Firworks
/

Hermes-4-70B-nvfp4

8-bit precision

compressed-tensors

Model card Files Files and versions

Hermes-4-70B-nvfp4 / recipe.yaml

Firworks's picture

Add NVFP4 quantized checkpoint

4fc1402 verified about 1 month ago

history blame contribute delete

130 Bytes

	default_stage:
	default_modifiers:
	QuantizationModifier:
	targets: [Linear]
	ignore: [lm_head]
	scheme: NVFP4