_runtime": 0.6238, |
|
"eval_samples_per_second": 171.538, |
|
"eval_steps_per_second": 6.413, |
|
"step": 532 |
|
}, |
|
{ |
|
"epoch": 72.0, |
|
"grad_norm": 8.855910301208496, |
|
"learning_rate": 1.2698412698412699e-05, |
|
"loss": 0.3363, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 72.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8305981159210205, |
|
"eval_runtime": 0.6209, |
|
"eval_samples_per_second": 172.332, |
|
"eval_steps_per_second": 6.442, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 72.93333333333334, |
|
"eval_accuracy": 0.7476635514018691, |
|
"eval_loss": 0.8875143527984619, |
|
"eval_runtime": 0.6229, |
|
"eval_samples_per_second": 171.771, |
|
"eval_steps_per_second": 6.421, |
|
"step": 547 |
|
}, |
|
{ |
|
"epoch": 73.33333333333333, |
|
"grad_norm": 8.251914978027344, |
|
"learning_rate": 1.1904761904761905e-05, |
|
"loss": 0.3494, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 74.0, |
|
"eval_accuracy": 0.7009345794392523, |
|
"eval_loss": 0.9151278138160706, |
|
"eval_runtime": 0.6911, |
|
"eval_samples_per_second": 154.833, |
|
"eval_steps_per_second": 5.788, |
|
"step": 555 |
|
}, |
|
{ |
|
"epoch": 74.66666666666667, |
|
"grad_norm": 6.429625988006592, |
|
"learning_rate": 1.1111111111111112e-05, |
|
"loss": 0.2989, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 74.93333333333334, |
|
"eval_accuracy": 0.7102803738317757, |
|
"eval_loss": 0.8605906963348389, |
|
"eval_runtime": 0.6304, |
|
"eval_samples_per_second": 169.733, |
|
"eval_steps_per_second": 6.345, |
|
"step": 562 |
|
}, |
|
{ |
|
"epoch": 76.0, |
|
"grad_norm": 26.255657196044922, |
|
"learning_rate": 1.0317460317460318e-05, |
|
"loss": 0.3157, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 76.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8639878630638123, |
|
"eval_runtime": 0.6362, |
|
"eval_samples_per_second": 168.197, |
|
"eval_steps_per_second": 6.288, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 76.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8531526327133179, |
|
"eval_runtime": 0.6397, |
|
"eval_samples_per_second": 167.268, |
|
"eval_steps_per_second": 6.253, |
|
"step": 577 |
|
}, |
|
{ |
|
"epoch": 77.33333333333333, |
|
"grad_norm": 7.6963958740234375, |
|
"learning_rate": 9.523809523809523e-06, |
|
"loss": 0.3013, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 78.0, |
|
"eval_accuracy": 0.7102803738317757, |
|
"eval_loss": 0.8478634357452393, |
|
"eval_runtime": 0.6402, |
|
"eval_samples_per_second": 167.14, |
|
"eval_steps_per_second": 6.248, |
|
"step": 585 |
|
}, |
|
{ |
|
"epoch": 78.66666666666667, |
|
"grad_norm": 6.458745956420898, |
|
"learning_rate": 8.73015873015873e-06, |
|
"loss": 0.2968, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 78.93333333333334, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8838663101196289, |
|
"eval_runtime": 0.6283, |
|
"eval_samples_per_second": 170.3, |
|
"eval_steps_per_second": 6.366, |
|
"step": 592 |
|
}, |
|
{ |
|
"epoch": 80.0, |
|
"grad_norm": 7.113694667816162, |
|
"learning_rate": 7.936507936507936e-06, |
|
"loss": 0.3013, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 80.0, |
|
"eval_accuracy": 0.719626168224299, |
|
"eval_loss": 0.8837152719497681, |
|
"eval_runtime": 0.6296, |
|
"eval_samples_per_second": 169.941, |
|
"eval_steps_per_second": 6.353, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 80.93333333333334, |
|
"eval_accuracy": 0.7102803738317757, |
|
"eval_loss": 0.8694174289703369, |
|
"eval_runtime": 0.6384, |
|
"eval_samples_per_second": 167.61, |
|
"eval_steps_per_second": 6.266, |
|
"step": 607 |
|
}, |
|
{ |
|
"epoch": 81.33333333333333, |
|
"grad_norm": 7.3130645751953125, |
|
"learning_rate": 7.142857142857143e-06, |
|
"loss": 0.3247, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 82.0, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8721389174461365, |
|
"eval_runtime": 0.6303, |
|
"eval_samples_per_second": 169.753, |
|
"eval_steps_per_second": 6.346, |
|
"step": 615 |
|
}, |
|
{ |
|
"epoch": 82.66666666666667, |
|
"grad_norm": 7.383735179901123, |
|
"learning_rate": 6.349206349206349e-06, |
|
"loss": 0.2515, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 82.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8605012893676758, |
|
"eval_runtime": 0.6316, |
|
"eval_samples_per_second": 169.406, |
|
"eval_steps_per_second": 6.333, |
