File size: 5,852 Bytes
40e0258 |
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 |
"""
curl -X GET http://localhost:7680/api/models
curl -X POST http://127.0.0.1:7680/hf/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"prompt": "你是什么模型?"
}'
"""
import json
import uuid # 用于生成随机用户 ID
import requests
from flask import Flask, request, Response
app = Flask(__name__)
# 全局字典用于保存用户的上下文对话
user_contexts = {}
MAX_HISTORY_LENGTH = 15 # 最大上下文历史长度
def get_models():
models = {
"object": "list",
"data": [
{"id": "Qwen2.5-72B", "object": "model", "created": 0, "owned_by": "Qwen"},
{"id": "Llama-3.1-Nemotron-70B", "object": "model", "created": 0, "owned_by": "Nemotron"},
{"id": "NVLM-D-72B", "object": "model", "created": 0, "owned_by": "NVDIA"},
{"id": "DeepSeek-Coder-V2", "object": "model", "created": 0, "owned_by": "DeepSeek"},
{"id": "Qwen2.5-Coder-32B", "object": "model", "created": 0, "owned_by": "Qwen"},
]
}
return json.dumps(models)
def chat_completion(
user_prompt, user_id: str = None, system_prompt="You are a helpful assistant.", model="Qwen2.5-72B",
project="DecentralGPT", stream=False, temperature=0.3, max_tokens=1024, top_p=0.5,
frequency_penalty=0, presence_penalty=0):
"""处理用户请求并保留上下文"""
url = 'https://usa-chat.degpt.ai/api/v0/chat/completion/proxy'
headers = {
'accept': 'application/json',
'accept-language': 'en-US,en;q=0.9,zh-CN;q=0.8,zh;q=0.7',
'content-type': 'application/json',
'dnt': '1',
'origin': 'https://www.degpt.ai',
'priority': 'u=1, i',
'referer': 'https://www.degpt.ai/',
'sec-ch-ua': 'Google Chrome";v="131", "Chromium";v="131", "Not_A Brand";v="24"',
'sec-ch-ua-mobile': '?0',
'sec-ch-ua-platform': '"macOS"',
'sec-fetch-dest': 'empty',
'sec-fetch-mode': 'cors',
'sec-fetch-site': 'same-site',
'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36'
}
# 初始化或更新用户的对话历史
if user_id is not None:
if user_id not in user_contexts:
user_contexts[user_id] = [{"role": "system", "content": system_prompt}]
user_contexts[user_id].append({"role": "user", "content": user_prompt})
# 检查是否需要修剪历史记录，保留 `system` 提示词
while len(user_contexts[user_id]) > MAX_HISTORY_LENGTH:
# 删除最早的用户问题和系统回复，但保留 `system` 提示词
if len(user_contexts[user_id]) > 2:
# 检查删除的条目是否有匹配的系统回复，如果没有，只删除用户输入
if user_contexts[user_id][2]["role"] == "user":
user_contexts[user_id] = [user_contexts[user_id][0]] + user_contexts[user_id][2:]
else:
user_contexts[user_id] = [user_contexts[user_id][0]] + user_contexts[user_id][2:]
else:
break
messages = user_contexts[user_id]
else:
# 如果没有提供 user_id，不保留上下文
messages = [{"role": "system", "content": system_prompt}, {"role": "user", "content": user_prompt}]
payload = {
"model": model,
"messages": messages,
"project": project,
"stream": stream,
"temperature": temperature,
"max_tokens": max_tokens,
"top_p": top_p,
"frequency_penalty": frequency_penalty,
"presence_penalty": presence_penalty
}
try:
response = requests.post(url, headers=headers, json=payload)
response.encoding = 'utf-8'
response.raise_for_status()
## print(response.text)
# 获取响应并添加到上下文
response_content = response.json()["choices"][0]["message"]["content"]
# print(
# f"=========== {user_id}:{user_prompt} ====================\r\n请求内容:{messages}\r\n完整响应:{response.text}")
# 将系统的回复添加到用户上下文中
if user_id is not None:
user_contexts[user_id].append({"role": "assistant", "content": response_content})
return response.text
except requests.exceptions.RequestException as e:
print(f"请求失败: {e}")
return "请求失败，请检查网络或参数配置。"
except (KeyError, IndexError) as e:
print(f"解析响应时出错: {e}")
return "解析响应内容失败。"
return {}
@app.route('/api/models', methods=['GET'])
@app.route('/api/v1/models', methods=['GET'])
@app.route('/hf/v1/models', methods=['GET'])
def models():
"""返回可用模型列表"""
return get_models()
@app.route('/api/chat/completion', methods=['POST'])
@app.route('/api/v1/chat/completions', methods=['POST'])
@app.route('/hf/v1/chat/completions', methods=['POST'])
def chat_completion_api():
"""处理用户请求并保留上下文"""
data = request.json
user_prompt = data.get("prompt")
user_id = data.get("user_id", str(uuid.uuid4())) # 如果未提供 user_id，生成随机值
response_content = chat_completion(
user_prompt,
user_id=user_id
)
# maybe \uxxxx
# return jsonify(response_content)
## maybe \"xxx\"
# return Response(
# json.dumps(response_content, ensure_ascii=False),
# content_type="application/json; charset=utf-8"
# )
# support Chinese
if isinstance(response_content, str): # 如果已经是 JSON 字符串
return Response(response_content, content_type="application/json; charset=utf-8")
if __name__ == '__main__':
app.run(host='0.0.0.0', port=7860)
|