PaddlePaddle
Official@paddlepaddle
Offers high-performance deep learning frameworks, cross-platform API synchronization, and specialized document extraction for enterprise-grade model deployment and research.
Agent Skills by PaddlePaddle
Showing 14 vetted skills indexed across 3 GitHub repositories.
PaddlePaddlewrite-fastdeploy-unittest
Generates CI-compliant pytest unit tests for FastDeploy source files.
PaddlePaddlefastdeploy-llm-integration
Automate adding inference deployment support for open-source LLMs to FastDeploy.
PaddlePaddlebenchmark
Benchmark FastDeploy versus SGLang with end-to-end tests and HTML reports.
PaddlePaddleresearch-report
Generate a self-contained HTML research report with embedded CSS and Mermaid diagrams.
PaddlePaddlensys-capture
Inject nsys profiling into FastDeploy GPU inference startup and collect .nsys-rep files.
PaddlePaddlepython-decorator
Apply Python decorators to unify parameter names and ordering between PyTorch and Paddle APIs.
PaddlePaddleapi-docs-updater
Synchronize Chinese API documentation with code changes by comparing docstrings and updating rst files.
PaddlePaddlepytorch-alignment-validator
Automate PyTorch API validation with pytest and custom test scripts.
PaddlePaddleapi-compatibility
Automate Paddle and PyTorch API synchronization across code, tests, and documentation.
PaddlePaddleapi-change-decider
Analyze PyTorch and Paddle API differences for migration decisions.
PaddlePaddlecpp-sink
Transform Python Paddle APIs into optimized C++ implementations.
PaddlePaddlecreate-pr
Automate git commits, pushes, and pull request creation across multiple repositories.
PaddlePaddlepaddleocr-text-recognition
Extract text and location data from images and PDFs via the PaddleOCR API.
PaddlePaddlepaddleocr-doc-parsing
Parse PDFs and images into Markdown and JSON with structured text, tables, and formulas.
Frequently Asked Questions About PaddlePaddle
FAQPage SchemaWhat specific document processing tasks does PaddlePaddle support?▼
PaddlePaddle enables high-accuracy extraction of text, tables, and mathematical formulas from images and PDFs. It converts unstructured visual document data into structured Markdown or JSON formats, facilitating data ingestion for enterprise research and reporting requirements.
Which engineers benefit from using these synchronization capabilities?▼
Machine learning engineers and infrastructure developers working on cross-framework model migration benefit from these capabilities. The system provides automated validation and synchronization between PyTorch and Paddle environments, ensuring consistent parameter ordering and documentation parity during code transitions.
What are the prerequisites for profiling inference performance?▼
To perform GPU inference profiling, users must have an environment configured with FastDeploy and the NVIDIA Nsight Systems (nsys) suite. The process involves injecting profiling hooks into the startup sequence to generate .nsys-rep files for detailed latency and resource utilization analysis.