wehub-resource-sync
e904b667c6
Build/Publish Develop Docs / deploy (push) Failing after 1s
PaddleOCR Code Style Check / check-code-style (push) Failing after 1s
PaddleOCR PR Tests GPU / detect-changes (push) Failing after 1s
PaddleOCR PR Tests / detect-changes (push) Failing after 1s
PaddleOCR PR Tests GPU / test-pr-gpu (push) Has been cancelled
PaddleOCR PR Tests / test-pr (push) Has been cancelled
PaddleOCR PR Tests GPU / test-pr-gpu-impl (push) Has been cancelled
PaddleOCR PR Tests / test-pr-python (3.13) (push) Has been cancelled
PaddleOCR PR Tests / test-pr-python (3.8) (push) Has been cancelled
PaddleOCR PR Tests / test-pr-python (3.9) (push) Has been cancelled
59 lines
2.1 KiB
Bash
59 lines
2.1 KiB
Bash
# PaddleOCR-VL High-Performance Serving Configuration
|
|
# Copy this file to .env and modify as needed
|
|
|
|
# =============================================================================
|
|
# Pipeline and SDK Configuration
|
|
# =============================================================================
|
|
|
|
# Pipeline name. Applies to the whole PaddleOCR-VL series, e.g.:
|
|
# PaddleOCR-VL, PaddleOCR-VL-1.5, PaddleOCR-VL-1.6
|
|
HPS_PIPELINE_NAME=PaddleOCR-VL-1.6
|
|
|
|
# PaddleX version (major.minor only, e.g. 3.6). Keeps the pipeline base image and
|
|
# the high-stability serving SDK in sync:
|
|
# 3.6 -> base image .../hps:paddlex3.6-gpu + SDK release dir v3.6
|
|
HPS_PADDLEX_VERSION=3.6
|
|
|
|
# Derived SDK directory name. Must match `paddlex_hps_${HPS_PIPELINE_NAME}_sdk`.
|
|
HPS_SDK_DIR=paddlex_hps_PaddleOCR-VL-1.6_sdk
|
|
|
|
# =============================================================================
|
|
# Gateway Configuration
|
|
# =============================================================================
|
|
|
|
# Maximum concurrent inference requests (layout-parsing / infer)
|
|
# Increase for higher throughput, decrease if device is overloaded
|
|
HPS_MAX_CONCURRENT_INFERENCE_REQUESTS=16
|
|
|
|
# Maximum concurrent non-inference requests (restructure-pages)
|
|
# Can be set higher since non-inference operations are lighter
|
|
HPS_MAX_CONCURRENT_NON_INFERENCE_REQUESTS=64
|
|
|
|
# Inference timeout in seconds
|
|
# Increase for complex documents, decrease for faster failure detection
|
|
HPS_INFERENCE_TIMEOUT=600
|
|
|
|
# Health check timeout in seconds
|
|
HPS_HEALTH_CHECK_TIMEOUT=5
|
|
|
|
# VLM server URL
|
|
# HPS_VLM_URL=http://paddleocr-vlm-server:8080
|
|
|
|
# Log level: DEBUG, INFO, WARNING, ERROR
|
|
HPS_LOG_LEVEL=INFO
|
|
|
|
# Filter health check endpoints from access logs (true/false)
|
|
# Set to false to see all access logs including health checks
|
|
HPS_FILTER_HEALTH_ACCESS_LOG=true
|
|
|
|
# Number of Uvicorn worker processes
|
|
# Recommended: 2-4 workers per CPU core
|
|
HPS_UVICORN_WORKERS=4
|
|
|
|
# =============================================================================
|
|
# Device Configuration
|
|
# =============================================================================
|
|
|
|
# Inference device ID to use (0, 1, 2, etc.)
|
|
HPS_DEVICE_ID=0
|