Popular repositories Loading

1. relax relax Public archive

   Python 193 56

2. libflash_attn libflash_attn Public

   Standalone Flash Attention v2 kernel without libtorch dependency

   C++ 108 15

3. tenset tenset Public

   Python 92 34

4. cutlass_fpA_intB_gemm cutlass_fpA_intB_gemm Public

   A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer

   C++ 92 23

5. TLCBench TLCBench Public

   Benchmark scripts for TVM

   Python 74 28

6. tophub tophub Public

   tophub autotvm log collections

   69 36