Pinned

1. microsoft/BitBLAS microsoft/BitBLAS Public

   BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.

   Python 124 12

2. apache/tvm apache/tvm Public

   Open deep learning compiler stack for cpu, gpu and specialized accelerators

   Python 11.2k 3.3k

3. ZYNQ-NVDLA ZYNQ-NVDLA Public

   NVDLA (An Opensource DL Accelerator Framework) implementation on FPGA.

   Verilog 264 55

4. microsoft/nnfusion microsoft/nnfusion Public

   A flexible and efficient deep neural network (DNN) compiler that generates high-performance executable from a DNN model description.

   C++ 923 154