nm-testing
/

Llama-3.2-90B-Vision-Instruct-quantized.w4a16

compressed-tensors

Model card Files Files and versions

Llama-3.2-90B-Vision-Instruct-quantized.w4a16 / recipe.yaml

shubhrapandit's picture

First model version

94d2c08 about 1 year ago

history blame contribute delete

205 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	GPTQModifier:
	dampening_frac: 0.05
	scheme: W4A16
	targets: Linear
	ignore: ['re:.lm_head', 're:multi_modal_projector.', 're:vision_model.*']