Zephyr-7B-Beta is a 7 billion parameter language model created by Hugging Face. It is a fine-tune of Mistral-7B using direct preference optimization (DPO) and synthetic data, resulting in strong performance on chat and instruction-following benchmarks. The model is provided with full open weights and a chat template, making it suitable for local deployment, research, and integration into applications via the Hugging Face ecosystem.
Zephyr 7b Beta is a Foundation models & chat product. It focuses on running a lightweight, high-performing open-source chat model that rivals larger proprietary assistants. It is built as an open-source project for developers. Zephyr 7b Beta is open source under the Apache-2.0 license. Zephyr 7b Beta is available on the web, the command line, and API.
Hugging Face builds and maintains Zephyr 7b Beta, and the product first shipped in 2023. Development happens publicly on GitHub with 5.6k stars. Key capabilities include chat template, instruction tuned, and open weights.
Latest indexed changes and source events
HuggingFaceH4/zephyr-7b-beta verified by the PulseGate indexer
Other apps tracked under the same category.
