Adding translation example using VAD, OPUS, ArgoTranslate, and a pret…

…ty printer for better readibility
dora-rs · Oct 9, 2024 · 15f123a · 15f123a
1 parent d572bd6
commit 15f123a
Show file tree

Hide file tree

Showing 31 changed files with 696 additions and 145 deletions.
diff --git a/examples/argostranslate/README.md b/examples/argostranslate/README.md
diff --git a/examples/argostranslate/argo_translate_op.py b/examples/argostranslate/argo_translate_op.py
diff --git a/examples/argostranslate/dataflow.yml b/examples/argostranslate/dataflow.yml
diff --git a/examples/argostranslate/write_op.py b/examples/argostranslate/write_op.py
diff --git a/examples/argostranslate/.gitignore → examples/translation/.gitignore b/examples/argostranslate/.gitignore → examples/translation/.gitignore
diff --git a/examples/translation/README.md b/examples/translation/README.md
@@ -0,0 +1,14 @@
+# Dora argo example
+
+Make sure to have, dora, pip and cargo installed.
+
+```bash
+# Install rerun if it's not done already
+dora up
+
+dora build dataflow_cn_terminal.yml
+dora start dataflow_cn_terminal.yml
+
+# In another terminal
+python pretty_print.py
+```
diff --git a/examples/translation/dataflow.yml b/examples/translation/dataflow.yml
@@ -0,0 +1,47 @@
+nodes:
+  - id: dora-microphone
+    build: pip install -e ../../node-hub/dora-microphone
+    path: dora-microphone
+    outputs:
+      - audio
+
+  - id: dora-vad
+    build: pip install -e ../../node-hub/dora-vad
+    path: dora-vad
+    inputs:
+      audio: dora-microphone/audio
+    outputs:
+      - audio
+
+  - id: dora-distil-whisper
+    build: pip install -e ../../node-hub/dora-distil-whisper
+    path: dora-distil-whisper
+    inputs:
+      input: dora-vad/audio
+    outputs:
+      - text
+    env:
+      TARGET_LANGUAGE: chinese
+      TRANSLATE: false
+
+  - id: dora-argotranslate
+    build: pip install -e ../../node-hub/dora-argotranslate
+    path: dora-argotranslate
+    inputs:
+      text:
+        source: dora-distil-whisper/text
+        queue_size: 1
+    outputs:
+      - text
+    env:
+      SOURCE_LANGUAGE: zh
+      TARGET_LANGUAGE: en
+
+  - id: plot
+    build: cargo build -p dora-rerun --release
+    path: dora-rerun
+    inputs:
+      text: dora-argotranslate/text
+    env:
+      IMAGE_WIDTH: 640
+      IMAGE_HEIGHT: 480
diff --git a/examples/translation/dataflow_cn.yml b/examples/translation/dataflow_cn.yml
@@ -0,0 +1,46 @@
+nodes:
+  - id: dora-microphone
+    build: pip install -e ../../node-hub/dora-microphone
+    path: dora-microphone
+    outputs:
+      - audio
+
+  - id: dora-vad
+    build: pip install -e ../../node-hub/dora-vad
+    path: dora-vad
+    inputs:
+      audio: dora-microphone/audio
+    outputs:
+      - audio
+
+  - id: dora-distil-whisper
+    build: pip install -e ../../node-hub/dora-distil-whisper
+    path: dora-distil-whisper
+    inputs:
+      input: dora-vad/audio
+    outputs:
+      - text
+    env:
+      TARGET_LANGUAGE: chinese
+      TRANSLATE: false
+
+  - id: dora-opus
+    build: pip install -e ../../node-hub/dora-opus
+    path: dora-opus
+    inputs:
+      text: dora-distil-whisper/text
+    outputs:
+      - text
+    env:
+      SOURCE_LANGUAGE: zh
+      TARGET_LANGUAGE: en
+
+  - id: plot
+    build: cargo build -p dora-rerun --release
+    path: dora-rerun
+    inputs:
+      translated_text: dora-opus/text
+      original_text: dora-distil-whisper/text
+    env:
+      IMAGE_WIDTH: 640
+      IMAGE_HEIGHT: 480
diff --git a/examples/translation/dataflow_cn_terminal.yml b/examples/translation/dataflow_cn_terminal.yml
@@ -0,0 +1,42 @@
+nodes:
+  - id: dora-microphone
+    build: pip install -e ../../node-hub/dora-microphone
+    path: dora-microphone
+    outputs:
+      - audio
+
+  - id: dora-vad
+    build: pip install -e ../../node-hub/dora-vad
+    path: dora-vad
+    inputs:
+      audio: dora-microphone/audio
+    outputs:
+      - audio
+
+  - id: dora-distil-whisper
+    build: pip install -e ../../node-hub/dora-distil-whisper
+    path: dora-distil-whisper
