List your company
ResearchFormerly WizWand

Conversational response generation on MD2Dial

31.2F1 ScoreChatR1-7b
9.67215.26120.8526.439Oct 15, 2025
Updated 6mo ago

Evaluation results

MethodLinks
ChatR1-7bRAG=RAG, LLM=Qwen-7b, Training Strategy=Supervised Fine-tuning or RL Training2025.1031.284.5
ChatR1 (w/o Rint.)RAG=RAG, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1026.477.4
ChatR1-3bRAG=RAG, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.102683.1
SFTRAG=No, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1025.484.2
QR Search R1RAG=RAG, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1023.182.1
ChatGPT (DI)RAG=No, LLM=GPT-3.5, Training Strategy=Zero-Shot2025.1021.681.7
Qwen-Instr. (RAG)RAG=RAG, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1018.875.1
CoT R1RAG=No, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.101880.2
IRCoTRAG=RAG, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1013.367.5
Qwen-Instr. (DI)RAG=No, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1013.264.4
Qwen-Instr. (CoT)RAG=No, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1010.563.6