{"id":91187,"name":"github.com/fluidinference/fluidaudio","ecosystem":"cocoapods","repository_url":null,"issues_count":3,"created_at":"2026-04-28T00:01:01.491Z","updated_at":"2026-04-28T00:01:01.491Z","purl":"pkg:cocoapods/github.com/fluidinference/fluidaudio","unique_repositories_count":1,"unique_repositories_count_past_30_days":1,"recent_issues":[{"uuid":"4576590889","node_id":"PR_kwDOSB81fs7iDSyU","number":12,"state":"open","title":"deps: bump github.com/fluidinference/fluidaudio from 0.13.6 to 0.14.8","user":"dependabot[bot]","labels":[],"assignees":[],"locked":false,"comments_count":1,"pull_request":true,"closed_at":null,"author_association":null,"state_reason":null,"created_at":"2026-06-03T03:17:43.000Z","updated_at":"2026-06-03T03:17:44.000Z","time_to_close":null,"merged_at":null,"merged_by":null,"closed_by":null,"dependency_metadata":{"prefix":"deps","packages":[{"name":"github.com/fluidinference/fluidaudio","old_version":"0.13.6","new_version":"0.14.8","repository_url":"https://github.com/FluidInference/FluidAudio"}],"path":null,"ecosystem":"cocoapods"},"body":"Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.13.6 to 0.14.8.\n\u003cdetails\u003e\n\u003csummary\u003eRelease notes\u003c/summary\u003e\n\u003cp\u003e\u003cem\u003eSourced from \u003ca href=\"https://github.com/FluidInference/FluidAudio/releases\"\u003egithub.com/fluidinference/fluidaudio's releases\u003c/a\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cblockquote\u003e\n\u003ch2\u003ev0.14.8\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efix(asr): honor caller modelNames in DownloadUtils for non-baseline files (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/524\"\u003e#524\u003c/a\u003e)  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/625\"\u003eFluidInference/FluidAudio#625\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eFix/word boost improvements by \u003ca href=\"https://github.com/smdesai\"\u003e\u003ccode\u003e@​smdesai\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/634\"\u003eFluidInference/FluidAudio#634\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(cli): expose missing Parakeet langs, add Greek script support by \u003ca href=\"https://github.com/rpkyle\"\u003e\u003ccode\u003e@​rpkyle\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/637\"\u003eFluidInference/FluidAudio#637\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eAdd timestamp support to EoU Streaming by \u003ca href=\"https://github.com/Kavi-Gupta\"\u003e\u003ccode\u003e@​Kavi-Gupta\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/629\"\u003eFluidInference/FluidAudio#629\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(asr): reduce English drift on French recordings via token blocklist by \u003ca href=\"https://github.com/Matth-93\"\u003e\u003ccode\u003e@​Matth-93\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/630\"\u003eFluidInference/FluidAudio#630\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(diarizer): expose per-chunk embeddings on DiarizationResult by \u003ca href=\"https://github.com/adamsro\"\u003e\u003ccode\u003e@​adamsro\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/633\"\u003eFluidInference/FluidAudio#633\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eTTS: store downloaded models in Application Support, not Caches (iOS)  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/642\"\u003eFluidInference/FluidAudio#642\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003ePocketTTS: stride-aware conditioning extraction for cloned voices  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/643\"\u003eFluidInference/FluidAudio#643\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eSupertonic-3: built-in voice enum + on-demand voice download  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/644\"\u003eFluidInference/FluidAudio#644\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/rpkyle\"\u003e\u003ccode\u003e@​rpkyle\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/637\"\u003eFluidInference/FluidAudio#637\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/Kavi-Gupta\"\u003e\u003ccode\u003e@​Kavi-Gupta\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/629\"\u003eFluidInference/FluidAudio#629\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/jonyoder\"\u003e\u003ccode\u003e@​jonyoder\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/639\"\u003eFluidInference/FluidAudio#639\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.7...v0.14.8\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.7...v0.14.8\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.7\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/pockettts): add 5 native-language voices + slim language-pack downloads ~40% in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/624\"\u003eFluidInference/FluidAudio#624\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(cli): expose all unexposed config fields across 6 commands by \u003ca href=\"https://github.com/barzhomi\"\u003e\u003ccode\u003e@​barzhomi\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/616\"\u003eFluidInference/FluidAudio#616\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(asr): add opt-in v3 no-mel decode arbitration by \u003ca href=\"https://github.com/vdt4534\"\u003e\u003ccode\u003e@​vdt4534\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/604\"\u003eFluidInference/FluidAudio#604\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/gmickel\"\u003e\u003ccode\u003e@​gmickel\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/618\"\u003eFluidInference/FluidAudio#618\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/vdt4534\"\u003e\u003ccode\u003e@​vdt4534\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/604\"\u003eFluidInference/FluidAudio#604\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.6...v0.14.7\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.6...v0.14.7\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.6\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts): Supertonic-3 multilingual CoreML TTS in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/617\"\u003eFluidInference/FluidAudio#617\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003erefactor(tts): async StyleTTS2 predict + drop non-native Magpie synthesizeStream  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/589\"\u003eFluidInference/FluidAudio#589\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eMake SpeakerManager a struct and de-async DiarizerManager by \u003ca href=\"https://github.com/panv-kw\"\u003e\u003ccode\u003e@​panv-kw\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/591\"\u003eFluidInference/FluidAudio#591\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/magpie): warmup API for cold-start mitigation (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/60\"\u003e#60\u003c/a\u003e Track 2) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/595\"\u003eFluidInference/FluidAudio#595\u003c/a\u003e* Fixed LS-EEND Memory Leak + Updated Docs