mirror of
https://github.com/volcengine/verl.git
synced 2025-10-20 13:43:50 +08:00
7aabfc437b
### What does this PR do? This is an initial effort to integrate LangGraph into agent loop: 1. add a LangGraph react agent loop implementation 2. add math expression example to demonstrate react agent loop usage. ### Design & Code Changes New components - ChatModel: [custom chat model](https://python.langchain.com/docs/how_to/custom_chat_model/) using LangChain abstractions, implementing following abstract method: - bind_tools: bind tools to the model - _generate: native async generate chat completion message - ReactAgentLoop: [LangGraph react agent](https://langchain-ai.github.io/langgraph/agents/overview/) which can use tools to perform tasks. <img width="593" height="467" alt="image" src="https://github.com/user-attachments/assets/d629b170-03c5-4810-a6b0-4dc27a285c0e" /> ### Checklist Before Submitting > [!IMPORTANT] > Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review. - [ ] Read the [Contribute Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md). - [ ] Apply [pre-commit checks](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md#code-linting-and-formatting): `pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always` - [ ] Add / Update [the documentation](https://github.com/volcengine/verl/tree/main/docs). - [ ] Add unit or end-to-end test(s) to [the CI workflow](https://github.com/volcengine/verl/tree/main/.github/workflows) to cover all the code. If not feasible, explain why: ... - [ ] Once your PR is ready for CI, send a message in [the `ci-request` channel](https://verl-project.slack.com/archives/C091TCESWB1) in [the `verl` Slack workspace](https://join.slack.com/t/verl-project/shared_invite/zt-3855yhg8g-CTkqXu~hKojPCmo7k_yXTQ).
40 lines
1.3 KiB
Python
40 lines
1.3 KiB
Python
# Copyright 2024 Bytedance Ltd. and/or its affiliates
|
|
#
|
|
# 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 langchain_core.tools import tool
|
|
|
|
from recipe.langgraph_agent.react_agent_loop import ReactAgentLoop
|
|
|
|
|
|
@tool(parse_docstring=True)
|
|
def calculate(a: int, b: int, operand: str) -> int:
|
|
"""
|
|
Compute the results using operand with two integers
|
|
|
|
Args:
|
|
a: the first operand
|
|
b: the second operand
|
|
operand: '+' or '-' or '*' or '@'
|
|
"""
|
|
assert operand in ["+", "-", "*", "@"], f"unknown operand {operand}"
|
|
if operand == "@":
|
|
return 3 * a - 2 * b
|
|
return eval(f"{a} {operand} {b}")
|
|
|
|
|
|
class MathExpressionReactAgentLoop(ReactAgentLoop):
|
|
@classmethod
|
|
def init_class(cls, config, tokenizer, **kwargs):
|
|
cls.tools = [calculate]
|
|
super().init_class(config, tokenizer)
|