nm-testing
/

Qwen3-Next-80B-A3B-Instruct-NVFP4

compressed-tensors

Model card Files Files and versions

Qwen3-Next-80B-A3B-Instruct-NVFP4 / recipe.yaml

dsikka's picture

Upload folder using huggingface_hub

40778be verified 8 months ago

history blame contribute delete

230 Bytes

	default_stage:
	default_modifiers:
	QuantizationModifier:
	targets: [Linear]
	ignore: [lm_head, 're:.mlp.gate$', 're:.mlp.shared_expert_gate$', 're:.linear_attn.',
	're:.self_attn.']
	scheme: NVFP4