+    inputs:
+      input: dora-vad/audio
+    outputs:
+      - text
+    env:
+      TARGET_LANGUAGE: chinese
+      TRANSLATE: false
+
+  - id: dora-opus
+    build: pip install -e ../../node-hub/dora-opus
+    path: dora-opus
+    inputs:
+      text: dora-distil-whisper/text
+    outputs:
+      - text
+    env:
+      SOURCE_LANGUAGE: zh
+      TARGET_LANGUAGE: en
+
+  - id: pretty-print
+    path: dynamic
+    inputs:
+      translated_text: dora-opus/text
+      original_text: dora-distil-whisper/text
diff --git a/examples/translation/dataflow_en.yml b/examples/translation/dataflow_en.yml
@@ -0,0 +1,45 @@
+nodes:
+  - id: dora-microphone
+    build: pip install -e ../../node-hub/dora-microphone
+    path: dora-microphone
+    outputs:
+      - audio
+
+  - id: dora-vad
+    build: pip install -e ../../node-hub/dora-vad
+    path: dora-vad
+    inputs:
+      audio: dora-microphone/audio
+    outputs:
+      - audio
+
+  - id: dora-distil-whisper
+    build: pip install -e ../../node-hub/dora-distil-whisper
+    path: dora-distil-whisper
+    inputs:
+      input: dora-vad/audio
+    outputs:
+      - text
+    env:
+      TARGET_LANGUAGE: english
+      TRANSLATE: false
+
+  - id: dora-argotranslate
+    build: pip install -e ../../node-hub/dora-argotranslate
+    path: dora-argotranslate
+    inputs:
+      text: dora-distil-whisper/text
+    outputs:
+      - text
+    env:
+      SOURCE_LANGUAGE: en
+      TARGET_LANGUAGE: zh
+
+  - id: dora-rerun
+    build: cargo build -p dora-rerun --release
+    path: dora-rerun
+    inputs:
+      text: dora-argotranslate/text
+    env:
+      IMAGE_WIDTH: 640
+      IMAGE_HEIGHT: 480
diff --git a/examples/translation/dataflow_en_terminal.yml b/examples/translation/dataflow_en_terminal.yml
@@ -0,0 +1,45 @@
+nodes:
+  - id: dora-microphone
+    build: pip install -e ../../node-hub/dora-microphone
+    path: dora-microphone
+    outputs:
+      - audio
+
+  - id: dora-vad
+    build: pip install -e ../../node-hub/dora-vad
+    path: dora-vad
+    inputs:
+      audio: dora-microphone/audio
+    outputs:
+      - audio
+
+  - id: dora-distil-whisper
+    build: pip install -e ../../node-hub/dora-distil-whisper
+    path: dora-distil-whisper
+    inputs:
+      input: dora-vad/audio
+    outputs:
+      - text
+    env:
+      TARGET_LANGUAGE: english
+      TRANSLATE: false
+
+  - id: dora-opus
+    build: pip install -e ../../node-hub/dora-opus
+    path: dora-opus
+    inputs:
+      text: dora-distil-whisper/text
+    outputs:
+      - text
+    env:
+      SOURCE_LANGUAGE: en
+      TARGET_LANGUAGE: zh
+
+  - id: pretty-print
+    path: dynamic
+    inputs:
+      translated_text: dora-opus/text
+      original_text: dora-distil-whisper/text
+    env:
+      IMAGE_WIDTH: 640
+      IMAGE_HEIGHT: 480
diff --git a/examples/translation/dataflow_en_terminal_argo.yml b/examples/translation/dataflow_en_terminal_argo.yml
@@ -0,0 +1,46 @@
+nodes:
+  - id: dora-microphone
+    build: pip install -e ../../node-hub/dora-microphone
+    path: dora-microphone
+    outputs:
+      - audio
+
+  - id: dora-vad
+    build: pip install -e ../../node-hub/dora-vad
+    path: dora-vad
+    inputs:
+      audio: dora-microphone/audio
+    outputs:
+      - audio
+
+  - id: dora-distil-whisper
+    build: pip install -e ../../node-hub/dora-distil-whisper
+    path: dora-distil-whisper
+    inputs:
+      input: dora-vad/audio
+    outputs:
+      - text
+    env:
+      TARGET_LANGUAGE: english
+      TRANSLATE: false
+
+  - id: dora-argotranslate
+    build: pip install -e ../../node-hub/dora-argotranslate
+    path: dora-argotranslate
+    inputs:
+      text: dora-distil-whisper/text
+    outputs:
+      - text
+    env:
+      SOURCE_LANGUAGE: en
+      TARGET_LANGUAGE: zh
+
+  - id: pretty-print
+    build: cargo build -p dora-rerun --release
+    path: dynamic
+    inputs:
+      translated_text: dora-argotranslate/text
+      original_text: dora-distil-whisper/text
+    env:
+      IMAGE_WIDTH: 640
+      IMAGE_HEIGHT: 480