|
"step": 622 |
|
}, |
|
{ |
|
"epoch": 84.0, |
|
"grad_norm": 10.514152526855469, |
|
"learning_rate": 5.555555555555556e-06, |
|
"loss": 0.3175, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 84.0, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8504555225372314, |
|
"eval_runtime": 0.6431, |
|
"eval_samples_per_second": 166.371, |
|
"eval_steps_per_second": 6.219, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 84.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8487720489501953, |
|
"eval_runtime": 0.7009, |
|
"eval_samples_per_second": 152.657, |
|
"eval_steps_per_second": 5.707, |
|
"step": 637 |
|
}, |
|
{ |
|
"epoch": 85.33333333333333, |
|
"grad_norm": 7.461284637451172, |
|
"learning_rate": 4.7619047619047615e-06, |
|
"loss": 0.3015, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 86.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8554015755653381, |
|
"eval_runtime": 0.63, |
|
"eval_samples_per_second": 169.848, |
|
"eval_steps_per_second": 6.349, |
|
"step": 645 |
|
}, |
|
{ |
|
"epoch": 86.66666666666667, |
|
"grad_norm": 7.623924732208252, |
|
"learning_rate": 3.968253968253968e-06, |
|
"loss": 0.2989, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 86.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8706844449043274, |
|
"eval_runtime": 0.619, |
|
"eval_samples_per_second": 172.858, |
|
"eval_steps_per_second": 6.462, |
|
"step": 652 |
|
}, |
|
{ |
|
"epoch": 88.0, |
|
"grad_norm": 8.086990356445312, |
|
"learning_rate": 3.1746031746031746e-06, |
|
"loss": 0.3155, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 88.0, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8711610436439514, |
|
"eval_runtime": 0.6221, |
|
"eval_samples_per_second": 171.99, |
|
"eval_steps_per_second": 6.43, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 88.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8659169673919678, |
|
"eval_runtime": 0.6827, |
|
"eval_samples_per_second": 156.733, |
|
"eval_steps_per_second": 5.859, |
|
"step": 667 |
|
}, |
|
{ |
|
"epoch": 89.33333333333333, |
|
"grad_norm": 7.784951210021973, |
|
"learning_rate": 2.3809523809523808e-06, |
|
"loss": 0.2871, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 90.0, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8573119044303894, |
|
"eval_runtime": 0.6276, |
|
"eval_samples_per_second": 170.489, |
|
"eval_steps_per_second": 6.373, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 90.66666666666667, |
|
"grad_norm": 5.8744587898254395, |
|
"learning_rate": 1.5873015873015873e-06, |
|
"loss": 0.2872, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 90.93333333333334, |
|
"eval_accuracy": 0.7289719626168224, |
|
"eval_loss": 0.8529960513114929, |
|
"eval_runtime": 0.6227, |
|
"eval_samples_per_second": 171.823, |
|
"eval_steps_per_second": 6.423, |
|
"step": 682 |
|
}, |
|
{ |
|
"epoch": 92.0, |
|
"grad_norm": 7.09710168838501, |
|
"learning_rate": 7.936507936507937e-07, |
|
"loss": 0.2587, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 92.0, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8516349196434021, |
|
"eval_runtime": 0.6363, |
|
"eval_samples_per_second": 168.151, |
|
"eval_steps_per_second": 6.286, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 92.93333333333334, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8501598238945007, |
|
"eval_runtime": 0.6733, |
|
"eval_samples_per_second": 158.923, |
|
"eval_steps_per_second": 5.941, |
|
"step": 697 |
|
}, |
|
{ |
|
"epoch": 93.33333333333333, |
|
"grad_norm": 9.111827850341797, |
|
"learning_rate": 0.0, |
|
"loss": 0.3133, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 93.33333333333333, |
|
"eval_accuracy": 0.7383177570093458, |
|
"eval_loss": 0.8501192331314087, |
|
"eval_runtime": 0.6288, |
|
"eval_samples_per_second": 170.154, |
|
"eval_steps_per_second": 6.361, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 93.33333333333333, |
|
"step": 700, |
|
"total_flos": 2.226634183539118e+18, |
|
"train_loss": 1.0035276814869472, |
|
"train_runtime": 1522.5733, |
|
"train_samples_per_second": 62.92, |
|
"train_steps_per_second": 0.46 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 700, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 100, |
|
"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": 2.226634183539118e+18, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|