Spaces:
Sleeping
Sleeping
| """Quick debug script to test pipeline step by step""" | |
| import os, sys | |
| os.chdir("c:/Saksham/Work/Projects/SupportRAG") | |
| sys.stdout.reconfigure(line_buffering=True) | |
| from src.core.dual_rag_pipeline import get_dual_rag_pipeline | |
| p = get_dual_rag_pipeline() | |
| p.load_vector_stores() | |
| print("Step 1: retrieve FAQ", flush=True) | |
| docs, sims = p.retrieve_with_scores("damaged product", "faq", 3) | |
| print(f" Got {len(docs)} docs, top sim: {sims[0]:.3f}", flush=True) | |
| print("Step 2: retrieve Tickets", flush=True) | |
| docs2, sims2 = p.retrieve_with_scores("damaged product", "ticket", 3) | |
| print(f" Got {len(docs2)} docs, top sim: {sims2[0]:.3f}", flush=True) | |
| print("Step 3: LLM call", flush=True) | |
| ans = p.llm._call("Say OK") | |
| print(f" LLM: {ans}", flush=True) | |
| print("Step 4: full query()", flush=True) | |
| r = p.query("The product I received is damaged") | |
| print(f" Source: {r['source']}, Confidence: {r['confidence']:.2f}", flush=True) | |
| print(f" Answer: {r['answer'][:150]}...", flush=True) | |
| print("\nStep 5: async aquery()", flush=True) | |
| import asyncio | |
| r2 = asyncio.run(p.aquery("The product I received is damaged")) | |
| print(f" Source: {r2['source']}, Confidence: {r2['confidence']:.2f}", flush=True) | |
| print(f" Answer: {r2['answer'][:150]}...", flush=True) | |
| print("\nALL TESTS PASSED", flush=True) | |