nm-testing
/

Moonlight-16B-A3B.w4a16

compressed-tensors

Model card Files Files and versions

Moonlight-16B-A3B.w4a16 / recipe.yaml

kylesayrs's picture

Upload folder using huggingface_hub

71548f5 verified about 1 year ago

history blame contribute delete

120 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	GPTQModifier:
	scheme: W4A16
	targets: Linear
	ignore: [lm_head]