Jinzhen Lin d4154c35a2

[Bugfix] fix moe marlin topk_weight loading (#18080) ...

Co-authored-by: mgoin <mgoin64@gmail.com>

2025-05-13 23:31:57 -07:00

..

attention

Implements dual-chunk-flash-attn backend for dual chunk attention with sparse attention support (#11844)

2025-05-12 19:52:47 -07:00

core

[Kernel] fp4 marlin kernel (#17687)

2025-05-10 19:58:49 -07:00

cpu

[Hardware][Power] Enable compressed tensor W8A8 INT8 quantization for POWER (#17153)

2025-05-07 22:35:03 -07:00

cutlass_extensions

Add cutlass support for blackwell fp8 blockwise gemm (#14383)

2025-05-08 15:09:55 -07:00

mamba

[BugFix] fix some typos found by typos. (#16314)

2025-04-09 03:43:59 -07:00

moe

[Bugfix] fix moe marlin topk_weight loading (#18080)

2025-05-13 23:31:57 -07:00

prepare_inputs

[Misc][Easy] Annotate unused vars in the csrc files (#14798)

2025-03-15 12:40:09 +08:00

quantization

[AMD][torch.compile] Enable silu+fp8_quant fusion for rocm (#18082)

2025-05-13 22:13:56 -07:00

rocm

[Easy] Eliminate c10::optional usage in vllm/csrc (#17819)

2025-05-08 03:05:10 -07:00

sparse/cutlass

[BugFix/Build] Fix sparse kernels not getting built on hopper (#14572)

2025-03-11 17:09:03 +00:00

activation_kernels.cu

[Kernel] Support MulAndSilu (#11624)

2025-01-15 02:29:53 +00:00

cache_kernels.cu

Allocate kv_cache with stride order (#16605)

2025-04-25 22:03:31 -07:00

cache.h

[Attention] MLA with chunked prefill (#12639)

2025-02-21 15:30:12 -08:00

cuda_compat.h

[Kernel][ROCm][AMD] enable fused topk_softmax kernel for moe layer (#4927)

2024-06-02 14:13:26 -07:00

cuda_utils_kernels.cu

[NVIDIA] Support nvfp4 quantization (#12784)

2025-02-12 19:51:51 -08:00

cuda_utils.h

[Attention] MLA with chunked prefill (#12639)

2025-02-21 15:30:12 -08:00

cuda_view.cu

[V1] Fully Transparent Implementation of CPU Offloading (#15354)

2025-03-31 20:22:34 +08:00

cumem_allocator.cpp

[core] improve error handling when wake up from sleep mode (#12981)

2025-02-10 09:38:57 +08:00

custom_all_reduce_test.cu

[Distributed] Add custom allreduce support for ROCM (#14125)

2025-03-31 22:49:12 -07:00

custom_all_reduce.cu

[Distributed] Add custom allreduce support for ROCM (#14125)

2025-03-31 22:49:12 -07:00

custom_all_reduce.cuh

fix: spelling (#16466)

2025-04-11 23:24:22 -07:00

dispatch_utils.h

dynamic distpatch of fp8 kernels (#14245)

2025-03-11 10:54:56 -04:00

layernorm_kernels.cu

[Kernel] Use fused rmsnorm for some models like qwen3 series (#17735)

2025-05-06 23:10:02 -07:00

layernorm_quant_kernels.cu

dynamic distpatch of fp8 kernels (#14245)

2025-03-11 10:54:56 -04:00

ops.h

Implements dual-chunk-flash-attn backend for dual chunk attention with sparse attention support (#11844)

2025-05-12 19:52:47 -07:00

permute_cols.cu

[Kernel] (2/N) Machete - Integrate into CompressedTensorsWNA16 and GPTQMarlin (#7701)

2024-09-23 13:46:26 -04:00

pos_encoding_kernels.cu

Make key optional for rotary embedding (#17566)

2025-05-07 00:11:46 -07:00

torch_bindings.cpp

Implements dual-chunk-flash-attn backend for dual chunk attention with sparse attention support (#11844)

2025-05-12 19:52:47 -07:00

type_convert.cuh

[torch.compile] Fuse RMSNorm with quant (#9138)

2024-11-08 21:20:08 +00:00