Baike.dev
Log in
Back to tool/Back to issues
#1325·fish-speech

Conversation.visualize exposes unsupported use_color and merge_audio_tokens parameters

Author: unsallabsCreated Aug 17, 2026Updated Sep 17, 2026
Labelsbugstale

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

OS: Windows 11 Python: 3.11 Project: fish-speech Component: Conversation.visualize()

Steps to Reproduce

  1. Create or use a Conversation instance.
  2. Call Conversation.visualize() with use_color=False:
python
conversation.visualize(
    tokenizer,
    use_color=False,
)
3. Observe that the output still contains ANSI color formatting.
Conversation.visualize() also exposes merge_audio_tokens:
conversation.visualize(
    tokenizer,
    merge_audio_tokens=True,
)
However, ContentSequence.visualize() does not currently implement this parameter.

### ✔️ Expected Behavior


```markdown
`use_color=False` should disable ANSI color formatting in the visualization output.

If `merge_audio_tokens` is intended to be a supported option, enabling it should affect the visualization accordingly.

If these parameters are not intended to be supported, they could instead be removed or documented as unsupported.

### ❌ Actual Behavior

`Conversation.visualize()` accepts `use_color` and `merge_audio_tokens`, but these parameters currently have no effect.

When `use_color=False` is passed, the visualization still uses ANSI color formatting.

`merge_audio_tokens=True` is also ignored because `ContentSequence.visualize()` currently only supports `ignore_loss_tokens` and `merge_semantic_tokens`.

This results in public API parameters that appear to be supported but do not currently change the behavior.

I would be happy to submit a PR if these parameters are intended to be supported.

Source: fishaudio/fish-speech

View original on GitHubView discussion on GitHub