Spaces:
Build error
Build error
File size: 6,502 Bytes
01523b5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 |
cnt_agents: &cnt_agents 4
max_turn: &max_turn 3
max_inner_turns: &max_inner_turns 2
task_description: |-
write a simple calculator GUI using Python3.
prompts:
role_assigner_prepend_prompt: &role_assigner_prepend_prompt |-
role_assigner_append_prompt: &role_assigner_append_prompt |-
You are the leader of a group of experts, now you are faced with a task:
${task_description}
You can recruit ${cnt_critic_agents} expert team members in different regions.
What experts will you recruit to better generate good ideas?
Output format 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
...
${advice}
You don't have to give the reason.
solver_prepend_prompt: &solver_prepend_prompt |-
You are faced with the task:
${task_description}
Below is the chat history among you and other teammates.
solver_append_prompt: &solver_append_prompt |-
Now you are going to give a new solution, based upon your former solution and the critics' opinions. Write the code step by step.
critic_prepend_prompt: &critic_prepend_prompt |-
Now you are ${role_description}
You are in a discussion group, aiming to ${task_description}.
Below is the chat history among you and other teammates.
critic_append_prompt: &critic_append_prompt |-
Now the group is asking your opinion about it. Based on your knowledge
in your field, do you agree that this solution can perfectly
solve the problem?
Or do you have any ideas to improve it?
- If you thinks it is perfect, use the following output format:
Action: Agree
Action Input: Agree.
(Do not output your reason for agreeing!)
- If you want to give complemented opinions to improve it or to contradict with it, use the following output format:
Action: Disagree
Action Input: (what you want to say in one line)
P.S. Always remember you are ${role_description}!
If no former solution or critic opinions are given, you can just disagree and output your idea freely, based on the expertise of your role.
Remember, the ideas should be specific and detailed enough, not just general opinions.
Please control output code in 2048 tokens! (Write concise code)
evaluator_prepend_prompt: &evaluator_prepend_prompt |-
evaluator_append_prompt: &evaluator_append_prompt |-
You are an professional and strict code reviewer. Your task is to evaluate the solution. The code is to ${task_description}. Your task is to evaluate the codes written by the code engineers.
Please not only give a general rating points (from 0 to 9) but also give detailed comments about where and how the code can be improved. Please consider the following aspects when you are evaluating the code:
1. The code should be able to run without any errors.
2. The code should be able to achieve the goal specified in the task description.
3. The code should be easy to read and understand, efficient, concise and elegant.
4. The code should be robust.
Please rate the code in the following dimensions:
1. Completeness: Is the code snippet complete enough without unimplemented functions of methods? Is it able to run without any errors?
2. Functionality: Is the code able to achieve the goal specified in the task description?
3. Readability: Is the code easy to read and understand, efficient, concise and elegant?
4. Robustness: Is the code snippet able to handle different unexpected input or other exceptions?
In the fifth line of output, give your detailed advice for the engineers to better generate good codes. Only give high grade when it is really great.
# Output format
1. Completeness: (a score between 0 and 9)
2. Functionality: (a score between 0 and 9)
3. Readability: (a score between 0 and 9)
4. Robustness: (a score between 0 and 9)
5. Advice: (your advice in one line)
Here is the content you have to evaluate:
${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
max_inner_turns: *max_inner_turns
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: 0.7
max_tokens: 256
output_parser:
type: role_assigner
- #solver_agent:
agent_type: solver
name: Summarizer
max_retry: 1000
max_history: 5
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: 0.7
max_tokens: 2048
output_parser:
type: dummy
- #critic_agents:
agent_type: critic
name: Reviewer
max_retry: 1000
max_history: 5
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: 0.7
max_tokens: 1024
output_parser:
type: critic
- #executor_agent:
agent_type: executor
name: Dummy Executor
max_retry: 1000
memory:
memory_type: chat_history
llm:
llm_type: gpt-4
model: "gpt-4"
temperature: 0.7
max_tokens: 512
output_parser:
type: dummy
- #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: 0.7
max_tokens: 1024
output_parser:
type: evaluator
dimensions:
- Completeness
- Functionality
- Readability
- Robustness
|