| { | |
| "best_global_step": 800, | |
| "best_metric": 9.927443504333496, | |
| "best_model_checkpoint": "./models/v-004/checkpoint-800", | |
| "epoch": 64.52173913043478, | |
| "eval_steps": 100, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 0.830869197845459, | |
| "learning_rate": 3.166666666666667e-06, | |
| "loss": 10.3478, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.2934782608695652, | |
| "grad_norm": 0.9053573608398438, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 10.3371, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.9456521739130435, | |
| "grad_norm": 0.9227997660636902, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 10.3133, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.5869565217391304, | |
| "grad_norm": 0.8464251160621643, | |
| "learning_rate": 1.3166666666666665e-05, | |
| "loss": 10.2729, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2282608695652173, | |
| "grad_norm": 0.6436305642127991, | |
| "learning_rate": 1.65e-05, | |
| "loss": 10.2092, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.2282608695652173, | |
| "eval_accuracy": 0.0004036802852141104, | |
| "eval_loss": 10.276432037353516, | |
| "eval_runtime": 23.4172, | |
| "eval_samples_per_second": 12.896, | |
| "eval_steps_per_second": 0.811, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.880434782608696, | |
| "grad_norm": 0.5253298282623291, | |
| "learning_rate": 1.9833333333333335e-05, | |
| "loss": 10.1241, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.5082888603210449, | |
| "learning_rate": 2.3166666666666666e-05, | |
| "loss": 10.0315, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.163043478260869, | |
| "grad_norm": 0.4747254252433777, | |
| "learning_rate": 2.6500000000000004e-05, | |
| "loss": 9.9378, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 5.815217391304348, | |
| "grad_norm": 0.45188403129577637, | |
| "learning_rate": 2.9833333333333335e-05, | |
| "loss": 9.8383, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 6.456521739130435, | |
| "grad_norm": 0.4596181809902191, | |
| "learning_rate": 3.316666666666667e-05, | |
| "loss": 9.7373, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.456521739130435, | |
| "eval_accuracy": 0.0019784330809998482, | |
| "eval_loss": 10.077253341674805, | |
| "eval_runtime": 49.8628, | |
| "eval_samples_per_second": 6.057, | |
| "eval_steps_per_second": 0.381, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.0978260869565215, | |
| "grad_norm": 0.43448299169540405, | |
| "learning_rate": 3.65e-05, | |
| "loss": 9.6192, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "grad_norm": 0.4260006546974182, | |
| "learning_rate": 3.983333333333333e-05, | |
| "loss": 9.5207, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 8.391304347826088, | |
| "grad_norm": 0.4230858385562897, | |
| "learning_rate": 4.316666666666667e-05, | |
| "loss": 9.409, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 9.032608695652174, | |
| "grad_norm": 0.6662802696228027, | |
| "learning_rate": 4.6500000000000005e-05, | |
| "loss": 9.3162, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 9.684782608695652, | |
| "grad_norm": 0.46965736150741577, | |
| "learning_rate": 4.9833333333333336e-05, | |
| "loss": 9.2158, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 9.684782608695652, | |
| "eval_accuracy": 0.0023421450211432546, | |
| "eval_loss": 9.940443992614746, | |
| "eval_runtime": 33.824, | |
| "eval_samples_per_second": 8.929, | |
| "eval_steps_per_second": 0.562, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.326086956521738, | |
| "grad_norm": 0.6454819440841675, | |
| "learning_rate": 5.316666666666667e-05, | |
| "loss": 9.1422, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 10.978260869565217, | |
| "grad_norm": 1.104384183883667, | |
| "learning_rate": 5.65e-05, | |
| "loss": 9.0368, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 11.619565217391305, | |
| "grad_norm": 0.6691080927848816, | |
| "learning_rate": 5.983333333333334e-05, | |
| "loss": 8.9517, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 12.26086956521739, | |
| "grad_norm": 0.669059157371521, | |
| "learning_rate": 6.316666666666668e-05, | |
| "loss": 8.8888, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 12.91304347826087, | |
| "grad_norm": 0.8564794659614563, | |
| "learning_rate": 6.65e-05, | |
| "loss": 8.8101, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 12.91304347826087, | |
| "eval_accuracy": 0.002657894947201816, | |
| "eval_loss": 9.944500923156738, | |
| "eval_runtime": 30.6925, | |
| "eval_samples_per_second": 9.84, | |
