Li Wang a5ae07a5d2 [Bugfix] Fix mm_merge (#5249) ...

### What this PR does / why we need it?
We should transfer the mm_embed to the dtype of input_embed before
performing the in-place assignment

- vLLM version: release/v0.13.0
- vLLM main:
ad32e3e19c

Signed-off-by: wangli <wangli858794774@gmail.com>

2025-12-31 09:49:55 +08:00

..

__init__.py

[CustomOp] Register AscendApplyRotaryEmb CustomOp and remove related patch (#4667)

2025-12-23 10:04:37 +08:00

patch_bert.py

[Model] Support pooling models (#3122)

2025-12-10 11:37:57 +08:00

patch_deepseek.py

upgrade vLLM to 0.12.0 tag (#4647)

2025-12-03 23:43:05 +08:00

patch_distributed.py

[Refactor] refactor patch module (#3555)

2025-10-21 20:19:46 +08:00

patch_minicpm.py

[Refactor] refactor patch module (#3555)

2025-10-21 20:19:46 +08:00

patch_module.py

[main][BugFix] Fixed an accuracy bug of Qwen3-next-MTP when batched inferring (#4932)

2025-12-15 13:22:30 +08:00

patch_multimodal_merge.py

[Bugfix] Fix mm_merge (#5249)

2025-12-31 09:49:55 +08:00

patch_qwen3_next_mtp.py

[BugFix][main] Adapted Qwen3-Next-MTP to chunked prefill (#4770)

2025-12-10 22:54:24 +08:00

patch_qwen3_next.py

[task] Add fused gdn gating triton kernel (#4304)

2025-12-22 14:09:19 +08:00

patch_rejection_sampler.py

[Feat] Refactor rejection sampler (#4975)

2025-12-16 11:32:26 +08:00

patch_rope.py

[MM][Model][Perf] Remove Qwen2.5-VL modeling files and add patch for VisionAttention (#4349)

2025-11-28 14:23:00 +08:00

patch_triton.py

[OPS] support triton causal_conv1d_fn ops (#4119)

2025-12-11 15:52:39 +08:00

patch_weight_loader.py

Drop 0.11.0 support (#4377)

2025-11-24 17:08:20 +08:00