| --- |
| license: cc-by-4.0 |
| base_model: moondream/parakeet-redux |
| language: [en, de, fr, es, it, pt, ru, uk, hr, sl, lv, lt, et, fi, sv, da, nl, pl, cs, sk, hu, ro, bg, el, mt] |
| pipeline_tag: automatic-speech-recognition |
| tags: [coreml, parakeet, tdt, ternary, speech-recognition, apple-silicon] |
| library_name: fluidaudio |
| --- |
| |
| # parakeet-redux-coreml |
|
|
| Core ML build of [moondream/parakeet-redux](https://huggingface.co/moondream/parakeet-redux), the ternary |
| (every encoder weight ∈ {-1, 0, +1}) re-training of `nvidia/parakeet-tdt-0.6b-v3`. Same 25 languages, tokenizer and |
| output conventions as v3. Encoder is **183 MB** (vs 445 MB for the 6-bit v3 encoder); whole model dir ~220 MB. |
|
|
| The encoder keeps the checkpoint's exact ternary weights: 2-bit palettized codes with the model's own per-row, |
| per-128 fp16 scales (iOS 18 / macOS 15 `constexpr_lut_to_dense` + `constexpr_blockwise_shift_scale`), so no |
| re-quantization noise is added. Decoder and JointDecision are re-exported from the redux checkpoint (its post-training |
| recovery also touched the joint). Preprocessor and vocabulary are the v3 ones. |
|
|
| ## Usage (FluidAudio ≥ next release) |
|
|
| ```swift |
| let models = try await AsrModels.downloadAndLoad(version: .redux) // iOS 18+ / macOS 15+ |
| let asr = AsrManager() |
| try await asr.initialize(models: models) |
| let result = try await asr.transcribe(audioFileURL) |
| ``` |
|
|
| ```bash |
| swift run fluidaudiocli transcribe audio.wav --model-version redux |
| swift run fluidaudiocli asr-benchmark --subset test-clean --max-files 100 --model-version redux |
| ``` |
|
|
| ## Files |
|
|
| | File | Notes | |
| |------|-------| |
| | `Encoder.mlmodelc` | 183 MB, 2-bit ternary, iOS 18+/macOS 15+, fixed 15 s window (`mel [1,128,1501]` → `encoder [1,1024,188]`) | |
| | `Decoder.mlmodelc` | RNNT prediction net, fp16, iOS 17+ | |
| | `JointDecisionv3.mlmodelc` | single-step joint + argmax/softmax + top-64, fp16, iOS 17+ | |
| | `Preprocessor.mlmodelc` | mel front-end, identical to parakeet-tdt-0.6b-v3-coreml | |
| | `parakeet_vocab.json` | identical to v3 | |
|
|
| ## Compute units |
|
|
| The encoder runs on the **Neural Engine** by default in FluidAudio, like v3, so iOS apps can keep transcribing in the |
| background (iOS does not allow GPU work there). The first ANE load compiles the 2-bit weights for several minutes |
| (~7 min measured on an M-series Mac); Core ML caches the result and later loads take seconds. Warm, it runs |
| 45–52 ms per 15 s window. |
|
|
| Pass `encoderComputeUnits: .cpuAndGPU` to skip the long first compile when background execution does not matter: the |
| GPU decompresses the 2-bit weights in-kernel, loads in about a second and runs ≈21 ms per window. |
|
|
| Requires iOS 18 / macOS 15 (the 2-bit encoding uses iOS 18 Core ML ops). On iOS 17 / macOS 14 FluidAudio refuses to |
| load Redux and points to |
| [parakeet-ultra-coreml](https://huggingface.co/FluidInference/parakeet-ultra-coreml) instead. |
|
|
| ## Accuracy |
|
|
| Full LibriSpeech, FluidAudio `asr-benchmark`, default compute units (ANE), both models back to back, M-series Mac. Corpus WER (total edit distance over |
| total reference words); RTFx = total audio / total processing time. |
|
|
| | Set | parakeet-tdt-0.6b-v3-coreml | parakeet-redux-coreml | |
| |-----|---------------------------:|----------------------:| |
| | test-clean (2620 files) | **2.27 %** | 2.71 % | |
| | test-other (2939 files) | **4.12 %** | 5.12 % | |
| | RTFx, test-clean | **128.6×** | 83.9× | |
|
|
| On English this build is 0.44 (clean) and 1.00 (other) points behind v3 — the same gap the upstream card reports |
| (+0.44 / +1.21). Absolute values are above the card's because FluidAudio decodes in 15 s windows and uses a simpler |
| text normalizer than the Open ASR Leaderboard; both models are scored identically here. |
|
|
| The Core ML transcripts match a PyTorch fp32 decode of the redux checkpoint to 0.19 % WER, so this is the |
| checkpoint's behaviour and not a conversion loss. |
|
|
| ### FLEURS, 24 languages, 100 samples each |
|
|
| | | parakeet-tdt-0.6b-v3-coreml | parakeet-redux-coreml | |
| |---|---:|---:| |
| | mean WER | 14.81 % | **13.06 %** | |
| | duration-weighted WER | 14.65 % | **12.89 %** | |
| | languages won | 11 | **13** | |
| | RTFx | **149×** | 134× | |
|
|
| (`es_es` is absent from FluidAudio's FLEURS set; both models are scored on the same 24.) |
|
|
| Redux wins the multilingual average by 1.74 points, taking the low-resource languages by wide margins — Latvian |
| −11.1, Maltese −7.7, Slovene −7.2, Estonian −7.0, Greek −5.2, Lithuanian −5.1 — and conceding the high-resource |
| ones: French +3.7, Russian +2.6, Dutch +1.9, English +1.9, Polish +1.7, Ukrainian +1.4. The per-language direction |
| matches the upstream card in 23 of 24 languages. |
|
|
| **Pick this model for multilingual work and for download size; pick |
| [parakeet-tdt-0.6b-v3-coreml](https://huggingface.co/FluidInference/parakeet-tdt-0.6b-v3-coreml) for English.** |
|
|
| ## License |
|
|
| CC-BY-4.0, same as the upstream checkpoint and nvidia/parakeet-tdt-0.6b-v3. Model by moondream (vik); Core ML |
| conversion by Fluid Inference (conversion recipe: `mobius/models/stt/parakeet-redux/coreml`). |
|
|
