Spaces:
Running
Running
File size: 4,981 Bytes
bf47268 |
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 |
import os
from typing import Optional
import asyncio
_llm_config = {
'provider': None,
'model': None
}
def setup_llm_fallback():
"""Setup LLM provider fallback chain"""
# Try OpenAI first
if os.getenv('OPENAI_API_KEY'):
_llm_config['provider'] = 'openai'
_llm_config['model'] = 'gpt-4o-mini'
return
# Fallback to Groq
if os.getenv('GROQ_API_KEY'):
_llm_config['provider'] = 'groq'
_llm_config['model'] = 'llama-3.3-70b-versatile'
return
# Fallback to Hyperbolic
if os.getenv('HYPERBOLIC_API_KEY'):
_llm_config['provider'] = 'hyperbolic'
_llm_config['model'] = 'meta-llama/Llama-3.3-70B-Instruct'
return
# Last resort: Hugging Face Inference API
if os.getenv('HF_TOKEN'):
_llm_config['provider'] = 'huggingface'
_llm_config['model'] = 'mistralai/Mixtral-8x7B-Instruct-v0.1'
return
raise ValueError("No LLM API keys configured. Please set at least one of: OPENAI_API_KEY, GROQ_API_KEY, HYPERBOLIC_API_KEY, HF_TOKEN")
async def get_llm_response(
prompt: str,
temperature: float = 0.7,
max_tokens: int = 2000
) -> str:
"""
Get LLM response using fallback chain
Args:
prompt: Input prompt
temperature: Sampling temperature
max_tokens: Maximum tokens to generate
Returns:
LLM response text
"""
provider = _llm_config.get('provider')
model = _llm_config.get('model')
if not provider:
setup_llm_fallback()
provider = _llm_config.get('provider')
model = _llm_config.get('model')
try:
if provider == 'openai':
return await _call_openai(prompt, model, temperature, max_tokens)
elif provider == 'groq':
return await _call_groq(prompt, model, temperature, max_tokens)
elif provider == 'hyperbolic':
return await _call_hyperbolic(prompt, model, temperature, max_tokens)
elif provider == 'huggingface':
return await _call_huggingface(prompt, model, temperature, max_tokens)
except Exception as e:
print(f"Error with {provider}: {e}")
# Try next provider in chain
if provider == 'openai' and os.getenv('GROQ_API_KEY'):
_llm_config['provider'] = 'groq'
return await get_llm_response(prompt, temperature, max_tokens)
raise
async def _call_openai(prompt: str, model: str, temperature: float, max_tokens: int) -> str:
"""Call OpenAI API"""
from openai import AsyncOpenAI
client = AsyncOpenAI(api_key=os.getenv('OPENAI_API_KEY'))
response = await client.chat.completions.create(
model=model,
messages=[{'role': 'user', 'content': prompt}],
temperature=temperature,
max_tokens=max_tokens
)
return response.choices[0].message.content
async def _call_groq(prompt: str, model: str, temperature: float, max_tokens: int) -> str:
"""Call Groq API"""
from groq import AsyncGroq
client = AsyncGroq(api_key=os.getenv('GROQ_API_KEY'))
response = await client.chat.completions.create(
model=model,
messages=[{'role': 'user', 'content': prompt}],
temperature=temperature,
max_tokens=max_tokens
)
return response.choices[0].message.content
async def _call_hyperbolic(prompt: str, model: str, temperature: float, max_tokens: int) -> str:
"""Call Hyperbolic API"""
import aiohttp
url = "https://api.hyperbolic.xyz/v1/chat/completions"
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer {os.getenv('HYPERBOLIC_API_KEY')}"
}
data = {
"model": model,
"messages": [{"role": "user", "content": prompt}],
"temperature": temperature,
"max_tokens": max_tokens
}
async with aiohttp.ClientSession() as session:
async with session.post(url, headers=headers, json=data) as response:
result = await response.json()
return result['choices'][0]['message']['content']
async def _call_huggingface(prompt: str, model: str, temperature: float, max_tokens: int) -> str:
"""Call Hugging Face Inference API"""
import aiohttp
url = f"https://api-inference.huggingface.co/models/{model}"
headers = {"Authorization": f"Bearer {os.getenv('HF_TOKEN')}"}
data = {
"inputs": prompt,
"parameters": {
"temperature": temperature,
"max_new_tokens": max_tokens,
"return_full_text": False
}
}
async with aiohttp.ClientSession() as session:
async with session.post(url, headers=headers, json=data) as response:
result = await response.json()
if isinstance(result, list) and len(result) > 0:
return result[0].get('generated_text', '')
return str(result)
|