This looks interesting for a model fine tuned with just 16 MB of data and 700 steps. Is there a way you could you share the code ??
#1
by
Aananda-giri
- opened
Aananda-giri
changed discussion title from
This looks interesting for a model fine tuned with just 16 MB of data. Is there a way you could you share the code ??
to This looks interesting for a model fine tuned with just 16 MB of data and 700 steps. Is there a way you could you share the code ??
hey, sorry for a late reply but there is a better model out there (https://huggingface.co/Telugu-LLM-Labs/Indic-gemma-2b-finetuned-sft-Navarasa-2.0) , but if you want to develop your own model then you can go through this and change it accordingly.(https://github.com/TeluguLLMLabs/Indic-gemma-7b-Navarasa)