wuba/dl_inference
A production-ready deep learning inference serving tool for TensorFlow, PyTorch, and Caffe models with automatic TensorRT optimization.
Not currently ranked — collecting fresh signals.
star history
dl_inference is a general deep learning inference tool developed by 58同城 that enables rapid deployment of trained models from TensorFlow, PyTorch, and Caffe frameworks into production environments. It provides unified RPC service interfaces, supports both GPU and CPU deployment modes, and implements load balancing for multi-node model serving. The tool can automatically convert SavedModel and PyTorch .pth models to TensorRT format to improve inference performance.
Frequently asked
- What is wuba/dl_inference?
- A production-ready deep learning inference serving tool for TensorFlow, PyTorch, and Caffe models with automatic TensorRT optimization.
- Is dl_inference open source?
- Yes — wuba/dl_inference is an open-source project tracked on heatdrop.
- What language is dl_inference written in?
- wuba/dl_inference is primarily written in Java.
- How popular is dl_inference?
- wuba/dl_inference has 417 stars on GitHub.
- Where can I find dl_inference?
- wuba/dl_inference is on GitHub at https://github.com/wuba/dl_inference.