Set COMPUTE_PREWARM_MODELS to false by default in runtime, docker-compose, and .env.example to avoid pre-downloading ONNX models during worker startup. Update Dockerfile and deployment documentation to reflect this change and clarify configuration for model prewarming.
58 lines
2.1 KiB
Text
58 lines
2.1 KiB
Text
# External compute-worker service env only.
|
|
# Keep COMPUTE_WORKER_TOKEN and S3_* aligned with app env.
|
|
# In external worker mode, worker runtime vars live here (or worker platform env),
|
|
# not in the app/root `.env`.
|
|
# App/root `.env` should only set:
|
|
# - COMPUTE_WORKER_URL
|
|
# - COMPUTE_WORKER_TOKEN
|
|
# - optional shared timeout/stale overrides
|
|
# Details: docs/deploy/compute-worker
|
|
|
|
# Compute worker bind
|
|
# Platform note:
|
|
# - Local/manual: keep PORT=8081
|
|
# - Railway/Render/Fly/etc: platform injects PORT
|
|
# COMPUTE_WORKER_HOST=0.0.0.0
|
|
# PORT=8081
|
|
# LOG_FORMAT=pretty
|
|
# COMPUTE_LOG_LEVEL=info
|
|
|
|
# Must match app env when app uses COMPUTE_WORKER_URL
|
|
COMPUTE_WORKER_TOKEN=local-compute-token
|
|
|
|
# NATS/JetStream
|
|
NATS_URL=nats://nats:4222
|
|
# Optional: NATS authentication credentials (e.g. for Synadia Cloud / NGS)
|
|
# You can specify the file path:
|
|
# NATS_CREDS_FILE=/path/to/NGS-Default-compute-worker.creds
|
|
# Or specify the raw credentials string content (excellent for container/cloud platforms):
|
|
# NATS_CREDS="-----BEGIN NATS USER JWT-----\n...\n-----BEGIN USER NKEY SEED-----\n..."
|
|
|
|
# Shared object storage (must be reachable from worker)
|
|
S3_BUCKET=openreader-documents
|
|
S3_REGION=us-east-1
|
|
S3_ACCESS_KEY_ID=devkey
|
|
S3_SECRET_ACCESS_KEY=devsecret
|
|
# S3_PREFIX=openreader
|
|
# Optional for non-AWS S3-compatible endpoints:
|
|
S3_ENDPOINT=http://host.docker.internal:8333
|
|
S3_FORCE_PATH_STYLE=true
|
|
|
|
# Optional tuning
|
|
# COMPUTE_PREWARM_MODELS=false
|
|
# COMPUTE_JOB_CONCURRENCY=1
|
|
# COMPUTE_WHISPER_TIMEOUT_MS=30000
|
|
# COMPUTE_PDF_TIMEOUT_MS=300000
|
|
# COMPUTE_PDF_JOB_ATTEMPTS=1
|
|
# COMPUTE_JOBS_STREAM_MAX_BYTES=268435456
|
|
# COMPUTE_EVENTS_STREAM_MAX_BYTES=134217728
|
|
# COMPUTE_JOB_STATES_MAX_BYTES=67108864
|
|
# COMPUTE_NATS_REPLICAS=1
|
|
# COMPUTE_OP_STALE_MS=1800000
|
|
# Optional model mirrors
|
|
# Default expected Whisper variant is q4:
|
|
# - onnx/encoder_model_q4.onnx
|
|
# - onnx/decoder_model_merged_q4.onnx
|
|
# - onnx/decoder_with_past_model_q4.onnx
|
|
# WHISPER_MODEL_BASE_URL=https://huggingface.co/onnx-community/whisper-base_timestamped/resolve/main
|
|
# PDF_LAYOUT_MODEL_BASE_URL=https://huggingface.co/Bei0001/PP-DocLayoutV3-ONNX/resolve/main
|