Card: FluidAudio usage via AsrModelVersion.bark3 / .bark3Clean (FluidAudio #1054)

README.md

@@ -75,7 +75,8 @@ corpus WER (total edits / total reference words, lowercase, punctuation stripped
 | Bark 3.0 Clean, upstream MLX (`bark_mlx.py`, fp16), same files and scorer | 2.39 % | — | 55× / — |
 | Parakeet TDT v3 (FluidInference, reference) | 2.27 % | 4.12 % | 150× / 140× |
 
-Long-form: a one-hour Earnings-22 recording (four concatenated calls) transcribes in **13.0 s (276×)** with Bark 3.0.
+Long-form: a one-hour Earnings-22 recording (four concatenated calls) transcribes in **6.5 s (558×)** with Bark 3.0 and
+7.0 s (514×) with Clean, models warm; Clean writes 7,787 words to Bark 3.0's 8,223 (fillers and repeats removed).
 
 ¹ Measured with a dense 6-bit-palette build of the same exact encoder weights; the shipped sparse `Encoder.mlmodelc`
 matched it on test-other (4.66 % both, 143× vs 142×).
@@ -86,22 +87,19 @@ Notes:
 - Bark 3.0 is evaluated upstream with 0.25 s of leading silence. Through FluidAudio's sliding-window pipeline it makes
   no difference (test-other 5.02 % with it vs 4.96 % without), so none is added.
 - Bark models emit `<unk>` (token 0) where their training text had a quotation mark. Upstream decoders skip special
-  tokens. Released FluidAudio versions print it, so strip it from `result.text` until FluidAudio drops it itself
-  (`result.text.replacingOccurrences(of: "<unk>", with: "")`); the WER numbers above are scored without it.
+  tokens; FluidAudio drops it from the transcript as of [#1054](https://github.com/FluidInference/FluidAudio/pull/1054).
+  On older releases strip it from `result.text` (`replacingOccurrences(of: "<unk>", with: "")`); the WER numbers above
+  are scored without it.
 
 ## Usage (Swift, FluidAudio)
 
-Download one folder, then load it with the v3 component layout:
-
-```bash
-hf download FluidInference/bark-3.0-coreml --include "bark-3.0/*" --local-dir bark
-```
+With [FluidAudio](https://github.com/FluidInference/FluidAudio) ([#1054](https://github.com/FluidInference/FluidAudio/pull/1054)
+adds `AsrModelVersion.bark3` / `.bark3Clean`, which download the right folder):
 
 ```swift
 import FluidAudio
 
-let dir = URL(fileURLWithPath: "bark/bark-3.0")
-let models = try AsrModels.loadLocal(from: dir, version: .ultra)   // same contract as Parakeet TDT v3
+let models = try await AsrModels.downloadAndLoad(version: .bark3)   // or .bark3Clean
 let asr = AsrManager(config: .default)
 try await asr.loadModels(models)
 var state = TdtDecoderState.make(decoderLayers: await asr.decoderLayerCount)
@@ -109,6 +107,20 @@ let result = try await asr.transcribe(URL(fileURLWithPath: "meeting.wav"), decod
 print(result.text)
 ```
 
+```bash
+swift run -c release fluidaudiocli transcribe meeting.wav --model-version bark3   # or bark3-clean
+```
+
+On a FluidAudio release without #1054, download one folder and load it with the v3 layout:
+
+```bash
+hf download FluidInference/bark-3.0-coreml --include "bark-3.0/*" --local-dir bark
+```
+
+```swift
+let models = try AsrModels.loadLocal(from: URL(fileURLWithPath: "bark/bark-3.0"), version: .ultra)
+```
+
 The first load compiles the encoder for the Neural Engine (up to a minute); later loads are cached by Core ML.
 Requires iOS 18 / macOS 15 or later. Not yet tested on iPhone.
 
@@ -125,4 +137,4 @@ Requires iOS 18 / macOS 15 or later. Not yet tested on iPhone.
 
 | Date | Files | Change | Issue | mobius PR | FluidAudio PR | HF commit |
 |---|---|---|---|---|---|---|
-| 2026-10-21 | `bark-3.0/*`, `bark-3.0-clean/*` | Initial release | — | — | — | (this commit) |
+| 2026-10-21 | `bark-3.0/*`, `bark-3.0-clean/*` | Initial release | — | — | [#1054](https://github.com/FluidInference/FluidAudio/pull/1054) | [`ca21bed2`](https://huggingface.co/FluidInference/bark-3.0-coreml/commit/ca21bed2ccd9c3fbcd104a6014b9a5638a764bc8) |