Multilingual LLM finetuning in Greek

Hello Team, I have a small dataset (1000 samples) written in Greek with mathematical questions about mathematical definitions and their corresponding answers. I would like to use a chat or a base (really I don’t know which is the right choice) LLM which supports Greek and finetune it further on this dataset using colab (please correct me if I am wrong but I don’t know what kind of model I should search for, text-generation, text2text-generation,…). For example can
“google/mt5-base” and “google/flan-t5-base” which are text2text-generation models finetuned to follow instructions like lama2-chat model?
So to summarize, the question is the following: do you know any small LLM which supports the Greek language in order to finetune it to follow instructions and is small enough so I can use colab?