by lmsys
Open source · 179k downloads · 391 likes
Vicuna 7B v1.5 is a language model specialized in conversational interactions, designed to respond naturally and contextually to a variety of queries. Fine-tuned from Llama 2 using over 125,000 user-shared conversations, it excels in understanding and generating fluid dialogues, whether for complex or creative exchanges. Its primary use cases include AI research, prototyping virtual assistants, and experimentation by enthusiasts in natural language processing. What sets it apart is its ability to mimic nuanced human-like responses, validated by standardized benchmarks and user preferences, while remaining practical for real-world applications. Its open-source approach fosters innovation and adaptability to specific needs.
Vicuna is a chat assistant trained by fine-tuning Llama 2 on user-shared conversations collected from ShareGPT.
The primary use of Vicuna is research on large language models and chatbots. The primary intended users of the model are researchers and hobbyists in natural language processing, machine learning, and artificial intelligence.
Vicuna v1.5 is fine-tuned from Llama 2 with supervised instruction fine-tuning. The training data is around 125K conversations collected from ShareGPT.com. See more details in the "Training Details of Vicuna Models" section in the appendix of this paper.
Vicuna is evaluated with standard benchmarks, human preference, and LLM-as-a-judge. See more details in this paper and leaderboard.
