Fine-Tuning Mistral Model With ChatRTX

[Atgh3]

Hello everyone,

Is it possible to directly train or fine-tune the model in ChatRTX using the local framework, or would I need to fine-tune the model externally (using Hugging Face, PyTorch, etc.) and then convert and integrate it back into ChatRTX's infrastructure?

Additionally, if external fine-tuning is required, could you provide some guidance on how to reformat and structure the fine-tuned model (such as converting it to TensorRT or preparing the correct folder structure) so that it can be used efficiently with ChatRTX?

Thank you for your support and guidance.

Best regards,
Louis C