What's Changed
- Add token-level transcription output with per-token log-probabilities by @visheshd in #103
- nemotron: add word/token-level timestamps by @sammcj in #108
- Dedup + api consistency by @altunenes in #110
- Make core ml compute units configurable by @altunenes in #112
- add ai contributing policy by @altunenes in #113
- Read nemotron streaming profile from onnx metadata by @altunenes in #114
- Fix dropped single-word utterances in the EOU streaming decoder by @bfulton in #115
- update ort rc13 by @altunenes in #116
New Contributors
Full Changelog: v0.3.6...v0.3.7