Benchmark for mistral models

[mingxuan-he]

Curious here if anyone has done/planned for running agentbench on any mistral models yet?

Mixtral 8x7b seems like one of the best open source models at the moment and the new mistral large supposedly matches gpt-4. Fast inference providers like Groq and fireworks also offers mixtral inference, so definitely great symergies with agents use cases.

很棒的项目! 期待更新!