Cache model in Dockerfile

Dockerfile

@@ -1,5 +1,11 @@
 FROM python:3.10
 
+# Pre-download Whisper model using ADD
+ARG WHISPER_MODEL_URL=https://openaipublic.azureedge.net/main/whisper/models/345ae4da62f9b3d59415adc60127b97c714f32e89e936602e85993674d08dcb1/medium.pt
+ARG WHISPER_MODEL=medium
+RUN mkdir -p /root/.cache/whisper
+ADD ${WHISPER_MODEL_URL} /root/.cache/whisper/${WHISPER_MODEL}.pt
+
 # Remove apt auto-clean hook to preserve cache
 RUN rm -f /etc/apt/apt.conf.d/docker-clean
 
@@ -17,13 +23,11 @@ WORKDIR /app
 COPY pyproject.toml uv.lock ./
 
 # Install dependencies with cache mount
+ENV UV_LINK_MODE=copy
 RUN --mount=type=cache,target=/root/.cache/uv \
     uv sync --locked
 
 # Copy application code
 COPY . .
 
-# Pre-download Whisper model to avoid runtime download
-RUN uv run python -c "import whisper; whisper.load_model('medium')"
-
 CMD ["uv", "run", "python", "app.py"]

README.md

@@ -13,6 +13,22 @@ git clone https://github.com/NINAnor/whisperUI
 docker build -t whisperui -f Dockerfile .
 ```
 
+To build with a different Whisper model:
+
+```bash
+# Get the model URL using the tool
+WHISPER_MODEL_URL=$(uv run python utils/get_model_url.py medium)
+
+# Build with custom model
+docker build \
+  --build-arg WHISPER_MODEL_URL=$WHISPER_MODEL_URL \
+  --build-arg WHISPER_MODEL=medium \
+  -t whisperui .
+```
+
+Available models: `tiny`, `base`, `small`, `medium`, `large-v1`, `large-v2`, `large-v3`, `turbo`
+(Also available with `.en` suffix for English-only versions)
+
 Run the application on `localhost`:
 
 ```

utils/get_model_url.py

@@ -0,0 +1,21 @@
+#!/usr/bin/env python3
+"""
+Tool to print Whisper model URLs for Docker builds.
+Usage: python get_model_url.py [model_name]
+"""
+import sys
+import whisper
+
+def main():
+    model_name = sys.argv[1] if len(sys.argv) > 1 else "medium"
+
+    try:
+        url = whisper._MODELS[model_name]
+        print(url)
+    except KeyError:
+        print(f"Error: Model '{model_name}' not found.", file=sys.stderr)
+        print("Available models:", ", ".join(whisper.available_models()), file=sys.stderr)
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()