Finetuning Multilingual Models
#1
by sanchitmonga22 - opened
Hello!
We are trying to fine-tune Qwen 2.5 and 3, 0.5-1.5B parameters model for different Indian languages (like Hindi, Tamil etc).
What is the best way to approach that? Seems like you tried doing it for a bunch of other languages?
Would appreciate your help!
Best,
San