NeoChen1024
/

Dolphin3.0-Llama3.1-8B-W8A8

8-bit precision

compressed-tensors

Model card Files Files and versions Community

Dolphin3.0-Llama3.1-8B-W8A8 / recipe.yaml

NeoChen1024's picture

Upload folder using huggingface_hub

ece6105 verified 4 months ago

history blame contribute delete

226 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	SmoothQuantModifier: {smoothing_strength: 0.8}
	GPTQModifier:
	dampening_frac: 0.1
	offload_hessians: false
	scheme: W8A8
	targets: Linear
	ignore: [lm_head]