horheynm
/

Llama-3.2-11B-Vision-Instruct-FP8-Dynamic

compressed-tensors

Model card Files Files and versions Community

Llama-3.2-11B-Vision-Instruct-FP8-Dynamic / recipe.yaml

horheynm's picture

Upload folder using huggingface_hub

41fb033 verified about 1 month ago

history blame contribute delete

192 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	QuantizationModifier:
	ignore: ['re:.lm_head', 're:multi_modal_projector.', 're:vision_model.*']
	targets: Linear
	scheme: FP8_DYNAMIC