| { | |
| "best_metric": 0.8066528066528067, | |
| "best_model_checkpoint": "test-hasy-5/checkpoint-18935", | |
| "epoch": 100.0, | |
| "eval_steps": 500, | |
| "global_step": 54100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.6960930824279785, | |
| "learning_rate": 1.9815157116451017e-05, | |
| "loss": 3.9645, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.3970893970893971, | |
| "eval_loss": 3.429507255554199, | |
| "eval_runtime": 2.0488, | |
| "eval_samples_per_second": 234.77, | |
| "eval_steps_per_second": 29.773, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 3.364806652069092, | |
| "learning_rate": 1.9630314232902035e-05, | |
| "loss": 3.4258, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.4781704781704782, | |
| "eval_loss": 2.879011392593384, | |
| "eval_runtime": 1.9708, | |
| "eval_samples_per_second": 244.059, | |
| "eval_steps_per_second": 30.951, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 3.35432767868042, | |
| "learning_rate": 1.944547134935305e-05, | |
| "loss": 3.04, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5467775467775468, | |
| "eval_loss": 2.4893012046813965, | |
| "eval_runtime": 2.0175, | |
| "eval_samples_per_second": 238.411, | |
| "eval_steps_per_second": 30.235, | |
| "step": 1623 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 4.459615230560303, | |
| "learning_rate": 1.9260628465804068e-05, | |
| "loss": 2.793, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5738045738045738, | |
| "eval_loss": 2.2005958557128906, | |
| "eval_runtime": 1.962, | |
| "eval_samples_per_second": 245.161, | |
| "eval_steps_per_second": 31.091, | |
| "step": 2164 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 4.481846332550049, | |
| "learning_rate": 1.9075785582255083e-05, | |
| "loss": 2.5551, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6340956340956341, | |
| "eval_loss": 1.9055824279785156, | |
| "eval_runtime": 2.0221, | |
| "eval_samples_per_second": 237.875, | |
| "eval_steps_per_second": 30.167, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 5.55, | |
| "grad_norm": 4.1784281730651855, | |
| "learning_rate": 1.88909426987061e-05, | |
| "loss": 2.3662, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6632016632016632, | |
| "eval_loss": 1.7023240327835083, | |
| "eval_runtime": 1.9666, | |
| "eval_samples_per_second": 244.588, | |
| "eval_steps_per_second": 31.018, | |
| "step": 3246 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "grad_norm": 6.06294059753418, | |
| "learning_rate": 1.8706099815157116e-05, | |
| "loss": 2.1965, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6798336798336798, | |
| "eval_loss": 1.5739575624465942, | |
| "eval_runtime": 1.9941, | |
| "eval_samples_per_second": 241.216, | |
| "eval_steps_per_second": 30.591, | |
| "step": 3787 | |
| }, | |
| { | |
| "epoch": 7.39, | |
| "grad_norm": 4.912960052490234, | |
| "learning_rate": 1.8521256931608135e-05, | |
| "loss": 2.1397, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6943866943866944, | |
| "eval_loss": 1.4560521841049194, | |
| "eval_runtime": 1.9835, | |
| "eval_samples_per_second": 242.507, | |
| "eval_steps_per_second": 30.754, | |
| "step": 4328 | |
| }, | |
| { | |
| "epoch": 8.32, | |
| "grad_norm": 5.236889362335205, | |
| "learning_rate": 1.833641404805915e-05, | |
| "loss": 1.9955, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7234927234927235, | |
| "eval_loss": 1.3202540874481201, | |
| "eval_runtime": 2.0536, | |
| "eval_samples_per_second": 234.218, | |
| "eval_steps_per_second": 29.703, | |
| "step": 4869 | |
| }, | |
| { | |
| "epoch": 9.24, | |
| "grad_norm": 5.675503253936768, | |
| "learning_rate": 1.8151571164510168e-05, | |
| "loss": 1.9282, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.738045738045738, | |
| "eval_loss": 1.2246184349060059, | |
| "eval_runtime": 2.0017, | |
| "eval_samples_per_second": 240.293, | |
| "eval_steps_per_second": 30.474, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 10.17, | |
| "grad_norm": 4.67825174331665, | |
| "learning_rate": 1.7966728280961186e-05, | |
| "loss": 1.8368, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.738045738045738, | |
| "eval_loss": 1.1823257207870483, | |
| "eval_runtime": 1.9774, | |
| "eval_samples_per_second": 243.246, | |
| "eval_steps_per_second": 30.848, | |
| "step": 5951 | |
| }, | |
| { | |
| "epoch": 11.09, | |
| "grad_norm": 4.809859275817871, | |
| "learning_rate": 1.77818853974122e-05, | |
| "loss": 1.812, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7214137214137214, | |
| "eval_loss": 1.1297953128814697, | |
| "eval_runtime": 2.0307, | |
| "eval_samples_per_second": 236.864, | |
| "eval_steps_per_second": 30.039, | |
| "step": 6492 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 5.255190849304199, | |
| "learning_rate": 1.759704251386322e-05, | |
| "loss": 1.7353, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.94, | |
| "grad_norm": 8.597217559814453, | |
| "learning_rate": 1.7412199630314234e-05, | |
| "loss": 1.7195, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7484407484407485, | |
| "eval_loss": 1.0423070192337036, | |
| "eval_runtime": 2.0193, | |
| "eval_samples_per_second": 238.201, | |
| "eval_steps_per_second": 30.208, | |
| "step": 7033 | |
| }, | |
| { | |
| "epoch": 13.86, | |
| "grad_norm": 6.453842639923096, | |
| "learning_rate": 1.7227356746765253e-05, | |
| "loss": 1.6314, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7422037422037422, | |
| "eval_loss": 1.0077309608459473, | |
| "eval_runtime": 2.0783, | |
| "eval_samples_per_second": 231.439, | |
| "eval_steps_per_second": 29.351, | |
| "step": 7574 | |
| }, | |
| { | |
| "epoch": 14.79, | |
| "grad_norm": 8.70645523071289, | |
| "learning_rate": 1.7042513863216268e-05, | |
| "loss": 1.5979, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7463617463617463, | |
| "eval_loss": 1.00509512424469, | |
| "eval_runtime": 1.9889, | |
| "eval_samples_per_second": 241.847, | |
| "eval_steps_per_second": 30.671, | |
| "step": 8115 | |
| }, | |
| { | |
| "epoch": 15.71, | |
| "grad_norm": 7.348147392272949, | |
| "learning_rate": 1.6857670979667286e-05, | |
| "loss": 1.5656, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7692307692307693, | |
| "eval_loss": 0.9325113296508789, | |
| "eval_runtime": 1.9923, | |
| "eval_samples_per_second": 241.43, | |
| "eval_steps_per_second": 30.618, | |
| "step": 8656 | |
| }, | |
| { | |
| "epoch": 16.64, | |
| "grad_norm": 6.420931816101074, | |
| "learning_rate": 1.66728280961183e-05, | |
| "loss": 1.5414, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7733887733887734, | |
| "eval_loss": 0.8889437913894653, | |
| "eval_runtime": 1.9727, | |
| "eval_samples_per_second": 243.822, | |
| "eval_steps_per_second": 30.921, | |
| "step": 9197 | |
| }, | |
| { | |
| "epoch": 17.56, | |
| "grad_norm": 8.127350807189941, | |
| "learning_rate": 1.6487985212569316e-05, | |
| "loss": 1.5342, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7484407484407485, | |
| "eval_loss": 0.9072721600532532, | |
| "eval_runtime": 1.9933, | |
| "eval_samples_per_second": 241.303, | |
| "eval_steps_per_second": 30.602, | |
| "step": 9738 | |
| }, | |
| { | |
| "epoch": 18.48, | |
| "grad_norm": 6.122061252593994, | |
| "learning_rate": 1.6303142329020334e-05, | |
| "loss": 1.4898, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7713097713097713, | |
| "eval_loss": 0.8425627946853638, | |
| "eval_runtime": 1.9868, | |
| "eval_samples_per_second": 242.099, | |
| "eval_steps_per_second": 30.703, | |
| "step": 10279 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "grad_norm": 6.640945911407471, | |
| "learning_rate": 1.611829944547135e-05, | |
| "loss": 1.4731, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7442827442827443, | |
| "eval_loss": 0.862506091594696, | |
| "eval_runtime": 1.9786, | |
| "eval_samples_per_second": 243.096, | |
| "eval_steps_per_second": 30.829, | |
| "step": 10820 | |
| }, | |
| { | |
| "epoch": 20.33, | |
| "grad_norm": 6.019400119781494, | |
| "learning_rate": 1.5933456561922367e-05, | |
| "loss": 1.451, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.762993762993763, | |
| "eval_loss": 0.8015209436416626, | |
| "eval_runtime": 1.9644, | |
| "eval_samples_per_second": 244.864, | |
| "eval_steps_per_second": 31.053, | |
| "step": 11361 | |
| }, | |
| { | |
| "epoch": 21.26, | |
| "grad_norm": 5.140503406524658, | |
| "learning_rate": 1.5748613678373382e-05, | |
| "loss": 1.4578, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.7588357588357588, | |
| "eval_loss": 0.8520306944847107, | |
| "eval_runtime": 2.0001, | |
| "eval_samples_per_second": 240.484, | |
| "eval_steps_per_second": 30.498, | |
| "step": 11902 | |
| }, | |
| { | |
| "epoch": 22.18, | |
| "grad_norm": 15.190984725952148, | |
| "learning_rate": 1.55637707948244e-05, | |
| "loss": 1.4126, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.7713097713097713, | |
| "eval_loss": 0.7928301692008972, | |
| "eval_runtime": 1.9822, | |
| "eval_samples_per_second": 242.66, | |
| "eval_steps_per_second": 30.774, | |
| "step": 12443 | |
| }, | |
| { | |
| "epoch": 23.11, | |
| "grad_norm": 11.220525741577148, | |
| "learning_rate": 1.5378927911275416e-05, | |
| "loss": 1.3626, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7837837837837838, | |
| "eval_loss": 0.754388689994812, | |
| "eval_runtime": 1.975, | |
| "eval_samples_per_second": 243.545, | |
| "eval_steps_per_second": 30.886, | |
| "step": 12984 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "grad_norm": 3.5185582637786865, | |
| "learning_rate": 1.5194085027726432e-05, | |
| "loss": 1.3905, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 24.95, | |
| "grad_norm": 8.19352912902832, | |
| "learning_rate": 1.5009242144177449e-05, | |
| "loss": 1.3694, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.7775467775467776, | |
| "eval_loss": 0.7698755860328674, | |
| "eval_runtime": 2.0179, | |
| "eval_samples_per_second": 238.368, | |
| "eval_steps_per_second": 30.23, | |
| "step": 13525 | |
| }, | |
| { | |
| "epoch": 25.88, | |
