Spaces:
Sleeping
Sleeping
File size: 9,718 Bytes
7644eac |
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 |
"""
Semantic caching layer using Redis for fast query result retrieval.
Semantic caching stores query embeddings and their results, allowing instant
responses for similar queries without re-running the expensive RAG pipeline.
"""
import os
import json
import hashlib
from typing import Optional, Any, Dict
import numpy as np
# Optional imports
try:
import redis
REDIS_AVAILABLE = True
except ImportError:
REDIS_AVAILABLE = False
print("⚠️ Redis not installed. Install with: pip install redis")
try:
from openai import OpenAI
OPENAI_AVAILABLE = True
except ImportError:
OPENAI_AVAILABLE = False
try:
from src.utils.config import EMBEDDING_MODEL
except Exception:
EMBEDDING_MODEL = "text-embedding-3-small"
class SemanticCache:
"""
Redis-based semantic cache for RAG queries.
Instead of exact string matching, this cache uses embedding similarity
to find semantically similar queries and return cached results.
"""
def __init__(
self,
redis_url: Optional[str] = None,
redis_host: str = "localhost",
redis_port: int = 6379,
redis_password: Optional[str] = None,
redis_db: int = 0,
ttl: int = 3600,
similarity_threshold: float = 0.95,
openai_api_key: Optional[str] = None
):
"""
Initialize semantic cache.
Args:
redis_url: Full Redis URL (e.g., redis://... or rediss://... for SSL)
redis_host: Redis server host (used if redis_url not provided)
redis_port: Redis server port (used if redis_url not provided)
redis_password: Redis password (used if redis_url not provided)
redis_db: Redis database number (used if redis_url not provided)
ttl: Time-to-live for cached entries (seconds)
similarity_threshold: Minimum cosine similarity for cache hit
openai_api_key: OpenAI API key for embeddings
"""
self.ttl = ttl
self.similarity_threshold = similarity_threshold
self.redis_client = None
self.openai_client = None
self.enabled = False
# Initialize Redis
if REDIS_AVAILABLE:
try:
# Prefer REDIS_URL if provided and valid (for Upstash, Render, etc.)
if redis_url and redis_url.strip() and redis_url.startswith(('redis://', 'rediss://', 'unix://')):
self.redis_client = redis.from_url(
redis_url,
decode_responses=False,
ssl_cert_reqs=None # Accept self-signed certs for Upstash
)
print(f"✅ Connecting to Redis via URL...")
else:
self.redis_client = redis.Redis(
host=redis_host,
port=redis_port,
password=redis_password if redis_password else None,
db=redis_db,
decode_responses=False
)
print(f"✅ Connecting to Redis at {redis_host}:{redis_port}...")
# Test connection
self.redis_client.ping()
self.enabled = True
print(f"✅ Semantic cache initialized successfully")
except Exception as e:
print(f"⚠️ Redis connection failed: {e}. Caching disabled.")
self.redis_client = None
else:
print("⚠️ Redis not available. Caching disabled.")
# Initialize OpenAI for embeddings
if OPENAI_AVAILABLE:
api_key = openai_api_key or os.getenv("OPENAI_API_KEY")
if api_key:
self.openai_client = OpenAI(api_key=api_key)
if not self.openai_client:
print("⚠️ OpenAI not available. Semantic caching disabled.")
self.enabled = False
def _get_embedding(self, text: str) -> Optional[np.ndarray]:
"""
Get embedding vector for text.
Args:
text: Text to embed
Returns:
Embedding vector or None if failed
"""
if not self.openai_client:
return None
try:
response = self.openai_client.embeddings.create(
model=EMBEDDING_MODEL,
input=text
)
embedding = np.array(response.data[0].embedding)
return embedding
except Exception as e:
print(f"⚠️ Embedding generation failed: {e}")
return None
def _cosine_similarity(self, vec1: np.ndarray, vec2: np.ndarray) -> float:
"""
Calculate cosine similarity between two vectors.
Args:
vec1: First vector
vec2: Second vector
Returns:
Cosine similarity (0-1)
"""
dot_product = np.dot(vec1, vec2)
norm1 = np.linalg.norm(vec1)
norm2 = np.linalg.norm(vec2)
if norm1 == 0 or norm2 == 0:
return 0.0
return float(dot_product / (norm1 * norm2))
def get(self, query: str, threshold: Optional[float] = None) -> Optional[Any]:
"""
Get cached result for a query.
Args:
query: Search query text
threshold: Optional override for similarity threshold
Returns:
Cached result or None if not found
"""
if not self.enabled or not self.redis_client:
return None
current_threshold = threshold if threshold is not None else self.similarity_threshold
try:
# Get query embedding
query_embedding = self._get_embedding(query)
if query_embedding is None:
return None
best_similarity = 0.0
best_result = None
# Iterate using SCAN to avoid blocking Redis on large keyspaces
for key in self.redis_client.scan_iter(match="semantic_cache:*"):
try:
cached_data = self.redis_client.get(key)
if not cached_data:
continue
cached_dict = json.loads(cached_data.decode("utf-8"))
cached_embedding = np.array(cached_dict["embedding"])
similarity = self._cosine_similarity(query_embedding, cached_embedding)
if similarity > best_similarity and similarity >= current_threshold:
best_similarity = similarity
best_result = cached_dict["result"]
except Exception:
continue
if best_result is not None:
print(f"💰 Cache hit! Similarity: {best_similarity:.3f}")
return best_result
return None
except Exception as e:
print(f"⚠️ Cache get failed: {e}")
return None
def set(self, query: str, result: Any) -> bool:
"""
Cache a query result.
Args:
query: Search query
result: Result to cache
Returns:
Success status
"""
if not self.enabled or not self.redis_client:
return False
try:
# Get query embedding
query_embedding = self._get_embedding(query)
if query_embedding is None:
return False
# Create cache key (hash of query for uniqueness)
query_hash = hashlib.md5(query.encode()).hexdigest()
cache_key = f"semantic_cache:{query_hash}"
# Prepare cache data
cache_data = {
'query': query,
'embedding': query_embedding.tolist(),
'result': result
}
# Store in Redis with TTL
self.redis_client.setex(
cache_key,
self.ttl,
json.dumps(cache_data).encode('utf-8')
)
print(f"💾 Cached query: '{query}' (TTL: {self.ttl}s)")
return True
except Exception as e:
print(f"⚠️ Cache set failed: {e}")
return False
def clear(self) -> int:
"""
Clear all cached entries.
Returns:
Number of entries cleared
"""
if not self.enabled or not self.redis_client:
return 0
try:
keys = self.redis_client.keys("semantic_cache:*")
if keys:
count = self.redis_client.delete(*keys)
print(f"🗑️ Cleared {count} cache entries")
return count
return 0
except Exception as e:
print(f"⚠️ Cache clear failed: {e}")
return 0
def get_stats(self) -> Dict[str, Any]:
"""
Get cache statistics.
Returns:
Dictionary with cache stats
"""
if not self.enabled or not self.redis_client:
return {"enabled": False}
try:
keys = self.redis_client.keys("semantic_cache:*")
return {
"enabled": True,
"entries": len(keys),
"ttl": self.ttl,
"similarity_threshold": self.similarity_threshold
}
except Exception as e:
return {"enabled": False, "error": str(e)}
|