Supporting the code model Codestral-22B-v0.1
Opened this issue · 0 comments
596192804 commented
MistralAI has unleashed a code model that sets a new benchmark in the performance-to-latency tradeoff for code generation, outperforming previous models commonly employed for coding. mistralai/Codestral-22B-v0.1
However, the model is currently not compatible for deployment with Refact. Could you provide support for integrating it with Refact?