yangjianfengo1 3754a9906d [Feature] block sparse attention (#3668) ...

* 支持稀疏attn

* fix bug

* code style

* fix moba attn get kv shape

* 修复a100编译

* codestyle

* code style

* code style

* code style

* fix conflict

* 增加单侧

* code style

* 增加eblite 加载时间

* fix bug

* for ci

* for ci

* for ci

* for ci

* 支持mlp block size 128

* 增加小算子单测

* fix 单测 mlp

* 将环境变量加入到config里面

* fix rollout config

* 修复显存

* add test server

* add test server

* fix mlp  最后一层使用full attn

2025-08-29 19:46:30 +08:00

..

graph_optimization

[CudaGraph] [SOT] Support spliting static graph into piecewise graph with cuda_graph (#3478)

2025-08-29 16:28:01 +08:00

guided_decoding

rename ernie_xxx to ernie4_5_xxx (#3621)

2025-08-26 19:29:27 +08:00

layers

[Feature] block sparse attention (#3668)

2025-08-29 19:46:30 +08:00

model_loader

[BugFix]fix dp&ep&tp and muti node infer (#3629)

2025-08-28 19:09:10 +08:00

models

[Model]support qwen2_5_vl (#3557)

2025-08-29 18:28:39 +08:00

ops

fix cpu __ini__.py (#3448)

2025-08-17 12:38:54 +08:00

__init__.py

polish code with new pre-commit rule (#2923)

2025-07-19 23:19:27 +08:00

forward_meta.py

enable dcu ci (#3402)

2025-08-29 10:23:08 +08:00

load_weight_utils.py

[BugFix]fix dp&ep&tp and muti node infer (#3629)

2025-08-28 19:09:10 +08:00

pre_and_post_process.py

add dtype int32 (#3692)

2025-08-29 14:56:35 +08:00

utils.py

[Precision] Support lm_head layer running in float32 (#3597)

2025-08-27 11:34:53 +08:00