| "grad_norm": 6.003907680511475, | |
| "learning_rate": 1.4824399260628467e-05, | |
| "loss": 1.3612, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.7775467775467776, | |
| "eval_loss": 0.7602183818817139, | |
| "eval_runtime": 1.9833, | |
| "eval_samples_per_second": 242.521, | |
| "eval_steps_per_second": 30.756, | |
| "step": 14066 | |
| }, | |
| { | |
| "epoch": 26.8, | |
| "grad_norm": 6.613931655883789, | |
| "learning_rate": 1.4639556377079484e-05, | |
| "loss": 1.2963, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.7713097713097713, | |
| "eval_loss": 0.7532169818878174, | |
| "eval_runtime": 2.0706, | |
| "eval_samples_per_second": 232.305, | |
| "eval_steps_per_second": 29.461, | |
| "step": 14607 | |
| }, | |
| { | |
| "epoch": 27.73, | |
| "grad_norm": 7.66683292388916, | |
| "learning_rate": 1.44547134935305e-05, | |
| "loss": 1.3009, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.7012535929679871, | |
| "eval_runtime": 1.9606, | |
| "eval_samples_per_second": 245.335, | |
| "eval_steps_per_second": 31.113, | |
| "step": 15148 | |
| }, | |
| { | |
| "epoch": 28.65, | |
| "grad_norm": 7.342077255249023, | |
| "learning_rate": 1.4269870609981517e-05, | |
| "loss": 1.2598, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.7796257796257796, | |
| "eval_loss": 0.7084705233573914, | |
| "eval_runtime": 1.9824, | |
| "eval_samples_per_second": 242.632, | |
| "eval_steps_per_second": 30.77, | |
| "step": 15689 | |
| }, | |
| { | |
| "epoch": 29.57, | |
| "grad_norm": 5.679790019989014, | |
| "learning_rate": 1.4085027726432534e-05, | |
| "loss": 1.2565, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.7775467775467776, | |
| "eval_loss": 0.7023281455039978, | |
| "eval_runtime": 1.9659, | |
| "eval_samples_per_second": 244.668, | |
| "eval_steps_per_second": 31.029, | |
| "step": 16230 | |
| }, | |
| { | |
| "epoch": 30.5, | |
| "grad_norm": 5.493412971496582, | |
| "learning_rate": 1.390018484288355e-05, | |
| "loss": 1.2735, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.7775467775467776, | |
| "eval_loss": 0.7047860026359558, | |
| "eval_runtime": 1.9718, | |
| "eval_samples_per_second": 243.937, | |
| "eval_steps_per_second": 30.936, | |
| "step": 16771 | |
| }, | |
| { | |
| "epoch": 31.42, | |
| "grad_norm": 6.2688093185424805, | |
| "learning_rate": 1.3715341959334567e-05, | |
| "loss": 1.2743, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6794067621231079, | |
| "eval_runtime": 1.9764, | |
| "eval_samples_per_second": 243.372, | |
| "eval_steps_per_second": 30.864, | |
| "step": 17312 | |
| }, | |
| { | |
| "epoch": 32.35, | |
| "grad_norm": 10.169917106628418, | |
| "learning_rate": 1.3530499075785584e-05, | |
| "loss": 1.2441, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.7858627858627859, | |
| "eval_loss": 0.693196713924408, | |
| "eval_runtime": 1.972, | |
| "eval_samples_per_second": 243.92, | |
| "eval_steps_per_second": 30.934, | |
| "step": 17853 | |
| }, | |
| { | |
| "epoch": 33.27, | |
| "grad_norm": 8.05045223236084, | |
| "learning_rate": 1.33456561922366e-05, | |
| "loss": 1.2282, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.7038751840591431, | |
| "eval_runtime": 1.992, | |
| "eval_samples_per_second": 241.466, | |
| "eval_steps_per_second": 30.623, | |
| "step": 18394 | |
| }, | |
| { | |
| "epoch": 34.2, | |
| "grad_norm": 5.410665035247803, | |
| "learning_rate": 1.3160813308687617e-05, | |
| "loss": 1.2204, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.8066528066528067, | |
| "eval_loss": 0.6860660910606384, | |
| "eval_runtime": 1.9773, | |
| "eval_samples_per_second": 243.258, | |
| "eval_steps_per_second": 30.85, | |
| "step": 18935 | |
| }, | |
| { | |
| "epoch": 35.12, | |
| "grad_norm": 11.123208045959473, | |
| "learning_rate": 1.2975970425138634e-05, | |
| "loss": 1.1808, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.6589930057525635, | |
| "eval_runtime": 1.9769, | |
| "eval_samples_per_second": 243.305, | |
| "eval_steps_per_second": 30.856, | |
| "step": 19476 | |
| }, | |
| { | |
| "epoch": 36.04, | |
| "grad_norm": 6.165465354919434, | |
| "learning_rate": 1.279112754158965e-05, | |
| "loss": 1.1933, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 36.97, | |
| "grad_norm": 6.407803535461426, | |
| "learning_rate": 1.2606284658040667e-05, | |
| "loss": 1.1928, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.7817047817047817, | |
| "eval_loss": 0.678415834903717, | |
| "eval_runtime": 1.9844, | |
| "eval_samples_per_second": 242.388, | |
| "eval_steps_per_second": 30.739, | |
| "step": 20017 | |
| }, | |
| { | |
| "epoch": 37.89, | |
| "grad_norm": 4.849668979644775, | |
| "learning_rate": 1.2421441774491683e-05, | |
| "loss": 1.1914, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.6559053659439087, | |
| "eval_runtime": 1.9912, | |
| "eval_samples_per_second": 241.56, | |
| "eval_steps_per_second": 30.634, | |
| "step": 20558 | |
| }, | |
| { | |
| "epoch": 38.82, | |
| "grad_norm": 9.1309232711792, | |
| "learning_rate": 1.2236598890942698e-05, | |
