apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-12-24 13:28:13 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[BugFix] qwen2.5vl enable_thinking=true and image_patch_id bug fix (#3921)

Browse Source
This commit is contained in:
CSWYF3634076
2025-09-11 15:08:24 +08:00
committed by GitHub
parent 2650f58740
commit e4c64a71cc
1 changed files with 11 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
fastdeploy/input/qwen_vl_processor/qwen_vl_processor.py
View File

@@ -69,7 +69,7 @@ class QwenVLProcessor(TextProcessor):
tokenizer=self.tokenizer,
**processor_kwargs,
)
self.image_patch_id = self.processor.image_token_id
self.limit_mm_per_prompt = self._parse_limits(limit_mm_per_prompt)
def process_request(self, request, max_model_len=None, **kwargs):
@@ -249,6 +249,16 @@ class QwenVLProcessor(TextProcessor):
# Handle continuation of previous generation by appending existing tokens
if metadata and metadata.get("generated_token_ids"):
self.append_generated_tokens(outputs, metadata["generated_token_ids"])
enable_thinking = False
if metadata:
enable_thinking = metadata.get("enable_thinking", False)
if request.get("chat_template_kwargs"):
chat_template_kwargs = request.get("chat_template_kwargs")
enable_thinking = chat_template_kwargs.get("enable_thinking", False)
request["enable_thinking"] = enable_thinking
outputs = self.pack_outputs(outputs)
request["prompt_token_ids"] = outputs["input_ids"].tolist()