Model description:

Saiga Mistral 7B zero-shot Mistral 7B (https://huggingface.co/mistralai/Mistral-7B-v0.1) was tuned on 5 Russian datasets: ru_turbo_saiga, ru_sharegpt_cleaned, oasst1_ru_main_branch, gpt_roleplay_realm, ru_instruct_gpt4 See more information on HuggingFace: https://huggingface.co/IlyaGusev/saiga_mistral_7b_lora Zero-shot evaluation script for the inference for Saiga Mistral model: https://github.com/IlyaGusev/rulm/blob/master/self_instruct/src/benchmarks/eval_zs_rsg.py

Parameter description: