IBM / onnx-mlir-servingLinks
ONNX Serving is a project written with C++ to serve onnx-mlir compiled models with GRPC and other protocols.Benefiting from C++ implementation, ONNX Serving has very low latency overhead and high throughput. ONNX Servring provides dynamic batch aggregation and workers pool to fully utilize AI accelerators on the machine.
☆25Updated 4 months ago
Alternatives and similar repositories for onnx-mlir-serving
Users that are interested in onnx-mlir-serving are comparing it to the libraries listed below
Sorting:
- Play with MLIR right in your browser☆139Updated 2 years ago
- Notes and artifacts from the ONNX steering committee☆28Updated this week
- An experimental CPU backend for Triton (https//github.com/openai/triton)☆48Updated 5 months ago
- Open deep learning compiler stack for cpu, gpu and specialized accelerators☆35Updated 3 years ago
- A lightweight, Pythonic, frontend for MLIR☆81Updated 2 years ago
- Unified compiler/runtime for interfacing with PyTorch Dynamo.☆104Updated last month
- ☆68Updated 2 years ago
- ☆50Updated last year
- Common utilities for ONNX converters☆293Updated last month
- This is a demo how to write a high performance convolution run on apple silicon☆57Updated 3 years ago
- Standalone Flash Attention v2 kernel without libtorch dependency☆113Updated last year
- C++ implementations for various tokenizers (sentencepiece, tiktoken etc).☆47Updated this week
- TORCH_TRACE parser for PT2☆72Updated 2 weeks ago
- MatMul Performance Benchmarks for a Single CPU Core comparing both hand engineered and codegen kernels.☆138Updated 2 years ago
- ☆171Updated last week
- Open source cross-platform compiler for compute-intensive loops used in AI algorithms, from Microsoft Research☆116Updated 2 years ago
- IREE's PyTorch Frontend, based on Torch Dynamo.☆105Updated this week
- Experiments and prototypes associated with IREE or MLIR☆56Updated last year
- Conversions to MLIR EmitC☆134Updated last year
- ☆322Updated last month
- MLIRX is now defunct. Please see PolyBlocks - https://docs.polymagelabs.com☆38Updated 2 years ago
- torch::deploy (multipy for non-torch uses) is a system that lets you get around the GIL problem by running multiple Python interpreters i…☆182Updated last month
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆127Updated last year
- Python interface for MLIR - the Multi-Level Intermediate Representation☆273Updated last year
- ☆24Updated last year
- TritonParse: A Compiler Tracer, Visualizer, and Reproducer for Triton Kernels☆189Updated this week
- MLIR-based partitioning system☆162Updated last week
- Header-only safetensors loader and saver in C++☆76Updated last month
- ONNX Command-Line Toolbox☆35Updated last year
- A Fusion Code Generator for NVIDIA GPUs (commonly known as "nvFuser")☆375Updated this week