--output checkpoints/fish-speech-1.5-yth-lora/

[rhfu]

Self Checks

• This template is only for bug reports. For questions, please visit Discussions.
• I have thoroughly reviewed the project documentation (installation, training, inference) but couldn't find information to solve my problem. English 中文 日本語 Portuguese (Brazil)
• I have searched for existing issues, including closed ones. Search issues
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• Please do not modify this template and fill in all required fields.

Cloud or Self Hosted

Self Hosted (Source)

Environment Details

macOS 13.5 python3.10 torch==2.4.1

Steps to Reproduce

训练好

✔️ Expected Behavior

python tools/llama/merge_lora.py
--lora-config r_8_alpha_16
--base-weight checkpoints/fish-speech-1.5
--lora-weight results/$project/checkpoints/step_000000010.ckpt
--output checkpoints/fish-speech-1.5-yth-lora/

After fine-tuning, how is the directory generated by checkpoints/fish-speech-1.5-yth-lora used in inference? Do you want to specify it during inference?

❌ Actual Behavior

After fine-tuning, the generated sound is similar to that without fine-tuning. Is the method used wrong? In which step of reasoning is the model generated by fine-tuning used?

[PoTaTo-Mika]

The lora file can be merged into the main checkpoint, and then you can use the merged one to infer.

[rhfu]

The lora file can be merged into the main checkpoint, and then you can use the merged one to infer.

When I merged the past and tried to reason again, the sound turned into noise. Thank you, I will try again.

[rhfu]

推荐是1个小时音频素材才去微调，如果只有几分钟音频，就换换推荐音频

[trackcan]

我这微调没效果（按步骤一步步操作的），结果都是噪音，有谁测验过微调吗？是不是哪里细节官方没说清楚？

[rhfu]

素材最好1个小时以上，训练时长估计得要8-9小时，