1
0
AlpinDale 0e6c400b13 feat: re-add GGUF (#600) 4 сар өмнө
..
all_reduce 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) 5 сар өмнө
attention dc1b59df9c fix: compiler warnings for _C and _moe 4 сар өмнө
backup f8dfac6372 chore: attention refactor and upstream sync apr01 (#365) 9 сар өмнө
core 141672a0d4 kernels: disambiguate quantized types via a new ScalarType 4 сар өмнө
cpu 141672a0d4 kernels: disambiguate quantized types via a new ScalarType 4 сар өмнө
hadamard 5d288aa76c feat: add fast hadamard transformation kernels (#232) 11 сар өмнө
mamba dc1b59df9c fix: compiler warnings for _C and _moe 4 сар өмнө
moe 141672a0d4 kernels: disambiguate quantized types via a new ScalarType 4 сар өмнө
prepare_inputs dd18c5042c move prepare_inputs to the GPU (#596) 5 сар өмнө
punica 848731f527 chore: add punica sizes for mistral nemo 4 сар өмнө
quantization 0e6c400b13 feat: re-add GGUF (#600) 4 сар өмнө
activation_kernels.cu 5fbd93797d fix: beta value in gelu_tanh kernel being divided by 0.5 4 сар өмнө
cache.h 32bdbd1ee4 chore: add fp8 support to `reshape_and_cache_flash` 4 сар өмнө
cache_kernels.cu 32bdbd1ee4 chore: add fp8 support to `reshape_and_cache_flash` 4 сар өмнө
cuda_compat.h 00acf371f9 rocm: fused topk softmax 5 сар өмнө
cuda_utils.h 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) 5 сар өмнө
cuda_utils_kernels.cu 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) 5 сар өмнө
dispatch_utils.h dc1b59df9c fix: compiler warnings for _C and _moe 4 сар өмнө
layernorm_kernels.cu dc1b59df9c fix: compiler warnings for _C and _moe 4 сар өмнө
ops.h dd18c5042c move prepare_inputs to the GPU (#596) 5 сар өмнө
pos_encoding_kernels.cu 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) 5 сар өмнө
reduction.cuh aba03b4756 feat: dynamic per-token activation quantization 5 сар өмнө
torch_bindings.cpp 0e6c400b13 feat: re-add GGUF (#600) 4 сар өмнө