nm-testing
/

Mixtral-8x7B-Instruct-v0.1-FP8-Dynamic

compressed-tensors

Model card Files Files and versions

Mixtral-8x7B-Instruct-v0.1-FP8-Dynamic / recipe.yaml

robertgshaw2's picture

Upload folder using huggingface_hub

310f2a5 verified about 1 year ago

history blame contribute delete

166 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	QuantizationModifier:
	ignore: [lm_head, 're:.*block_sparse_moe.gate']
	targets: [Linear]
	scheme: FP8_DYNAMIC