File size: 7,006 Bytes
37da0c0 2190187 37da0c0 2190187 37da0c0 2190187 37da0c0 2190187 37da0c0 2190187 37da0c0 2190187 37da0c0 2190187 37da0c0 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 |
import json
import httpx
from fastapi import FastAPI, Request, HTTPException
from fastapi.responses import StreamingResponse
from fastapi.middleware.cors import CORSMiddleware
from stream import openai, anthropic, google, huggingface
app = FastAPI()
app.include_router(openai.router)
app.include_router(anthropic.router)
app.include_router(google.router)
app.include_router(huggingface.router)
# Allow all origins for testing (adjust for production)
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
# Replace these with secure methods in production
import os
from collections import defaultdict
@app.post("/summarize_openai")
async def summarize_openai(request: Request):
try:
body = await request.json()
except Exception as e:
raise HTTPException(status_code=400, detail="Invalid JSON payload") from e
previous_summary = body.get("previous_summary", "")
latest_conversation = body.get("latest_conversation", "")
persona = body.get("persona", "helpful assistant")
temperature = body.get("temperature", 0.7)
max_tokens = body.get("max_tokens", 1024)
model = body.get("model", MODEL_NAME)
# Load the prompt from prompts.toml
import tomli
with open("../../configs/prompts.toml", "rb") as f:
prompts_config = tomli.load(f)
# Get the prompt and system prompt
prompt_template = prompts_config["summarization"]["prompt"]
system_prompt = prompts_config["summarization"]["system_prompt"]
# Replace variables in the prompt
prompt = prompt_template.replace("$previous_summary", previous_summary).replace("$latest_conversation", latest_conversation)
system_prompt = system_prompt.replace("$persona", persona)
# Using OpenAI's SDK
from openai import AsyncOpenAI
# Initialize the client with the API key
client = AsyncOpenAI(api_key=OPENAI_API_KEY)
try:
print(f"Starting OpenAI summarization for model: {model}")
# Use the SDK to create a completion
response = await client.chat.completions.create(
model=model,
messages=[
{"role": "system", "content": system_prompt},
{"role": "user", "content": prompt}
],
temperature=temperature,
max_tokens=max_tokens
)
summary = response.choices[0].message.content
print("OpenAI summarization completed successfully")
return {"summary": summary}
except Exception as e:
print(f"Error during OpenAI summarization: {str(e)}")
raise HTTPException(status_code=500, detail=f"Error during summarization: {str(e)}")
@app.post("/summarize_anthropic")
async def summarize_anthropic(request: Request):
try:
body = await request.json()
except Exception as e:
raise HTTPException(status_code=400, detail="Invalid JSON payload") from e
previous_summary = body.get("previous_summary", "")
latest_conversation = body.get("latest_conversation", "")
persona = body.get("persona", "helpful assistant")
temperature = body.get("temperature", 0.7)
max_tokens = body.get("max_tokens", 1024)
model = body.get("model", "claude-3-opus-20240229")
# Load the prompt from prompts.toml
import tomli
with open("../../configs/prompts.toml", "rb") as f:
prompts_config = tomli.load(f)
# Get the prompt and system prompt
prompt_template = prompts_config["summarization"]["prompt"]
system_prompt = prompts_config["summarization"]["system_prompt"]
# Replace variables in the prompt
prompt = prompt_template.replace("$previous_summary", previous_summary).replace("$latest_conversation", latest_conversation)
system_prompt = system_prompt.replace("$persona", persona)
try:
import anthropic
# Initialize Anthropic client
client = anthropic.Anthropic(api_key=ANTHROPIC_API_KEY)
print(f"Starting Anthropic summarization for model: {model}")
# Create the response
response = client.messages.create(
model=model,
messages=[
{"role": "user", "content": prompt}
],
system=system_prompt,
max_tokens=max_tokens,
temperature=temperature
)
summary = response.content[0].text
print("Anthropic summarization completed successfully")
return {"summary": summary}
except Exception as e:
print(f"Error during Anthropic summarization: {str(e)}")
raise HTTPException(status_code=500, detail=f"Error during summarization: {str(e)}")
@app.post("/summarize_google")
async def summarize_google(request: Request):
try:
body = await request.json()
except Exception as e:
raise HTTPException(status_code=400, detail="Invalid JSON payload") from e
previous_summary = body.get("previous_summary", "")
latest_conversation = body.get("latest_conversation", "")
persona = body.get("persona", "helpful assistant")
temperature = body.get("temperature", 0.7)
max_tokens = body.get("max_tokens", 1024)
model = body.get("model", "gemini-1.5-pro")
# Load the prompt from prompts.toml
import tomli
with open("../../configs/prompts.toml", "rb") as f:
prompts_config = tomli.load(f)
# Get the prompt and system prompt
prompt_template = prompts_config["summarization"]["prompt"]
system_prompt = prompts_config["summarization"]["system_prompt"]
# Replace variables in the prompt
prompt = prompt_template.replace("$previous_summary", previous_summary).replace("$latest_conversation", latest_conversation)
system_prompt = system_prompt.replace("$persona", persona)
try:
import google.generativeai as genai
# Configure the Google API
genai.configure(api_key=GOOGLE_API_KEY)
# Initialize the model
model_obj = genai.GenerativeModel(model_name=model)
print(f"Starting Google summarization for model: {model}")
# Combine system prompt and user prompt for Google's API
combined_prompt = f"{system_prompt}\n\n{prompt}"
# Generate the response
response = model_obj.generate_content(
contents=combined_prompt,
generation_config=genai.types.GenerationConfig(
temperature=temperature,
max_output_tokens=max_tokens
)
)
summary = response.text
print("Google summarization completed successfully")
return {"summary": summary}
except Exception as e:
print(f"Error during Google summarization: {str(e)}")
raise HTTPException(status_code=500, detail=f"Error during summarization: {str(e)}")
|