Base model

[GuuD]

Hello,
Thank you guys for this great model, it is quite capable even in real-world usage scenarios.
However, as every coding model aimed at mainstream languages, this one also has room for improvement when working with more niche languages like Haskell/OCaml/F# etc. I have prepared some experimental datasets and wanted to finetune it, both to boost it performance in the languages I use, and evaluate impact on model's reasoning performance after exposure to functional and logic programming. I've been waiting for the release since reading #3, so is there still a chance to get it?

Also, if the answer is yes, could you may be share any kind of details about your training process and stages, and may be some samples of the datasets, so there are more examples for people who aren't AI/ML professionals, but want to have a mental model about impact of the different kinds of data on various stages on the final results