📊 Evaluation Method: CUSTOM

🎯 AutoRAG Optimization Report

📅 Generated: 2026-01-11T00:16:22.180806 🔬 8 Configurations Tested
Best Score
0.270
Custom Aggregate
Configs Tested
8
Total experiments
Improvement
+7.7%
vs. worst config
Top Parameters
k=5, t=1.0
Optimal settings
🏆

Best Configuration

Chunk Size
500
Chunk Overlap
50
Embedding Model
all-MiniLM-L6-v2
Top K Documents
5
Temperature
1.0
Overall Score
0.270

Custom Metrics Breakdown

Context Recall 0.000
Faithfulness 0.000
Answer Relevancy 0.500
Answer Similarity 0.481
📊

All Configurations

Rank Chunk Overlap Embedding top_k Temp Score
1 500 50 all-MiniLM-L 5 1.0 0.270
2 500 100 all-MiniLM-L 10 0.3 0.270
3 500 50 all-MiniLM-L 10 0.7 0.267
4 500 100 all-MiniLM-L 3 0.7 0.263
5 256 100 paraphrase-M 10 1.0 0.258
6 256 50 paraphrase-M 5 0.3 0.254
7 256 50 paraphrase-M 5 0.3 0.254
8 256 50 all-MiniLM-L 5 1.0 0.251
💡

Recommendations