0.2.5

What's Changed

Everything! since this is the first release on github. But I think since 0.2.4:

• fix: require torch>=2.5.1 for CUDA graph reliability by @andimarafioti in #53
• Fix sample-rate mismatch causing low/slow audio in OpenAI server by @andimarafioti in #54
• feat: add instruct parameter to generate_voice_clone by @andimarafioti in #40
• Fix generate_with_embedding speech tokenizer access by @andimarafioti in #61
• feat: accept precomputed voice_clone_prompt in public API by @andimarafioti in #56
• fix(base): align voice clone defaults with upstream by @andimarafioti in #69
• Add StreamPlayer helper and clarify streaming playback docs by @andimarafioti in #78
• docs(setup): update Blackwell guidance to cu128 by @andimarafioti in #67
• Add demo text-feed toggle for non_streaming_mode by @andimarafioti in #90
• Fix PyPI license metadata by @andimarafioti in #91

Full Changelog: https://github.com/andimarafioti/faster-qwen3-tts/commits/0.2.5