File size: 5,038 Bytes
01f199c | 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 | from typing import List
import tiktoken
import os
import copy
import time
from models.Base import BaseModel
from datasets.Dataset import Dataset
from results.Results import Results
from utils.parse import parse_response
class BaseStrategy(object):
def __init__(
self,
model: BaseModel,
data: Dataset,
language: str,
pass_at_k: int,
results: Results,
verbose: bool = True,
):
self.model = model
self.data = data
self.pass_at_k = pass_at_k
self.results = results
self.language = language
self.verbose = verbose
def gpt_chat(self, processed_input: List[dict]) -> (str, int, int):
return self.model.prompt(processed_input=processed_input)
def run_single_pass(self, item: dict):
pass
def run(self):
num_items = len(self.data)
num_success = 0
for i, item in enumerate(self.data):
print("", flush=True, end="")
# if i < len(self.results):
# is_passing = self.results[i]["is_solved"]
# """
# if not is_passing:
# for response in self.results[i]["source_codes"]:
# cur_imp = response
# # parse_response(
# # response,
# # item["entry_point"]
# # )
# is_passing = self.data.evaluate(
# item=item,
# cur_imp=cur_imp,
# language=self.language
# )
# if is_passing:
# break
# """
# if is_passing:
# num_success += 1
# if self.verbose:
# print(f'completed {i+1}/{num_items}, Solved: {is_passing}, number of success = {num_success}/{i+1}, acc = {round(num_success/(i+1)*100, 2)}')
# continue
if i < len(self.results):
item = copy.deepcopy(self.results[i])
cur_pass = len(item["source_codes"])
is_solved = item["is_solved"]
# if not item["source_codes"]:
# print(f"Error Debug: item content -> {item}")
cur_imp = item["source_codes"][-1]
# if item["source_codes"]:
# # 如果有代码，取最新生成的
# cur_imp = item["source_codes"][-1]
# else:
# # 如果列表为空（生成失败），打印警告并跳过
# print(f"⚠️ Warning: Task {item.get('task_id')} has no source codes. Skipping generation check.")
# cur_imp = "" # 设置为空字符串，避免后面报错
else:
item = copy.deepcopy(item)
item["source_codes"] = []
item["responses"] = []
item["prompt_tokens"] = []
item["completion_tokens"] = []
item["no_of_try"] = 0
cur_pass = 0
is_solved = False
cur_imp = ""
while cur_pass < self.pass_at_k and not is_solved:
try:
response, prompt_tokens, completion_tokens = self.run_single_pass(
item)
except Exception as e:
print(f"Error processing item {item.get('task_id', 'unknown')}: {e}")
cur_pass += 1
continue
if hasattr(self, "parse_code"):
cur_imp = self.parse_code(response)
else:
cur_imp = parse_response(response)
# cur_imp = parse_response(response, item.get("entry_point", None))
item["source_codes"].append(cur_imp)
item["responses"].append(response)
item["prompt_tokens"].append(prompt_tokens)
item["completion_tokens"].append(completion_tokens)
item["no_of_try"] += 1
is_solved = self.data.evaluate(
item=item,
cur_imp=cur_imp,
language=self.language
)
cur_pass += 1
if is_solved:
num_success += 1
item["is_solved"] = is_solved
item["language"] = self.language
item["task_id"] = item[self.data.id_key]
if i < len(self.results):
self.results.results[i] = item
self.results.save_results()
else:
self.results.add_result(item)
if self.verbose:
print(
f'completed {i+1}/{num_items}, Solved: {self.results[i]["is_solved"]}, number of success = {num_success}/{i+1}, acc = {round(num_success/(i+1)*100, 2)}')
# break
|