apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 00:17:25 +08:00
Code Issues Actions 19 Packages Projects Releases Wiki Activity

[BugFix] Fix model loading error for 300B FP8 EP parallel test case (#6382)

Browse Source 
* fix fp8 bug

* fix

* fix comment, cn to en

* fix ci

* del else in utils

* fix review
This commit is contained in:
yuxuan
2026-02-10 11:32:57 +08:00
committed by GitHub
parent a8ffcaa068
commit 83b4b082ab
3 changed files with 3 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastdeploy/model_executor/pre_and_post_process.py
+2 -2
View File
@@ -139,7 +139,7 @@ def limit_thinking_content_length(
step_idx,
limit_think_status,
stop_flags,
eos_token_ids, # 处理由于模型效果问题导致思考过程中输出eos token的问题
eos_token_ids, # Address the issue where the model outputs EOS tokens during the reasoning process due to model performance problems
think_end_id,
)
elif limit_strategy == "\n</think>\n\n":
@@ -179,7 +179,7 @@ def speculate_limit_thinking_content_length(
limit_think_status,
accept_num,
stop_flags,
eos_token_ids, # 处理由于模型效果问题导致思考过程中输出eos token的问题
eos_token_ids, # Address the issue where the model outputs EOS tokens during the reasoning process due to model performance problems
think_end_id,
)
elif limit_strategy == "\n</think>\n\n":