[EPLB][Bugfix] policy_swift_balancer bugfix and renaming (#5897)

### What this PR does / why we need it?
1. Rename dynamic_ep to default_eplb.
2. Rename dynamic_ep_v2 to swift_balancer
3. Discard func compose_expert_update_info_bipartite.

- vLLM version: v0.13.0
- vLLM main:
https://github.com/vllm-project/vllm/commit/bde38c11df0ea066a740efe9b77fff5418be45df

Signed-off-by: shenchuxiaofugui <1311027364@qq.com>
This commit is contained in:
LI SHENGYONG authored and GitHub committed 2026-01-19 05:47:40 +00:00
1 parent b27774dbd6
commit 83de5385b4
8 files changed
+39 -149

No files matched your search

@@ -25,8 +25,8 @@ vllm_ascend
│ ├── core
│ │ ├── policy
│ │ │ ├── policy_abstract.py
│ │ │ ├── policy_dynamic_ep.py
│ │ │ ├── policy_dynamic_ep_v2.py
│ │ │ ├── policy_default_eplb.py
│ │ │ ├── policy_swift_balancer.py
│ │ │ ├── policy_factory.py
│ │ │ ├── policy_flashlb.py
│ │ ├── eplb_device_transfer_loader.py
@@ -52,9 +52,9 @@ vllm_ascend
*Load balancing algorithms with factory pattern instantiation*
- `policy_abstract.py`
Abstract class for load balancing strategy interfaces
- `policy_dynamic_ep.py`
- `policy_default_eplb.py`
Default implementation of open-source EPLB paper algorithm
- `policy_dynamic_ep_v2.py`
- `policy_swift_balancer.py`
Enhanced version optimizing expert swaps for low-bandwidth devices (e.g., A2)
- `policy_flashlb.py`
Threshold-based adjustment reducing operational costs through layer-wise fluctuation detection