apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-21 07:40:37 +08:00
Code Issues Actions 6 Packages Projects Releases Wiki Activity

[Cherry-Pick][Bug Fix]fix the bug for real size 0 in cudagraph (#3888)

Browse Source 
* fix the bug for real size 0 in cudagraph

* fix cache_messager

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
This commit is contained in:
Zero Rains
2025-09-08 14:06:10 +08:00
committed by GitHub
parent c7c1627456
commit d43549953c
2 changed files with 5 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
fastdeploy/cache_manager/cache_messager.py
View File

@@ -163,7 +163,7 @@ class CacheMessager:
try:
prefilled_step_idx_data = np.zeros(shape=[1], dtype=np.int32)
prefilled_layer_idx_data = np.zeros(shape=[1], dtype=np.int32)
prefilled_layer_name = f"splitwise_complete_prefilled_step_{self.dp_rank_id}.{self.gpu_id}"
prefilled_layer_name = f"splitwise_complete_prefilled_layer_{self.dp_rank_id}.{self.gpu_id}"
prefilled_step_name = f"splitwise_complete_prefilled_step_{self.dp_rank_id}.{self.gpu_id}"
step_shm_value = IPCSignal(
name=f"splitwise_complete_prefilled_step_{self.dp_rank_id}",