mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-10-06 00:57:33 +08:00
Add token processor plugin support (#4059)
* Add token processor plugin support * fix import * fix import
This commit is contained in:
@@ -42,10 +42,16 @@ from fastdeploy.inter_communicator import (
|
||||
from fastdeploy.metrics.metrics import main_process_metrics
|
||||
from fastdeploy.metrics.trace_util import start_span, start_span_request
|
||||
from fastdeploy.model_executor.guided_decoding import schema_checker
|
||||
from fastdeploy.output.token_processor import TokenProcessor
|
||||
from fastdeploy.plugins.token_processor import load_token_processor_plugins
|
||||
from fastdeploy.splitwise.splitwise_connector import SplitwiseConnector
|
||||
from fastdeploy.utils import EngineError, envs, llm_logger
|
||||
|
||||
try:
|
||||
TokenProcessor = load_token_processor_plugins()
|
||||
llm_logger.info(f"TokenProcessor plugin {TokenProcessor} loaded")
|
||||
except:
|
||||
from fastdeploy.output.token_processor import TokenProcessor
|
||||
|
||||
|
||||
class EngineService:
|
||||
"""
|
||||
|
@@ -18,10 +18,12 @@ from .input_processor import load_input_processor_plugins
|
||||
from .model_register import load_model_register_plugins
|
||||
from .model_runner import load_model_runner_plugins
|
||||
from .reasoning_parser import load_reasoning_parser_plugins
|
||||
from .token_processor import load_token_processor_plugins
|
||||
|
||||
__all__ = [
|
||||
"load_model_register_plugins",
|
||||
"load_model_runner_plugins",
|
||||
"load_input_processor_plugins",
|
||||
"load_reasoning_parser_plugins",
|
||||
"load_token_processor_plugins",
|
||||
]
|
||||
|
27
fastdeploy/plugins/token_processor/__init__.py
Normal file
27
fastdeploy/plugins/token_processor/__init__.py
Normal file
@@ -0,0 +1,27 @@
|
||||
"""
|
||||
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License"
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
"""
|
||||
|
||||
from fastdeploy.plugins.utils import load_plugins_by_group
|
||||
|
||||
# make sure one process only loads plugins once
|
||||
PLUGINS_GROUP = "fastdeploy.token_processor_plugins"
|
||||
|
||||
|
||||
def load_token_processor_plugins():
|
||||
"""load_token_processor_plugins"""
|
||||
plugins = load_plugins_by_group(group=PLUGINS_GROUP)
|
||||
assert len(plugins) <= 1, "Most one plugin is allowed to be loaded."
|
||||
return next(iter(plugins.values()))()
|
Reference in New Issue
Block a user