vixy/orpheus-tts
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Use GitHub orpheus-tts (supports max_model_len) to fix OOM on Jetson

Browse Source
This commit is contained in:
vixy
2026-01-11 17:39:55 -06:00
parent 86cf77d2d9
commit 0e43b76204
3 changed files with 11 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
Dockerfile
View File

@@ -21,9 +21,9 @@ COPY requirements.txt /app/
# Install Python dependencies (FastAPI, etc - but NOT torch/vllm)
RUN pip3 install --no-cache-dir -r requirements.txt
# Install orpheus-speech from regular PyPI (not Jetson index) WITHOUT dependencies
# Install orpheus-speech from GitHub repo (supports max_model_len) WITHOUT dependencies
# to avoid overwriting vllm/torch. Then install snac audio codec.
RUN pip3 install --no-cache-dir --no-deps --index-url https://pypi.org/simple/ orpheus-speech && \
RUN pip3 install --no-cache-dir --no-deps git+https://github.com/canopyai/Orpheus-TTS.git#subdirectory=orpheus_tts_pypi && \
pip3 install --no-cache-dir --index-url https://pypi.org/simple/ snac
# Copy application code