RagLeapSmall follow-up to 0.1.0. Pre-alpha, for testing. No code change.
What changed: a Documentation project URL (it points to the package's design notes on GitHub, not a docs site), and a note that ragleap-voice[whisper] installed from PyPI on a fresh Colab T4 runtime and transcribed one clip each in fr, pt, ru, tr, id, vi and ja (see #960).
What it does not claim: every model is unverified and none is a default, so calls need allow_unverified=True. One clip per language is not an accuracy measurement. There is no comparison with any other service. Not built: streaming, a voice-activity gate, resampling, voice cloning, emotion names.
Help wanted: #960 (test Whisper), #990 (test Parler, gated model), #970 (roadmap).