Inferece result has noise

[ChengsongLu]

Self Checks

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

1. Is this request related to a challenge you're experiencing? Tell us your story.

I followed the steps that inference with an specific speaker (text.wav), and the result (output.wav) quality doesn's sound good, it have some noises there.

inference.zip

And here are the steps I have done:

python tools/vqgan/inference.py -i "text.wav" --checkpoint-path "checkpoints/fish-speech-1.5/firefly-gan-vq-fsq-8x1024-21hz-generator.pth"

python tools/llama/generate.py --text "你们这个是什么群啊，你们这是害人不浅啊你们这个群！谁是群主，出来！真的太过分了。" --prompt-text "人间灯火倒映湖中，她的渴望让静水泛起涟漪。若代价只是孤独，那就让这份愿望肆意流淌。流入她所注视的世间，也流入她如湖水般澄澈的目光。" --prompt-tokens "fake.npy" --checkpoint-path "checkpoints/fish-speech-1.5" --num-samples 2

python tools/vqgan/inference.py -i "codes_0.npy" --checkpoint-path "checkpoints/fish-speech-1.5/firefly-gan-vq-fsq-8x1024-21hz-generator.pth"

2. What is your suggested solution?

Any suggestions on how to improve the sound quality?

3. Additional context or comments

No response

4. Can you help us with this feature?

• [X] I am interested in contributing to this feature.