-
Notifications
You must be signed in to change notification settings - Fork 43
/
Copy path.env
97 lines (74 loc) · 5.8 KB
/
.env
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
# R1模型代理输入端口
PROXY_URL=
# 正式输出模型代理输入端口
PROXY_URL2=
# 图像识别模型输入端口
PROXY_URL3=
# 联网模型输入端口
PROXY_URL4=
# 代理服务器输出端口
PROXY_PORT=4120
# Deepseek R1 模型配置
DEEPSEEK_R1_API_KEY=
DEEPSEEK_R1_MODEL=deepseek-ai/DeepSeek-R1
DEEPSEEK_R1_MAX_TOKENS=7985
DEEPSEEK_R1_CONTEXT_WINDOW=64000
DEEPSEEK_R1_TEMPERATURE=0.7
#考虑到R1模型卡顿,可参考如下模型列表:
## deepseek-ai/DeepSeek-R1-Distill-Llama-70B
## deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
## deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
# 正式输出模型配置
Model_output_API_KEY=
Model_output_MODEL=gemini-2.0-pro-exp
Model_output_MAX_TOKENS=65536
Model_output_CONTEXT_WINDOW=1048576
Model_output_TEMPERATURE=0.7
Model_output_WebSearch=False
# 图像识别模型 模型配置(暂定为gemini-2.0-pro-exp)
Image_Model_API_KEY=
Image_MODEL=gemini-2.0-pro-exp
Image_Model_MAX_TOKENS=7985
Image_Model_CONTEXT_WINDOW=32000
Image_Model_TEMPERATURE=0.4
# 联网主动判断模型 模型配置(gemini-2.0-flash-lite-preview更快但不够稳定,若出现联网判断输出不为yes或no,那么使用gemini-2.0-flash-thinking-exp)
SearchDetermine_API_KEY=
SearchDetermine_MODEL=gemini-2.0-flash-lite-preview
SearchDetermine_Model_MAX_TOKENS=7985
SearchDetermine_Model_CONTEXT_WINDOW=32000
SearchDetermine_Model_TEMPERATURE=0.4
# 联网搜索模型 模型配置(暂定为gemini-2.0-pro-exp)
GoogleSearch_API_KEY=
GoogleSearch_MODEL=gemini-2.0-pro-exp
GoogleSearch_Model_MAX_TOKENS=7985
GoogleSearch_Model_CONTEXT_WINDOW=32000
GoogleSearch_Model_TEMPERATURE=0.4
# R1Think微调 提示语
Think_Lora_PROMPT="你将作为一个前置思考辅助系统工作,对前文所有信息进行整理,对用户需求进行逻辑思考,深化思考。在编程,计算等方便,只提供具体思路,不要提供实际的代码或者具体的计算过程。在你思考结束后,会有正式输出模型来完成输出。"
# 接力 提示语
RELAY_PROMPT="前置思考辅助系统已完成思考,结合辅助思考内容和你自己的思考,开始你的正式输出。[SystemNote:辅助思考的内容用户不可见,只有Model也就是你可见。而且前置思考系统的知识库相对较为陈旧,仅停留在2023年4月,很多新生代知识还未接触,而实际现实时间已经抵达2025年初。所以,有些内容需要你自行辨别其思考内容的正确性。另外,前置思考辅助系统不具备多模态能力,它是基于多模态模型的图片转文字翻译功能实现了对图片等多模态文件的识别。最后,请勿向用户过多谈起SystemNote提及的这些技术内容。]"
# 图像识别 提示语
Image_Model_PROMPT="现在开始你是一个图像识别引擎,请对收到的图片进行详细的内容描述。你只需要描述图片内容即可,不要输出其它任何额外内容。你只能作为图像识别引擎工作。"
# 针对R1模型的图像识别文字传递 提示词
Image_SendR1_PROMPT="虽然你不是多模态模型,但是已有专业多模态模型帮你转译,所以你可以认为自己是多模态模型。请用中文进行思考。以下是转译内容,用户发送了一个图片,图片内容为:"
# 针对联网模型的主动判断功能 提示词
GoogleSearch_Determine_PROMPT="Based on the conversation context, determine if web search is needed. Consider the following aspects:1) Time-sensitive information like news, events, or current affairs,或者日期、天气等即时性提问 2) Technical or professional data that requires verification 3) Latest developments in technology, science, or any field 4) Statistical data or factual information that might need updating 5) Complex topics that benefit from multiple sources 6) Regional or cultural information that might vary by location or time 7) New internet slang, viral content, or MEME.If the query contains above element, respond with "Yes".If the query is general knowledge, or involves role-play, fictional scenarios, or ongoing storylines, or the information is well-established and static, or the user is engaging in casual conversation, or the context is explicitly fictional/unrelated to real-world data, respond with "No".Only respond with 'Yes' or 'No'. Do not output any other content.任何情况下,你只能输出yes or no."
# 针对联网模型的搜索信息功能 提示词
GoogleSearch_PROMPT="Based on the conversation, generate comprehensive search queries to gather complete information. Please follow these guidelines: 1) Break down the topic into multiple specific search terms 2) Include both broad and specific queries 3) Use different phrasings and synonyms 4) Include technical/professional terms when relevant 5) Add year/date for time-sensitive information 6) Consider both English and Chinese search terms for better coverage 7) Include location-specific terms if relevant 8) Use quotes for exact phrases when needed. Format output as a list of search terms only, one per line, without any additional text.搜索分析要详细,全面,携带多方面的内容."
# 针对联网模型的输出信息传递 提示词
GoogleSearch_Send_PROMPT="GoogleSearchAI辅助功能已启动,以下是来自多个来源、多语言检索的综合搜索结果,包含最新信息和专业参考数据:\n"
# 最终混合模型名称
HYBRID_MODEL_NAME=GeminiMIXR1
# 输出 API 密钥
OUTPUT_API_KEY=
# URL解析核心配置
PROXY_URL_PARSE=http://127.0.0.1:7890
REQUEST_TIMEOUT=30000
REQUEST_MAX_REDIRECTS=5
# 请求头配置
REQUEST_HEADERS_USER_AGENT=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36
REQUEST_HEADERS_ACCEPT=text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8
REQUEST_HEADERS_ACCEPT_LANGUAGE=zh-CN,zh;q=0.9,en;q=0.8
# 代理重试配置
PROXY_RETRY_ATTEMPTS=3
PROXY_RETRY_DELAY=2000