Skip to content

deps: bump github.com/fluidinference/fluidaudio from 0.15.5 to 0.18.3 - #10

Open
dependabot[bot] wants to merge 1 commit into
mainfrom
dependabot/swift/github.com/fluidinference/fluidaudio-0.18.3
Open

dependabot[bot] wants to merge 1 commit into
mainfrom
dependabot/swift/github.com/fluidinference/fluidaudio-0.18.3

Conversation

@dependabot

@dependabot dependabot Bot commented on behalf of github Nov 11, 2026

Copy link
Copy Markdown
Contributor

Bumps github.com/fluidinference/fluidaudio from 0.15.5 to 0.18.3.

Release notes

Sourced from github.com/fluidinference/fluidaudio's releases.

v0.18.3 — Kokoro Mandarin punctuation pauses and misaki parity, streaming encoder-outputs callback

Kokoro Mandarin

  • Punctuation pauses (#1131, #1105): KokoroAneManager.setPunctuationPauses(["、": .milliseconds(300)]), or --pause 、=300 in the CLI. Keys match the mark as written, so 、 and , can be set separately. A longer pause is rendered as silence, on the legacy chain and ANE-v3, for every language.
  • Phonemes match misaki v1.1 (#1129): the space misaki puts after fullwidth punctuation (, → , ), / between words, and erhua as R (小孩儿 → ㄒ要3ㄏㄞR2), the tokens Kokoro v1.1-zh was trained on. Mandarin strings from phonemes(for:) change to match, and custom-lexicon @ escapes accept R.
  • Mainland polyphone readings (#1130): g2pW reads 和 as hé instead of the Taiwan hàn, and 吨, 厂 and 咋 as dūn, chǎng and zǎ.

Kokoro ANE-v3 follows the OS 27 default routing instead of pinning Noise, Tail and SourceGenerator to the GPU. That GPU route aborted in MPSGraph on macOS 27.0.1 (#1104).

ASR

  • Opt-in raw encoder-outputs callback for StreamingUnifiedAsrManager and StreamingNemotronAsrManager (#1109).
  • CtcModels downloads report download and compile progress (#1124).
  • Streaming: a word re-spelled at a window seam ("thirty" → "30") replaces the kept one (#1128). tdtJa's seam re-decode takes a time tail instead of the whole previous window, which dropped whole windows (#1125).
  • Vocabulary boosting reads tokenizer.json next to CtcModels loaded from a custom directory (#1126), and scores a lone stopword that is an exact alias (#1127).
  • Language hints treat currency symbols such as € and £ as script-neutral (#1120).

Pocket TTS: closing quotes and brackets stay with their sentence (#1121), and a period inside a word (www.kyutai.org) no longer ends a sentence (#1122).

Docs: tdtJa's multi-sentence window limitation (#1133).

What's Changed

New Contributors

Full Changelog: FluidInference/FluidAudio@v0.18.2...v0.18.3

v0.18.2 — CAM++ v2, Pocket TTS 2026-10, long-form ASR repairs

Models

  • Pocket TTS english_2026-10 is now the English pack (#1089). It uses Kyutai's default temperature of 0.3; .english2026Apr keeps the previous pack.
  • CAM++ _v2 speaker embeddings (#1094, #1099) now match 3D-Speaker's reference at every length: AMI EER 4.67% → 2.85%, equal to PyTorch.

... (truncated)

Commits
  • c8b41f9 fix(build): bump podspec to 0.18.3
  • 4b53afa fix(tts/kokoro-ane/zh): keep g2pW to the mainland dict's readings (#1130)
  • d98ec82 feat(tts/kokoro-ane): per-punctuation pause durations (#1105) (#1131)
  • d2bfd78 fix(tts/kokoro-ane/zh): emit misaki v1.1's spaces, word slashes and erhua R (...
  • a99f5e4 docs(asr/tdtJa): document the multi-sentence window limitation (#1117) (#1133)
  • 70f4405 fix(asr/tdtJa): streaming seam re-decodes a time tail, not the whole previous...
  • 9e6723c feat(asr/vocab): report download and compile progress from CtcModels (#1124)
  • 3037913 fix(asr/streaming): a re-spelled word at a window seam replaces the kept one ...
  • bc02577 fix(asr/vocab): score a lone stopword that is an exact alias (#1115) (#1127)
  • 02bcb9a fix(asr/vocab): read CTC tokenizer next to the loaded CtcModels (#1114) (#1126)
  • Additional commits viewable in compare view

Dependabot compatibility score

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.


Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

  • @dependabot rebase will rebase this PR
  • @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
  • @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
  • @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
  • @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
  • @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.15.5 to 0.18.3.
- [Release notes](https://github.com/FluidInference/FluidAudio/releases)
- [Commits](FluidInference/FluidAudio@v0.15.5...v0.18.3)

---
updated-dependencies:
- dependency-name: github.com/fluidinference/fluidaudio
  dependency-version: 0.18.3
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
@dependabot dependabot Bot added dependencies Pull requests that update a dependency file swift_package_manager Pull requests that update swift_package_manager code labels Nov 11, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

dependencies Pull requests that update a dependency file swift_package_manager Pull requests that update swift_package_manager code

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants