mirror of
https://github.com/InternLM/InternBootcamp.git
synced 2026-04-22 16:49:04 +00:00
255 lines
9.2 KiB
Python
Executable file
255 lines
9.2 KiB
Python
Executable file
"""# 谜题训练场开发任务
|
||
|
||
## 任务概述
|
||
你是一位资深程序员，我需要你帮我实现一个特定谜题的训练场环境类。这个类继承自`Basebootcamp`，用于生成谜题实例并验证解答。
|
||
|
||
## 背景说明
|
||
我正在开发一系列谜题训练场，每个训练场对应一个特定类型的谜题。训练场类命名为`{PuzzleName}bootcamp`，其中`PuzzleName`是谜题的名称。
|
||
|
||
每个训练场类主要提供两个核心功能：
|
||
1. 生成该谜题类型的问题实例
|
||
2. 验证用户对问题的回答是否正确
|
||
|
||
## 技术接口规范
|
||
|
||
### 类方法实现要求
|
||
|
||
```python
|
||
class {PuzzleName}bootcamp(Basebootcamp):
|
||
def __init__(self, **params):
|
||
\"\"\"
|
||
请你自定义params，以保存该puzzle相关的参数，例如网格大小等，参数配有默认值
|
||
\"\"\"
|
||
pass
|
||
|
||
def case_generator(self):
|
||
\"\"\"
|
||
生成谜题实例，提示：为保证谜题有解，可以先生成结果再对结果处理得到谜题
|
||
返回：一个可JSON序列化的字典（避免包含set等无法通过json.dumps处理的数据结构）
|
||
\"\"\"
|
||
pass
|
||
|
||
@staticmethod
|
||
def prompt_func(question_case) -> str:
|
||
\"\"\"
|
||
将case_generator生成的谜题实例转换为文本形式的问题，问题中包含问题背景、对谜题规则的介绍、具体要解决的谜题实例、期望最终答案的格式，
|
||
例如：你是xxxx，请你解答yyyy，规则如下：yyyy，最终答案放置在：zzzzz
|
||
|
||
参数:
|
||
question_case: 由case_generator生成的谜题实例
|
||
|
||
返回:
|
||
str: 格式化的问题字符串
|
||
|
||
注意:
|
||
1. 需考虑问题的格式，以便后续能正确提取
|
||
2. 问题描述中应包含期望的答案格式说明，以便后续能正确提取，为了避免抽取时匹配出干扰项，请要求模型将答案放在特定标签，如[answer] [/answer]内
|
||
\"\"\"
|
||
pass
|
||
|
||
@staticmethod
|
||
def extract_output(output):
|
||
\"\"\"
|
||
从LLM的回复中提取符合格式要求的答案，如有多个，请抽取最后一个，避免使用re.search等只抽取第一个结果的方式。
|
||
|
||
参数:
|
||
output: LLM的完整输出（包含原始问题和回答）
|
||
|
||
返回:
|
||
提取的答案，若未找到符合格式的答案则返回None
|
||
\"\"\"
|
||
pass
|
||
|
||
@classmethod
|
||
def _verify_correction(cls, solution, identity):
|
||
\"\"\"
|
||
验证提取的答案是否正确，注意一个问题可以能有多个解，按照谜题规则进行检验，不要直接匹配可能的答案。
|
||
|
||
参数:
|
||
solution: extract_output提取的答案
|
||
identity: case_generator生成的谜题实例
|
||
|
||
返回:
|
||
bool: 答案是否正确
|
||
\"\"\"
|
||
pass
|
||
```
|
||
|
||
### 验证评分方法（基类已实现）
|
||
|
||
```python
|
||
@classmethod
|
||
def verify_score(cls, model_output, identity:dict, format_score=0.1) -> float:
|
||
\"\"\"
|
||
验证输出结果并评分。
|
||
|
||
参数:
|
||
model_output: 模型的完整输出
|
||
identity: 谜题实例（由case_generator生成）
|
||
format_score: 答案格式正确时的基础分数
|
||
|
||
返回:
|
||
float: 评分结果（0-1之间）
|
||
\"\"\"
|
||
score = 0.
|
||
try:
|
||
extract_solution = cls.extract_output(model_output)
|
||
if extract_solution is None:
|
||
return score
|
||
else:
|
||
score = format_score # 格式正确时的基础分数
|
||
if cls._verify_correction(extract_solution, identity):
|
||
score = 1. # 答案完全正确时的满分
|
||
except Exception as e:
|
||
# 处理异常情况
|
||
pass
|
||
return score
|
||
```
|
||
|
||
### 使用示例
|
||
|
||
```python
|
||
# 初始化谜题训练场
|
||
bootcamp = Puzzlebootcamp()
|
||
|
||
# 生成谜题实例
|
||
case = bootcamp.case_generator()
|
||
|
||
# 将谜题转换为文本问题
|
||
prompt = Puzzlebootcamp.prompt_func(case)
|
||
|
||
# 获取LLM对问题的解答
|
||
response = get_response(prompt, \"LLM\")
|
||
|
||
# 从完整对话中提取答案
|
||
extracted_output = Puzzlebootcamp.extract_output(prompt + response)
|
||
|
||
# 验证答案并评分
|
||
score = Puzzlebootcamp.verify_score(extracted_output, case)
|
||
```
|
||
|
||
## 你的任务
|
||
请根据以下谜题描述（谜题描述可能不完整，请先结合你的知识澄清规则），实现一个完整的谜题训练场类：
|
||
|
||
### 谜题描述
|
||
|
||
**Objective**: Clear a rectangular grid of hidden cells without detonating any mines. Cells contain either mines or numbers indicating adjacent mines.