| "loss": 1.1856, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.6769025325775146, | |
| "eval_runtime": 2.0066, | |
| "eval_samples_per_second": 239.713, | |
| "eval_steps_per_second": 30.4, | |
| "step": 21099 | |
| }, | |
| { | |
| "epoch": 39.74, | |
| "grad_norm": 5.001546382904053, | |
| "learning_rate": 1.2051756007393715e-05, | |
| "loss": 1.1585, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.8004158004158004, | |
| "eval_loss": 0.64976966381073, | |
| "eval_runtime": 1.9804, | |
| "eval_samples_per_second": 242.874, | |
| "eval_steps_per_second": 30.801, | |
| "step": 21640 | |
| }, | |
| { | |
| "epoch": 40.67, | |
| "grad_norm": 14.044866561889648, | |
| "learning_rate": 1.1866913123844732e-05, | |
| "loss": 1.1713, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6447434425354004, | |
| "eval_runtime": 1.9973, | |
| "eval_samples_per_second": 240.829, | |
| "eval_steps_per_second": 30.542, | |
| "step": 22181 | |
| }, | |
| { | |
| "epoch": 41.59, | |
| "grad_norm": 10.289350509643555, | |
| "learning_rate": 1.1682070240295748e-05, | |
| "loss": 1.1183, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.7713097713097713, | |
| "eval_loss": 0.6748064756393433, | |
| "eval_runtime": 1.9672, | |
| "eval_samples_per_second": 244.509, | |
| "eval_steps_per_second": 31.008, | |
| "step": 22722 | |
| }, | |
| { | |
| "epoch": 42.51, | |
| "grad_norm": 12.7116117477417, | |
| "learning_rate": 1.1497227356746765e-05, | |
| "loss": 1.1564, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6545261740684509, | |
| "eval_runtime": 1.9659, | |
| "eval_samples_per_second": 244.669, | |
| "eval_steps_per_second": 31.029, | |
| "step": 23263 | |
| }, | |
| { | |
| "epoch": 43.44, | |
| "grad_norm": 3.0720624923706055, | |
| "learning_rate": 1.1312384473197783e-05, | |
| "loss": 1.1215, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.7879417879417879, | |
| "eval_loss": 0.6690270900726318, | |
| "eval_runtime": 1.9635, | |
| "eval_samples_per_second": 244.971, | |
| "eval_steps_per_second": 31.067, | |
| "step": 23804 | |
| }, | |
| { | |
| "epoch": 44.36, | |
| "grad_norm": 7.927094459533691, | |
| "learning_rate": 1.11275415896488e-05, | |
| "loss": 1.1008, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_accuracy": 0.7879417879417879, | |
| "eval_loss": 0.659792423248291, | |
| "eval_runtime": 1.9747, | |
| "eval_samples_per_second": 243.578, | |
| "eval_steps_per_second": 30.89, | |
| "step": 24345 | |
| }, | |
| { | |
| "epoch": 45.29, | |
| "grad_norm": 8.912357330322266, | |
| "learning_rate": 1.0942698706099817e-05, | |
| "loss": 1.1344, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_accuracy": 0.8024948024948025, | |
| "eval_loss": 0.6550182104110718, | |
| "eval_runtime": 2.0112, | |
| "eval_samples_per_second": 239.156, | |
| "eval_steps_per_second": 30.33, | |
| "step": 24886 | |
| }, | |
| { | |
| "epoch": 46.21, | |
| "grad_norm": 9.598004341125488, | |
| "learning_rate": 1.0757855822550833e-05, | |
| "loss": 1.126, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "eval_accuracy": 0.7858627858627859, | |
| "eval_loss": 0.6521425247192383, | |
| "eval_runtime": 1.9713, | |
| "eval_samples_per_second": 244.004, | |
| "eval_steps_per_second": 30.944, | |
| "step": 25427 | |
| }, | |
| { | |
| "epoch": 47.13, | |
| "grad_norm": 4.670881271362305, | |
| "learning_rate": 1.057301293900185e-05, | |
| "loss": 1.125, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_accuracy": 0.7817047817047817, | |
| "eval_loss": 0.6812848448753357, | |
| "eval_runtime": 2.016, | |
| "eval_samples_per_second": 238.588, | |
| "eval_steps_per_second": 30.258, | |
| "step": 25968 | |
| }, | |
| { | |
| "epoch": 48.06, | |
| "grad_norm": 8.11451244354248, | |
| "learning_rate": 1.0388170055452866e-05, | |
| "loss": 1.0682, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 48.98, | |
| "grad_norm": 8.960821151733398, | |
| "learning_rate": 1.0203327171903883e-05, | |
| "loss": 1.0855, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "eval_accuracy": 0.7858627858627859, | |
| "eval_loss": 0.6419298052787781, | |
| "eval_runtime": 1.974, | |
| "eval_samples_per_second": 243.673, | |
| "eval_steps_per_second": 30.902, | |
| "step": 26509 | |
| }, | |
| { | |
| "epoch": 49.91, | |
| "grad_norm": 3.053118944168091, | |
| "learning_rate": 1.00184842883549e-05, | |
| "loss": 1.0452, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_accuracy": 0.8004158004158004, | |
| "eval_loss": 0.6550863981246948, | |
| "eval_runtime": 2.0504, | |
| "eval_samples_per_second": 234.587, | |
| "eval_steps_per_second": 29.75, | |
| "step": 27050 | |
| }, | |
| { | |
| "epoch": 50.83, | |
| "grad_norm": 5.4594340324401855, | |
| "learning_rate": 9.833641404805916e-06, | |
| "loss": 1.0626, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6675499081611633, | |
| "eval_runtime": 1.9991, | |
| "eval_samples_per_second": 240.605, | |
| "eval_steps_per_second": 30.513, | |
| "step": 27591 | |
| }, | |
| { | |
| "epoch": 51.76, | |
| "grad_norm": 8.158236503601074, | |
