mirror of
https://github.com/InternLM/InternBootcamp.git
synced 2026-04-24 17:05:00 +00:00
init-commit
This commit is contained in:
commit
18a552597a
3461 changed files with 1150579 additions and 0 deletions
319
internbootcamp/bootcamp/kor_puzzle_crypto_math/kor_puzzle_crypto_math.py
Executable file
319
internbootcamp/bootcamp/kor_puzzle_crypto_math/kor_puzzle_crypto_math.py
Executable file
|
|
@ -0,0 +1,319 @@
|
|||
"""# 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
|
||||
from bootcamp import Basebootcamp
|
||||
|
||||
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
|
||||
注意:请参照提供的谜题描述进行复述,规则应当描述详细,包括任务背景、具体任务操作规则、对题目格式和答案格式的含义介绍等,
|
||||
|
||||
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(如双括号)内,例如[[your answer here]]
|
||||
\"\"\"
|
||||
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
|
||||
1.The game gives a formula of letters, each representing a unique number (0-9).
|
||||
2.Different letters cannot represent the same number.
|
||||
3.The first letter of any multi-digit number cannot represent 0.Example questions are as follows:
|
||||
|
||||
<example 0>
|
||||
SEND+MORE=MONEY.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 0>
|
||||
|
||||
<example 1>
|
||||
TO+GO=OUT.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 1>
|
||||
|
||||
<example 2>
|
||||
ONE+ONE+TWO=FOUR.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 2>
|
||||
|
||||
<example 3>
|
||||
TT+TT=LTC
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 3>
|
||||
|
||||
<example 4>
|
||||
FORTY+TEN+TEN=SIXTY.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 4>
|
||||
|
||||
<example 5>
|
||||
FIFTY+STATES=AMERICA.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 5>
|
||||
|
||||
<example 6>
|
||||
I+BB=ILL.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 6>
|
||||
|
||||
<example 7>
|
||||
EAT+THAT=APPLE.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 7>
|
||||
|
||||
<example 8>
|
||||
HERE+SHE=COMES.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 8>
|
||||
|
||||
<example 9>
|
||||
NUMBER+NUMBER=PUZZLE.
|
||||
Please provide your answer in the form: letter=number, and make sure to enclose your answer in double square brackets, like this: [[A=1,B=2,...]].
|
||||
</example 9>
|
||||
|
||||
|
||||
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 re
|
||||
import random
|
||||
from itertools import permutations
|
||||
from bootcamp import Basebootcamp
|
||||
|
||||
class KorPuzzleCryptoMathbootcamp(Basebootcamp):
|
||||
def __init__(self, **params):
|
||||
"""
|
||||
参数:
|
||||
min_terms: 加数最小数量 (默认2)
|
||||
max_terms: 加数最大数量 (默认3)
|
||||
term_length: 项的数字位数 (默认3)
|
||||
result_length: 结果的数字位数 (默认4)
|
||||
"""
|
||||
self.min_terms = params.get('min_terms', 2)
|
||||
self.max_terms = params.get('max_terms', 3)
|
||||
self.term_length = params.get('term_length', 3)
|
||||
self.result_length = params.get('result_length', 4)
|
||||
|
||||
def _generate_valid_equation(self):
|
||||
"""动态生成有效等式"""
|
||||
# 生成随机加法结构:A + B + ... = SUM
|
||||
num_terms = random.randint(self.min_terms, self.max_terms)
|
||||
|
||||
while True:
|
||||
# 生成随机数字组合
|
||||
digits = random.sample(range(0, 10), self.term_length)
|
||||
terms = [random.randint(10**(self.term_length-1), 10**self.term_length-1)
|
||||
for _ in range(num_terms)]
|
||||
total = sum(terms)
|
||||
|
||||
if len(str(total)) == self.result_length:
|
||||
# 转换为字母模式
|
||||
letters = set()
|
||||
equation_parts = []
|
||||
for term in terms + [total]:
|
||||
term_str = str(term)
|
||||
if len(term_str) < self.term_length:
|
||||
term_str = term_str.zfill(self.term_length)
|
||||
equation_parts.append(term_str)
|
||||
letters.update(term_str)
|
||||
|
||||
# 确保结果首位非零
|
||||
if equation_parts[-1][0] == '0':
|
||||
continue
|
||||
|
||||
# 转换为字母方程
|
||||
char_map = {}
|
||||
unique_chars = list(letters)
|
||||
random.shuffle(unique_chars)
|
||||
for c in unique_chars:
|
||||
char_map[c] = chr(65 + len(char_map)) # 映射到不同字母
|
||||
|
||||
equation = []
|
||||
for part in equation_parts[:-1]:
|
||||
equation.append(''.join([char_map[c] for c in part]))
|
||||
result = ''.join([char_map[c] for c in equation_parts[-1]])
|
||||
|
||||
return f"{'+'.join(equation)}={result}"
|
||||
|
||||
def case_generator(self):
|
||||
"""生成动态有效的字母算术方程"""
|
||||
equation = self._generate_valid_equation()
|
||||
return {"equation": equation}
|
||||
|
||||
@staticmethod
|
||||
def prompt_func(question_case) -> str:
|
||||
equation = question_case["equation"]
|
||||
prompt = f"""Solve this cryptarithmetic puzzle where each letter represents a unique digit (0-9).
|
||||
Different letters must have different values. Leading letters cannot be zero.
|
||||
|
||||
Equation: {equation}
|
||||
|
||||
Provide your answer as comma-separated letter=number pairs enclosed in double square brackets.
|
||||
Example: [[A=5,B=3,...,Z=9]]"""
|
||||
return prompt
|
||||
|
||||
@staticmethod
|
||||
def extract_output(output):
|
||||
matches = re.findall(r'\[\[(.*?)\]\]', output)
|
||||
if not matches:
|
||||
return None
|
||||
|
||||
solution = {}
|
||||
for pair in matches[-1].split(','):
|
||||
pair = pair.strip()
|
||||
if '=' not in pair:
|
||||
continue
|
||||
letter, value = pair.split('=', 1)
|
||||
letter = letter.strip().upper()
|
||||
try:
|
||||
num = int(value.strip())
|
||||
if 0 <= num <= 9:
|
||||
solution[letter] = num
|
||||
except ValueError:
|
||||
continue
|
||||
return solution if solution else None
|
||||
|
||||
@classmethod
|
||||
def _verify_correction(cls, solution, identity):
|
||||
equation = identity["equation"]
|
||||
try:
|
||||
left, right = equation.split('=')
|
||||
terms = left.split('+')
|
||||
sum_terms = 0
|
||||
for term in terms:
|
||||
if len(term) > 1 and solution[term[0]] == 0:
|
||||
return False
|
||||
sum_terms += int(''.join(str(solution[c]) for c in term))
|
||||
|
||||
result = right.strip()
|
||||
if len(result) > 1 and solution[result[0]] == 0:
|
||||
return False
|
||||
result_num = int(''.join(str(solution[c]) for c in result))
|
||||
|
||||
# 验证唯一性
|
||||
values = list(solution.values())
|
||||
if len(values) != len(set(values)):
|
||||
return False
|
||||
|
||||
return sum_terms == result_num
|
||||
except (KeyError, ValueError):
|
||||
return False
|
||||
Loading…
Add table
Add a link
Reference in a new issue