by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/605\"\u003eFluidInference/FluidAudio#605\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(tts/pocket-tts): repair v1 voice cloning for pocket-tts 2.0.0 (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/592\"\u003e#592\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/601\"\u003eFluidInference/FluidAudio#601\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eTimestamping RTTN decoder by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/608\"\u003eFluidInference/FluidAudio#608\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(asr/nemotron): seed cache_len=1 to avoid ios17.slice_by_index zero-shape warning (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/607\"\u003e#607\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/609\"\u003eFluidInference/FluidAudio#609\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(tts/pockettts): normalize French text and preserve mid-sentence chunks (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/584\"\u003e#584\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/606\"\u003eFluidInference/FluidAudio#606\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(asr): expose tdtConfig in SlidingWindowAsrConfig by \u003ca href=\"https://github.com/execsumo\"\u003e\u003ccode\u003e@​execsumo\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/611\"\u003eFluidInference/FluidAudio#611\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003edeprecate: remove CosyVoice3 and mono Kokoro (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/571\"\u003e#571\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/602\"\u003eFluidInference/FluidAudio#602\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eDiarization progress by \u003ca href=\"https://github.com/nburns\"\u003e\u003ccode\u003e@​nburns\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/615\"\u003eFluidInference/FluidAudio#615\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.5...v0.14.6\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.5...v0.14.6\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.5\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/magpie): nanocodec v1/v2/v3 + decoder_step ANE pin + dual-precision API n \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/580\"\u003eFluidInference/FluidAudio#580\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c!-- raw HTML omitted --\u003e\n\u003c/blockquote\u003e\n\u003cp\u003e... (truncated)\u003c/p\u003e\n\u003c/details\u003e\n\u003cdetails\u003e\n\u003csummary\u003eCommits\u003c/summary\u003e\n\u003cul\u003e\n\u003cli\u003eSee full diff in \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/0.13.6...56607d90b97de7d95a731200563bf49aa5beef20\"\u003ecompare view\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/details\u003e\n\u003cbr /\u003e\n\n\n[![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=github.com/fluidinference/fluidaudio\u0026package-manager=swift\u0026previous-version=0.13.6\u0026new-version=0.14.8)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)\n\nDependabot 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`.\n\n[//]: # (dependabot-automerge-start)\n[//]: # (dependabot-automerge-end)\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003eDependabot commands and options\u003c/summary\u003e\n\u003cbr /\u003e\n\nYou can trigger Dependabot actions by commenting on this PR:\n- `@dependabot rebase` will rebase this PR\n- `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it\n- `@dependabot show \u003cdependency name\u003e ignore conditions` will show all of the ignore conditions of the specified dependency\n- `@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)\n- `@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)\n- `@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)\n\n\n\u003c/details\u003e","html_url":"https://github.com/j0elmiller/lmnh/pull/12","url":"https://dependabot.ecosyste.ms/api/v1/hosts/GitHub/repositories/j0elmiller%2Flmnh/issues/12","packages_url":"https://dependabot.ecosyste.ms/api/v1/issues/12/packages"},{"uuid":"4425489375","node_id":"PR_kwDOSB81fs7afGM0","number":9,"state":"open","title":"deps: bump github.com/fluidinference/fluidaudio from 0.13.6 to 0.14.5","user":"dependabot[bot]","labels":[],"assignees":[],"locked":false,"comments_count":2,"pull_request":true,"closed_at":null,"author_association":null,"state_reason":null,"created_at":"2026-05-12T01:30:37.000Z","updated_at":"2026-05-19T05:02:59.361Z","time_to_close":null,"merged_at":null,"merged_by":null,"closed_by":null,"dependency_metadata":{"prefix":"deps","packages":[{"name":"github.com/fluidinference/fluidaudio","old_version":"0.13.6","new_version":"0.14.5","repository_url":"https://github.com/FluidInference/FluidAudio"}],"path":null,"ecosystem":"cocoapods"},"body":"Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.13.6 to 0.14.5.\n\u003cdetails\u003e\n\u003csummary\u003eRelease notes\u003c/summary\u003e\n\u003cp\u003e\u003cem\u003eSourced from \u003ca href=\"https://github.com/FluidInference/FluidAudio/releases\"\u003egithub.com/fluidinference/fluidaudio's releases\u003c/a\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cblockquote\u003e\n\u003ch2\u003ev0.14.5\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/magpie): nanocodec v1/v2/v3 + decoder_step ANE pin + dual-precision API n \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/580\"\u003eFluidInference/FluidAudio#580\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/magpie): nanocodec v4 (fp32 + int8 palettize) precision in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/581\"\u003eFluidInference/FluidAudio#581\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(tts): guard direct Float16 reads with #if arch(arm64) (CosyVoice3, StyleTTS2) by \u003ca href=\"https://github.com/Beingpax\"\u003e\u003ccode\u003e@​Beingpax\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/582\"\u003eFluidInference/FluidAudio#582\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eAdd Utter app to showcase in README.md by \u003ca href=\"https://github.com/joepetrakovich\"\u003e\u003ccode\u003e@​joepetrakovich\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/585\"\u003eFluidInference/FluidAudio#585\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eFix: Prevent Metal crash when targetTokens is 0 in Kokoro TTS by \u003ca href=\"https://github.com/gregyoung14\"\u003e\u003ccode\u003e@​gregyoung14\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/586\"\u003eFluidInference/FluidAudio#586\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts): StyleTTS2 LibriTTS (iteration_3) CoreML backend in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/588\"\u003eFluidInference/FluidAudio#588\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/joepetrakovich\"\u003e\u003ccode\u003e@​joepetrakovich\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/585\"\u003eFluidInference/FluidAudio#585\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/gregyoung14\"\u003e\u003ccode\u003e@​gregyoung14\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/586\"\u003eFluidInference/FluidAudio#586\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.4...v0.14.5\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.4...v0.14.5\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.4\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003eFeat/pocket