| "learning_rate": 9.648798521256933e-06, | |
| "loss": 1.0155, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6945971846580505, | |
| "eval_runtime": 1.9873, | |
| "eval_samples_per_second": 242.042, | |
| "eval_steps_per_second": 30.696, | |
| "step": 28132 | |
| }, | |
| { | |
| "epoch": 52.68, | |
| "grad_norm": 5.626604080200195, | |
| "learning_rate": 9.46395563770795e-06, | |
| "loss": 1.0319, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "eval_accuracy": 0.7796257796257796, | |
| "eval_loss": 0.6942130923271179, | |
| "eval_runtime": 1.966, | |
| "eval_samples_per_second": 244.665, | |
| "eval_steps_per_second": 31.028, | |
| "step": 28673 | |
| }, | |
| { | |
| "epoch": 53.6, | |
| "grad_norm": 6.82182502746582, | |
| "learning_rate": 9.279112754158966e-06, | |
| "loss": 1.0488, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "eval_accuracy": 0.7983367983367984, | |
| "eval_loss": 0.6496003866195679, | |
| "eval_runtime": 2.0069, | |
| "eval_samples_per_second": 239.673, | |
| "eval_steps_per_second": 30.395, | |
| "step": 29214 | |
| }, | |
| { | |
| "epoch": 54.53, | |
| "grad_norm": 7.865675926208496, | |
| "learning_rate": 9.094269870609981e-06, | |
| "loss": 1.0558, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "eval_accuracy": 0.8045738045738046, | |
| "eval_loss": 0.6465332508087158, | |
| "eval_runtime": 1.9938, | |
| "eval_samples_per_second": 241.25, | |
| "eval_steps_per_second": 30.595, | |
| "step": 29755 | |
| }, | |
| { | |
| "epoch": 55.45, | |
| "grad_norm": 7.172035217285156, | |
| "learning_rate": 8.909426987060998e-06, | |
| "loss": 0.9913, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6654109954833984, | |
| "eval_runtime": 1.9693, | |
| "eval_samples_per_second": 244.248, | |
| "eval_steps_per_second": 30.975, | |
| "step": 30296 | |
| }, | |
| { | |
| "epoch": 56.38, | |
| "grad_norm": 6.30518102645874, | |
| "learning_rate": 8.724584103512016e-06, | |
| "loss": 1.0555, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 57.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.656141996383667, | |
| "eval_runtime": 1.9741, | |
| "eval_samples_per_second": 243.66, | |
| "eval_steps_per_second": 30.901, | |
| "step": 30837 | |
| }, | |
| { | |
| "epoch": 57.3, | |
| "grad_norm": 3.0917370319366455, | |
| "learning_rate": 8.539741219963033e-06, | |
| "loss": 0.9803, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.673220157623291, | |
| "eval_runtime": 1.993, | |
| "eval_samples_per_second": 241.346, | |
| "eval_steps_per_second": 30.607, | |
| "step": 31378 | |
| }, | |
| { | |
| "epoch": 58.23, | |
| "grad_norm": 8.285308837890625, | |
| "learning_rate": 8.35489833641405e-06, | |
| "loss": 1.0393, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 59.0, | |
| "eval_accuracy": 0.7817047817047817, | |
| "eval_loss": 0.6892696619033813, | |
| "eval_runtime": 1.9798, | |
| "eval_samples_per_second": 242.949, | |
| "eval_steps_per_second": 30.811, | |
| "step": 31919 | |
| }, | |
| { | |
| "epoch": 59.15, | |
| "grad_norm": 3.1396327018737793, | |
| "learning_rate": 8.170055452865066e-06, | |
| "loss": 0.9677, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_accuracy": 0.8045738045738046, | |
| "eval_loss": 0.6823599934577942, | |
| "eval_runtime": 2.0127, | |
| "eval_samples_per_second": 238.985, | |
| "eval_steps_per_second": 30.308, | |
| "step": 32460 | |
| }, | |
| { | |
| "epoch": 60.07, | |
| "grad_norm": 12.875879287719727, | |
| "learning_rate": 7.985212569316083e-06, | |
| "loss": 1.0366, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 61.0, | |
| "grad_norm": 8.770364761352539, | |
| "learning_rate": 7.8003696857671e-06, | |
| "loss": 1.0082, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 61.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.6618274450302124, | |
| "eval_runtime": 2.0008, | |
| "eval_samples_per_second": 240.403, | |
| "eval_steps_per_second": 30.488, | |
| "step": 33001 | |
| }, | |
| { | |
| "epoch": 61.92, | |
| "grad_norm": 6.0600972175598145, | |
| "learning_rate": 7.615526802218115e-06, | |
| "loss": 1.0096, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "eval_accuracy": 0.7837837837837838, | |
| "eval_loss": 0.6691136360168457, | |
| "eval_runtime": 1.9817, | |
| "eval_samples_per_second": 242.719, | |
| "eval_steps_per_second": 30.781, | |
| "step": 33542 | |
| }, | |
| { | |
| "epoch": 62.85, | |
| "grad_norm": 10.777630805969238, | |
| "learning_rate": 7.430683918669132e-06, | |
| "loss": 0.9685, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 63.0, | |
| "eval_accuracy": 0.8024948024948025, | |
| "eval_loss": 0.6792653203010559, | |
| "eval_runtime": 2.0645, | |
| "eval_samples_per_second": 232.985, | |
| "eval_steps_per_second": 29.547, | |
| "step": 34083 | |
| }, | |
| { | |
| "epoch": 63.77, | |
| "grad_norm": 3.9615447521209717, | |
| "learning_rate": 7.245841035120148e-06, | |
| "loss": 0.9847, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "eval_accuracy": 0.7837837837837838, | |
| "eval_loss": 0.6894533634185791, | |
| "eval_runtime": 2.0054, | |
