请问训练到位的指标是啥

[tailangjun]

我今天试着训练了一下Obama，打开TensorBoard后发现参数太多了，不知道看哪个，有知道的兄弟没

[tailangjun]

感觉是看控制台中的 total_loss

| Validation results@18000: {'total_loss': 26.5918741798, 'mse_loss': 0.0003416548, 'sr_mse_loss': 0.0004718732, 'lambda_ambient': 26.5910606384} 03/16 01:51:34 PM Epoch 00000@18000: saving model to checkpoints/motion2video_nerf/Obama_head/model_ckpt_steps_18000.ckpt 03/16 01:51:34 PM Delete ckpt: model_ckpt_steps_16000.ckpt Valid: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 100/100 [00:03<00:00, 32.34step/s] | Validation results@20000: {'total_loss': 2.5853313923, 'mse_loss': 0.0003031579, 'sr_mse_loss': 0.0004077603, 'lambda_ambient': 2.5846204758} 03/16 01:53:22 PM Epoch 00000@20000: saving model to checkpoints/motion2video_nerf/Obama_head/model_ckpt_steps_20000.ckpt 03/16 01:53:22 PM Delete ckpt: model_ckpt_steps_18000.ckpt

[tailangjun]

另外发现 PM Delete ckpt只保留了最后一个，一不注意最小的那个就被删掉了

[xiao-keeplearning]

问下你的loss最后训练到多少了