fix: resolve sync timeout with longer timeout and hf_transfer

Default HF Hub read timeout (10s) causes failures on large state uploads.
Set HF_HUB_DOWNLOAD_TIMEOUT=300s and enable hf_transfer for faster
multipart transfers. Install hf_transfer package in Dockerfile.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Dockerfile

@@ -28,7 +28,7 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
     fonts-liberation \
     fonts-noto-color-emoji \
     && rm -rf /var/lib/apt/lists/* \
-    && uv pip install --python /opt/hermes/.venv/bin/python --no-cache-dir huggingface_hub
+    && uv pip install --python /opt/hermes/.venv/bin/python --no-cache-dir huggingface_hub hf_transfer
 
 COPY --chown=hermes:hermes start.sh /opt/huggingmes/start.sh
 COPY --chown=hermes:hermes health-server.js /opt/huggingmes/health-server.js

hermes-sync.py

@@ -17,6 +17,8 @@ from pathlib import Path
 
 os.environ.setdefault("HF_HUB_DISABLE_PROGRESS_BARS", "1")
 os.environ.setdefault("HF_HUB_VERBOSITY", "error")
+os.environ.setdefault("HF_HUB_DOWNLOAD_TIMEOUT", "300")   # 5 min — default 10s causes timeout on large state
+os.environ.setdefault("HF_HUB_ENABLE_HF_TRANSFER", "1")   # faster multipart upload/download
 
 from huggingface_hub import HfApi, snapshot_download, upload_folder
 from huggingface_hub.errors import HfHubHTTPError, RepositoryNotFoundError