Duplicate from mefengl/SenseVoiceSmall-coreml

Co-authored-by: mefengl <mefengl@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,72 @@
+---
+library_name: coreml
+pipeline_tag: automatic-speech-recognition
+tags:
+- coreml
+- speech-recognition
+- sensevoice
+- asr
+license: other
+license_name: funasr-model-license
+license_link: https://github.com/modelscope/FunASR/blob/main/MODEL_LICENSE
+language:
+- zh
+- en
+- ja
+- ko
+- yue
+---
+
+# SenseVoiceSmall-coreml
+
+[Hugging Face](https://huggingface.co/mefengl/SenseVoiceSmall-coreml) | [GitHub](https://github.com/mefengl/SenseVoiceSmall-coreml)
+
+CoreML artifact for **SenseVoiceSmall**.
+
+- Upstream model: https://huggingface.co/FunAudioLLM/SenseVoiceSmall
+- Upstream code: https://github.com/FunAudioLLM/SenseVoice
+
+## Download
+
+```bash
+uvx hf download mefengl/SenseVoiceSmall-coreml --local-dir . \
+  --include "coreml/SenseVoiceSmall.mlmodelc.zip" \
+  --include "config.json" \
+  --include "checksums.sha256"
+```
+
+## Verify
+
+```bash
+make verify
+```
+
+## Optional: generate CoreML `.mlpackage`
+
+A CoreML conversion of SenseVoiceSmall, using the upstream model definition.
+
+```bash
+# clones upstream code into ./.upstream/SenseVoice automatically
+make convert DEPLOYMENT_TARGET=macOS15
+# output: ./.coreml-build/SenseVoiceSmall.mlpackage
+```
+
+## Optional: rebuild `coreml/SenseVoiceSmall.mlmodelc.zip` from a `.mlpackage`
+
+macOS + Xcode required.
+
+```bash
+make build
+make inspect
+make verify
+```
+
+## Optional: compare (numeric)
+
+Compare the CoreML output against the original PyTorch model.
+
+```bash
+make compare \
+  MLPACKAGE=.coreml-build/SenseVoiceSmall.mlpackage
+# SENSEVOICE_REPO defaults to ./.upstream/SenseVoice (auto-cloned by `make upstream`)
+```

checksums.sha256

@@ -0,0 +1 @@
+880711fa03577363e6c1b1b6e9321f130ea1a53d5c065d92e1abd8a431bad6be  coreml/SenseVoiceSmall.mlmodelc.zip

config.json

@@ -0,0 +1,82 @@
+{
+  "name": "SenseVoiceSmall-coreml",
+  "upstream": {
+    "model": "FunAudioLLM/SenseVoiceSmall",
+    "code": "https://github.com/FunAudioLLM/SenseVoice",
+    "model_revision": "3eb3b4eeffc2f2dde6051b853983753db33e35c3",
+    "code_commit": "b97857e5c807a466d0a062cf84dba9fe38a1d66a",
+    "assets": [
+      {
+        "name": "cmvn_am.mvn",
+        "url": "https://modelscope.cn/api/v1/models/iic/SenseVoiceSmall/repo?Revision=master&FilePath=am.mvn",
+        "sha256": "29b3c740a2c0cfc6b308126d31d7f265fa2be74f3bb095cd2f143ea970896ae5",
+        "bytes": 11203
+      },
+      {
+        "name": "spm",
+        "url": "https://modelscope.cn/api/v1/models/iic/SenseVoiceSmall/repo?Revision=master&FilePath=chn_jpn_yue_eng_ko_spectok.bpe.model",
+        "sha256": "aa87f86064c3730d799ddf7af3c04659151102cba548bce325cf06ba4da4e6a8",
+        "bytes": 377341
+      }
+    ]
+  },
+  "artifacts": {
+    "coreml_zip": "coreml/SenseVoiceSmall.mlmodelc.zip"
+  },
+  "coreml": {
+    "format": "mlmodelc",
+    "inputs": [
+      {
+        "name": "speech",
+        "dtype": "float32",
+        "shape": [
+          1,
+          "T",
+          560
+        ]
+      },
+      {
+        "name": "speech_lengths",
+        "dtype": "int32",
+        "shape": [
+          1
+        ]
+      },
+      {
+        "name": "language",
+        "dtype": "int32",
+        "shape": [
+          1
+        ]
+      },
+      {
+        "name": "textnorm",
+        "dtype": "int32",
+        "shape": [
+          1
+        ]
+      }
+    ],
+    "outputs": [
+      {
+        "name": "ctc_logits"
+      },
+      {
+        "name": "encoder_out_lens"
+      }
+    ]
+  },
+  "decoding": {
+    "ctc_blank_id": 0,
+    "token_offset": 0
+  },
+  "notes": [
+    "Do not run a single forward pass on extremely long audio; use VAD/chunking."
+  ],
+  "build": {
+    "artifact": "coreml/SenseVoiceSmall.mlmodelc.zip",
+    "sha256": "880711fa03577363e6c1b1b6e9321f130ea1a53d5c065d92e1abd8a431bad6be",
+    "bytes": 432164139,
+    "xcodebuild_version": "Xcode 26.0\nBuild version 17A324"
+  }
+}

coreml/SenseVoiceSmall.mlmodelc.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:880711fa03577363e6c1b1b6e9321f130ea1a53d5c065d92e1abd8a431bad6be
+size 432164139