Error when load model in 4bit

[rin2401]

Don't need kwargs['load_in_4bit'] = True when use quantization_config

LLaVA/llava/model/builder.py

Lines 34 to 40 in c121f04

	elif load_4bit:
	kwargs['load_in_4bit'] = True
	kwargs['quantization_config'] = BitsAndBytesConfig(
	load_in_4bit=True,
	bnb_4bit_compute_dtype=torch.float16,
	bnb_4bit_use_double_quant=True,
	bnb_4bit_quant_type='nf4'

[drzraf]

Happens with transformers 4.44.2 . The ValueError was introduced here in transformers (huggingface/transformers#21579) (instead of a slow deprecation)

This was merged in huggingface/transformers@3668ec1 (part of v4.27.0 and onward)

and since this project requires transformers==4.37.2, there is no reason to keep passing deprecated booleans which trigger this error.