Upload README.md with huggingface_hub

README.md

@@ -12,25 +12,55 @@ license: mit
 
 # Nuremberg Trials AI
 
-An AI-powered question-answering system for the Nuremberg Trials (1945-1946).
+The most comprehensive AI-powered question-answering system for the Nuremberg Trials (1945-1949).
 
 ## Features
 
-- **RAG-powered answers**: Retrieves relevant passages from 12,000+ document chunks
+- **RAG-powered answers**: Searches through 120,000+ document chunks
 - **Source citations**: Every answer shows the original source documents
-- **Comprehensive data**: Covers IMT transcripts, judgments, charter, and defendant biographies
+- **Complete coverage**: All 13 trials (IMT + 12 NMT), judgments, charter, and biographies
 
 ## Data Sources
 
-- **Harvard Law School Nuremberg Trials Project**: Full IMT transcript (17,268 pages)
-- **Yale Avalon Project**: Judgments, indictments, London Charter
-- **Wikipedia**: Defendant biographies
+### Harvard Law School Nuremberg Trials Project (518MB)
+Complete transcripts from all 13 trials:
+- Trial 1: Medical Case (Karl Brandt et al.)
+- Trial 2: Milch Case
+- Trial 3: Justice Case
+- Trial 4: Hostages Case
+- Trial 5: Pohl Case
+- Trial 6: Einsatzgruppen Case
+- Trial 7: IMT - Main Trial (Goering et al.)
+- Trial 8: Flick Case
+- Trial 9: IG Farben Case
+- Trial 10: RuSHA Case
+- Trial 11: Krupp Case
+- Trial 12: Ministries Case
+- Trial 13: High Command Case
+
+**Total: 153,010 pages**
+
+### Yale Avalon Project (192MB)
+- London Charter & IMT Constitution
+- Indictments and Judgments (all defendants)
+- Trial Proceedings (22 volumes)
+- Nazi Conspiracy & Aggression (4 volumes)
+
+**Total: 857 documents, 11.3 million words**
+
+### Wikipedia
+- 24 defendant biographies
+- 11 trial overview pages
+- 8 key prosecutors/judges
+- 6 Nazi organization pages
+
+**Total: 49 pages**
 
 ## Technology
 
 - **Embeddings**: all-MiniLM-L6-v2 (sentence-transformers)
-- **Vector Search**: FAISS
-- **LLM**: Mistral-7B-Instruct (HuggingFace Inference API)
+- **Vector Search**: FAISS with 120,385 vectors
+- **LLM**: Zephyr-7B (HuggingFace Inference API)
 - **UI**: Gradio
 
 ## Example Questions
@@ -39,3 +69,11 @@ An AI-powered question-answering system for the Nuremberg Trials (1945-1946).
 - What were the four counts in the Nuremberg indictment?
 - Who was the chief prosecutor for the United States?
 - What happened to Hermann Goering?
+- What were the Nazi medical experiments?
+- What was the Einsatzgruppen trial about?
+- Who was prosecuted in the IG Farben trial?
+
+## Related Resources
+
+- **Raw Dataset**: [Adherence/nuremberg-trials-complete](https://huggingface.co/datasets/Adherence/nuremberg-trials-complete)
+- **RAG Dataset**: [Adherence/nuremberg-trials-rag](https://huggingface.co/datasets/Adherence/nuremberg-trials-rag)