Update model card for all StemDeck Ultra assets including native F16 Turbo ASR

README.md

@@ -1,38 +1,107 @@
 ---
 language:
-  - en
-  - zh
+  - multilingual
 tags:
   - audio
   - music-source-separation
+  - automatic-speech-recognition
+  - audio-super-resolution
   - gguf
+  - whisper
+  - native-cpp
+  - stemdeck
   - openstems
 pipeline_tag: audio-to-audio
+inference: false
 ---
 
-# OpenStems Models
+# StemDeck Ultra model resources
 
-Model resources for the OpenStems native C++ separation engine. CPU and NVIDIA editions use the same model weights. This release preserves the FP32 GGUF files used by OpenStems for Windows on 2026-09-21.
+Updated **2026-11-05**. This repository contains the complete current model-weight and model-constant set used by **StemDeck Ultra for Windows**: **17 files, 3,705,369,284 bytes**. OpenStems is the project/repository name; the application is StemDeck Ultra.
 
-## Download files
+**ASR Turbo is uploaded here:** [ggml-large-v3-turbo.bin](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69/ggml-large-v3-turbo.bin) — **1,624,555,275 bytes**, original upstream **F16 GGML** artifact. [Download ASR Turbo](https://huggingface.co/OpenStems/OpenStems-Models/resolve/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69/ggml-large-v3-turbo.bin?download=true).
 
-Each ZIP contains one GGUF file. ZIP compression reduces download size without changing the model bytes or precision.
+[Browse all current files](https://huggingface.co/OpenStems/OpenStems-Models/tree/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105) · [Exact manifest](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/models.json) · [Version-folder documentation](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/README.md)
 
-| Model | Download | Extracted file | Extracted bytes |
-| --- | --- | --- | ---: |
-| T400 | [T400-478978733353.zip](T400-478978733353.zip) | T400.gguf | 270145408 |
-| B400 | [B400-aa162a2eca9f.zip](B400-aa162a2eca9f.zip) | B400.gguf | 206407520 |
-| BS Karaoke | [karaoke-4804181ac76c.zip](karaoke-4804181ac76c.zip) | karaoke.gguf | 204258848 |
-| BS SW | [sw-68bed29111b6.zip](sw-68bed29111b6.zip) | sw.gguf | 698768832 |
+## Current application mapping
 
-[model-files.json](model-files.json) records the exact archive and extracted-file sizes and SHA-256 checksums. Separate `.sha256` files are also supplied. Installers should pin a repository commit and verify both the download and extracted model against the manifest.
+| Application choice | Model/checkpoint | Stored format | CPU / NVIDIA GPU |
+|---|---|---|---|
+| Fast | B400, selected step **38200** | GGUF | Same weight file and SHA256 |
+| Pro (**default**) | B62, selected step **27000** | GGUF | Same weight file and SHA256 |
+| Ensemble | Karaoke followed by SW | Two GGUF files | Same two weight files and SHA256 |
+| Lyrics ASR | **Whisper large-v3-turbo only** | Original **F16 GGML `.bin`** | Same ASR weight file and SHA256 |
+| Optional 48kHz enhancement | LavaSR v2 | Original FP32 `.pt` asset | Matching native enhancement integration |
+| Beat analysis | Beat-this final0 | `.ckpt` | Original application analysis asset |
+| Song-section analysis | All-In-One Harmonix, eight folds | Eight `.pth` files | Original application analysis assets |
 
-## Use
+CPU/GPU selection changes the matching native execution backend and configuration. It does not require a second set of separation or ASR weights. GGUF is a container format and does not itself imply integer quantization. **This release adds no INT8 models.**
 
-The OpenStems installer downloads and verifies the required resources automatically. Fast uses T400, Pro uses B400, and Ultra selects the required Karaoke/SW recipe. These are model resources for the OpenStems native runtime; a `.gguf` extension alone does not guarantee compatibility with another runtime.
+Separation and ASR are integrated through native inference libraries. The `.gguf` files require the matching audio-model engine and manifests; the extension alone does not establish compatibility with an unrelated GGUF runtime. Analysis assets retain the original application analysis path.
 
-The model files remain subject to their respective applicable upstream terms. Packaging and format conversion do not replace those terms or establish a new license for the weights.
+The 48kHz option is independent of Fast/Pro/Ensemble and is **off by default**. The application follows its export sample-rate settings.
+
+## Complete downloadable inventory
+
+All links below point to the immutable model-content revision **`f2da66683f7a0fddcd772ccd5fbf0b460378daef`**. File SHA256 values and relative installation destinations are recorded in [`models.json`](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/models.json).
+
+| Purpose | File | Bytes | SHA256 |
+|---|---|---:|---|
+| Fast separation | [B400-step38200.gguf](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/293af89efaf9650be656eeac3567ecf1368effae9b9e8660dda0561c44c68690/B400-step38200.gguf) | 206,407,584 | `293af89efaf9650be656eeac3567ecf1368effae9b9e8660dda0561c44c68690` |
+| Pro separation | [B62-step27000.gguf](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/1ded8095f4f99602766d8f60cf692eebd48da6bb9e8bcafb92b2c3b17468fbc3/B62-step27000.gguf) | 812,481,984 | `1ded8095f4f99602766d8f60cf692eebd48da6bb9e8bcafb92b2c3b17468fbc3` |
+| Ensemble first stage | [karaoke.gguf](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/4804181ac76c3d4fc9b282884cb7b6785dd517ea5e547ee90a4d9bb5eccea338/karaoke.gguf) | 204,258,848 | `4804181ac76c3d4fc9b282884cb7b6785dd517ea5e547ee90a4d9bb5eccea338` |
+| Ensemble second stage | [sw.gguf](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/68bed29111b6414a5b441bd147b341dbf311a5bbaf7eac1c1073a6a82e21eb72/sw.gguf) | 698,768,832 | `68bed29111b6414a5b441bd147b341dbf311a5bbaf7eac1c1073a6a82e21eb72` |
+| ASR / lyrics | [ggml-large-v3-turbo.bin](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69/ggml-large-v3-turbo.bin) | 1,624,555,275 | `1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69` |
+| Optional 48kHz enhancement | [lavasr_v2.pt](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/8ca5a4c9114761419de20a1d8231337c3ec77e5759a206b3cfa57fbe401e0911/lavasr_v2.pt) | 56,389,364 | `8ca5a4c9114761419de20a1d8231337c3ec77e5759a206b3cfa57fbe401e0911` |
+| Beat analysis | [beat_this-final0.ckpt](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/8c328b45f59d8dd3dff219253ff6a8d6482be57d0133a29140e2febbf8eb8331/beat_this-final0.ckpt) | 81,058,141 | `8c328b45f59d8dd3dff219253ff6a8d6482be57d0133a29140e2febbf8eb8331` |
+| B62 analysis constants | [b62-hamming.safetensors](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/20e08e4596bd9e4b07245b749b58410821a43fb037d232d7e07c87ac4e2d0da0/b62-hamming.safetensors) | 1,411,288 | `20e08e4596bd9e4b07245b749b58410821a43fb037d232d7e07c87ac4e2d0da0` |
+| Karaoke / SW constants | [bs_constants.safetensors](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/15380c3bb19d15872e4934232079bd6c07a4bdf828bbc40d8ce3a9bc08fb5c4d/bs_constants.safetensors) | 8,833,400 | `15380c3bb19d15872e4934232079bd6c07a4bdf828bbc40d8ce3a9bc08fb5c4d` |
+| Song-section analysis fold 0 | [harmonix-fold0-0vra4ys2.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/0db596dfb0995f41d62f6267d76a9d54c046f1649bd35e1dbeca0c5f9a7b8acd/harmonix-fold0-0vra4ys2.pth) | 1,400,571 | `0db596dfb0995f41d62f6267d76a9d54c046f1649bd35e1dbeca0c5f9a7b8acd` |
+| Song-section analysis fold 1 | [harmonix-fold1-3ozjhtsj.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/607b4864eaabccd4f6951410d5ea6676666ad1cfa6ce2e18bc99f3824b77c3b7/harmonix-fold1-3ozjhtsj.pth) | 1,400,571 | `607b4864eaabccd4f6951410d5ea6676666ad1cfa6ce2e18bc99f3824b77c3b7` |
+| Song-section analysis fold 2 | [harmonix-fold2-gmgo0nsy.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/67c28ec7ca612d365e5fcb152f6530eb53e34c4303940afb30fd17a295c958a4/harmonix-fold2-gmgo0nsy.pth) | 1,400,571 | `67c28ec7ca612d365e5fcb152f6530eb53e34c4303940afb30fd17a295c958a4` |
+| Song-section analysis fold 3 | [harmonix-fold3-i92b7m8p.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/446d05672b2669cdb7d9b8bbbc228a4f3592484335a7995cca89a744e3950ddc/harmonix-fold3-i92b7m8p.pth) | 1,400,571 | `446d05672b2669cdb7d9b8bbbc228a4f3592484335a7995cca89a744e3950ddc` |
+| Song-section analysis fold 4 | [harmonix-fold4-1bql5qo0.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/96327222ab9b33c005388b788ba602f529aa03cc4dde1c859c011f1004490493/harmonix-fold4-1bql5qo0.pth) | 1,400,571 | `96327222ab9b33c005388b788ba602f529aa03cc4dde1c859c011f1004490493` |
+| Song-section analysis fold 5 | [harmonix-fold5-x4z5zeef.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/c87f162a1090290bcbb863688746784919cf912c56744fb3d81a1a7dba3b114e/harmonix-fold5-x4z5zeef.pth) | 1,400,571 | `c87f162a1090290bcbb863688746784919cf912c56744fb3d81a1a7dba3b114e` |
+| Song-section analysis fold 6 | [harmonix-fold6-x7t226rq.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/13e7e639bf30fcd3fa049ad863088773a4e57b06d7c68581385b3ef625c60688/harmonix-fold6-x7t226rq.pth) | 1,400,571 | `13e7e639bf30fcd3fa049ad863088773a4e57b06d7c68581385b3ef625c60688` |
+| Song-section analysis fold 7 | [harmonix-fold7-qwwskhg6.pth](https://huggingface.co/OpenStems/OpenStems-Models/blob/f2da66683f7a0fddcd772ccd5fbf0b460378daef/ultra-20261105/assets/b888b7fba95efa3d5a37ca013fe7efe2741621ea9ca335e6b2723e4fd0fade57/harmonix-fold7-qwwskhg6.pth) | 1,400,571 | `b888b7fba95efa3d5a37ca013fe7efe2741621ea9ca335e6b2723e4fd0fade57` |
+
+## ASR identity and provenance
+
+The ASR file is **not a GGUF conversion and not an INT8 variant**. It is the unchanged upstream F16 large-v3-turbo GGML asset mirrored for the application's native ASR integration.
+
+- File: `ggml-large-v3-turbo.bin`
+- Bytes: **1,624,555,275**
+- SHA256: **`1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69`**
+- Upstream artifact provenance: [ggerganov/whisper.cpp, fixed source revision](https://huggingface.co/ggerganov/whisper.cpp/blob/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-large-v3-turbo.bin).
+
+The application uses this one Turbo model for CPU/GPU lyrics recognition. No additional ASR model or separate alignment model is included in this release.
+
+## Download and integrity
+
+The application installer downloads the required resources and checks the complete file identities. Integrations should pin **`f2da66683f7a0fddcd772ccd5fbf0b460378daef`**, use the manifest, and verify SHA256 before passing files to the matching native engine.
+
+For example, the official Hugging Face CLI can download the ASR artifact directly:
+
+```sh
+hf download OpenStems/OpenStems-Models ultra-20261105/assets/1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69/ggml-large-v3-turbo.bin --revision f2da66683f7a0fddcd772ccd5fbf0b460378daef
+```
+
+Checkpoint step counts identify the selected saved separation checkpoints. This resource card does not claim a new quality score or a hardware-independent speedup. Hardware-specific performance and numerical validation belong to the application benchmark records.
+
+## Historical resources
+
+Older root-level T400/B400/Karaoke/SW ZIPs and `model-files.json` are retained for reproducibility of previous application releases. **T400 is not the current Fast choice**, and the old Fast/Pro/Ultra mapping does not describe StemDeck Ultra. Current Fast/Pro/Ensemble and all ASR/analysis resources are listed above.
+
+## Licensing and repository scope
+
+Original model and component terms continue to apply. Format conversion, mirroring, and packaging do not relicense the weights. Consult the corresponding upstream model terms and the application's component notices. A single blanket license is not assigned to this mixed resource collection.
+
+This repository contains model weights, model constants, and their documentation. It does not contain user recordings, training datasets, private inference SDK binaries, machine configuration, or authentication material.
 
 ## 中文说明
 
-本仓库存放 OpenStems 原生 C++ 分离引擎使用的四个模型。CPU 与 NVIDIA 版本共用同一份权重，保留当前 FP32 GGUF 精度。ZIP 仅用于无损压缩下载，解压后的模型字节保持不变。安装器应使用固定提交版本，并按清单校验下载包及解压后的模型文件。
+本仓库已包含 StemDeck Ultra 当前使用的**全部 17 项模型权重及常量**，其中包括 **ASR 的原版 F16 large-v3-turbo**。ASR 文件和直接下载入口已在本页顶部列出，不只在版本目录中说明。
+
+当前档位：**Fast = B400 38200 步；Pro = B62 27000 步，默认 Pro；Ensemble = Karaoke → SW**。分离模型的 CPU/GPU 共用同一份 GGUF 权重；ASR 的 CPU/GPU 共用同一份 F16 GGML `.bin`。没有新增 INT8 模型或独立对齐模型。
+
+LavaSR 48kHz 为独立可选开关，默认关闭。节拍和歌曲结构分析所需的 Beat-this、八个 All-In-One fold，以及两份模型常量，也都已上传并列入清单。旧根目录 ZIP 为历史资源，当前版本请使用本页固定版本链接和 `models.json` 校验。