[作者大大你好，我在使用自己的数据集训练的时候遇到了问题，训练数据大概是120张，推理了300轮，，最后送入sample.py，设置s=0的时候输出的全是噪声，请问这是模型没有学习到吗，训练日志如下所示
[2025-11-18 13:56:25,051][train_diffusion.py][line: 303][ INFO] * Loss_sum 0.01308
[2025-11-18 13:56:32,883][train_diffusion.py][line: 237][ INFO] Epoch: [141/1000] Iter: [7021/50000] Loss 0.02448 (0.04316)
[2025-11-18 13:56:38,934][train_diffusion.py][line: 237][ INFO] Epoch: [142/1000] Iter: [7051/50000] Loss 0.05771 (0.05771)
[2025-11-18 13:56:44,710][train_diffusion.py][line: 237][ INFO] Epoch: [142/1000] Iter: [7081/50000] Loss 0.02813 (0.06932)
[2025-11-18 13:56:50,744][train_diffusion.py][line: 237][ INFO] Epoch: [143/1000] Iter: [7111/50000] Loss 0.02488 (0.05089)
[2025-11-18 13:56:56,513][train_diffusion.py][line: 237][ INFO] Epoch: [143/1000] Iter: [7141/50000] Loss 0.06132 (0.07150)
[2025-11-18 13:57:02,545][train_diffusion.py][line: 237][ INFO] Epoch: [144/1000] Iter: [7171/50000] Loss 0.02471 (0.04264)
[2025-11-18 13:57:08,555][train_diffusion.py][line: 237][ INFO] Epoch: [145/1000] Iter: [7201/50000] Loss 0.02558 (0.02558)
[2025-11-18 13:57:14,328][train_diffusion.py][line: 237][ INFO] Epoch: [145/1000] Iter: [7231/50000] Loss 0.03769 (0.03124)
[2025-11-18 13:57:20,359][train_diffusion.py][line: 237][ INFO] Epoch: [146/1000] Iter: [7261/50000] Loss 0.02431 (0.02542)
[2025-11-18 13:57:26,204][train_diffusion.py][line: 237][ INFO] Epoch: [146/1000] Iter: [7291/50000] Loss 0.04465 (0.04257)
[2025-11-18 13:57:32,265][train_diffusion.py][line: 237][ INFO] Epoch: [147/1000] Iter: [7321/50000] Loss 0.02421 (0.04816)
[作者大大你好，我在使用自己的数据集训练的时候遇到了问题，训练数据大概是120张，推理了300轮，，最后送入sample.py，设置s=0的时候输出的全是噪声，请问这是模型没有学习到吗，训练日志如下所示
[2025-11-18 13:56:25,051][train_diffusion.py][line: 303][ INFO] * Loss_sum 0.01308
[2025-11-18 13:56:32,883][train_diffusion.py][line: 237][ INFO] Epoch: [141/1000] Iter: [7021/50000] Loss 0.02448 (0.04316)
[2025-11-18 13:56:38,934][train_diffusion.py][line: 237][ INFO] Epoch: [142/1000] Iter: [7051/50000] Loss 0.05771 (0.05771)
[2025-11-18 13:56:44,710][train_diffusion.py][line: 237][ INFO] Epoch: [142/1000] Iter: [7081/50000] Loss 0.02813 (0.06932)
[2025-11-18 13:56:50,744][train_diffusion.py][line: 237][ INFO] Epoch: [143/1000] Iter: [7111/50000] Loss 0.02488 (0.05089)
[2025-11-18 13:56:56,513][train_diffusion.py][line: 237][ INFO] Epoch: [143/1000] Iter: [7141/50000] Loss 0.06132 (0.07150)
[2025-11-18 13:57:02,545][train_diffusion.py][line: 237][ INFO] Epoch: [144/1000] Iter: [7171/50000] Loss 0.02471 (0.04264)
[2025-11-18 13:57:08,555][train_diffusion.py][line: 237][ INFO] Epoch: [145/1000] Iter: [7201/50000] Loss 0.02558 (0.02558)
[2025-11-18 13:57:14,328][train_diffusion.py][line: 237][ INFO] Epoch: [145/1000] Iter: [7231/50000] Loss 0.03769 (0.03124)
[2025-11-18 13:57:20,359][train_diffusion.py][line: 237][ INFO] Epoch: [146/1000] Iter: [7261/50000] Loss 0.02431 (0.02542)
[2025-11-18 13:57:26,204][train_diffusion.py][line: 237][ INFO] Epoch: [146/1000] Iter: [7291/50000] Loss 0.04465 (0.04257)
[2025-11-18 13:57:32,265][train_diffusion.py][line: 237][ INFO] Epoch: [147/1000] Iter: [7321/50000] Loss 0.02421 (0.04816)