leon-se
/

gemma-3-27b-it-qat-W4A16-G128

Image-Text-to-Text

compressed-tensors

Model card Files Files and versions

gemma-3-27b-it-qat-W4A16-G128 / recipe.yaml

leon-se's picture

Upload folder using huggingface_hub

c7c688f verified 8 months ago

history blame contribute delete

225 Bytes

	default_stage:
	default_modifiers:
	GPTQModifier:
	sequential_targets: [Gemma3DecoderLayer]
	scheme: W4A16
	targets: Linear
	ignore: ['re:.lm_head', 're:vision_tower.', 're:multi_modal_projector.*']