Spaces:
Build error
Build error
File size: 9,204 Bytes
7640b67 a6df84a 7640b67 a6df84a 7640b67 a6df84a 7640b67 a6df84a 7640b67 | 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 | """
End-to-End Test Script for RAG Pipeline
This script demonstrates and tests the complete RAG pipeline:
1. Document loading
2. Chunking
3. Embedding and storage
4. Retrieval
5. Generation with LLM
Run this script to verify the entire pipeline works correctly.
"""
from src.rag.loaders import MultiFormatDocumentLoader
from src.rag.chunking import DocumentChunker
from src.rag.embeddings import get_embeddings, EmbeddingsManager
from src.rag.vectorstore import create_vector_store
from src.rag.retrieval import Retriever
from src.rag.llm import LLMManager
from src.rag.pipeline import create_rag_pipeline
import sys
import logging
from pathlib import Path
# Add project root to path
project_root = Path(__file__).parent.parent
sys.path.insert(0, str(project_root))
# Configure logging
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
)
logger = logging.getLogger(__name__)
def print_section(title: str):
"""Print a section header."""
print("\n" + "=" * 60)
print(f" {title}")
print("=" * 60)
def test_document_loading():
"""Test document loading functionality."""
print_section("1. Testing Document Loading")
loader = MultiFormatDocumentLoader()
# Test loading sample documents
sample_dir = project_root / "data" / "sample_docs"
if not sample_dir.exists():
print("β οΈ Sample documents directory not found. Creating sample docs...")
sample_dir.mkdir(parents=True, exist_ok=True)
# Create a sample document
sample_file = sample_dir / "sample.txt"
sample_file.write_text("""
RAG (Retrieval-Augmented Generation) Documentation
What is RAG?
RAG is a technique that combines information retrieval with text generation.
It retrieves relevant documents from a knowledge base and uses them as context
for generating accurate, grounded responses.
Key Components:
1. Document Loader - Ingests documents from various sources
2. Chunker - Splits documents into smaller pieces
3. Embeddings - Converts text to vector representations
4. Vector Store - Stores and indexes embeddings for fast retrieval
5. Retriever - Finds relevant documents for a query
6. LLM - Generates responses based on retrieved context
Benefits of RAG:
- Reduces hallucination by grounding responses in real documents
- Enables up-to-date knowledge without retraining
- Provides traceable sources for generated answers
- Cost-effective compared to fine-tuning
Refund Policy:
Our refund policy allows returns within 30 days of purchase.
All refunds are processed within 7 business days.
Items must be in original condition with receipt.
""")
print(f"β
Created sample document: {sample_file}")
# Load documents
documents = loader.load_directory(sample_dir)
print(f"β
Loaded {len(documents)} document(s)")
for doc in documents:
print(f" - {doc.metadata.get('file_name', 'unknown')}: {len(doc.page_content)} chars")
return documents
def test_chunking(documents):
"""Test document chunking."""
print_section("2. Testing Document Chunking")
chunker = DocumentChunker(chunk_size=500, chunk_overlap=50)
chunks = chunker.chunk_documents(documents)
print(f"β
Created {len(chunks)} chunks")
stats = chunker.get_chunk_stats(chunks)
print(f" - Average chunk size: {stats['avg_chunk_size']:.0f} chars")
print(f" - Min chunk size: {stats['min_chunk_size']} chars")
print(f" - Max chunk size: {stats['max_chunk_size']} chars")
return chunks
def test_embeddings():
"""Test embedding functionality."""
print_section("3. Testing Embeddings")
manager = EmbeddingsManager()
embeddings = manager.get_embeddings()
# Test embedding a sample text
sample_text = "What is RAG?"
vector = manager.embed_text(sample_text)
print(f"β
Embedding model loaded: {manager.model_name}")
print(f" - Vector dimensions: {len(vector)}")
print(f" - Sample vector (first 5): {vector[:5]}")
return embeddings
def test_vector_store(chunks, embeddings):
"""Test vector store operations."""
