muhtasham
/

TowerInstruct-7B-v0.1-INT8

8-bit precision

Model card Files Files and versions Community

TowerInstruct-7B-v0.1-INT8 / recipe.yaml

muhtasham's picture

Upload folder using huggingface_hub

0e4749e verified 3 months ago

history blame contribute delete

170 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	SmoothQuantModifier: {smoothing_strength: 0.8}
	GPTQModifier:
	targets: Linear
	ignore: [lm_head]
	scheme: W8A8