aleiko
/

NVLM-D-72B-w4a16

compressed-tensors

Model card Files Files and versions Community

NVLM-D-72B-w4a16 / recipe.yaml

aleiko

tokenizer to lfs

daba653 21 days ago

169 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	QuantizationModifier:
	ignore: ['re:.lm_head', 're:mlp1.', 're:vision_model.*']
	targets: Linear
	scheme: W4A16