模型自定义参数
Offen
enhancement
- Vorherrschende Sprache
- Dart
- Sterne
- 8.8k
- Forks
- 626
- Ø Merge
- 9 T. 22 Std.
- Gemergte PRs (30 T.)
- 1
Beschreibung
因为模型进化很快，而且不同供应商的参数不太一致，希望能够自己自定义参数。
比如，希望能够自定义参数把模型的思考关闭。Qwen3系列的很多模型都支持参数完全关闭思考。这样可以提高阅读时查询单词的响应速度。
感谢！
> This has been implemented in v1.14.0. The release includes provider-level `reasoning_effort` configuration with auto mode and advanced settings. You can now configure reasoning effort per provider in the AI Provider Configuration Center. Please update to the latest version and check Settings > AI > Provider Configuration.
_Originally posted by @Anxcye in [#840](https://github.com/Anxcye/anx-reader/issues/840#issuecomment-4350865485)_
Beitragsleitfaden
Bewertung
Dieses Issue wurde noch nicht bewertet.