|
||
|
||
**Grid Setup**:
|
||
1. The grid consists of hidden cells, some containing mines (randomly placed).
|
||
2. Non-mine cells reveal a number when uncovered, representing the total mines in the 8 adjacent cells (vertically, horizontally, and diagonally).
|
||
|
||
**Gameplay**:
|
||
1. **Uncover a Cell**: Click/select a cell to reveal its content:
|
||
- If it contains a mine, the game ends (loss).
|
||
- If it shows a number, use this to deduce nearby mine locations.
|
||
- If it shows **0** (no adjacent mines), all adjacent cells automatically uncover recursively until numbered cells are reached.
|
||
|
||
2. **Flagging Mines**: Right-click/mark a cell to flag it as a suspected mine (prevents accidental uncovering). Flags help track potential mines but do not affect gameplay logic otherwise.
|
||
|
||
**Win Condition**:
|
||
- All non-mine cells are uncovered, and all mines are correctly flagged.
|
||
|
||
**Logic Rules**:
|
||
- Numbers on the grid are **static hints**, not live updates. Flagging a mine does not change existing numbers.
|
||
- Use the numbers to infer mine positions: e.g., a cell labeled \"3\" must have exactly 3 mines in its 8 neighboring cells.
|
||
|
||
**Loss Condition**:
|
||
- Uncovering any cell containing a mine.
|
||
|
||
|
||
请完成上述谜题的训练场环境类实现，包括所有必要的方法。
|
||
"""
|
||
|
||
from bootcamp import Basebootcamp
|
||
import random
|
||
import re
|
||
import ast
|
||
|
||
class Minesweeperbootcamp(Basebootcamp):
|
||
def __init__(self, rows=8, cols=8, mines_count=10):
|
||
if mines_count > rows * cols:
|
||
raise ValueError("Number of mines cannot exceed grid size.")
|
||
self.rows = rows
|
||
self.cols = cols
|
||
self.mines_count = mines_count
|
||
|
||
def case_generator(self):
|
||
all_cells = [(i, j) for i in range(self.rows) for j in range(self.cols)]
|
||
if self.mines_count > len(all_cells):
|
||
raise ValueError("mines_count exceeds valid cells count.")
|
||
mines = random.sample(all_cells, self.mines_count)
|
||
mines_list = [list(coord) for coord in mines]
|
||
return {
|
||
'rows': self.rows,
|
||
'cols': self.cols,
|
||
'mines': mines_list
|
||
}
|
||
|
||
@staticmethod
|
||
def prompt_func(question_case):
|
||
rows = question_case['rows']
|
||
cols = question_case['cols']
|
||
mines_count = len(question_case['mines'])
|
||
mines_set = set(tuple(coord) for coord in question_case['mines'])
|
||
|
||
grid_info = []
|
||
for i in range(rows):
|
||
for j in range(cols):
|
||
count = 0
|
||
for dx in (-1, 0, 1):
|
||
for dy in (-1, 0, 1):
|
||
if dx == 0 and dy == 0:
|
||
continue
|
||
x, y = i + dx, j + dy
|
||
if 0 <= x < rows and 0 <= y < cols and (x, y) in mines_set:
|
||
count += 1
|
||
grid_info.append((i, j, count))
|
||
|
||
prompt = (
|
||
f"You are playing Minesweeper on a {rows}x{cols} grid with {mines_count} mines.\n"
|
||
"Each number below represents the count of adjacent mines for a cell. "
|
||
"Find all the mine locations and provide them in the specified format.\n\n"
|
||
"Revealed cells (format: row, column: count):\n"
|
||
)
|
||
for i, j, num in grid_info:
|
||
prompt += f"- ({i}, {j}): {num}\n"
|
||
prompt += (
|
||
"\nYour answer must be a list of mine coordinates formatted as [[row1, col1], [row2, col2], ...]. "
|
||
"Place your final answer between [answer] and [/answer]."
|
||
)
|
||
return prompt
|
||
|
||
@staticmethod
|
||
def extract_output(output):
|
||
pattern = r'\[answer\](.*?)\[/answer\]'
|
||
matches = re.findall(pattern, output, re.DOTALL)
|
||
if not matches:
|
||
return None
|
||
last_match = matches[-1].strip()
|
||
try:
|
||
solution = ast.literal_eval(last_match)
|
||
if (isinstance(solution, list) and
|
||
all(isinstance(coord, list) and len(coord) == 2 for coord in solution)):
|
||
return solution
|
||
return None
|
||
except (SyntaxError, ValueError):
|
||
return None
|
||
|
||
@classmethod
|
||
def _verify_correction(cls, solution, identity):
|
||
if not isinstance(solution, list):
|
||
return False
|
||
try:
|
||
solution_set = {tuple(coord) for coord in solution}
|
||
except TypeError:
|
||
return False
|
||
mines = identity.get('mines', [])
|
||
mines_set = {tuple(mine) for mine in mines}
|
||
if len(solution_set) != len(mines_set):
|
||
return False
|
||
rows, cols = identity['rows'], identity['cols']
|
||
for (r, c) in solution_set:
|
||
if not (0 <= r < rows and 0 <= c < cols):
|
||
return False
|
||
return solution_set == mines_set
|