{ "config": { "epochs": 30, "lr": 0.0003, "warmup_steps": 667, "grad_accum_steps": 4, "grad_clip": 1.0 }, "train": [ { "epoch": 1, "avg_loss": 7.425536, "lr": 0.0002998625141045411, "step": 1113, "time_s": 5022.7 }, { "epoch": 2, "avg_loss": 6.084101, "lr": 0.000298322988234269, "step": 2226, "time_s": 5041.5 }, { "epoch": 3, "avg_loss": 5.6477, "lr": 0.00029509154537521725, "step": 3339, "time_s": 5047.8 }, { "epoch": 4, "avg_loss": 5.394878, "lr": 0.00029020504652335535, "step": 4452, "time_s": 5048.9 }, { "epoch": 5, "avg_loss": 5.222645, "lr": 0.0002837192318611158, "step": 5565, "time_s": 5051.5 }, { "epoch": 6, "avg_loss": 5.080133, "lr": 0.00027570808493038764, "step": 6678, "time_s": 5294.4 }, { "epoch": 7, "avg_loss": 4.958657, "lr": 0.0002662629887039187, "step": 7791, "time_s": 4913.8 }, { "epoch": 8, "avg_loss": 4.863998, "lr": 0.0002554916831818026, "step": 8904, "time_s": 4915.7 }, { "epoch": 9, "avg_loss": 4.794975, "lr": 0.00024351703640372173, "step": 10017, "time_s": 4916.2 }, { "epoch": 10, "avg_loss": 4.742931, "lr": 0.00023047564289598436, "step": 11130, "time_s": 4915.9 }, { "epoch": 11, "avg_loss": 4.69845, "lr": 0.00021651626554083667, "step": 12243, "time_s": 4915.4 }, { "epoch": 12, "avg_loss": 4.662304, "lr": 0.00020179813864161316, "step": 13356, "time_s": 4915.9 }, { "epoch": 13, "avg_loss": 4.629976, "lr": 0.00018648915154062174, "step": 14469, "time_s": 4918.6 }, { "epoch": 14, "avg_loss": 4.601584, "lr": 0.00017076393350919474, "step": 15582, "time_s": 5164.3 }, { "epoch": 15, "avg_loss": 4.576644, "lr": 0.00015480186175552063, "step": 16695, "time_s": 5177.1 }, { "epoch": 16, "avg_loss": 4.553974, "lr": 0.00013878501527286722, "step": 17808, "time_s": 5179.6 }, { "epoch": 17, "avg_loss": 4.533928, "lr": 0.00012289609786860295, "step": 18921, "time_s": 5178.9 }, { "epoch": 18, "avg_loss": 4.51544, "lr": 0.0001073163540659783, "step": 20034, "time_s": 5179.3 }, { "epoch": 19, "avg_loss": 4.499081, "lr": 9.222350165192764e-05, "step": 21147, "time_s": 5182.2 }, { "epoch": 20, "avg_loss": 4.483817, "lr": 7.778970445427346e-05, "step": 22260, "time_s": 5511.1 }, { "epoch": 21, "avg_loss": 4.470338, "lr": 6.417960847281683e-05, "step": 23373, "time_s": 5518.2 }, { "epoch": 22, "avg_loss": 4.458165, "lr": 5.154846376612302e-05, "step": 24486, "time_s": 5509.7 }, { "epoch": 23, "avg_loss": 4.447244, "lr": 4.0040353517597394e-05, "step": 25599, "time_s": 5505.8 }, { "epoch": 24, "avg_loss": 4.437949, "lr": 2.9786550481854855e-05, "step": 26712, "time_s": 5518.2 }, { "epoch": 25, "avg_loss": 4.429744, "lr": 2.0904019559362385e-05, "step": 27825, "time_s": 5518.6 }, { "epoch": 26, "avg_loss": 4.423089, "lr": 1.4999999999999999e-05, "step": 28938, "time_s": 5200.8 }, { "epoch": 27, "avg_loss": 4.419632, "lr": 1.4999999999999999e-05, "step": 30051, "time_s": 5203.6 }, { "epoch": 28, "avg_loss": 4.417868, "lr": 1.4999999999999999e-05, "step": 31164, "time_s": 5205.5 }, { "epoch": 29, "avg_loss": 4.416322, "lr": 1.4999999999999999e-05, "step": 32277, "time_s": 5201.4 }, { "epoch": 30, "avg_loss": 4.414624, "lr": 1.4999999999999999e-05, "step": 33390, "time_s": 5202.3 } ], "val": [ { "epoch": 1, "avg_loss": 6.385433 }, { "epoch": 2, "avg_loss": 5.7099 }, { "epoch": 3, "avg_loss": 5.391578 }, { "epoch": 4, "avg_loss": 5.185818 }, { "epoch": 5, "avg_loss": 5.030019 }, { "epoch": 6, "avg_loss": 4.92016 }, { "epoch": 7, "avg_loss": 4.794157 }, { "epoch": 8, "avg_loss": 4.722416 }, { "epoch": 9, "avg_loss": 4.666241 }, { "epoch": 10, "avg_loss": 4.629954 }, { "epoch": 11, "avg_loss": 4.589608 }, { "epoch": 12, "avg_loss": 4.56284 }, { "epoch": 13, "avg_loss": 4.541597 }, { "epoch": 14, "avg_loss": 4.519729 }, { "epoch": 15, "avg_loss": 4.498085 }, { "epoch": 16, "avg_loss": 4.483384 }, { "epoch": 17, "avg_loss": 4.466313 }, { "epoch": 18, "avg_loss": 4.456282 }, { "epoch": 19, "avg_loss": 4.444423 }, { "epoch": 20, "avg_loss": 4.433954 }, { "epoch": 21, "avg_loss": 4.422863 }, { "epoch": 22, "avg_loss": 4.415679 }, { "epoch": 23, "avg_loss": 4.407862 }, { "epoch": 24, "avg_loss": 4.400964 }, { "epoch": 25, "avg_loss": 4.395205 }, { "epoch": 26, "avg_loss": 4.392105 }, { "epoch": 27, "avg_loss": 4.391044 }, { "epoch": 28, "avg_loss": 4.389972 }, { "epoch": 29, "avg_loss": 4.389403 }, { "epoch": 30, "avg_loss": 4.3887 } ], "sanity": [] }