tts int8 precision swap by \u003ca href=\"https://github.com/Gaozhongpai\"\u003e\u003ccode\u003e@​Gaozhongpai\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/558\"\u003eFluidInference/FluidAudio#558\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eOptimized LS-EEND API by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/526\"\u003eFluidInference/FluidAudio#526\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eAdded Back the Old LS-EEND Constructors by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/563\"\u003eFluidInference/FluidAudio#563\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(asr/parakeet-v3): introduce int4 encoder in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/560\"\u003eFluidInference/FluidAudio#560\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/benchmark): tts-benchmark CLI covering all TTS backends in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/557\"\u003eFluidInference/FluidAudio#557\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(asr/cohere): long-form transcribeLong + cold/warm docs \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/564\"\u003eFluidInference/FluidAudio#564\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eFix DiarizerTimeline Short Segment Filter by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/565\"\u003eFluidInference/FluidAudio#565\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/kokoro-ane): add Mandarin (v1.1-zh) variant by in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/570\"\u003eFluidInference/FluidAudio#570\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/Gaozhongpai\"\u003e\u003ccode\u003e@​Gaozhongpai\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/558\"\u003eFluidInference/FluidAudio#558\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/alamparelli\"\u003e\u003ccode\u003e@​alamparelli\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/561\"\u003eFluidInference/FluidAudio#561\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.3...v0.14.4\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.3...v0.14.4\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.3\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/magpie): add NVIDIA Magpie TTS Multilingual 357M Swift port in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/541\"\u003eFluidInference/FluidAudio#541\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/styletts2): scaffold StyleTTS2 4-stage pipeline integration in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/554\"\u003eFluidInference/FluidAudio#554\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix: remove module_map from MachTaskSelfWrapper subspec by \u003ca href=\"https://github.com/dianshu\"\u003e\u003ccode\u003e@​dianshu\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/546\"\u003eFluidInference/FluidAudio#546\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/dianshu\"\u003e\u003ccode\u003e@​dianshu\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/546\"\u003eFluidInference/FluidAudio#546\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.2...v0.14.3\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.2...v0.14.3\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.2\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003eCosyVoice3 Mandarin zero-shot TTS port: experimental backend gated to macOS 15 / iOS 18 (CoreML \u003ccode\u003eMLState\u003c/code\u003e). Package baseline stays at macOS 14 / iOS 17, so ASR / Diarization / VAD / Kokoro / PocketTTS consumers are unaffected (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/536\"\u003e#536\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003emulti-language support EN + 5 new languages: Each pack also ships an optional selective-int8 \u003ccode\u003eflowlm_stepv2.mlmodelc\u003c/code\u003e  for opt-in size reduction: \u003cstrong\u003e6L 145 MB → 74 MB, 24L 1.1 GB → 291 MB\u003c/strong\u003e. Swift continues to load fp16 by default; clients can swap in the int8 variant (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/549\"\u003e#549\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003eNew Kokoro ANE model  7-stage CoreML chain support (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/547\"\u003e#547\u003c/a\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c!-- raw HTML omitted --\u003e\n\u003c/blockquote\u003e\n\u003cp\u003e... (truncated)\u003c/p\u003e\n\u003c/details\u003e\n\u003cdetails\u003e\n\u003csummary\u003eCommits\u003c/summary\u003e\n\u003cul\u003e\n\u003cli\u003eSee full diff in \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/0.13.6...ce59fb14b8b8978b196f6a34282e20ea6762d164\"\u003ecompare view\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/details\u003e\n\u003cbr /\u003e\n\n\n[![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=github.com/fluidinference/fluidaudio\u0026package-manager=swift\u0026previous-version=0.13.6\u0026new-version=0.14.5)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)\n\nDependabot 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`.\n\n[//]: # (dependabot-automerge-start)\n[//]: # (dependabot-automerge-end)\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003eDependabot commands and options\u003c/summary\u003e\n\u003cbr /\u003e\n\nYou can trigger Dependabot actions by commenting on this PR:\n- `@dependabot rebase` will rebase this PR\n- `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it\n- `@dependabot show \u003cdependency name\u003e ignore conditions` will show all of the ignore conditions of the specified dependency\n- `@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)\n- `@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)\n- `@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)\n\n\n\u003c/details\u003e","html_url":"https://github.com/j0elmiller/lmnh/pull/9","url":"https://dependabot.ecosyste.ms/api/v1/hosts/GitHub/repositories/j0elmiller%2Flmnh/issues/9","packages_url":"https://dependabot.ecosyste.ms/api/v1/issues/9/packages"},{"uuid":"4339215893","node_id":"PR_kwDOSB81fs7WIfF9","number":7,"state":"open","title":"deps: bump github.com/fluidinference/fluidaudio from 0.13.6 to 0.14.1","user":"dependabot[bot]","labels":[],"assignees":[],"locked":false,"comments_count":1,"pull_request":true,"closed_at":null,"author_association":null,"state_reason":null,"created_at":"2026-04-27T22:21:27.000Z","updated_at":"2026-04-27T22:21:28.000Z","time_to_close":null,"merged_at":null,"merged_by":null,"closed_by":null,"dependency_metadata":{"prefix":"deps","packages":[{"name":"github.com/fluidinference/fluidaudio","old_version":"0.13.6","new_version":"0.14.1","repository_url":"https://github.com/FluidInference/FluidAudio"}],"path":null,"ecosystem":"cocoapods"},"body":"Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.13.6 to 0.14.1.