otherhalf-dev
/

Llama-3.3-70B-Instruct-abliterated-fp8

compressed-tensors

Model card Files Files and versions Community

Llama-3.3-70B-Instruct-abliterated-fp8 / recipe.yaml

otherhalf-dev's picture

Upload folder using huggingface_hub

36e30c0 verified 2 months ago

history blame contribute delete

136 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	QuantizationModifier:
	ignore: [lm_head]
	targets: [Linear]
	scheme: FP8_DYNAMIC