print_section("4. Testing Vector Store")
# Create vector store with test collection
store = create_vector_store(
embedding_function=embeddings,
collection_name="test_collection"
)
# Add documents
ids = store.add_documents(chunks)
print(f"β
Added {len(ids)} chunks to vector store")
# Get stats
stats = store.get_collection_stats()
print(f" - Collection: {stats['collection_name']}")
print(f" - Document count: {stats['document_count']}")
return store
def test_retrieval(store):
"""Test retrieval functionality."""
print_section("5. Testing Retrieval")
retriever = Retriever(vector_store=store, default_k=3)
# Test queries
test_queries = [
"What is RAG?",
"What is the refund policy?",
"What are the key components?",
]
for query in test_queries:
result = retriever.retrieve(query)
print(f"\nπ Query: '{query}'")
print(f" Retrieved {result.num_results} documents")
if result.scores:
print(f" Top score: {result.scores[0]:.4f}")
print(f" Context preview: {result.get_context()[:100]}...")
return retriever
def test_llm():
"""Test LLM functionality."""
print_section("6. Testing LLM")
try:
manager = LLMManager()
info = manager.get_model_info()
if not info.get("api_key_configured"):
print("β οΈ No API key configured. Set GOOGLE_API_KEY in .env")
print(" Skipping LLM generation test...")
return None
print(f"β
LLM configured: {info['model_name']}")
print(f" - Temperature: {info['temperature']}")
# Test simple generation
response = manager.generate("Say 'Hello, RAG!' in exactly 3 words.")
print(f" - Test response: {response[:50]}...")
return manager
except Exception as e:
print(f"β οΈ LLM test failed: {e}")
return None
def test_rag_pipeline(llm_available=True):
"""Test complete RAG pipeline."""
print_section("7. Testing Complete RAG Pipeline")
if not llm_available:
print("β οΈ Skipping RAG pipeline test (no LLM available)")
return
try:
# Create pipeline
pipeline = create_rag_pipeline(collection_name="test_collection")
print("β
RAG Pipeline created")
info = pipeline.get_pipeline_info()
print(f" - LLM: {info['llm_model']}")
print(f" - Top K: {info['top_k']}")
# Test query
question = "What is RAG and what are its benefits?"
print(f"\nπ Testing query: '{question}'")
result = pipeline.query(question, include_sources=True)
print("\n㪠Answer:")
print(f" {result['answer'][:300]}...")
if result.get("sources"):
print(f"\nπ Sources ({result['num_sources']}):")
for source in result["sources"][:2]:
print(f" - {source['metadata'].get('file_name', 'unknown')}")
return True
except Exception as e:
print(f"β RAG pipeline test failed: {e}")
return False
def cleanup_test_collection():
"""Clean up test collection."""
print_section("8. Cleanup")
try:
embeddings = get_embeddings()
store = create_vector_store(
embedding_function=embeddings,
collection_name="test_collection"
)
store.delete_collection()
print("β
Test collection cleaned up")
except Exception as e:
print(f"β οΈ Cleanup warning: {e}")
def main():
"""Run all end-to-end tests."""
print("\n" + "=" * 60)
print(" π RAG Pipeline End-to-End Test Suite")
print("=" * 60)
try:
# Run tests in order
documents = test_document_loading()
chunks = test_chunking(documents)
embeddings = test_embeddings()
store = test_vector_store(chunks, embeddings)
test_retrieval(store)
llm = test_llm()
if llm:
success = test_rag_pipeline(llm_available=True)
else:
test_rag_pipeline(llm_available=False)
success = True # Partial success
# Cleanup
cleanup_test_collection()
# Summary
print_section("Test Summary")
if success:
print("β
All tests passed!")
print("\nπ Phase 1 Complete: Core RAG Pipeline is working!")
else:
print("β οΈ Some tests skipped (LLM not configured)")
print(" Configure GOOGLE_API_KEY in .env to enable full testing")
return 0
except Exception as e:
print(f"\nβ Test failed with error: {e}")
import traceback
traceback.print_exc()
return 1
if __name__ == "__main__":
sys.exit(main())
|