\n\u003cdetails\u003e\n\u003csummary\u003eRelease notes\u003c/summary\u003e\n\u003cp\u003e\u003cem\u003eSourced from \u003ca href=\"https://github.com/FluidInference/FluidAudio/releases\"\u003egithub.com/fluidinference/fluidaudio's releases\u003c/a\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cblockquote\u003e\n\u003ch2\u003ev0.14.1\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efix(diarizer): convert SpeakerManager to actor, Speaker to struct (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/539\"\u003e#539\u003c/a\u003e) — fixes heap corruption and Swift 6 concurrency warnings on iOS 26.4 (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/528\"\u003e#528\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003eASR: fix Parakeet TDT v3 emitting Cyrillic for short Latin-script utterances (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/515\"\u003e#515\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003efeat(asr/cohere): ANE-friendly static-shape decoder v2 (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/537\"\u003e#537\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003efix: abbreviation handling for Kokoro TTS (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/538\"\u003e#538\u003c/a\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.0...v0.14.1\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.0...v0.14.1\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.0 Cohere Transcribe\u003c/h2\u003e\n\u003ch2\u003eWhat's New in v0.14.0\u003c/h2\u003e\n\u003ch3\u003eFeatures\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eCohere Transcribe ASR engine\u003c/strong\u003e (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/487\"\u003e#487\u003c/a\u003e) — new 14-language encoder-decoder ASR backed by \u003ca href=\"https://huggingface.co/CohereLabs/cohere-transcribe-03-2026\"\u003eCohereLabs/cohere-transcribe-03-2026\u003c/a\u003e, converted to CoreML as a mixed-precision INT8 encoder + FP16 cache-external decoder hybrid.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4\u003eModel\u003c/h4\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eINT8 hybrid (recommended)\u003c/strong\u003e: \u003ca href=\"https://huggingface.co/FluidInference/cohere-transcribe-q8-cache-external-coreml\"\u003eFluidInference/cohere-transcribe-q8-cache-external-coreml\u003c/a\u003e — 1.8 GB encoder, iOS 18+\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eFP16 reference\u003c/strong\u003e: \u003ca href=\"https://huggingface.co/FluidInference/cohere-transcribe-cache-external-coreml\"\u003eFluidInference/cohere-transcribe-cache-external-coreml\u003c/a\u003e — 3.6 GB encoder, iOS 17+\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4\u003eArchitecture\u003c/h4\u003e\n\u003cul\u003e\n\u003cli\u003e48-layer Conformer encoder, fixed input \u003ccode\u003e[1, 128, 3500]\u003c/code\u003e (35 s @ 10 ms hop)\u003c/li\u003e\n\u003cli\u003e8-layer decoder with host-managed (Parakeet-style) external KV cache, 108-token window\u003c/li\u003e\n\u003cli\u003e16,384 SentencePiece tokens\u003c/li\u003e\n\u003cli\u003eMax 35 s per call (matches upstream \u003ccode\u003emax_audio_clip_s: 35\u003c/code\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4\u003eSupported languages\u003c/h4\u003e\n\u003cp\u003eEnglish, French, German, Spanish, Italian, Portuguese, Dutch, Polish, Greek, Arabic, Japanese, Chinese, Korean, Vietnamese.\u003c/p\u003e\n\u003ch4\u003eAPI\u003c/h4\u003e\n\u003cpre lang=\"swift\"\u003e\u003ccode\u003eimport CoreML\r\nimport FluidAudio\r\n\u003cp\u003elet models = try await CoherePipeline.loadModels(\nencoderDir: encoderDir, decoderDir: decoderDir, vocabDir: decoderDir)\nlet result = try await CoherePipeline().transcribe(\naudio: samples, models: models, language: .english)\nprint(result.text)\n\u003c/code\u003e\u003c/pre\u003e\u003c/p\u003e\n\u003cp\u003eLanguage must be specified explicitly — Cohere Transcribe uses a conditioned prompt that hard-codes the language; wrong language produces degenerate output.\u003c/p\u003e\n\u003ch4\u003eBenchmarks\u003c/h4\u003e\n\u003cp\u003e\u003cstrong\u003eLibriSpeech test-clean\u003c/strong\u003e, full split (2,620 utterances), Apple M2 (2022), Tahoe 26.0:\u003c/p\u003e\n\u003ctable\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth\u003eSubset\u003c/th\u003e\n\u003cth align=\"right\"\u003eSamples\u003c/th\u003e\n\u003cth align=\"right\"\u003eWER\u003c/th\u003e\n\u003cth align=\"right\"\u003eCER\u003c/th\u003e\n\u003cth align=\"right\"\u003eRTFx (per-file mean)\u003c/th\u003e\n\u003cth align=\"right\"\u003eRTFx (total audio/compute)\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003etest-clean\u003c/td\u003e\n\u003ctd align=\"right\"\u003e2,620\u003c/td\u003e\n\u003ctd align=\"right\"\u003e\u003cstrong\u003e1.77%\u003c/strong\u003e\u003c/td\u003e\n\u003ctd align=\"right\"\u003e\u003cstrong\u003e0.60%\u003c/strong\u003e\u003c/td\u003e\n\u003ctd align=\"right\"\u003e2.04×\u003c/td\u003e\n\u003ctd align=\"right\"\u003e1.72×\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c!-- raw HTML omitted --\u003e\n\u003c/blockquote\u003e\n\u003cp\u003e... (truncated)\u003c/p\u003e\n\u003c/details\u003e\n\u003cdetails\u003e\n\u003csummary\u003eCommits\u003c/summary\u003e\n\u003cul\u003e\n\u003cli\u003eSee full diff in \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/0.13.6...d302273d49ef4d8914b27f20d342be482e8810f1\"\u003ecompare view\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/details\u003e\n\u003cbr /\u003e\n\n\n[![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=github.com/fluidinference/fluidaudio\u0026package-manager=swift\u0026previous-version=0.13.6\u0026new-version=0.14.1)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)\n\nDependabot 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`.\n\n[//]: # (dependabot-automerge-start)\n[//]: # (dependabot-automerge-end)\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003eDependabot commands and options\u003c/summary\u003e\n\u003cbr /\u003e\n\nYou can trigger Dependabot actions by commenting on this PR:\n- `@dependabot rebase` will rebase this PR\n- `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it\n- `@dependabot show \u003cdependency name\u003e ignore conditions` will show all of the ignore conditions of the specified dependency\n- `@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)\n- `@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)\n- `@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)\n\n\n\u003c/details\u003e","html_url":"https://github.com/j0elmiller/lmnh/pull/7","url":"https://dependabot.ecosyste.ms/api/v1/hosts/GitHub/repositories/j0elmiller%2Flmnh/issues/7","packages_url":"https://dependabot.ecosyste.ms/api/v1/issues/7/packages"}],"issue_packages":[{"old_version":"0.13.6","new_version":"0.14.8","update_type":"minor","path":null,"pr_created_at":"2026-06-03T03:17:43.000Z","version_change":"0.13.6 → 0.14.8","issue":{"uuid":"4576590889","node_id":"PR_kwDOSB81fs7iDSyU","number":12,"state":"open","title":"deps: bump github.com/fluidinference/fluidaudio from 0.13.6 to 0.14.8","user":"dependabot[bot]","labels":[],"assignees":[],"locked":false,"comments_count":1,"pull_request":true,"closed_at":null,"author_association":null,"state_reason":null,"created_at":"2026-06-03T03:17:43.000Z","updated_at":"2026-06-03T03:17:44.000Z","time_to_close":null,"merged_at":null,"merged_by":null,"closed_by":null,"dependency_metadata":{"prefix":"deps","packages":[{"name":"github.com/fluidinference/fluidaudio","old_version":"0.13.6","new_version":"0.14.8","repository_url":"https://github.com/FluidInference/FluidAudio"}],"path":null,"ecosystem":"cocoapods"},"body":"Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.13.6 to 0.14.8.