Upload new sense-voice models (#2580)

@@ -156,6 +156,27 @@ jobs:
             ls -lh *.tar.bz2
           done
 
+      - name: sense-voice
+        if: false
+        shell: bash
+        run: |
+          git lfs install
+          d=sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2025-09-09
+          f=sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09
+          git clone https://huggingface.co/csukuangfj/$d
+          git clone https://huggingface.co/csukuangfj/$f
+
+          rm -rf $d/.git
+          rm -rf $f/.gi*
+
+          rm -rf $d/.gitattributes
+          rm -rf $f/.gitattributes
+
+          tar cjfv $d.tar.bz2 $d
+          tar cjfv $f.tar.bz2 $f
+
+          ls -lh *.tar.bz2
+
       - name: Vietnamese (zipformer)
         if: false
         shell: bash

@@ -117,7 +117,7 @@ def get_2nd_models():
             model_name="sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17",
             idx=15,
             lang="zh_en_ko_ja_yue",
-            short_name="sense_voice",
+            short_name="sense_voice_2024_07_17_int8",
             cmd="""
             pushd $model_name
 
@@ -175,6 +175,23 @@ def get_2nd_models():
             popd
             """,
         ),
+        Model(
+            model_name="sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09",
+            idx=41,
+            lang="zh_en_ko_ja_yue",
+            short_name="sense_voice_2025_09_09_int8",
+            cmd="""
+            pushd $model_name
+
+            rm -rfv test_wavs
+            rm -fv model.onnx
+            rm -fv *.py
+
+            ls -lh
+
+            popd
+            """,
+        ),
     ]
     return models
 
@@ -380,6 +397,7 @@ def get_models():
         "sherpa-onnx-paraformer-zh-2023-09-14",
         "icefall-asr-zipformer-wenetspeech-20230615",
         "sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17",
+        "sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09",
         "sherpa-onnx-dolphin-base-ctc-multi-lang-int8-2025-04-02",
     ]
     for first_m in first_zh:
@@ -403,6 +421,10 @@ def get_models():
             "sherpa-onnx-streaming-zipformer-en-20M-2023-02-17",
             "sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17",
         ),
+        (
+            "sherpa-onnx-streaming-zipformer-en-20M-2023-02-17",
+            "sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09",
+        ),
     ]
     models = []
     for f, s in combinations:

@@ -99,7 +99,7 @@ def get_models():
             idx=15,
             lang="zh_en_ko_ja_yue",
             lang2="中英粤日韩",
-            short_name="sense_voice",
+            short_name="sense_voice_2024_07_17_int8",
             use_hr=True,
             cmd="""
             pushd $model_name
@@ -697,6 +697,23 @@ def get_models():
             popd
             """,
         ),
+        Model(
+            model_name="sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09",
+            idx=41,
+            lang="zh_en_ko_ja_yue",
+            lang2="中英粤日韩",
+            short_name="sense_voice_2025_09_09_int8",
+            use_hr=True,
+            cmd="""
+            pushd $model_name
+
+            rm -rfv test_wavs
+
+            ls -lh
+
+            popd
+            """,
+        ),
     ]
     return models
 

@@ -695,6 +695,16 @@ fun getOfflineModelConfig(type: Int): OfflineModelConfig? {
                 modelType = "nemo_transducer",
             )
         }
+
+        41 -> {
+            val modelDir = "sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09"
+            return OfflineModelConfig(
+                senseVoice = OfflineSenseVoiceModelConfig(
+                    model = "$modelDir/model.int8.onnx",
+                ),
+                tokens = "$modelDir/tokens.txt",
+            )
+        }
     }
     return null
 }