Add ReazonSpeech Japanese pre-trained model (#1203)

k2-fsa · Aug 2, 2024 · 35c1b4a · 35c1b4a
1 parent 53484fc
commit 35c1b4a
Show file tree

Hide file tree

Showing 5 changed files with 92 additions and 14 deletions.
diff --git a/.github/scripts/test-offline-transducer.sh b/.github/scripts/test-offline-transducer.sh
@@ -15,6 +15,50 @@ echo "PATH: $PATH"
 
 which $EXE
 
+log "------------------------------------------------------------------------"
+log "Run zipformer transducer models (Japanese from ReazonSpeech)                              "
+log "------------------------------------------------------------------------"
+url=https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-zipformer-ja-reazonspeech-2024-08-01.tar.bz2
+
+name=$(basename $url)
+curl -SL -O $url
+tar xvf $name
+rm $name
+repo=$(basename -s .tar.bz2 $name)
+ls -lh $repo
+
+cat $repo/test_wavs/*.txt
+
+log "test $repo"
+test_wavs=(
+1.wav
+2.wav
+3.wav
+4.wav
+5.wav
+)
+
+for w in ${test_wavs[@]}; do
+  time $EXE \
+    --tokens=$repo/tokens.txt \
+    --encoder=$repo/encoder-epoch-99-avg-1.onnx \
+    --decoder=$repo/decoder-epoch-99-avg-1.onnx \
+    --joiner=$repo/joiner-epoch-99-avg-1.onnx \
+    --debug=1 \
+    $repo/test_wavs/$w
+done
+
+for w in ${test_wavs[@]}; do
+  time $EXE \
+    --tokens=$repo/tokens.txt \
+    --encoder=$repo/encoder-epoch-99-avg-1.int8.onnx \
+    --decoder=$repo/decoder-epoch-99-avg-1.onnx \
+    --joiner=$repo/joiner-epoch-99-avg-1.int8.onnx \
+    --debug=1 \
+    $repo/test_wavs/$w
+done
+rm -rf $repo
+
 log "------------------------------------------------------------------------"
 log "Run Nemo fast conformer hybrid transducer ctc models (transducer branch)"
 log "------------------------------------------------------------------------"

diff --git a/.github/workflows/linux.yaml b/.github/workflows/linux.yaml
@@ -141,6 +141,17 @@ jobs:
           name: release-${{ matrix.build_type }}-with-shared-lib-${{ matrix.shared_lib }}-with-tts-${{ matrix.with_tts }}
           path: install/*
 
+      - name: Test offline transducer
+        shell: bash
+        run: |
+          du -h -d1 .
+          export PATH=$PWD/build/bin:$PATH
+          export EXE=sherpa-onnx-offline
+
+          .github/scripts/test-offline-transducer.sh
+          du -h -d1 .
+
+
       - name: Test offline CTC
         shell: bash
         run: |
@@ -191,16 +202,6 @@ jobs:
           .github/scripts/test-online-ctc.sh
           du -h -d1 .
 
-      - name: Test offline transducer
-        shell: bash
-        run: |
-          du -h -d1 .
-          export PATH=$PWD/build/bin:$PATH
-          export EXE=sherpa-onnx-offline
-
-          .github/scripts/test-offline-transducer.sh
-          du -h -d1 .
-
       - name: Test C API
         shell: bash
         run: |

diff --git a/.github/workflows/macos.yaml b/.github/workflows/macos.yaml
@@ -113,21 +113,22 @@ jobs:
           otool -L build/bin/sherpa-onnx
           otool -l build/bin/sherpa-onnx
 
-      - name: Test offline CTC
+      - name: Test offline transducer
         shell: bash
         run: |
           export PATH=$PWD/build/bin:$PATH
           export EXE=sherpa-onnx-offline
 
-          .github/scripts/test-offline-ctc.sh
+          .github/scripts/test-offline-transducer.sh
 
-      - name: Test offline transducer
+
+      - name: Test offline CTC
         shell: bash
         run: |
           export PATH=$PWD/build/bin:$PATH
           export EXE=sherpa-onnx-offline
 
-          .github/scripts/test-offline-transducer.sh
+          .github/scripts/test-offline-ctc.sh
 
       - name: Test online CTC
         shell: bash

diff --git a/scripts/apk/generate-vad-asr-apk-script.py b/scripts/apk/generate-vad-asr-apk-script.py
@@ -275,6 +275,25 @@ def get_models():
 
             ls -lh
 
+            popd
+            """,
+        ),
+        Model(
+            model_name="sherpa-onnx-zipformer-ja-reazonspeech-2024-08-01",
+            idx=16,
+            lang="ja",
+            short_name="zipformer_reazonspeech",
+            cmd="""
+            pushd $model_name
+
+            rm -rfv test_wavs
+
+            rm -fv encoder-epoch-99-avg-1.onnx
+            rm -fv decoder-epoch-99-avg-1.int8.onnx
+            rm -fv joiner-epoch-99-avg-1.onnx
+
+            ls -lh
+
             popd
             """,
         ),

diff --git a/sherpa-onnx/kotlin-api/OfflineRecognizer.kt b/sherpa-onnx/kotlin-api/OfflineRecognizer.kt
@@ -338,6 +338,19 @@ fun getOfflineModelConfig(type: Int): OfflineModelConfig? {
                 tokens = "$modelDir/tokens.txt",
             )
         }
+
+        16 -> {
+            val modelDir = "sherpa-onnx-zipformer-ja-reazonspeech-2024-08-01"
+            return OfflineModelConfig(
+                transducer = OfflineTransducerModelConfig(
+                    encoder = "$modelDir/encoder-epoch-99-avg-1.int8.onnx",
+                    decoder = "$modelDir/decoder-epoch-99-avg-1.onnx",
+                    joiner = "$modelDir/joiner-epoch-99-avg-1.int8.onnx",
+                ),
+                tokens = "$modelDir/tokens.txt",
+                modelType = "transducer",
+            )
+        }
     }
     return null
 }