| "eval_samples_per_second": 239.847, | |
| "eval_steps_per_second": 30.417, | |
| "step": 34624 | |
| }, | |
| { | |
| "epoch": 64.7, | |
| "grad_norm": 9.38687801361084, | |
| "learning_rate": 7.060998151571166e-06, | |
| "loss": 0.9639, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "eval_accuracy": 0.7733887733887734, | |
| "eval_loss": 0.7297117114067078, | |
| "eval_runtime": 2.0147, | |
| "eval_samples_per_second": 238.744, | |
| "eval_steps_per_second": 30.277, | |
| "step": 35165 | |
| }, | |
| { | |
| "epoch": 65.62, | |
| "grad_norm": 12.292973518371582, | |
| "learning_rate": 6.876155268022182e-06, | |
| "loss": 0.9776, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.6561179757118225, | |
| "eval_runtime": 1.9845, | |
| "eval_samples_per_second": 242.381, | |
| "eval_steps_per_second": 30.739, | |
| "step": 35706 | |
| }, | |
| { | |
| "epoch": 66.54, | |
| "grad_norm": 14.023015022277832, | |
| "learning_rate": 6.691312384473199e-06, | |
| "loss": 1.0074, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 67.0, | |
| "eval_accuracy": 0.7775467775467776, | |
| "eval_loss": 0.6998913884162903, | |
| "eval_runtime": 1.9686, | |
| "eval_samples_per_second": 244.338, | |
| "eval_steps_per_second": 30.987, | |
| "step": 36247 | |
| }, | |
| { | |
| "epoch": 67.47, | |
| "grad_norm": 13.870222091674805, | |
| "learning_rate": 6.506469500924215e-06, | |
| "loss": 0.9466, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.6880961656570435, | |
| "eval_runtime": 1.9686, | |
| "eval_samples_per_second": 244.34, | |
| "eval_steps_per_second": 30.987, | |
| "step": 36788 | |
| }, | |
| { | |
| "epoch": 68.39, | |
| "grad_norm": 6.1949639320373535, | |
| "learning_rate": 6.321626617375231e-06, | |
| "loss": 0.9425, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 69.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.6805587410926819, | |
| "eval_runtime": 1.9709, | |
| "eval_samples_per_second": 244.052, | |
| "eval_steps_per_second": 30.95, | |
| "step": 37329 | |
| }, | |
| { | |
| "epoch": 69.32, | |
| "grad_norm": 7.145143508911133, | |
| "learning_rate": 6.136783733826248e-06, | |
| "loss": 0.9594, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "eval_accuracy": 0.7900207900207901, | |
| "eval_loss": 0.7202461361885071, | |
| "eval_runtime": 2.0125, | |
| "eval_samples_per_second": 239.001, | |
| "eval_steps_per_second": 30.31, | |
| "step": 37870 | |
| }, | |
| { | |
| "epoch": 70.24, | |
| "grad_norm": 9.215810775756836, | |
| "learning_rate": 5.951940850277265e-06, | |
| "loss": 0.9311, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 71.0, | |
| "eval_accuracy": 0.7754677754677755, | |
| "eval_loss": 0.7161967754364014, | |
| "eval_runtime": 1.977, | |
| "eval_samples_per_second": 243.297, | |
| "eval_steps_per_second": 30.855, | |
| "step": 38411 | |
| }, | |
| { | |
| "epoch": 71.16, | |
| "grad_norm": 6.461187362670898, | |
| "learning_rate": 5.767097966728281e-06, | |
| "loss": 0.9429, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.7284368276596069, | |
| "eval_runtime": 2.0663, | |
| "eval_samples_per_second": 232.783, | |
| "eval_steps_per_second": 29.521, | |
| "step": 38952 | |
| }, | |
| { | |
| "epoch": 72.09, | |
| "grad_norm": 11.850204467773438, | |
| "learning_rate": 5.582255083179298e-06, | |
| "loss": 0.9666, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 73.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.6871474981307983, | |
| "eval_runtime": 1.97, | |
| "eval_samples_per_second": 244.162, | |
| "eval_steps_per_second": 30.964, | |
| "step": 39493 | |
| }, | |
| { | |
| "epoch": 73.01, | |
| "grad_norm": 8.0579252243042, | |
| "learning_rate": 5.3974121996303146e-06, | |
| "loss": 0.932, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 73.94, | |
| "grad_norm": 1.1972132921218872, | |
| "learning_rate": 5.212569316081332e-06, | |
| "loss": 0.945, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.6778899431228638, | |
| "eval_runtime": 2.0376, | |
| "eval_samples_per_second": 236.067, | |
| "eval_steps_per_second": 29.938, | |
| "step": 40034 | |
| }, | |
| { | |
| "epoch": 74.86, | |
| "grad_norm": 5.484439849853516, | |
| "learning_rate": 5.027726432532349e-06, | |
| "loss": 0.9387, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.735752522945404, | |
| "eval_runtime": 1.9762, | |
| "eval_samples_per_second": 243.395, | |
| "eval_steps_per_second": 30.867, | |
| "step": 40575 | |
| }, | |
| { | |
| "epoch": 75.79, | |
| "grad_norm": 2.0908420085906982, | |
| "learning_rate": 4.8428835489833645e-06, | |
| "loss": 0.9132, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.7043733596801758, | |
| "eval_runtime": 1.9764, | |
| "eval_samples_per_second": 243.367, | |
| "eval_steps_per_second": 30.864, | |
| "step": 41116 | |
| }, | |
| { | |
| "epoch": 76.71, | |
| "grad_norm": 10.380330085754395, | |
| "learning_rate": 4.658040665434381e-06, | |
| "loss": 0.9181, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 77.0, | |
| "eval_accuracy": 0.7962577962577962, | |