\n\u003cdetails\u003e\n\u003csummary\u003eRelease notes\u003c/summary\u003e\n\u003cp\u003e\u003cem\u003eSourced from \u003ca href=\"https://github.com/FluidInference/FluidAudio/releases\"\u003egithub.com/fluidinference/fluidaudio's releases\u003c/a\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cblockquote\u003e\n\u003ch2\u003ev0.14.8\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efix(asr): honor caller modelNames in DownloadUtils for non-baseline files (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/524\"\u003e#524\u003c/a\u003e)  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/625\"\u003eFluidInference/FluidAudio#625\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eFix/word boost improvements by \u003ca href=\"https://github.com/smdesai\"\u003e\u003ccode\u003e@​smdesai\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/634\"\u003eFluidInference/FluidAudio#634\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(cli): expose missing Parakeet langs, add Greek script support by \u003ca href=\"https://github.com/rpkyle\"\u003e\u003ccode\u003e@​rpkyle\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/637\"\u003eFluidInference/FluidAudio#637\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eAdd timestamp support to EoU Streaming by \u003ca href=\"https://github.com/Kavi-Gupta\"\u003e\u003ccode\u003e@​Kavi-Gupta\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/629\"\u003eFluidInference/FluidAudio#629\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(asr): reduce English drift on French recordings via token blocklist by \u003ca href=\"https://github.com/Matth-93\"\u003e\u003ccode\u003e@​Matth-93\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/630\"\u003eFluidInference/FluidAudio#630\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(diarizer): expose per-chunk embeddings on DiarizationResult by \u003ca href=\"https://github.com/adamsro\"\u003e\u003ccode\u003e@​adamsro\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/633\"\u003eFluidInference/FluidAudio#633\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eTTS: store downloaded models in Application Support, not Caches (iOS)  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/642\"\u003eFluidInference/FluidAudio#642\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003ePocketTTS: stride-aware conditioning extraction for cloned voices  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/643\"\u003eFluidInference/FluidAudio#643\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eSupertonic-3: built-in voice enum + on-demand voice download  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/644\"\u003eFluidInference/FluidAudio#644\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/rpkyle\"\u003e\u003ccode\u003e@​rpkyle\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/637\"\u003eFluidInference/FluidAudio#637\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/Kavi-Gupta\"\u003e\u003ccode\u003e@​Kavi-Gupta\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/629\"\u003eFluidInference/FluidAudio#629\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/jonyoder\"\u003e\u003ccode\u003e@​jonyoder\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/639\"\u003eFluidInference/FluidAudio#639\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.7...v0.14.8\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.7...v0.14.8\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.7\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/pockettts): add 5 native-language voices + slim language-pack downloads ~40% in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/624\"\u003eFluidInference/FluidAudio#624\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(cli): expose all unexposed config fields across 6 commands by \u003ca href=\"https://github.com/barzhomi\"\u003e\u003ccode\u003e@​barzhomi\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/616\"\u003eFluidInference/FluidAudio#616\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(asr): add opt-in v3 no-mel decode arbitration by \u003ca href=\"https://github.com/vdt4534\"\u003e\u003ccode\u003e@​vdt4534\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/604\"\u003eFluidInference/FluidAudio#604\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/gmickel\"\u003e\u003ccode\u003e@​gmickel\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/618\"\u003eFluidInference/FluidAudio#618\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/vdt4534\"\u003e\u003ccode\u003e@​vdt4534\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/604\"\u003eFluidInference/FluidAudio#604\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.6...v0.14.7\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.6...v0.14.7\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.6\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts): Supertonic-3 multilingual CoreML TTS in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/617\"\u003eFluidInference/FluidAudio#617\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003erefactor(tts): async StyleTTS2 predict + drop non-native Magpie synthesizeStream  in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/589\"\u003eFluidInference/FluidAudio#589\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eMake SpeakerManager a struct and de-async DiarizerManager by \u003ca href=\"https://github.com/panv-kw\"\u003e\u003ccode\u003e@​panv-kw\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/591\"\u003eFluidInference/FluidAudio#591\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/magpie): warmup API for cold-start mitigation (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/60\"\u003e#60\u003c/a\u003e Track 2) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/595\"\u003eFluidInference/FluidAudio#595\u003c/a\u003e* Fixed LS-EEND Memory Leak + Updated Docs by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/605\"\u003eFluidInference/FluidAudio#605\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(tts/pocket-tts): repair v1 voice cloning for pocket-tts 2.0.0 (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/592\"\u003e#592\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/601\"\u003eFluidInference/FluidAudio#601\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eTimestamping RTTN decoder by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/608\"\u003eFluidInference/FluidAudio#608\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(asr/nemotron): seed