3 people 9da56fd18b [Bugfix] Add swiglu limits to deepgemm fp8 methods (#41986) ...

Cherry-picked from https://github.com/vllm-project/vllm/pull/41986

Plumb SwiGLU clamp limit through DeepGemm FP8/W4A8 MoE quant configs
and experts. Extend silu_mul_per_token_group_quant_fp8_colmajor with
clamp support and forward the limit on all FP8/MXFP8/MXFP4 paths.

Co-authored-by: Yongye Zhu <zyy1102000@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>

Signed-off-by: khluu <khluu000@gmail.com>

2026-05-14 12:38:36 -07:00

..

attention

Add nvfp4 kv cache support (#40177)

2026-05-01 04:55:16 +00:00

core

[CI] fix test_rotary_embedding_opcheck format error (#41202)

2026-04-29 00:32:37 -07:00

helion

[Kernel][Helion] Optimize Helion config parsing latency (#40850)

2026-05-07 20:27:34 -07:00

ir

[vLLM IR] 2/N fused_add_rms_norm and maybe_inplace overload (#36823)

2026-05-01 23:41:15 -04:00

mamba

[Misc] Replace mamba_type string literals with MambaAttentionBackendEnum (#41188)

2026-05-11 03:59:36 +00:00

moe

[Bugfix] Add swiglu limits to deepgemm fp8 methods (#41986)

2026-05-14 12:38:36 -07:00

quantization

[CUDA][CUTLASS] Enable cutlass scaled mm for non-compatible sizes (#41868)

2026-05-08 15:58:05 -07:00

__init__.py

[CI/Build] Move test_utils.py to tests/utils.py (#4425)

2024-05-13 23:50:09 +09:00

allclose_default.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

quant_utils.py

[Perf] Create TMA-aligned input scale tensor for DeepGemm on Hopper (#32619)

2026-01-22 15:47:04 -05:00

test_apply_repetition_penalties.py

[Platform] Deprecate seed_everything (#31659)

2026-01-04 18:34:04 -08:00

test_awq_int4_to_int8.py

[CPU] Bump up to the latest CPU kernels (#41924)

2026-05-07 05:45:59 -07:00

test_cache_kernels.py

[Hardware] Replace torch.cuda.device_count/current_device/set_device API (#36145)

2026-03-12 07:57:47 -07:00

test_compressor_kv_cache.py

[DSv4] Use cvt PTX for FP32->FP4 conversion (#41015)

2026-04-29 16:16:40 -07:00

test_concat_mla_q.py

[Attention][Perf][Kernel] Replace torch.cat with vectorized CUDA kernel MLA query concat - DeepSeek-V3.2 (#34917)

2026-03-09 09:50:36 -07:00

test_cp_gather_fp8.py

[Attention][Perf] Optimize cp_gather_and_upconvert_fp8_kv_cache - DeepSeek-v3.2 (#35290)

2026-03-09 09:46:57 -07:00

test_fla_layernorm_guard.py

[Bugfix] Fix dtype mismatch in RMSNormGated.forward_native() during torch.compile (#35256)

2026-03-01 15:14:46 -05:00

test_flex_attention.py

full cudagraph for flex-attn (#36298)

2026-04-02 21:15:01 -07:00

test_fused_deepseek_v4_qnorm_rope_kv_insert.py

[Feat] DeepSeek V4 Rebased (#40860)

2026-04-26 18:31:08 -07:00

test_fused_gdn_post_conv.py

[Perf] fuse kernels in gdn (#37813)

2026-04-02 11:52:18 +00:00

test_fused_indexer_q_rope_quant.py

[DSv4] Improved fused Indexer Q quant kernel (#41428)

2026-05-09 01:20:32 -07:00

test_fused_inv_rope_fp8_quant.py

[Feat] DeepSeek V4 Rebased (#40860)

2026-04-26 18:31:08 -07:00

test_fused_quant_activation.py

Add platform manual_seed_all API (#38468)

2026-04-10 13:43:50 +08:00

test_fused_recurrent_packed_decode.py

[Bugfix] fix main branch pre-commit error (1 line change) (#36897)

2026-03-12 09:08:37 -07:00

test_fused_sigmoid_gating_delta_rule.py

[Kernel] Add fused_sigmoid_gating_delta_rule_update kernel for Qwen3 Next (#35777)

2026-03-08 23:41:01 -07:00

test_kda.py

[Bugfix] correct h matrix layout in chunk_kda output kernel (#40956)

2026-04-30 16:22:41 +08:00

test_mhc_kernels.py

add fused mhc_post_pre kernel (#41536)

2026-05-10 19:56:52 -07:00

test_onednn.py

[CPU] Refactor CPU attention backend (#27954)

2025-11-12 09:43:06 +08:00

test_shuffle_rows.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

test_top_k_per_row.py

[Feat] DeepSeek V4 Rebased (#40860)

2026-04-26 18:31:08 -07:00

utils.py

[ROCm] Fix MoE kernel test failures on gfx950 (#37833)

2026-03-25 13:46:40 -05:00