lidenghui1110 0f3939e5a9 [Feature]cpu offload connector (#1659) ...

This PR implements cpu offload connector to enable NPU kv cache offload
to host DRAM.

- vLLM version: v0.10.2
- vLLM main:
5aeb925452

Signed-off-by: lidenghui <lidenghui1110@gmail.com>
Signed-off-by: AlvisGong <gwly0401@163.com>
Signed-off-by: CalvinXKY <kyxiezju@163.com>
Co-authored-by: AlvisGong <gwly0401@163.com>

2025-09-23 14:25:05 +08:00

..

attention

[Feature]cpu offload connector (#1659)

2025-09-23 14:25:05 +08:00

compilation

[Refactor][Graph] Move graph parameter logic to acl_graph module (#3101)

2025-09-22 22:23:14 +08:00

core

[CI] Update vllm version to 20250922(5aeb925) (#3091)

2025-09-22 22:18:13 +08:00

device_allocator

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

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

distributed

[Feature]cpu offload connector (#1659)

2025-09-23 14:25:05 +08:00

eplb

FlashLB algorithm (#3042)

2025-09-23 10:27:14 +08:00

lora

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

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

models

[Feature]cpu offload connector (#1659)

2025-09-23 14:25:05 +08:00

multistream

[Quickfix] update CachedRequestState as NewRequestData changed (#2367)

2025-08-15 07:35:27 +08:00

ops

[CI] Update vllm version to 20250922(5aeb925) (#3091)

2025-09-22 22:18:13 +08:00

patch

[CI] Upgrade vLLM to 20250919 (6d8246aa) and fix some broken issue (#2907)

2025-09-20 17:37:57 +08:00

quantization

[CI] Upgrade vLLM to 20250919 (6d8246aa) and fix some broken issue (#2907)

2025-09-20 17:37:57 +08:00

sample

[CI] Upgrade vLLM to 20250920 (c60e613) and address config break (#3067)

2025-09-21 09:49:17 +08:00

spec_decode

[Refactor] Adjustments to moe_comm_method selection process (#3001)

2025-09-22 19:12:58 +08:00

torchair

[CI] Update vllm version to 20250922(5aeb925) (#3091)

2025-09-22 22:18:13 +08:00

worker

[Refactor] Rename cudagraph_support to aclgraph_support (#3104)

2025-09-23 11:30:31 +08:00

__init__.py

Bump torch version to 2.7.1 (#1562)

2025-08-05 08:43:24 +08:00

ascend_config.py

[Feature] Support moe multi-stream for aclgraph. (#2946)

2025-09-19 11:06:45 +08:00

ascend_forward_context.py

[Refactor] Adjustments to moe_comm_method selection process (#3001)

2025-09-22 19:12:58 +08:00

envs.py

[Bugfix][PD] Auto-clear producer KV cache if no pull notification (#2174)

2025-09-23 09:53:34 +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

[CI] Update vllm version to 20250922(5aeb925) (#3091)

2025-09-22 22:18:13 +08:00

utils.py

[Refactor][Graph] Move graph parameter logic to acl_graph module (#3101)

2025-09-22 22:23:14 +08:00