Repository navigation
speech, speech.cpp's one executable (tts, asr, vad, voice, info, devices, models, pull, rm, quantize, update, serve and worker), and the shared library libspeech with its C API, speech.h, in one archive for each of macOS arm64 (Metal), Windows x64 (Vulkan) and Linux x64 (Vulkan, and the CPU alone; glibc 2.34 or later). The README lists the model families and where their converted weights are on Hugging Face.
What's Changed
- Look for OpenAI's name of an input before naming a model's option by @nyosegawa in #201
- Name Haas-ASR's custom vocabulary keywords, as OpenAI's transcription does, and refuse timestamps with it by @nyosegawa in #202
- Release 0.10.0, with Haas-ASR's custom vocabulary under OpenAI's name keywords by @nyosegawa in #203
Full Changelog: v0.9.4...v0.10.0