czhu-cohere cdd7025961

[kernel] Improve FP8 PTPC on Hopper for larger shapes (#28692) ...

Signed-off-by: czhu-cohere <conway.zhu@cohere.com>

2025-11-14 09:59:11 -08:00

..

awq

…

cutlass_w4a8

…

fp4

[Bugfix] Use latency MOE backend as default for Flashinfer and other misc fixes (#27439)

2025-11-07 04:18:39 -08:00

fused_kernels

[torch.compile] Enable attention and allreduce fusion without custom ops enabled (#24604)

2025-10-17 08:10:23 -06:00

gguf

…

gptq

[Kernel] Add GPTQv2 format support for low-bit or asymmetric quantization, by adapting gptq_gemm (#26092)

2025-10-23 23:26:13 -04:00

gptq_allspark

…

gptq_marlin

Rewrite C++ meta funcs to Python (#28595)

2025-11-14 00:52:50 +08:00

hadamard/hadacore

…

machete

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

marlin/sparse

…

w8a8

[kernel] Improve FP8 PTPC on Hopper for larger shapes (#28692)

2025-11-14 09:59:11 -08:00

activation_kernels.cu

[Performance][B200] silu_mul_quant: pack scales in int32 (#28358)

2025-11-13 10:16:55 -08:00

utils.cuh

…

vectorization_utils.cuh

…

vectorization.cuh

…