| "eval_steps_per_second": 0.619, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 13.554347826086957, | |
| "grad_norm": 0.9737327098846436, | |
| "learning_rate": 6.983333333333334e-05, | |
| "loss": 8.7212, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 14.195652173913043, | |
| "grad_norm": 0.8952248096466064, | |
| "learning_rate": 7.316666666666668e-05, | |
| "loss": 8.6471, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 14.847826086956522, | |
| "grad_norm": 1.1972002983093262, | |
| "learning_rate": 7.65e-05, | |
| "loss": 8.5526, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 15.48913043478261, | |
| "grad_norm": 1.041368842124939, | |
| "learning_rate": 7.983333333333334e-05, | |
| "loss": 8.4403, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 16.130434782608695, | |
| "grad_norm": 1.307642936706543, | |
| "learning_rate": 8.316666666666666e-05, | |
| "loss": 8.3439, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 16.130434782608695, | |
| "eval_accuracy": 0.0016307084788847234, | |
| "eval_loss": 9.954625129699707, | |
| "eval_runtime": 28.708, | |
| "eval_samples_per_second": 10.52, | |
| "eval_steps_per_second": 0.662, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 16.782608695652176, | |
| "grad_norm": 1.4131497144699097, | |
| "learning_rate": 8.65e-05, | |
| "loss": 8.2149, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 17.42391304347826, | |
| "grad_norm": 1.2959651947021484, | |
| "learning_rate": 8.983333333333334e-05, | |
| "loss": 8.1152, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 18.065217391304348, | |
| "grad_norm": 1.1717864274978638, | |
| "learning_rate": 9.316666666666666e-05, | |
| "loss": 7.9669, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 18.717391304347824, | |
| "grad_norm": 1.3156694173812866, | |
| "learning_rate": 9.65e-05, | |
| "loss": 7.8164, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 19.358695652173914, | |
| "grad_norm": 1.3559956550598145, | |
| "learning_rate": 9.983333333333334e-05, | |
| "loss": 7.7001, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 19.358695652173914, | |
| "eval_accuracy": 0.0012270281936706128, | |
| "eval_loss": 9.93155574798584, | |
| "eval_runtime": 59.927, | |
| "eval_samples_per_second": 5.039, | |
| "eval_steps_per_second": 0.317, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.7204277515411377, | |
| "learning_rate": 9.995456138403733e-05, | |
| "loss": 7.5688, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 20.652173913043477, | |
| "grad_norm": 1.426324725151062, | |
| "learning_rate": 9.980864681729001e-05, | |
| "loss": 7.4108, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 21.293478260869566, | |
| "grad_norm": 1.4127748012542725, | |
| "learning_rate": 9.956242426451834e-05, | |
| "loss": 7.2349, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 21.945652173913043, | |
| "grad_norm": 1.3751271963119507, | |
| "learning_rate": 9.921638958517565e-05, | |
| "loss": 7.1472, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 22.58695652173913, | |
| "grad_norm": 1.4017088413238525, | |
| "learning_rate": 9.877123964705497e-05, | |
| "loss": 6.9652, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 22.58695652173913, | |
| "eval_accuracy": 0.0007234270457797424, | |
| "eval_loss": 9.93958568572998, | |
| "eval_runtime": 59.9271, | |
| "eval_samples_per_second": 5.039, | |
| "eval_steps_per_second": 0.317, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 23.22826086956522, | |
| "grad_norm": 1.25152587890625, | |
| "learning_rate": 9.822787092288991e-05, | |
| "loss": 6.8018, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 23.880434782608695, | |
| "grad_norm": 1.3836736679077148, | |
| "learning_rate": 9.758737768497802e-05, | |
| "loss": 6.6247, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 24.52173913043478, | |
| "grad_norm": 1.4601590633392334, | |
| "learning_rate": 9.685104980146193e-05, | |
| "loss": 6.5264, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 25.16304347826087, | |
| "grad_norm": 1.3552601337432861, | |
| "learning_rate": 9.60203701387066e-05, | |
| "loss": 6.3179, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 25.815217391304348, | |
| "grad_norm": 1.391802191734314, | |
| "learning_rate": 9.509701157500376e-05, | |
| "loss": 6.2067, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 25.815217391304348, | |
| "eval_accuracy": 0.0007274238802868129, | |
| "eval_loss": 9.927443504333496, | |
| "eval_runtime": 32.7444, | |
| "eval_samples_per_second": 9.223, | |
| "eval_steps_per_second": 0.58, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 26.456521739130434, | |
| "grad_norm": 1.4579627513885498, | |
| "learning_rate": 9.408283363161774e-05, | |
| "loss": 5.9989, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 27.097826086956523, | |
| "grad_norm": 1.6435540914535522, | |
| "learning_rate": 9.297987872795705e-05, | |
| "loss": 5.9059, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 27.75, | |
| "grad_norm": 1.4759621620178223, | |
| "learning_rate": 9.179036806841353e-05, | |
| "loss": 5.7574, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 28.391304347826086, | |
| "grad_norm": 1.960924506187439, | |
| "learning_rate": 9.051669716915227e-05, | |
| "loss": 5.6674, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 29.032608695652176, | |
| "grad_norm": 1.803576111793518, | |
| "learning_rate": 8.916143103386093e-05, | |
| "loss": 5.5185, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 29.032608695652176, | |
| "eval_accuracy": 0.000659477693666616, | |
| "eval_loss": 9.943469047546387, | |
| "eval_runtime": 44.712, | |
| "eval_samples_per_second": 6.754, | |
| "eval_steps_per_second": 0.425, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 29.684782608695652, | |
| "grad_norm": 1.7185384035110474, | |
| "learning_rate": 8.77272989881736e-05, | |
| "loss": 5.3861, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 30.32608695652174, | |
| "grad_norm": 1.8554881811141968, | |
| "learning_rate": 8.621718918317225e-05, | |
| "loss": 5.1988, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 30.97826086956522, | |
| "grad_norm": 1.824589490890503, | |
| "learning_rate": 8.463414277903475e-05, | |
| "loss": 5.1284, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 31.619565217391305, | |
| "grad_norm": 1.7106842994689941, | |
| "learning_rate": 8.298134782054305e-05, | |
| "loss": 4.9489, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 32.26086956521739, | |
| "grad_norm": 2.0789778232574463, | |
| "learning_rate": 8.126213281678526e-05, | |
| "loss": 4.8318, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 32.26086956521739, | |
| "eval_accuracy": 0.0004436486302848144, | |
| "eval_loss": 9.991825103759766, | |
| "eval_runtime": 36.8715, | |
| "eval_samples_per_second": 8.191, | |
| "eval_steps_per_second": 0.515, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 32.91304347826087, | |
| "grad_norm": 1.6994128227233887, | |
| "learning_rate": 7.94799600379813e-05, | |
| "loss": 4.7451, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 33.55434782608695, | |
| "grad_norm": 1.636572241783142, | |
| "learning_rate": 7.763841854293145e-05, | |
| "loss": 4.7058, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 34.19565217391305, | |
| "grad_norm": 1.6913096904754639, | |
| "learning_rate": 7.574121695112954e-05, | |
| "loss": 4.5013, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 34.84782608695652, | |
| "grad_norm": 1.8612340688705444, | |
| "learning_rate": 7.379217597409688e-05, | |
| "loss": 4.3998, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 35.48913043478261, | |
| "grad_norm": 1.8019436597824097, | |
| "learning_rate": 7.179522072097774e-05, | |
| "loss": 4.343, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 35.48913043478261, | |
| "eval_accuracy": 0.0004476454647918848, | |
| "eval_loss": 10.02548885345459, | |
| "eval_runtime": 43.794, | |
| "eval_samples_per_second": 6.896, | |
| "eval_steps_per_second": 0.434, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 36.130434782608695, | |
| "grad_norm": 2.059352159500122, | |
| "learning_rate": 6.975437279389181e-05, | |
| "loss": 4.231, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 36.78260869565217, | |
| "grad_norm": 1.8938504457473755, | |
| "learning_rate": 6.767374218896286e-05, | |
| "loss": 4.1478, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 37.42391304347826, | |
| "grad_norm": 1.8322842121124268, | |
| "learning_rate": 6.555751901933342e-05, | |
| "loss": 4.0147, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 38.06521739130435, | |
| "grad_norm": 1.706099271774292, | |
| "learning_rate": 6.340996507683458e-05, | |
| "loss": 3.9337, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 38.71739130434783, | |
| "grad_norm": 1.79011070728302, | |
| "learning_rate": 6.123540524930442e-05, | |
| "loss": 3.9477, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 38.71739130434783, | |
| "eval_accuracy": 0.00040767711972118084, | |
| "eval_loss": 10.079157829284668, | |
| "eval_runtime": 47.0549, | |
| "eval_samples_per_second": 6.418, | |
| "eval_steps_per_second": 0.404, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 39.358695652173914, | |