cache_len=1 to avoid ios17.slice_by_index zero-shape warning (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/607\"\u003e#607\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/609\"\u003eFluidInference/FluidAudio#609\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(tts/pockettts): normalize French text and preserve mid-sentence chunks (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/584\"\u003e#584\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/606\"\u003eFluidInference/FluidAudio#606\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(asr): expose tdtConfig in SlidingWindowAsrConfig by \u003ca href=\"https://github.com/execsumo\"\u003e\u003ccode\u003e@​execsumo\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/611\"\u003eFluidInference/FluidAudio#611\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003edeprecate: remove CosyVoice3 and mono Kokoro (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/571\"\u003e#571\u003c/a\u003e) in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/602\"\u003eFluidInference/FluidAudio#602\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eDiarization progress by \u003ca href=\"https://github.com/nburns\"\u003e\u003ccode\u003e@​nburns\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/615\"\u003eFluidInference/FluidAudio#615\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.5...v0.14.6\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.5...v0.14.6\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.5\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/magpie): nanocodec v1/v2/v3 + decoder_step ANE pin + dual-precision API n \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/580\"\u003eFluidInference/FluidAudio#580\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c!-- raw HTML omitted --\u003e\n\u003c/blockquote\u003e\n\u003cp\u003e... (truncated)\u003c/p\u003e\n\u003c/details\u003e\n\u003cdetails\u003e\n\u003csummary\u003eCommits\u003c/summary\u003e\n\u003cul\u003e\n\u003cli\u003eSee full diff in \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/0.13.6...56607d90b97de7d95a731200563bf49aa5beef20\"\u003ecompare view\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/details\u003e\n\u003cbr /\u003e\n\n\n[![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=github.com/fluidinference/fluidaudio\u0026package-manager=swift\u0026previous-version=0.13.6\u0026new-version=0.14.8)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)\n\nDependabot 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`.\n\n[//]: # (dependabot-automerge-start)\n[//]: # (dependabot-automerge-end)\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003eDependabot commands and options\u003c/summary\u003e\n\u003cbr /\u003e\n\nYou can trigger Dependabot actions by commenting on this PR:\n- `@dependabot rebase` will rebase this PR\n- `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it\n- `@dependabot show \u003cdependency name\u003e ignore conditions` will show all of the ignore conditions of the specified dependency\n- `@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)\n- `@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)\n- `@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)\n\n\n\u003c/details\u003e","html_url":"https://github.com/j0elmiller/lmnh/pull/12","url":"https://dependabot.ecosyste.ms/api/v1/hosts/GitHub/repositories/j0elmiller%2Flmnh/issues/12","packages_url":"https://dependabot.ecosyste.ms/api/v1/issues/12/packages"}},{"old_version":"0.13.6","new_version":"0.14.5","update_type":"minor","path":null,"pr_created_at":"2026-05-12T01:30:37.000Z","version_change":"0.13.6 → 0.14.5","issue":{"uuid":"4425489375","node_id":"PR_kwDOSB81fs7afGM0","number":9,"state":"open","title":"deps: bump github.com/fluidinference/fluidaudio from 0.13.6 to 0.14.5","user":"dependabot[bot]","labels":[],"assignees":[],"locked":false,"comments_count":2,"pull_request":true,"closed_at":null,"author_association":null,"state_reason":null,"created_at":"2026-05-12T01:30:37.000Z","updated_at":"2026-05-19T05:02:59.361Z","time_to_close":null,"merged_at":null,"merged_by":null,"closed_by":null,"dependency_metadata":{"prefix":"deps","packages":[{"name":"github.com/fluidinference/fluidaudio","old_version":"0.13.6","new_version":"0.14.5","repository_url":"https://github.com/FluidInference/FluidAudio"}],"path":null,"ecosystem":"cocoapods"},"body":"Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.13.6 to 0.14.5.\n\u003cdetails\u003e\n\u003csummary\u003eRelease notes\u003c/summary\u003e\n\u003cp\u003e\u003cem\u003eSourced from \u003ca href=\"https://github.com/FluidInference/FluidAudio/releases\"\u003egithub.com/fluidinference/fluidaudio's releases\u003c/a\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cblockquote\u003e\n\u003ch2\u003ev0.14.5\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/magpie): nanocodec v1/v2/v3 + decoder_step ANE pin + dual-precision API n \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/580\"\u003eFluidInference/FluidAudio#580\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/magpie): nanocodec v4 (fp32 + int8 palettize) precision in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/581\"\u003eFluidInference/FluidAudio#581\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix(tts): guard direct Float16 reads with #if arch(arm64) (CosyVoice3, StyleTTS2) by \u003ca href=\"https://github.com/Beingpax\"\u003e\u003ccode\u003e@​Beingpax\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/582\"\u003eFluidInference/FluidAudio#582\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eAdd Utter app to showcase in README.md by \u003ca href=\"https://github.com/joepetrakovich\"\u003e\u003ccode\u003e@​joepetrakovich\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/585\"\u003eFluidInference/FluidAudio#585\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eFix: Prevent Metal crash when targetTokens is 0 in Kokoro TTS by \u003ca href=\"https://github.com/gregyoung14\"\u003e\u003ccode\u003e@​gregyoung14\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/586\"\u003eFluidInference/FluidAudio#586\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts): StyleTTS2 LibriTTS (iteration_3) CoreML backend in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/588\"\u003eFluidInference/FluidAudio#588\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/joepetrakovich\"\u003e\u003ccode\u003e@​joepetrakovich\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/585\"\u003eFluidInference/FluidAudio#585\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/gregyoung14\"\u003e\u003ccode\u003e@​gregyoung14\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/586\"\u003eFluidInference/FluidAudio#586\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.4...v0.14.5\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.4...v0.14.5\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.4\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003eFeat/pocket tts int8 precision swap by \u003ca