File size: 10,733 Bytes
7dfe46c e1b6749 7dfe46c e1b6749 7dfe46c f9e1fac 7dfe46c |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 |
import logging
import requests
import time
import os
import sys
from typing import Dict, List, Any, Optional
from dataclasses import dataclass
from dotenv import load_dotenv
import json
# Load environment variables
load_dotenv()
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
try:
from logger.custom_logger import CustomLoggerTracker
custom_log = CustomLoggerTracker()
logger = custom_log.get_logger("embedding_system")
except ImportError:
# Fallback to standard logging if custom logger not available
logger = logging.getLogger("embedding_system")
SILICONFLOW_API_KEY = os.getenv('SILICONFLOW_API_KEY', 'sk-mamyyymhoyklygepxyaazxpxiaphjjbbynxgdrzebbmusmwl')
@dataclass
class EmbeddingResult:
"""Result of embedding generation."""
embeddings: List[List[float]]
model_name: str
processing_time: float
token_count: int
success: bool
error_message: Optional[str] = None
@dataclass
class RerankResult:
"""Result of reranking operation."""
text: str
score: float
index: int
class EmbeddingSystem:
def __init__(self, config: Dict[str, Any]):
self.config = config
# Get API configuration
self.api_key = SILICONFLOW_API_KEY
if not self.api_key:
raise ValueError("SiliconFlow API key is required")
# API endpoints
self.base_url = "https://api.siliconflow.com/v1"
self.session = requests.Session()
self.session.headers.update({
'Authorization': f'Bearer {self.api_key}',
'Content-Type': 'application/json'
})
# Model configuration from your config
self.embedding_model = config.get('embedding_model', 'Qwen/Qwen3-Embedding-8B')
self.reranker_model = config.get('reranker_model', 'Qwen/Qwen3-Reranker-8B')
# Rate limiting
self.max_requests_per_minute = 60
self.request_timestamps = []
logger.info(f"EmbeddingSystem initialized with model: {self.embedding_model}")
def generate_embeddings(self, texts: List[str]) -> List[List[float]]:
if isinstance(texts, str):
texts = [texts]
if not texts:
logger.warning("No texts provided for embedding generation")
return []
try:
self._check_rate_limit()
payload = {
"model": self.embedding_model,
"input": texts,
"encoding_format": "float"
}
response = self.session.post(
f"{self.base_url}/embeddings",
json=payload,
timeout=30
)
if response.status_code == 200:
data = response.json()
embeddings = [item['embedding'] for item in data.get('data', [])]
if len(embeddings) != len(texts):
logger.warning(f"Expected {len(texts)} embeddings, got {len(embeddings)}")
logger.debug(f"Generated {len(embeddings)} embeddings")
return embeddings
else:
error_msg = f"SiliconFlow API error {response.status_code}: {response.text}"
logger.error(error_msg)
return []
except Exception as e:
logger.error(f"Embedding generation failed: {e}")
return []
def generate_query_embedding(self, query: str) -> List[float]:
embeddings = self.generate_embeddings([query])
return embeddings[0] if embeddings else []
def rerank_documents(self, query: str, documents: List[str],
top_k: Optional[int] = None) -> List[RerankResult]:
if not documents:
return []
try:
self._check_rate_limit()
payload = {
"model": self.reranker_model,
"query": query,
"documents": documents,
"top_k": top_k or len(documents),
"return_documents": True
}
response = self.session.post(
f"{self.base_url}/rerank",
json=payload,
timeout=30
)
if response.status_code == 200:
data = response.json()
results = []
for item in data.get('results', []):
results.append(RerankResult(
text=item.get('document', {}).get('text', ''),
score=item.get('relevance_score', 0.0),
index=item.get('index', 0)
))
# Sort by score (descending)
results.sort(key=lambda x: x.score, reverse=True)
logger.debug(f"Reranked {len(results)} documents")
return results
else:
error_msg = f"SiliconFlow rerank API error {response.status_code}: {response.text}"
logger.error(error_msg)
return []
except Exception as e:
logger.error(f"Reranking failed: {e}")
return []
def rerank_results(self, query: str, documents: List[str], top_k: Optional[int] = None) -> List[RerankResult]:
"""Alias for rerank_documents to match the interface expected by rag_engine."""
