python_backend
ФоркИзбранное
0
1
Описание
Triton backend that enables pre-process, post-processing and other logic to be implemented in Python.
Языки
- C++86,6%
- Python8%
- Shell3,2%
- CMake2,2%
3 года назад
3 года назад
2 года назад
2 года назад
2 года назад
3 года назад
5 лет назад
3 года назад
2 года назад
6 лет назад
2 года назад
3 года назад
Python Backend
User Documentation
Quick Start
Building from Source
Usage
Model Config File
Inference Request Parameters
Managing Python Runtime and Libraries
Error Handling
Managing Shared Memory
Multiple Model Instance Support
Running Multiple Instances of Triton Server
Business Logic Scripting
Using BLS with Decoupled Models
Model Loading API
Using BLS with Stateful Models
Limitation
Interoperability and GPU Support
`pbutils.Tensor.todlpack() -> PyCapsule`
`pbutils.Tensor.fromdlpack() -> Tensor`
`pbutils.Tensor.iscpu() -> bool`
Input Tensor Device Placement
Frameworks
PyTorch
TensorFlow
Custom Metrics
Examples
AddSub in NumPy
AddSubNet in PyTorch
AddSub in JAX
Business Logic Scripting
Preprocessing
Decoupled Models
Model Instance Kind
Auto-complete config
Custom Metrics
Running with Inferentia
Logging
Adding Custom Parameters in the Model Configuration
Reporting problems, asking questions
README.md