Julien Denize a5d06dc557 Add 320 dimension size support to MLA (#36161) ...

Signed-off-by: Julien Denize <julien.denize@mistral.ai>

2026-03-11 10:21:22 -07:00

..

conftest.py

[Chore] Clean up pytorch helper functions in vllm.utils (#26908)

2025-10-18 09:48:22 -07:00

test_aiter_flash_attn.py

[ROCm][AITER] Fix AITER import regression for explicit backend selection (#33749)

2026-02-06 15:08:16 +00:00

test_attention_selector.py

Reapply [Attention] Refactor check_and_update_config (#35122)

2026-03-09 07:17:14 -07:00

test_attention.py

[Bugfix][CI] fix typos (#34934)

2026-03-05 17:05:46 +00:00

test_cache.py

Add 320 dimension size support to MLA (#36161)

2026-03-11 10:21:22 -07:00

test_cascade_flash_attn.py

[Platform] Deprecate seed_everything (#31659)

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

test_cpu_attn.py

[CPU] Split attention dispatch by head_dim alignment (#32161)

2026-02-03 19:37:15 -08:00

test_cutlass_mla_decode.py

[Platform] Add current_platform.num_compute_units interface (#35042)

2026-02-24 22:22:49 -08:00

test_deepgemm_attention.py

[Perf] Disable clean_logits in deepgemm fp8_mqa_logits kernel (#33568)

2026-02-05 20:34:00 -05:00

test_flash_attn.py

[Platform] Deprecate seed_everything (#31659)

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

test_flashinfer_mla_decode.py

[CI/Build][AMD] Fix import errors in tests/kernels/attention (#29032)

2025-11-20 17:48:09 +08:00

test_flashinfer_trtllm_attention.py

Revert "[Kernels][FI] Skip trtllm attention when num_kv_heads=1 (#308… (#31617)

2026-01-10 12:39:59 -08:00

test_flashinfer.py

[Update] Use FlashInfer fast_decode_plan directly instead of replication (#34687)

2026-02-26 16:31:43 -08:00

test_flashmla_sparse.py

Update FlashMLA (#32491)

2026-01-21 13:03:37 +08:00

test_flashmla.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

test_lightning_attn.py

[Platform] Deprecate seed_everything (#31659)

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

test_merge_attn_states.py

[Hardware] Replace torch.cuda.synchronize() api with torch.accelerator.synchronize (#36085)

2026-03-05 10:36:39 +00:00

test_mha_attn.py

[MM Encoder] Default to use TORCH_SDPA backend for ViT on Volta/Turing GPU (#36472)

2026-03-09 03:43:44 -07:00

test_mla_decode_cpu.py

[Chore]:Extract math and argparse utilities to separate modules (#27188)

2025-10-26 04:03:32 -07:00

test_pack_unpack_triton.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

test_prefix_prefill.py

[Hardware] Replace torch.cuda.synchronize() api with torch.accelerator.synchronize (#36085)

2026-03-05 10:36:39 +00:00

test_rocm_attention_selector.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

test_triton_decode_attention.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

test_triton_prefill_attention.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

test_triton_unified_attention.py

Add unit tests for fp8 output fusion of triton_attn (#34228)

2026-02-18 06:22:49 -05:00

test_use_trtllm_attention.py

TRTLLM gen-full attn Test Coverage (#34986)

2026-03-03 11:35:34 -05:00

test_xpu_mla_sparse.py

Add XPU MLA Sparse backend for DeepSeek v3.2 (#33230)

2026-03-11 19:19:15 +08:00