wandb: WARNING The get_url method is deprecated and will be removed in a future release. Please use `run.url` instead. [wandb] logging to https://wandb.ai/seungeun-goodgang-labs/animasync-v3-face/runs/86prm5ac epoch 0 train l1=0.1695 vel=0.0076 val l1=0.0775 vel=0.0018 lr=3.74e-04 4.1s → saved best (val l1=0.0775) epoch 1 train l1=0.0822 vel=0.0018 val l1=0.0781 vel=0.0018 lr=7.48e-04 3.4s epoch 2 train l1=0.0826 vel=0.0017 val l1=0.0783 vel=0.0017 lr=1.00e-03 3.4s epoch 3 train l1=0.0825 vel=0.0016 val l1=0.0781 vel=0.0016 lr=9.98e-04 3.5s epoch 4 train l1=0.0828 vel=0.0016 val l1=0.0782 vel=0.0016 lr=9.94e-04 3.3s epoch 5 train l1=0.0822 vel=0.0016 val l1=0.0784 vel=0.0016 lr=9.88e-04 3.5s epoch 6 train l1=0.0824 vel=0.0016 val l1=0.0784 vel=0.0016 lr=9.80e-04 3.5s epoch 7 train l1=0.0828 vel=0.0016 val l1=0.0783 vel=0.0016 lr=9.69e-04 3.4s epoch 8 train l1=0.0825 vel=0.0016 val l1=0.0783 vel=0.0016 lr=9.57e-04 3.4s epoch 9 train l1=0.0824 vel=0.0016 val l1=0.0784 vel=0.0016 lr=9.42e-04 3.5s epoch 10 train l1=0.0821 vel=0.0016 val l1=0.0781 vel=0.0016 lr=9.26e-04 3.4s epoch 11 train l1=0.0823 vel=0.0016 val l1=0.0784 vel=0.0016 lr=9.07e-04 3.4s epoch 12 train l1=0.0833 vel=0.0016 val l1=0.0784 vel=0.0016 lr=8.87e-04 3.6s epoch 13 train l1=0.0827 vel=0.0016 val l1=0.0787 vel=0.0016 lr=8.65e-04 3.4s epoch 14 train l1=0.0829 vel=0.0016 val l1=0.0782 vel=0.0016 lr=8.42e-04 3.5s epoch 15 train l1=0.0825 vel=0.0016 val l1=0.0784 vel=0.0016 lr=8.17e-04 3.5s epoch 16 train l1=0.0822 vel=0.0016 val l1=0.0781 vel=0.0016 lr=7.90e-04 3.4s epoch 17 train l1=0.0824 vel=0.0016 val l1=0.0782 vel=0.0016 lr=7.63e-04 3.5s epoch 18 train l1=0.0822 vel=0.0016 val l1=0.0782 vel=0.0016 lr=7.34e-04 3.4s epoch 19 train l1=0.0820 vel=0.0016 val l1=0.0784 vel=0.0016 lr=7.04e-04 3.4s epoch 20 train l1=0.0824 vel=0.0017 val l1=0.0783 vel=0.0016 lr=6.73e-04 3.5s epoch 21 train l1=0.0824 vel=0.0016 val l1=0.0784 vel=0.0016 lr=6.42e-04 3.5s epoch 22 train l1=0.0827 vel=0.0016 val l1=0.0783 vel=0.0016 lr=6.10e-04 3.5s epoch 23 train l1=0.0815 vel=0.0017 val l1=0.0785 vel=0.0016 lr=5.77e-04 3.4s epoch 24 train l1=0.0823 vel=0.0017 val l1=0.0782 vel=0.0016 lr=5.44e-04 3.5s epoch 25 train l1=0.0825 vel=0.0017 val l1=0.0781 vel=0.0018 lr=5.11e-04 3.8s epoch 26 train l1=0.0827 vel=0.0017 val l1=0.0767 vel=0.0022 lr=4.78e-04 4.3s → saved best (val l1=0.0767) epoch 27 train l1=0.0822 vel=0.0017 val l1=0.0756 vel=0.0025 lr=4.45e-04 3.7s → saved best (val l1=0.0756) epoch 28 train l1=0.0798 vel=0.0023 val l1=0.0716 vel=0.0023 lr=4.12e-04 3.7s → saved best (val l1=0.0716) epoch 29 train l1=0.0767 vel=0.0027 val l1=0.0688 vel=0.0028 lr=3.80e-04 3.6s → saved best (val l1=0.0688) epoch 30 train l1=0.0717 vel=0.0037 val l1=0.0728 vel=0.0026 lr=3.48e-04 3.4s epoch 31 train l1=0.0719 vel=0.0032 val l1=0.0661 vel=0.0025 lr=3.16e-04 3.5s → saved best (val l1=0.0661) epoch 32 train l1=0.0678 vel=0.0033 val l1=0.0607 vel=0.0033 lr=2.86e-04 3.5s → saved best (val l1=0.0607) epoch 33 train l1=0.0606 vel=0.0039 val l1=0.0574 vel=0.0032 lr=2.56e-04 3.5s → saved best (val l1=0.0574) epoch 34 train l1=0.0594 vel=0.0041 val l1=0.0516 vel=0.0035 lr=2.28e-04 3.5s → saved best (val l1=0.0516) epoch 35 train l1=0.0542 vel=0.0043 val l1=0.0492 vel=0.0036 lr=2.01e-04 3.4s → saved best (val l1=0.0492) epoch 36 train l1=0.0519 vel=0.0044 val l1=0.0453 vel=0.0039 lr=1.75e-04 3.5s → saved best (val l1=0.0453) epoch 37 train l1=0.0475 vel=0.0050 val l1=0.0431 vel=0.0045 lr=1.50e-04 3.5s → saved best (val l1=0.0431) epoch 38 train l1=0.0439 vel=0.0054 val l1=0.0394 vel=0.0044 lr=1.27e-04 3.5s → saved best (val l1=0.0394) epoch 39 train l1=0.0420 vel=0.0053 val l1=0.0384 vel=0.0041 lr=1.06e-04 3.5s → saved best (val l1=0.0384) epoch 40 train l1=0.0400 vel=0.0052 val l1=0.0373 vel=0.0041 lr=8.66e-05 3.4s → saved best (val l1=0.0373) epoch 41 train l1=0.0394 vel=0.0051 val l1=0.0360 vel=0.0040 lr=6.89e-05 3.5s → saved best (val l1=0.0360) epoch 42 train l1=0.0384 vel=0.0050 val l1=0.0358 vel=0.0039 lr=5.30e-05 3.4s → saved best (val l1=0.0358) epoch 43 train l1=0.0378 vel=0.0050 val l1=0.0350 vel=0.0039 lr=3.91e-05 3.4s → saved best (val l1=0.0350) epoch 44 train l1=0.0374 vel=0.0049 val l1=0.0350 vel=0.0038 lr=2.73e-05 3.5s → saved best (val l1=0.0350) epoch 45 train l1=0.0372 vel=0.0049 val l1=0.0344 vel=0.0038 lr=1.75e-05 3.5s → saved best (val l1=0.0344) epoch 46 train l1=0.0366 vel=0.0049 val l1=0.0344 vel=0.0037 lr=9.88e-06 3.5s → saved best (val l1=0.0344) epoch 47 train l1=0.0365 vel=0.0048 val l1=0.0342 vel=0.0037 lr=4.40e-06 3.4s → saved best (val l1=0.0342) epoch 48 train l1=0.0365 vel=0.0048 val l1=0.0341 vel=0.0037 lr=1.10e-06 3.4s → saved best (val l1=0.0341) epoch 49 train l1=0.0364 vel=0.0048 val l1=0.0339 vel=0.0037 lr=0.00e+00 3.4s → saved best (val l1=0.0339) Done. best val l1: 0.0339 checkpoints: /dataset/kemix-engine/package/face/animasync-face-v3/models/v3_face/checkpoints