| "grad_norm": 1.7470873594284058, | |
| "learning_rate": 5.903821881083942e-05, | |
| "loss": 3.7965, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 2.18630313873291, | |
| "learning_rate": 5.682283060251932e-05, | |
| "loss": 3.6986, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 40.65217391304348, | |
| "grad_norm": 1.515807032585144, | |
| "learning_rate": 5.4593702121365955e-05, | |
| "loss": 3.6499, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 41.29347826086956, | |
| "grad_norm": 1.7721681594848633, | |
| "learning_rate": 5.235532253548213e-05, | |
| "loss": 3.559, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 41.94565217391305, | |
| "grad_norm": 1.5675290822982788, | |
| "learning_rate": 5.0112199643464376e-05, | |
| "loss": 3.5394, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 41.94565217391305, | |
| "eval_accuracy": 0.00018785122183230882, | |
| "eval_loss": 10.124716758728027, | |
| "eval_runtime": 92.8739, | |
| "eval_samples_per_second": 3.252, | |
| "eval_steps_per_second": 0.205, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 42.58695652173913, | |
| "grad_norm": 1.6255255937576294, | |
| "learning_rate": 4.7868850796296495e-05, | |
| "loss": 3.5101, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 43.22826086956522, | |
| "grad_norm": 1.5351227521896362, | |
| "learning_rate": 4.5629793800005945e-05, | |
| "loss": 3.3166, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 43.880434782608695, | |
| "grad_norm": 1.4253877401351929, | |
| "learning_rate": 4.339953781740363e-05, | |
| "loss": 3.3786, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 44.52173913043478, | |
| "grad_norm": 1.4174528121948242, | |
| "learning_rate": 4.1182574287230224e-05, | |
| "loss": 3.2175, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 45.16304347826087, | |
| "grad_norm": 1.6721420288085938, | |
| "learning_rate": 3.898336787899612e-05, | |
| "loss": 3.2964, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 45.16304347826087, | |
| "eval_accuracy": 0.000319746760565632, | |
| "eval_loss": 10.182437896728516, | |
| "eval_runtime": 76.0884, | |
| "eval_samples_per_second": 3.969, | |
| "eval_steps_per_second": 0.25, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 45.81521739130435, | |
| "grad_norm": 1.4447053670883179, | |
| "learning_rate": 3.680634750173137e-05, | |
| "loss": 3.1779, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 46.45652173913044, | |
| "grad_norm": 1.295456051826477, | |
| "learning_rate": 3.4655897384752146e-05, | |
| "loss": 3.1888, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 47.09782608695652, | |
| "grad_norm": 1.1828927993774414, | |
| "learning_rate": 3.2536348248406534e-05, | |
| "loss": 3.0698, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 47.75, | |
| "grad_norm": 1.3669146299362183, | |
| "learning_rate": 3.0451968582579915e-05, | |
| "loss": 3.0414, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 48.391304347826086, | |
| "grad_norm": 1.146458387374878, | |
| "learning_rate": 2.840695605052458e-05, | |
| "loss": 3.0237, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 48.391304347826086, | |
| "eval_accuracy": 0.00031574992605856164, | |
| "eval_loss": 10.224663734436035, | |
| "eval_runtime": 83.422, | |
| "eval_samples_per_second": 3.62, | |
| "eval_steps_per_second": 0.228, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 49.03260869565217, | |
| "grad_norm": 1.1331515312194824, | |
| "learning_rate": 2.6405429035324403e-05, | |
| "loss": 3.0009, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 49.68478260869565, | |
| "grad_norm": 1.286600112915039, | |
| "learning_rate": 2.4451418346019573e-05, | |
| "loss": 2.9825, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 50.32608695652174, | |
| "grad_norm": 1.0778050422668457, | |
| "learning_rate": 2.2548859100093407e-05, | |
| "loss": 2.9051, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 50.97826086956522, | |
| "grad_norm": 1.0530340671539307, | |
| "learning_rate": 2.0701582798669676e-05, | |
| "loss": 2.8988, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 51.619565217391305, | |
| "grad_norm": 1.0122408866882324, | |
| "learning_rate": 1.8913309610379015e-05, | |
| "loss": 2.8621, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 51.619565217391305, | |
| "eval_accuracy": 0.0002917689190161392, | |
| "eval_loss": 10.255624771118164, | |
| "eval_runtime": 76.4508, | |
| "eval_samples_per_second": 3.95, | |
| "eval_steps_per_second": 0.249, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 52.26086956521739, | |
| "grad_norm": 1.008852243423462, | |
| "learning_rate": 1.7187640879434553e-05, | |
| "loss": 2.845, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 52.91304347826087, | |
| "grad_norm": 1.016990303993225, | |
| "learning_rate": 1.552805187300389e-05, | |
| "loss": 2.8895, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 53.55434782608695, | |
| "grad_norm": 0.9810470342636108, | |
| "learning_rate": 1.3937884782483484e-05, | |
| "loss": 2.7809, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 54.19565217391305, | |
| "grad_norm": 1.027442216873169, | |
| "learning_rate": 1.242034199277008e-05, | |
| "loss": 2.8908, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 54.84782608695652, | |
| "grad_norm": 1.0243561267852783, | |
| "learning_rate": 1.097847963308351e-05, | |
| "loss": 2.765, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 54.84782608695652, | |
| "eval_accuracy": 0.0001438860422545344, | |
| "eval_loss": 10.27558422088623, | |
| "eval_runtime": 50.0398, | |
| "eval_samples_per_second": 6.035, | |
| "eval_steps_per_second": 0.38, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 55.48913043478261, | |
| "grad_norm": 0.9682217240333557, | |
| "learning_rate": 9.615201422329406e-06, | |
| "loss": 2.726, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 56.130434782608695, | |
| "grad_norm": 0.9422224760055542, | |
| "learning_rate": 8.333252821395526e-06, | |
| "loss": 2.7822, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 56.78260869565217, | |
| "grad_norm": 0.9879501461982727, | |
| "learning_rate": 7.135215504159115e-06, | |
| "loss": 2.8115, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 57.42391304347826, | |
| "grad_norm": 0.938834011554718, | |
| "learning_rate": 6.023502158339078e-06, | |
| "loss": 2.7159, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 58.06521739130435, | |
| "grad_norm": 0.9302940368652344, | |
| "learning_rate": 5.000351626664207e-06, | |
| "loss": 2.7383, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 58.06521739130435, | |
| "eval_accuracy": 0.0003037594225373504, | |
| "eval_loss": 10.28667163848877, | |
| "eval_runtime": 26.8193, | |
| "eval_samples_per_second": 11.261, | |
| "eval_steps_per_second": 0.708, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 58.71739130434783, | |
| "grad_norm": 0.9569380879402161, | |
| "learning_rate": 4.067824398141701e-06, | |
| "loss": 2.7457, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 59.358695652173914, | |
| "grad_norm": 0.9395641088485718, | |
| "learning_rate": 3.2277984585066366e-06, | |
| "loss": 2.7463, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 1.2450119256973267, | |
| "learning_rate": 2.4819655082085835e-06, | |
| "loss": 2.7273, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 60.65217391304348, | |
| "grad_norm": 1.0745036602020264, | |
| "learning_rate": 1.8318275555520237e-06, | |
| "loss": 2.6887, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 61.29347826086956, | |
| "grad_norm": 0.9726008176803589, | |
| "learning_rate": 1.2786938918515568e-06, | |
| "loss": 2.7324, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 61.29347826086956, | |
| "eval_accuracy": 0.0003037594225373504, | |
| "eval_loss": 10.290609359741211, | |
| "eval_runtime": 42.7369, | |
| "eval_samples_per_second": 7.067, | |
| "eval_steps_per_second": 0.445, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 61.94565217391305, | |
| "grad_norm": 0.9399953484535217, | |
| "learning_rate": 8.236784546933718e-07, | |
| "loss": 2.7602, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 62.58695652173913, | |
| "grad_norm": 0.9184156656265259, | |
| "learning_rate": 4.676975846132692e-07, | |
| "loss": 2.7586, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 63.22826086956522, | |
| "grad_norm": 0.9342102408409119, | |
| "learning_rate": 2.1146817970871258e-07, | |
| "loss": 2.6681, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 63.880434782608695, | |
| "grad_norm": 0.8547150492668152, | |
| "learning_rate": 5.550625190150483e-08, | |
| "loss": 2.7384, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 64.52173913043478, | |
| "grad_norm": 0.9187237024307251, | |
| "learning_rate": 1.2588775841204658e-10, | |
| "loss": 2.6918, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 64.52173913043478, | |
| "eval_accuracy": 0.0003037594225373504, | |
| "eval_loss": 10.291138648986816, | |
| "eval_runtime": 106.7866, | |
| "eval_samples_per_second": 2.828, | |
| "eval_steps_per_second": 0.178, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 65, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4054390970018304.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |