Does this support PEFT API for fine-tuning?
#16
by
larry5
- opened
Is there any guidance on applying PEFT on this model?
Yes it does. I'd recommend taking a look at the example script here: https://github.com/huggingface/trl/blob/main/examples/scripts/vsft_llava.py and adapt for LLaVa-v1.6 accordingly
Mismatch between tokenizer and model vocabulary sizes. Got an error message during training:"../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [256,0,0], thread: [24,0,0] Assertion srcIndex < srcSelectDimSize failed."
Feel frustrated. Any suggestions?
