vllmellm d260f799a9 [FEAT] [ROCm] Upgrade AITER Fused MoE kernels. (#18271) ...

Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>

2025-05-26 23:14:07 -07:00

..

configs

Add files via uploadAdd fused MoE kernel tuning configs (fp8_w8a8) for DeepSeek V3/R1 on a single-node 8x NVIDIA H20 96GB setup (#18337)

2025-05-19 09:49:57 -07:00

__init__.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

cutlass_moe.py

[ModelOpt] Introduce VLLM_MAX_TOKENS_PER_EXPERT_FP4_MOE env var to control blockscale tensor allocation (#18160)

2025-05-23 15:52:20 -07:00

deep_gemm_moe.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

fused_batched_moe.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

fused_marlin_moe.py

[Kernel] fp4 marlin kernel (#17687)

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

fused_moe.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

layer.py

[FEAT] [ROCm] Upgrade AITER Fused MoE kernels. (#18271)

2025-05-26 23:14:07 -07:00

modular_kernel.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

moe_align_block_size.py

Update deprecated type hinting in model_executor/layers (#18056)

2025-05-13 04:17:23 -07:00

moe_pallas.py

[TPU] Re-enable the Pallas MoE kernel (#18025)

2025-05-20 19:52:27 -07:00

moe_permute_unpermute.py

[Build/CI] Fix CUDA 11.8 build (#17679)

2025-05-22 12:13:54 -07:00

moe_torch_iterative.py

Expert Parallelism (EP) Support for DeepSeek V2 (#12583)

2025-02-24 07:33:20 -08:00

pplx_prepare_finalize.py

Refactor pplx init logic to make it modular (prepare for deepep) (#18200)

2025-05-23 23:43:43 +08:00

prepare_finalize.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

rocm_aiter_fused_moe.py

[FEAT] [ROCm] Upgrade AITER Fused MoE kernels. (#18271)

2025-05-26 23:14:07 -07:00

triton_deep_gemm_moe.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00

utils.py

Modularize fused experts and integrate PPLX kernels (#15956)

2025-05-14 13:11:54 -07:00