{ "config": { "epochs": 30, "lr": 0.0003, "warmup_steps": 500, "grad_accum_steps": 1, "grad_clip": 1.0 }, "train": [ { "epoch": 1, "avg_loss": 6.375119, "lr": 0.0002994896488977018, "step": 2296, "time_s": 2555.0 }, { "epoch": 2, "avg_loss": 5.113232, "lr": 0.0002973570158693121, "step": 4592, "time_s": 2558.0 }, { "epoch": 3, "avg_loss": 4.698285, "lr": 0.0002935862338662239, "step": 6888, "time_s": 2558.8 }, { "epoch": 4, "avg_loss": 4.461017, "lr": 0.00028821922218651554, "step": 9184, "time_s": 2502.3 }, { "epoch": 5, "avg_loss": 4.27398, "lr": 0.0002813156452080821, "step": 11480, "time_s": 2504.1 }, { "epoch": 6, "avg_loss": 4.111808, "lr": 0.0002729522491074273, "step": 13776, "time_s": 2508.4 }, { "epoch": 7, "avg_loss": 3.960994, "lr": 0.00026322200868229975, "step": 16072, "time_s": 2509.8 }, { "epoch": 8, "avg_loss": 3.824394, "lr": 0.0002522330937628358, "step": 18368, "time_s": 2515.9 }, { "epoch": 9, "avg_loss": 3.704921, "lr": 0.000240107666701484, "step": 20664, "time_s": 2519.0 }, { "epoch": 10, "avg_loss": 3.603755, "lr": 0.00022698052430985955, "step": 22960, "time_s": 2516.9 }, { "epoch": 11, "avg_loss": 3.514683, "lr": 0.00021299759933994378, "step": 25256, "time_s": 2518.8 }, { "epoch": 12, "avg_loss": 3.435752, "lr": 0.00019831433816846762, "step": 27552, "time_s": 2507.9 }, { "epoch": 13, "avg_loss": 3.366267, "lr": 0.00018309397271955332, "step": 29848, "time_s": 2508.7 }, { "epoch": 14, "avg_loss": 3.30577, "lr": 0.000167505705836429, "step": 32144, "time_s": 2513.6 }, { "epoch": 15, "avg_loss": 3.249759, "lr": 0.0001517228302752031, "step": 34440, "time_s": 2516.2 }, { "epoch": 16, "avg_loss": 3.198617, "lr": 0.00013592080223160462, "step": 36736, "time_s": 2518.5 }, { "epoch": 17, "avg_loss": 3.151068, "lr": 0.00012027529081704365, "step": 39032, "time_s": 2515.4 }, { "epoch": 18, "avg_loss": 3.106923, "lr": 0.0001049602251677148, "step": 41328, "time_s": 2510.8 }, { "epoch": 19, "avg_loss": 3.093604, "lr": 9.014586089678252e-05, "step": 43624, "time_s": 2436.8 }, { "epoch": 20, "avg_loss": 3.051561, "lr": 7.599688738464758e-05, "step": 45920, "time_s": 2437.1 }, { "epoch": 21, "avg_loss": 3.015202, "lr": 6.267059694830767e-05, "step": 48216, "time_s": 2434.4 }, { "epoch": 22, "avg_loss": 2.982195, "lr": 5.031513624291888e-05, "step": 50512, "time_s": 2433.5 }, { "epoch": 23, "avg_loss": 2.953464, "lr": 3.9067859334502825e-05, "step": 52808, "time_s": 2435.8 }, { "epoch": 24, "avg_loss": 2.941402, "lr": 2.9053800752477626e-05, "step": 55104, "time_s": 2542.3 }, { "epoch": 25, "avg_loss": 2.917992, "lr": 2.0384285496891105e-05, "step": 57400, "time_s": 2553.6 }, { "epoch": 26, "avg_loss": 2.903834, "lr": 1.4999999999999999e-05, "step": 59696, "time_s": 2503.9 }, { "epoch": 27, "avg_loss": 2.893638, "lr": 1.4999999999999999e-05, "step": 61992, "time_s": 2503.6 }, { "epoch": 28, "avg_loss": 2.888698, "lr": 1.4999999999999999e-05, "step": 64288, "time_s": 2502.7 }, { "epoch": 29, "avg_loss": 2.883319, "lr": 1.4999999999999999e-05, "step": 66584, "time_s": 2505.1 }, { "epoch": 30, "avg_loss": 2.883806, "lr": 1.4999999999999999e-05, "step": 68880, "time_s": 2541.8 } ], "val": [ { "epoch": 1, "avg_loss": 5.357941 }, { "epoch": 2, "avg_loss": 4.749041 }, { "epoch": 3, "avg_loss": 4.459847 }, { "epoch": 4, "avg_loss": 4.238065 }, { "epoch": 5, "avg_loss": 4.085373 }, { "epoch": 6, "avg_loss": 3.930117 }, { "epoch": 7, "avg_loss": 3.791228 }, { "epoch": 8, "avg_loss": 3.664075 }, { "epoch": 9, "avg_loss": 3.561065 }, { "epoch": 10, "avg_loss": 3.480822 }, { "epoch": 11, "avg_loss": 3.412537 }, { "epoch": 12, "avg_loss": 3.360075 }, { "epoch": 13, "avg_loss": 3.307848 }, { "epoch": 14, "avg_loss": 3.26252 }, { "epoch": 15, "avg_loss": 3.222749 }, { "epoch": 16, "avg_loss": 3.197138 }, { "epoch": 17, "avg_loss": 3.16499 }, { "epoch": 18, "avg_loss": 3.147003 }, { "epoch": 19, "avg_loss": 2.865002 }, { "epoch": 20, "avg_loss": 2.865283 }, { "epoch": 21, "avg_loss": 2.853566 }, { "epoch": 22, "avg_loss": 2.851498 }, { "epoch": 23, "avg_loss": 2.844062 }, { "epoch": 24, "avg_loss": 2.699958 }, { "epoch": 25, "avg_loss": 2.700966 }, { "epoch": 26, "avg_loss": 2.659113 }, { "epoch": 27, "avg_loss": 2.665918 }, { "epoch": 28, "avg_loss": 2.673086 }, { "epoch": 29, "avg_loss": 2.680465 }, { "epoch": 30, "avg_loss": 2.635866 } ], "sanity": [] }