href=\"https://github.com/Gaozhongpai\"\u003e\u003ccode\u003e@​Gaozhongpai\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/558\"\u003eFluidInference/FluidAudio#558\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eOptimized LS-EEND API by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/526\"\u003eFluidInference/FluidAudio#526\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eAdded Back the Old LS-EEND Constructors by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/563\"\u003eFluidInference/FluidAudio#563\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(asr/parakeet-v3): introduce int4 encoder in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/560\"\u003eFluidInference/FluidAudio#560\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/benchmark): tts-benchmark CLI covering all TTS backends in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/557\"\u003eFluidInference/FluidAudio#557\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(asr/cohere): long-form transcribeLong + cold/warm docs \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/564\"\u003eFluidInference/FluidAudio#564\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eFix DiarizerTimeline Short Segment Filter by \u003ca href=\"https://github.com/SGD2718\"\u003e\u003ccode\u003e@​SGD2718\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/565\"\u003eFluidInference/FluidAudio#565\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/kokoro-ane): add Mandarin (v1.1-zh) variant by in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/570\"\u003eFluidInference/FluidAudio#570\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/Gaozhongpai\"\u003e\u003ccode\u003e@​Gaozhongpai\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/558\"\u003eFluidInference/FluidAudio#558\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/alamparelli\"\u003e\u003ccode\u003e@​alamparelli\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/561\"\u003eFluidInference/FluidAudio#561\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.3...v0.14.4\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.3...v0.14.4\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.3\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efeat(tts/magpie): add NVIDIA Magpie TTS Multilingual 357M Swift port in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/541\"\u003eFluidInference/FluidAudio#541\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efeat(tts/styletts2): scaffold StyleTTS2 4-stage pipeline integration in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/554\"\u003eFluidInference/FluidAudio#554\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003efix: remove module_map from MachTaskSelfWrapper subspec by \u003ca href=\"https://github.com/dianshu\"\u003e\u003ccode\u003e@​dianshu\u003c/code\u003e\u003c/a\u003e in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/546\"\u003eFluidInference/FluidAudio#546\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003eNew Contributors\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/dianshu\"\u003e\u003ccode\u003e@​dianshu\u003c/code\u003e\u003c/a\u003e made their first contribution in \u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/pull/546\"\u003eFluidInference/FluidAudio#546\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.2...v0.14.3\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.2...v0.14.3\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.2\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003eCosyVoice3 Mandarin zero-shot TTS port: experimental backend gated to macOS 15 / iOS 18 (CoreML \u003ccode\u003eMLState\u003c/code\u003e). Package baseline stays at macOS 14 / iOS 17, so ASR / Diarization / VAD / Kokoro / PocketTTS consumers are unaffected (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/536\"\u003e#536\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003emulti-language support EN + 5 new languages: Each pack also ships an optional selective-int8 \u003ccode\u003eflowlm_stepv2.mlmodelc\u003c/code\u003e  for opt-in size reduction: \u003cstrong\u003e6L 145 MB → 74 MB, 24L 1.1 GB → 291 MB\u003c/strong\u003e. Swift continues to load fp16 by default; clients can swap in the int8 variant (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/549\"\u003e#549\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003eNew Kokoro ANE model  7-stage CoreML chain support (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/547\"\u003e#547\u003c/a\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c!-- raw HTML omitted --\u003e\n\u003c/blockquote\u003e\n\u003cp\u003e... (truncated)\u003c/p\u003e\n\u003c/details\u003e\n\u003cdetails\u003e\n\u003csummary\u003eCommits\u003c/summary\u003e\n\u003cul\u003e\n\u003cli\u003eSee full diff in \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/0.13.6...ce59fb14b8b8978b196f6a34282e20ea6762d164\"\u003ecompare view\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/details\u003e\n\u003cbr /\u003e\n\n\n[![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=github.com/fluidinference/fluidaudio\u0026package-manager=swift\u0026previous-version=0.13.6\u0026new-version=0.14.5)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)\n\nDependabot 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`.\n\n[//]: # (dependabot-automerge-start)\n[//]: # (dependabot-automerge-end)\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003eDependabot commands and options\u003c/summary\u003e\n\u003cbr /\u003e\n\nYou can trigger Dependabot actions by commenting on this PR:\n- `@dependabot rebase` will rebase this PR\n- `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it\n- `@dependabot show \u003cdependency name\u003e ignore conditions` will show all of the ignore conditions of the specified dependency\n- `@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)\n- `@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)\n- `@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)\n\n\n\u003c/details\u003e","html_url":"https://github.com/j0elmiller/lmnh/pull/9","url":"https://dependabot.ecosyste.ms/api/v1/hosts/GitHub/repositories/j0elmiller%2Flmnh/issues/9","packages_url":"https://dependabot.ecosyste.ms/api/v1/issues/9/packages"}},{"old_version":"0.13.6","new_version":"0.14.1","update_type":"minor","path":null,"pr_created_at":"2026-04-27T22:21:27.000Z","version_change":"0.13.6 → 0.14.1","issue":{"uuid":"4339215893","node_id":"PR_kwDOSB81fs7WIfF9","number":7,"state":"open","title":"deps: bump github.com/fluidinference/fluidaudio from 0.13.6 to 0.14.1","user":"dependabot[bot]","labels":[],"assignees":[],"locked":false,"comments_count":1,"pull_request":true,"closed_at":null,"author_association":null,"state_reason":null,"created_at":"2026-04-27T22:21:27.000Z","updated_at":"2026-04-27T22:21:28.000Z","time_to_close":null,"merged_at":null,"merged_by":null,"closed_by":null,"dependency_metadata":{"prefix":"deps","packages":[{"name":"github.com/fluidinference/fluidaudio","old_version":"0.13.6","new_version":"0.14.1","repository_url":"https://github.com/FluidInference/FluidAudio"}],"path":null,"ecosystem":"cocoapods"},"body":"Bumps [github.com/fluidinference/fluidaudio](https://github.com/FluidInference/FluidAudio) from 0.13.6 to 0.14.1.