Ronald 916a9a1913 fix synchronize error of exceeds_max_model_len d2h copy (#4708) ...

### What this PR does / why we need it?
there is d2h copy blocking cpu operations in mtp propose method, which
make host bound issue. this pr refactor it and use cpu tensor to
implement it.

### Does this PR introduce _any_ user-facing change?
no

### How was this patch tested?
vllm main f5d3d93c40417c296c20dc301100e55708a17f3f

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

Signed-off-by: Ronald1995 <ronaldautomobile@163.com>
Co-authored-by: wangxiyuan <wangxiyuan1007@gmail.com>

2025-12-08 09:07:59 +08:00

..

_cann_ops_custom

[Kernel] add custom op GmmSwigluQuantWeightNzTensorList (#3804)

2025-11-28 18:06:39 +08:00

attention

Support DeepSeekV3.2 with MLAPO operator (#4753)

2025-12-07 12:40:24 +08:00

compilation

Adopt inductor fusion and define quantization fusion pass (#4168)

2025-12-04 10:29:48 +08:00

core

Drop ascend scheduler (#4623)

2025-12-05 09:03:45 +08:00

device_allocator

[Misc]Clean up useless import from vllm (#2049)

2025-07-28 16:01:59 +08:00

distributed

[P/D] check kv extra config and del hccl backend (#4547)

2025-12-07 15:19:42 +08:00

eplb

[Bugifx] fix quant_apply_mlp w1_scale type error & fix getting num_local_expert (#4632)

2025-12-05 16:04:24 +08:00

kv_offload

upgrade vLLM to main (#4608)

2025-12-02 22:10:52 +08:00

lora

[refact] unified soc_version code (#4359)

2025-11-26 14:28:55 +08:00

model_loader

Drop 0.11.0 support (#4377)

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

ops

[Feat]enable sfa cp for dsv3.2 (#4702)

2025-12-06 19:46:41 +08:00

patch

[MM][Patch] Remove patch for cos/sin cache (#4672)

2025-12-04 22:30:06 +08:00

quantization

[Bugifx] fix quant_apply_mlp w1_scale type error & fix getting num_local_expert (#4632)

2025-12-05 16:04:24 +08:00

sample

[BugFix][Triton] Fix ub overflow bug of sample_recover_tokens_kernel (#4673)

2025-12-05 15:16:19 +08:00

spec_decode

fix synchronize error of exceeds_max_model_len d2h copy (#4708)

2025-12-08 09:07:59 +08:00

torchair

Drop ascend scheduler (#4623)

2025-12-05 09:03:45 +08:00

worker

[BugFix] Refactor ACL graph size adjustment for speculative decoding (#4640)

2025-12-07 17:32:45 +08:00

xlite

[Feat] Add Euler xlite graph wrapper support (#4526)

2025-12-08 08:27:46 +08:00

__init__.py

clean up model module (#4611)

2025-12-02 17:35:47 +08:00

ascend_config.py

[Feat] Add Euler xlite graph wrapper support (#4526)

2025-12-08 08:27:46 +08:00

ascend_forward_context.py

add dispatch_gmm_combine kernel (#3532)

2025-12-04 23:00:59 +08:00

cpu_binding.py

[main] support cpu binding (#3546)

2025-10-21 09:17:03 +08:00

envs.py

[refact] unified soc_version code (#4359)

2025-11-26 14:28:55 +08:00

meta_registration.py

Fix the bugs about operator registration by PyTorch Dispatcher (#2786)

2025-09-13 11:58:52 +08:00

platform.py

[Feat] Add Euler xlite graph wrapper support (#4526)

2025-12-08 08:27:46 +08:00

profiling_config.py

Drop ascend scheduler (#4623)

2025-12-05 09:03:45 +08:00

utils.py

[BugFix] Refactor ACL graph size adjustment for speculative decoding (#4640)

2025-12-07 17:32:45 +08:00