| "eval_loss": 0.7041053771972656, | |
| "eval_runtime": 2.0006, | |
| "eval_samples_per_second": 240.43, | |
| "eval_steps_per_second": 30.491, | |
| "step": 41657 | |
| }, | |
| { | |
| "epoch": 77.63, | |
| "grad_norm": 9.135781288146973, | |
| "learning_rate": 4.473197781885398e-06, | |
| "loss": 0.9218, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.6986111998558044, | |
| "eval_runtime": 1.9688, | |
| "eval_samples_per_second": 244.309, | |
| "eval_steps_per_second": 30.983, | |
| "step": 42198 | |
| }, | |
| { | |
| "epoch": 78.56, | |
| "grad_norm": 17.338001251220703, | |
| "learning_rate": 4.288354898336414e-06, | |
| "loss": 0.8621, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 79.0, | |
| "eval_accuracy": 0.8004158004158004, | |
| "eval_loss": 0.6909247040748596, | |
| "eval_runtime": 1.997, | |
| "eval_samples_per_second": 240.86, | |
| "eval_steps_per_second": 30.546, | |
| "step": 42739 | |
| }, | |
| { | |
| "epoch": 79.48, | |
| "grad_norm": 6.793923854827881, | |
| "learning_rate": 4.103512014787431e-06, | |
| "loss": 0.9236, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "eval_accuracy": 0.7983367983367984, | |
| "eval_loss": 0.7135599851608276, | |
| "eval_runtime": 1.9949, | |
| "eval_samples_per_second": 241.121, | |
| "eval_steps_per_second": 30.579, | |
| "step": 43280 | |
| }, | |
| { | |
| "epoch": 80.41, | |
| "grad_norm": 3.9345781803131104, | |
| "learning_rate": 3.918669131238448e-06, | |
| "loss": 0.8667, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 81.0, | |
| "eval_accuracy": 0.8024948024948025, | |
| "eval_loss": 0.7008742094039917, | |
| "eval_runtime": 1.992, | |
| "eval_samples_per_second": 241.461, | |
| "eval_steps_per_second": 30.622, | |
| "step": 43821 | |
| }, | |
| { | |
| "epoch": 81.33, | |
| "grad_norm": 16.883420944213867, | |
| "learning_rate": 3.7338262476894642e-06, | |
| "loss": 0.8856, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.7127683162689209, | |
| "eval_runtime": 2.0451, | |
| "eval_samples_per_second": 235.201, | |
| "eval_steps_per_second": 29.828, | |
| "step": 44362 | |
| }, | |
| { | |
| "epoch": 82.26, | |
| "grad_norm": 7.969069480895996, | |
| "learning_rate": 3.548983364140481e-06, | |
| "loss": 0.917, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 83.0, | |
| "eval_accuracy": 0.7983367983367984, | |
| "eval_loss": 0.7134777307510376, | |
| "eval_runtime": 2.0529, | |
| "eval_samples_per_second": 234.298, | |
| "eval_steps_per_second": 29.714, | |
| "step": 44903 | |
| }, | |
| { | |
| "epoch": 83.18, | |
| "grad_norm": 1.545163631439209, | |
| "learning_rate": 3.3641404805914975e-06, | |
| "loss": 0.8835, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "eval_accuracy": 0.7900207900207901, | |
| "eval_loss": 0.7295302748680115, | |
| "eval_runtime": 1.9747, | |
| "eval_samples_per_second": 243.583, | |
| "eval_steps_per_second": 30.891, | |
| "step": 45444 | |
| }, | |
| { | |
| "epoch": 84.1, | |
| "grad_norm": 5.072544097900391, | |
| "learning_rate": 3.1792975970425146e-06, | |
| "loss": 0.8879, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "eval_accuracy": 0.7900207900207901, | |
| "eval_loss": 0.7449509501457214, | |
| "eval_runtime": 2.0759, | |
| "eval_samples_per_second": 231.71, | |
| "eval_steps_per_second": 29.385, | |
| "step": 45985 | |
| }, | |
| { | |
| "epoch": 85.03, | |
| "grad_norm": 5.645694732666016, | |
| "learning_rate": 2.9944547134935308e-06, | |
| "loss": 0.9114, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 85.95, | |
| "grad_norm": 5.065194129943848, | |
| "learning_rate": 2.8096118299445474e-06, | |
| "loss": 0.8764, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 86.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.7362204194068909, | |
| "eval_runtime": 1.9869, | |
| "eval_samples_per_second": 242.081, | |
| "eval_steps_per_second": 30.7, | |
| "step": 46526 | |
| }, | |
| { | |
| "epoch": 86.88, | |
| "grad_norm": 5.654088020324707, | |
| "learning_rate": 2.624768946395564e-06, | |
| "loss": 0.8674, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 87.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.7232093811035156, | |
| "eval_runtime": 2.0232, | |
| "eval_samples_per_second": 237.746, | |
| "eval_steps_per_second": 30.151, | |
| "step": 47067 | |
| }, | |
| { | |
| "epoch": 87.8, | |
| "grad_norm": 12.72859001159668, | |
| "learning_rate": 2.4399260628465807e-06, | |
| "loss": 0.8583, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 88.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.7407870888710022, | |
| "eval_runtime": 1.9934, | |
| "eval_samples_per_second": 241.296, | |
| "eval_steps_per_second": 30.601, | |
| "step": 47608 | |
| }, | |
| { | |
| "epoch": 88.72, | |
| "grad_norm": 6.526777744293213, | |
| "learning_rate": 2.2550831792975973e-06, | |
| "loss": 0.881, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 89.0, | |
| "eval_accuracy": 0.8004158004158004, | |
| "eval_loss": 0.7377821803092957, | |
| "eval_runtime": 1.9802, | |
| "eval_samples_per_second": 242.901, | |
