Refactor Fused Moe Module (#1309)

Author: wenscarl

.gitignore

@@ -7,6 +7,7 @@ compile_commands.json
 
 # Generated files
 csrc/generated/
+csrc/nv_internal/tensorrt_llm/cutlass_instantiations/
 docs/generated/
 flashinfer/_build_meta.py
 flashinfer/data/

README.md

@@ -88,7 +88,7 @@ python -m flashinfer.aot
 # Build AOT wheel
 python -m build --no-isolation --wheel
 # Install AOT wheel
-python -m pip install dist/flashinfer-*.whl
+python -m pip install dist/flashinfer_*.whl
 ```
 
 For more details, refer to the [Install from Source documentation](https://docs.flashinfer.ai/installation.html#install-from-source).

csrc/nv_internal/cpp/kernels/quantization.cu

@@ -363,5 +363,7 @@ template void invokeBatchedFP4Quantization<__nv_fp8_e4m3, 32>(
     int32_t* SFOuput, bool useUE8M0, int multiProcessorCount, cudaStream_t stream);
 #endif
 
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
 }  // namespace kernels
 }  // namespace tensorrt_llm