return self.rerank_documents(query, documents, top_k)
def _check_rate_limit(self):
"""Check and enforce rate limiting."""
current_time = time.time()
# Remove timestamps older than 1 minute
self.request_timestamps = [
ts for ts in self.request_timestamps
if current_time - ts < 60
]
# Check if we're at the rate limit
if len(self.request_timestamps) >= self.max_requests_per_minute:
sleep_time = 60 - (current_time - self.request_timestamps[0])
if sleep_time > 0:
logger.warning(f"Rate limit reached, sleeping for {sleep_time:.2f} seconds")
time.sleep(sleep_time)
# Add current request timestamp
self.request_timestamps.append(current_time)
def test_api_connection(self) -> Dict[str, Any]:
"""Test the API connection."""
if not self.api_key:
return {
'success': False,
'error': 'API key not set',
'details': 'Please set the SILICONFLOW_API_KEY environment variable'
}
try:
# Test with a simple embedding request
test_payload = {
"model": self.embedding_model,
"input": ["test connection"],
"encoding_format": "float"
}
response = self.session.post(
f"{self.base_url}/embeddings",
json=test_payload,
timeout=10
)
if response.status_code == 200:
return {
'success': True,
'message': 'API connection successful',
'status_code': response.status_code,
'model': self.embedding_model
}
else:
return {
'success': False,
'error': f'API error {response.status_code}',
'details': response.text[:200],
'status_code': response.status_code
}
except Exception as e:
return {
'success': False,
'error': 'Connection failed',
'details': str(e)
}
def get_cache_stats(self) -> dict:
"""Get cache statistics (placeholder for compatibility)."""
return {
"caching_disabled": True,
"note": "Caching not implemented in this version"
}
# Test function
def test_embedding_system():
"""Test the embedding system with your configuration."""
print("🧪 Testing SiliconFlow Embedding System")
print("-" * 40)
# Test configuration
config = {
'siliconflow_api_key': os.getenv('SILICONFLOW_API_KEY'),
'embedding_model': 'Qwen/Qwen3-Embedding-8B',
'reranker_model': 'Qwen/Qwen3-Reranker-8B'
}
try:
# Initialize system
embedding_system = EmbeddingSystem(config)
print("✅ System initialized")
# Test API connection
connection_test = embedding_system.test_api_connection()
if connection_test['success']:
print("✅ API connection successful")
else:
print(f"❌ API connection failed: {connection_test['error']}")
return
# Test embedding generation
test_texts = [
"What is the production yield?",
"How is quality controlled in manufacturing?",
"What safety measures are in place?"
]
print(f"🔄 Generating embeddings for {len(test_texts)} texts...")
embeddings = embedding_system.generate_embeddings(test_texts)
if embeddings and len(embeddings) == len(test_texts):
print(f"✅ Generated {len(embeddings)} embeddings of size {len(embeddings[0])}")
else:
print(f"❌ Embedding generation failed. Got {len(embeddings)} embeddings")
return
# Test reranking
query = "manufacturing quality control"
documents = [
"Quality control processes ensure product reliability",
"Manufacturing efficiency can be improved through automation",
"Safety protocols are essential in industrial settings"
]
print(f"🔄 Testing reranking with query: '{query}'")
rerank_results = embedding_system.rerank_documents(query, documents)
if rerank_results:
print(f"✅ Reranking successful. Top result score: {rerank_results[0].score:.3f}")
for i, result in enumerate(rerank_results):
print(f" {i+1}. Score: {result.score:.3f} - {result.text[:50]}...")
else:
print("❌ Reranking failed")
return
print("\n🎉 All tests passed successfully!")
except Exception as e:
print(f"❌ Test failed: {e}")
if __name__ == "__main__":
test_embedding_system() |