openchat-3.6-8b-20240522

🏆 The Overall Best Performing Open-source 8B Model 🏆 🚀 Outperforms Llama-3-8B-Instruct and open-source finetunes/merges 🚀 To use this model, we highly recommend installing the OpenChat package by following the installation guide in our repository and using the OpenChat OpenAI-compatible API server by running the serving command from the table below. The server is optimized for high-throughput deployment using vLLM and can run on a consumer GPU with 24GB RAM. To enable tensor parallelism, append `--tensor-parallel-size N` to the serving command. Once started, the server listens at `localhost:18888` for requests and is compatible with the OpenAI ChatCompletion API specifications. Please refer to the example request below for reference. Additionally, you can use the OpenChat Web UI for a user-friendly experience.

译文尚未就绪，显示来源原文