whx 3393d53b36 [Scheduler][MTP] Add support for speculative decoding in AsecendScheduler. (#943) ...

This PR adds support for speculative decoding in AsecendScheduler.
Also inculde part of support for disaggregated prefill, full support
will be merged in follow-up PR.

---------

Signed-off-by: whx-sjtu <2952154980@qq.com>

2025-06-11 20:55:44 +08:00

..

attention

Enable kvcache_nz for the decode process in torchair graph mode (#1098)

2025-06-11 14:09:28 +08:00

compilation

[CI] Upgrade vllm to 0.9.1 (#1165)

2025-06-11 16:33:11 +08:00

core

[Scheduler][MTP] Add support for speculative decoding in AsecendScheduler. (#943)

2025-06-11 20:55:44 +08:00

device_allocator

[bugfix] Improve log level and info for custom ops build (#937)

2025-05-23 10:05:57 +08:00

distributed

[bugfix] some bugs maybe fail to run (#896)

2025-06-03 11:07:33 +08:00

lora

[Bugfix] fix import error (#600)

2025-04-22 08:57:25 +08:00

models

Support multistream of shared experts in FusedMoE (#997)

2025-06-11 09:18:38 +08:00

multistream

[perf]: support dual-batch overlap(dbo) for deepseek (#941)

2025-06-07 16:46:58 +08:00

ops

Support multistream of shared experts in FusedMoE (#997)

2025-06-11 09:18:38 +08:00

patch

[CI] Upgrade vllm to 0.9.1 (#1165)

2025-06-11 16:33:11 +08:00

quantization

Support multistream of shared experts in FusedMoE (#997)

2025-06-11 09:18:38 +08:00

sample

Spec decode support for V1 Engine (#874)

2025-05-23 14:25:46 +08:00

worker

[CI] Upgrade vllm to 0.9.1 (#1165)

2025-06-11 16:33:11 +08:00

__init__.py

[CI] Patch torch.library.infer_schema for fused moe ops to fix CI (#854)

2025-05-14 19:49:09 +08:00

ascend_config.py

Enable kvcache_nz for the decode process in torchair graph mode (#1098)

2025-06-11 14:09:28 +08:00

envs.py

[Bugfix] Set ACL_OP_INIT_MODE env var default to 0 (#1123)

2025-06-09 14:07:37 +08:00

platform.py

[BugFix] Fix data parallel (#940)

2025-06-09 14:08:18 +08:00

utils.py

Support multistream of shared experts in FusedMoE (#997)

2025-06-11 09:18:38 +08:00