maxtext icon indicating copy to clipboard operation
maxtext copied to clipboard

Support LoRA training

Open hxssgaa opened this issue 10 months ago • 2 comments

Is there a plan to support PEFT methods like LoRA training in maxtext to support larger model fine-tuning / continue pretraining so that bigger models like LLaMA-3-70B can be trainined even with small amount of TPU/GPUs?

hxssgaa avatar Apr 20 '24 04:04 hxssgaa

Any updates on when LoRA support would be available?

sbhavani avatar Jun 10 '24 17:06 sbhavani

This is on our roadmap with high priority, wil update here once we start working on it

gobbleturk avatar Aug 28 '24 02:08 gobbleturk