\n\u003cdetails\u003e\n\u003csummary\u003eRelease notes\u003c/summary\u003e\n\u003cp\u003e\u003cem\u003eSourced from \u003ca href=\"https://github.com/FluidInference/FluidAudio/releases\"\u003egithub.com/fluidinference/fluidaudio's releases\u003c/a\u003e.\u003c/em\u003e\u003c/p\u003e\n\u003cblockquote\u003e\n\u003ch2\u003ev0.14.1\u003c/h2\u003e\n\u003ch2\u003eWhat's Changed\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003efix(diarizer): convert SpeakerManager to actor, Speaker to struct (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/539\"\u003e#539\u003c/a\u003e) — fixes heap corruption and Swift 6 concurrency warnings on iOS 26.4 (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/528\"\u003e#528\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003eASR: fix Parakeet TDT v3 emitting Cyrillic for short Latin-script utterances (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/515\"\u003e#515\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003efeat(asr/cohere): ANE-friendly static-shape decoder v2 (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/537\"\u003e#537\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003efix: abbreviation handling for Kokoro TTS (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/538\"\u003e#538\u003c/a\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eFull Changelog\u003c/strong\u003e: \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/v0.14.0...v0.14.1\"\u003ehttps://github.com/FluidInference/FluidAudio/compare/v0.14.0...v0.14.1\u003c/a\u003e\u003c/p\u003e\n\u003ch2\u003ev0.14.0 Cohere Transcribe\u003c/h2\u003e\n\u003ch2\u003eWhat's New in v0.14.0\u003c/h2\u003e\n\u003ch3\u003eFeatures\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eCohere Transcribe ASR engine\u003c/strong\u003e (\u003ca href=\"https://redirect.github.com/FluidInference/FluidAudio/issues/487\"\u003e#487\u003c/a\u003e) — new 14-language encoder-decoder ASR backed by \u003ca href=\"https://huggingface.co/CohereLabs/cohere-transcribe-03-2026\"\u003eCohereLabs/cohere-transcribe-03-2026\u003c/a\u003e, converted to CoreML as a mixed-precision INT8 encoder + FP16 cache-external decoder hybrid.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4\u003eModel\u003c/h4\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eINT8 hybrid (recommended)\u003c/strong\u003e: \u003ca href=\"https://huggingface.co/FluidInference/cohere-transcribe-q8-cache-external-coreml\"\u003eFluidInference/cohere-transcribe-q8-cache-external-coreml\u003c/a\u003e — 1.8 GB encoder, iOS 18+\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eFP16 reference\u003c/strong\u003e: \u003ca href=\"https://huggingface.co/FluidInference/cohere-transcribe-cache-external-coreml\"\u003eFluidInference/cohere-transcribe-cache-external-coreml\u003c/a\u003e — 3.6 GB encoder, iOS 17+\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4\u003eArchitecture\u003c/h4\u003e\n\u003cul\u003e\n\u003cli\u003e48-layer Conformer encoder, fixed input \u003ccode\u003e[1, 128, 3500]\u003c/code\u003e (35 s @ 10 ms hop)\u003c/li\u003e\n\u003cli\u003e8-layer decoder with host-managed (Parakeet-style) external KV cache, 108-token window\u003c/li\u003e\n\u003cli\u003e16,384 SentencePiece tokens\u003c/li\u003e\n\u003cli\u003eMax 35 s per call (matches upstream \u003ccode\u003emax_audio_clip_s: 35\u003c/code\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4\u003eSupported languages\u003c/h4\u003e\n\u003cp\u003eEnglish, French, German, Spanish, Italian, Portuguese, Dutch, Polish, Greek, Arabic, Japanese, Chinese, Korean, Vietnamese.\u003c/p\u003e\n\u003ch4\u003eAPI\u003c/h4\u003e\n\u003cpre lang=\"swift\"\u003e\u003ccode\u003eimport CoreML\r\nimport FluidAudio\r\n\u003cp\u003elet models = try await CoherePipeline.loadModels(\nencoderDir: encoderDir, decoderDir: decoderDir, vocabDir: decoderDir)\nlet result = try await CoherePipeline().transcribe(\naudio: samples, models: models, language: .english)\nprint(result.text)\n\u003c/code\u003e\u003c/pre\u003e\u003c/p\u003e\n\u003cp\u003eLanguage must be specified explicitly — Cohere Transcribe uses a conditioned prompt that hard-codes the language; wrong language produces degenerate output.\u003c/p\u003e\n\u003ch4\u003eBenchmarks\u003c/h4\u003e\n\u003cp\u003e\u003cstrong\u003eLibriSpeech test-clean\u003c/strong\u003e, full split (2,620 utterances), Apple M2 (2022), Tahoe 26.0:\u003c/p\u003e\n\u003ctable\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth\u003eSubset\u003c/th\u003e\n\u003cth align=\"right\"\u003eSamples\u003c/th\u003e\n\u003cth align=\"right\"\u003eWER\u003c/th\u003e\n\u003cth align=\"right\"\u003eCER\u003c/th\u003e\n\u003cth align=\"right\"\u003eRTFx (per-file mean)\u003c/th\u003e\n\u003cth align=\"right\"\u003eRTFx (total audio/compute)\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003etest-clean\u003c/td\u003e\n\u003ctd align=\"right\"\u003e2,620\u003c/td\u003e\n\u003ctd align=\"right\"\u003e\u003cstrong\u003e1.77%\u003c/strong\u003e\u003c/td\u003e\n\u003ctd align=\"right\"\u003e\u003cstrong\u003e0.60%\u003c/strong\u003e\u003c/td\u003e\n\u003ctd align=\"right\"\u003e2.04×\u003c/td\u003e\n\u003ctd align=\"right\"\u003e1.72×\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c!-- raw HTML omitted --\u003e\n\u003c/blockquote\u003e\n\u003cp\u003e... (truncated)\u003c/p\u003e\n\u003c/details\u003e\n\u003cdetails\u003e\n\u003csummary\u003eCommits\u003c/summary\u003e\n\u003cul\u003e\n\u003cli\u003eSee full diff in \u003ca href=\"https://github.com/FluidInference/FluidAudio/compare/0.13.6...d302273d49ef4d8914b27f20d342be482e8810f1\"\u003ecompare view\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/details\u003e\n\u003cbr /\u003e\n\n\n[![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=github.com/fluidinference/fluidaudio\u0026package-manager=swift\u0026previous-version=0.13.6\u0026new-version=0.14.1)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)\n\nDependabot 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`.\n\n[//]: # (dependabot-automerge-start)\n[//]: # (dependabot-automerge-end)\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003eDependabot commands and options\u003c/summary\u003e\n\u003cbr /\u003e\n\nYou can trigger Dependabot actions by commenting on this PR:\n- `@dependabot rebase` will rebase this PR\n- `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it\n- `@dependabot show \u003cdependency name\u003e ignore conditions` will show all of the ignore conditions of the specified dependency\n- `@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)\n- `@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)\n- `@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)\n\n\n\u003c/details\u003e","html_url":"https://github.com/j0elmiller/lmnh/pull/7","url":"https://dependabot.ecosyste.ms/api/v1/hosts/GitHub/repositories/j0elmiller%2Flmnh/issues/7","packages_url":"https://dependabot.ecosyste.ms/api/v1/issues/7/packages"}}]}