AlpinDale 0e6c400b13 feat: re-add GGUF (#600) il y a 6 mois
..
all_reduce 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) il y a 7 mois
attention dc1b59df9c fix: compiler warnings for _C and _moe il y a 6 mois
backup f8dfac6372 chore: attention refactor and upstream sync apr01 (#365) il y a 11 mois
core 141672a0d4 kernels: disambiguate quantized types via a new ScalarType il y a 6 mois
cpu 141672a0d4 kernels: disambiguate quantized types via a new ScalarType il y a 6 mois
hadamard 5d288aa76c feat: add fast hadamard transformation kernels (#232) il y a 1 an
mamba dc1b59df9c fix: compiler warnings for _C and _moe il y a 6 mois
moe 141672a0d4 kernels: disambiguate quantized types via a new ScalarType il y a 6 mois
prepare_inputs dd18c5042c move prepare_inputs to the GPU (#596) il y a 7 mois
punica 848731f527 chore: add punica sizes for mistral nemo il y a 6 mois
quantization 0e6c400b13 feat: re-add GGUF (#600) il y a 6 mois
activation_kernels.cu 5fbd93797d fix: beta value in gelu_tanh kernel being divided by 0.5 il y a 7 mois
cache.h 32bdbd1ee4 chore: add fp8 support to `reshape_and_cache_flash` il y a 7 mois
cache_kernels.cu 32bdbd1ee4 chore: add fp8 support to `reshape_and_cache_flash` il y a 7 mois
cuda_compat.h 00acf371f9 rocm: fused topk softmax il y a 7 mois
cuda_utils.h 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) il y a 7 mois
cuda_utils_kernels.cu 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) il y a 7 mois
dispatch_utils.h dc1b59df9c fix: compiler warnings for _C and _moe il y a 6 mois
layernorm_kernels.cu dc1b59df9c fix: compiler warnings for _C and _moe il y a 6 mois
ops.h dd18c5042c move prepare_inputs to the GPU (#596) il y a 7 mois
pos_encoding_kernels.cu 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) il y a 7 mois
reduction.cuh aba03b4756 feat: dynamic per-token activation quantization il y a 7 mois
torch_bindings.cpp 0e6c400b13 feat: re-add GGUF (#600) il y a 6 mois