| "eval_steps_per_second": 30.804, | |
| "step": 48149 | |
| }, | |
| { | |
| "epoch": 89.65, | |
| "grad_norm": 8.497318267822266, | |
| "learning_rate": 2.070240295748614e-06, | |
| "loss": 0.8668, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "eval_accuracy": 0.7900207900207901, | |
| "eval_loss": 0.7473007440567017, | |
| "eval_runtime": 2.0137, | |
| "eval_samples_per_second": 238.867, | |
| "eval_steps_per_second": 30.293, | |
| "step": 48690 | |
| }, | |
| { | |
| "epoch": 90.57, | |
| "grad_norm": 6.455136775970459, | |
| "learning_rate": 1.8853974121996305e-06, | |
| "loss": 0.8779, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 91.0, | |
| "eval_accuracy": 0.7983367983367984, | |
| "eval_loss": 0.7438368201255798, | |
| "eval_runtime": 1.9731, | |
| "eval_samples_per_second": 243.774, | |
| "eval_steps_per_second": 30.915, | |
| "step": 49231 | |
| }, | |
| { | |
| "epoch": 91.5, | |
| "grad_norm": 5.713993072509766, | |
| "learning_rate": 1.700554528650647e-06, | |
| "loss": 0.8717, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 92.0, | |
| "eval_accuracy": 0.8004158004158004, | |
| "eval_loss": 0.7389739751815796, | |
| "eval_runtime": 1.9686, | |
| "eval_samples_per_second": 244.34, | |
| "eval_steps_per_second": 30.987, | |
| "step": 49772 | |
| }, | |
| { | |
| "epoch": 92.42, | |
| "grad_norm": 5.342690467834473, | |
| "learning_rate": 1.5157116451016638e-06, | |
| "loss": 0.8781, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 93.0, | |
| "eval_accuracy": 0.7983367983367984, | |
| "eval_loss": 0.7473535537719727, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 242.926, | |
| "eval_steps_per_second": 30.808, | |
| "step": 50313 | |
| }, | |
| { | |
| "epoch": 93.35, | |
| "grad_norm": 9.870634078979492, | |
| "learning_rate": 1.3308687615526802e-06, | |
| "loss": 0.8845, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "eval_accuracy": 0.7900207900207901, | |
| "eval_loss": 0.7445840835571289, | |
| "eval_runtime": 1.9776, | |
| "eval_samples_per_second": 243.222, | |
| "eval_steps_per_second": 30.845, | |
| "step": 50854 | |
| }, | |
| { | |
| "epoch": 94.27, | |
| "grad_norm": 8.909347534179688, | |
| "learning_rate": 1.1460258780036969e-06, | |
| "loss": 0.8623, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.7315581440925598, | |
| "eval_runtime": 1.9728, | |
| "eval_samples_per_second": 243.814, | |
| "eval_steps_per_second": 30.92, | |
| "step": 51395 | |
| }, | |
| { | |
| "epoch": 95.19, | |
| "grad_norm": 10.748625755310059, | |
| "learning_rate": 9.611829944547135e-07, | |
| "loss": 0.8341, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "eval_accuracy": 0.7879417879417879, | |
| "eval_loss": 0.7457364201545715, | |
| "eval_runtime": 2.0017, | |
| "eval_samples_per_second": 240.29, | |
| "eval_steps_per_second": 30.473, | |
| "step": 51936 | |
| }, | |
| { | |
| "epoch": 96.12, | |
| "grad_norm": 3.179774761199951, | |
| "learning_rate": 7.763401109057302e-07, | |
| "loss": 0.8766, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 97.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.743617832660675, | |
| "eval_runtime": 1.999, | |
| "eval_samples_per_second": 240.625, | |
| "eval_steps_per_second": 30.516, | |
| "step": 52477 | |
| }, | |
| { | |
| "epoch": 97.04, | |
| "grad_norm": 12.243720054626465, | |
| "learning_rate": 5.914972273567468e-07, | |
| "loss": 0.8101, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 97.97, | |
| "grad_norm": 18.670886993408203, | |
| "learning_rate": 4.066543438077634e-07, | |
| "loss": 0.8681, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 98.0, | |
| "eval_accuracy": 0.7900207900207901, | |
| "eval_loss": 0.7483807802200317, | |
| "eval_runtime": 2.0039, | |
| "eval_samples_per_second": 240.035, | |
| "eval_steps_per_second": 30.441, | |
| "step": 53018 | |
| }, | |
| { | |
| "epoch": 98.89, | |
| "grad_norm": 8.483085632324219, | |
| "learning_rate": 2.2181146025878005e-07, | |
| "loss": 0.8635, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 99.0, | |
| "eval_accuracy": 0.7941787941787942, | |
| "eval_loss": 0.7391884922981262, | |
| "eval_runtime": 1.9875, | |
| "eval_samples_per_second": 242.013, | |
| "eval_steps_per_second": 30.692, | |
| "step": 53559 | |
| }, | |
| { | |
| "epoch": 99.82, | |
| "grad_norm": 10.068202018737793, | |
| "learning_rate": 3.696857670979668e-08, | |
| "loss": 0.8091, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_accuracy": 0.7920997920997921, | |
| "eval_loss": 0.7390549182891846, | |
| "eval_runtime": 2.0448, | |
| "eval_samples_per_second": 235.228, | |
| "eval_steps_per_second": 29.831, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 54100, | |
| "total_flos": 3.355193271048192e+19, | |
| "train_loss": 1.2517024893769495, | |
| "train_runtime": 5380.7558, | |
| "train_samples_per_second": 80.379, | |
| "train_steps_per_second": 10.054 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 54100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 3.355193271048192e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |