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

[Feature] Pass through the chat_template_kwargs to the data processing module (#3421)

Browse Source 
* fix chat_template_args

* fix args

* add offline

* add offline

* fix

* fix

* fix default enable_thinking value

* fix default enable_thinking value

* modify condition

* Revert "modify condition"

This reverts commit 26430bdeb1.

* fix unit test
This commit is contained in:
luukunn
2025-08-19 10:50:01 +08:00
committed by GitHub
parent a053ab889b
commit 3a7a20d191
6 changed files with 50 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
fastdeploy/input/ernie_processor.py
View File

@@ -108,7 +108,16 @@ class ErnieProcessor(BaseDataProcessor):
request.prompt_token_ids = token_ids
data_processor_logger.info(f"req_id:{request.request_id}, tokens:{tokens}, token_ids: {token_ids}")
else:
request.prompt_token_ids = self.messages2ids(request.to_dict())
task = request.to_dict()
chat_template_kwargs = kwargs.get("chat_template_kwargs")
if chat_template_kwargs:
if isinstance(chat_template_kwargs, dict):
for k, v in chat_template_kwargs.items():
if k not in task:
task[k] = v
else:
raise ValueError("Invalid input: chat_template_kwargs must be a dict")
request.prompt_token_ids = self.messages2ids(task)
if len(request.prompt_token_ids) == 0:
raise ValueError("Invalid input: prompt_token_ids must be a non-empty sequence of token IDs")
@@ -163,6 +172,14 @@ class ErnieProcessor(BaseDataProcessor):
req_id = request.get("request_id", None)
data_processor_logger.info(f"req_id:{req_id}, tokens:{tokens}, token_ids: {token_ids}")
else:
chat_template_kwargs = request.get("chat_template_kwargs")
if chat_template_kwargs:
if isinstance(chat_template_kwargs, dict):
for k, v in chat_template_kwargs.items():
if k not in request:
request[k] = v
else:
raise ValueError("Invalid input: chat_template_kwargs must be a dict")
request["prompt_token_ids"] = self.messages2ids(request)
if len(request["prompt_token_ids"]) == 0:
raise ValueError("Invalid input: prompt_token_ids must be a non-empty sequence of token IDs")

11
fastdeploy/input/ernie_vl_processor.py
View File

@@ -111,7 +111,7 @@ class ErnieMoEVLProcessor(ErnieProcessor):
"""process the input data"""
request.chat_template = kwargs.get("chat_template")
task = request.to_dict()
task["enable_thinking"] = kwargs.get("enable_thinking", True)
task["chat_template_kwargs"] = kwargs.get("chat_template_kwargs")
self.process_request_dict(task, max_model_len)
request = Request.from_dict(task)
request = self._apply_default_parameters(request)
@@ -218,6 +218,15 @@ class ErnieMoEVLProcessor(ErnieProcessor):
elif request.get("messages"):
messages = request["messages"]
self._check_mm_limits(messages)
chat_template_kwargs = request.get("chat_template_kwargs")
if chat_template_kwargs:
if isinstance(chat_template_kwargs, dict):
for k, v in chat_template_kwargs.items():
if k not in request:
request[k] = v
else:
raise ValueError("Invalid input: chat_template_kwargs must be a dict")
request.setdefault("enable_thinking", True)
outputs = self.ernie_processor.request2ids(request)
else:
raise ValueError(f"Request must contain 'prompt', or 'messages': {request}")

20
fastdeploy/input/text_processor.py
View File

@@ -208,7 +208,6 @@ class DataProcessor(BaseDataProcessor):
request = self._apply_default_parameters(request)
if request.get("eos_token_ids") is None or len(request.eos_token_ids) == 0:
request.eos_token_ids = self.eos_token_ids
stop_sequences = request.get("stop", [])
if stop_sequences is not None and len(stop_sequences) != 0:
stop_seqs, stop_seqs_len = self.update_stop_seq(stop_sequences)
@@ -222,7 +221,15 @@ class DataProcessor(BaseDataProcessor):
if self.tokenizer.chat_template is None:
raise ValueError("This model does not support chat_template.")
task = request.to_dict()
task["enable_thinking"] = kwargs.get("enable_thinking", True)
chat_template_kwargs = kwargs.get("chat_template_kwargs")
if chat_template_kwargs:
if isinstance(chat_template_kwargs, dict):
for k, v in chat_template_kwargs.items():
if k not in task:
task[k] = v
else:
raise ValueError("Invalid input: chat_template_kwargs must be a dict")
task.setdefault("enable_thinking", True)
request.prompt_token_ids = self.messages2ids(task)
else:
raise ValueError(f"The request should have `input_ids`, `text` or `messages`: {request}.")
@@ -272,6 +279,15 @@ class DataProcessor(BaseDataProcessor):
elif "messages" in request:
if self.tokenizer.chat_template is None:
raise ValueError("This model does not support chat_template.")
chat_template_kwargs = request.get("chat_template_kwargs")
if chat_template_kwargs:
if isinstance(chat_template_kwargs, dict):
for k, v in chat_template_kwargs.items():
if k not in request:
request[k] = v
else:
raise ValueError("Invalid input: chat_template_kwargs must be a dict")
request.setdefault("enable_thinking", True)
request["prompt_token_ids"] = self.messages2ids(request)
else:
raise ValueError(f"Request must contain 'prompt_token_ids', 'prompt', or 'messages': {request}")