AgentVerse's picture
bump version to 0.1.8
01523b5
cnt_agents: &cnt_agents 2
max_turn: &max_turn 3
max_criticizing_rounds: 3
prompts:
role_assigner_prepend_prompt: &role_assigner_prepend_prompt |-
# Role Description
You are the leader of a group, now you are facing a problem:
```
${task_description}
```
You can recruit ${cnt_critic_agents} people to solve the logic problem. What people will you recruit?
Here are some suggestion:
${advice}
role_assigner_append_prompt: &role_assigner_append_prompt |-
# Response Format Guidance
You should respond with a list of ${cnt_critic_agents} people description. For example:
1. an electrical engineer specified in the filed of xxx
2. an economist who is good at xxx
3. a lawyer with a good knowledge of xxx
...
Only respond with the description of each role. Do not include your reason.
solver_prepend_prompt: &solver_prepend_prompt |-
${task_description}
# Messages from the solver and critics will be filled here in the code.
solver_append_prompt: &solver_append_prompt |-
Using these information, can you provide the correct solution to the math problem? Explain your reasoning and solve the problem step by step. Your final answer should be a single integer, which is the number of choice, in the form \boxed{answer}, at the end of your response.
critic_prepend_prompt: &critic_prepend_prompt |-
You are in a discussion group, aiming to collaborative solve the following logic problem:
```
${task_description}
```
# Messages from the solver and critics will be filled here in the code.
critic_append_prompt: &critic_append_prompt |-
You are ${role_description}. Based on your knowledge, can you check the correctness of the solutions given above? You should give your correct solution to the problem step by step. When responding, you should follow the following rules:
1. Double-check the above solutions, give your critics, then generate the correct solution step by step.
2. If the final answer in your solution is the same as the final answer in the above provided solution, end your response with a special token "[Agree]".
3. You must highlight your final answer in the form \boxed{answer} at the end of your response. The answer must be a single integer.
Now give your response.
evaluator_prepend_prompt: &evaluator_prepend_prompt |-
Problem:
```
${task_description}
```
Solution:
```
${solution}
```
You are a logic problem lover. Above is a logic problem and a solution. Check whether the solution and the deduction is correct. If the deduction is wrong, you should explain why it is wrong, but do not give your solution. When it is correct, output a correctness of 1 and why it is correct.
evaluator_append_prompt: &evaluator_append_prompt |-
You should respond in the following format:
Correctness: (0 or 1, 0 is wrong, and 1 is correct)
Response: (explain in details why it is wrong or correct. do not provide your solution)
name: pipeline
environment:
env_type: task-basic
max_turn: *max_turn
rule:
role_assigner:
type: role_description
cnt_agents: *cnt_agents
decision_maker:
type: vertical-solver-first
executor:
type: none
evaluator:
type: basic
agents:
- #role_assigner_agent:
agent_type: role_assigner
name: role assigner
max_retry: 1000
prepend_prompt_template: *role_assigner_prepend_prompt
append_prompt_template: *role_assigner_append_prompt
memory:
memory_type: chat_history
llm:
llm_type: gpt-4
model: gpt-4
temperature: 0
max_tokens: 512
output_parser:
type: role_assigner
- #solver_agent:
agent_type: solver
name: Planner
max_retry: 1000
prepend_prompt_template: *solver_prepend_prompt
append_prompt_template: *solver_append_prompt
memory:
memory_type: chat_history
llm:
llm_type: gpt-4
model: gpt-4
temperature: 0
max_tokens: 1024
output_parser:
type: mgsm
- #critic_agents:
agent_type: critic
name: Critic 1
max_retry: 1000
role_description: |-
Waiting to be assigned.
prepend_prompt_template: *critic_prepend_prompt
append_prompt_template: *critic_append_prompt
memory:
memory_type: chat_history
llm:
llm_type: gpt-4
model: "gpt-4"
temperature: 0
max_tokens: 1024
output_parser:
type: mgsm-critic-agree
- #executor_agent:
agent_type: executor
name: Executor
max_retry: 1000
memory:
memory_type: chat_history
llm:
llm_type: gpt-4
model: gpt-4
temperature: 0
max_tokens: 1024
output_parser:
type: mgsm
- #evaluator_agent:
agent_type: evaluator
name: Evaluator
max_retry: 1000
role_description: |-
Evaluator
prepend_prompt_template: *evaluator_prepend_prompt
append_prompt_template: *evaluator_append_prompt
memory:
memory_type: chat_history
llm:
llm_type: gpt-4
model: gpt-4
temperature: 0.3
max_tokens: 1024
output_parser:
type: mgsm-evaluator
dimensions:
- Correctness
tools: