Benchmark: add LLM-as-a-Judge + BERTScore (hackathon 30% accuracy criterion) 5d58764 muthuk1 commited on 5 days ago
Improve latency: parallel LLM calls, embedding cache, client reuse 90b36cb muthuk1 commited on 5 days ago
Add .gitignore, dataset metadata, retrieval layer, and latest web/graphrag updates 577adc4 muthuk1 commited on 5 days ago
Add benchmark API route for live benchmark runs from dashboard fe0369c verified muthuk1 commited on 12 days ago