jeffcookio
/

Qwen2.5-VL-W4A16-G128

compressed-tensors

Model card Files Files and versions Community

Qwen2.5-VL-W4A16-G128 / recipe.yaml

jeffcookio's picture

Upload folder using huggingface_hub

d40da64 verified 9 days ago

history blame contribute delete

186 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	GPTQModifier:
	sequential_targets: [Qwen2_5_VLDecoderLayer]
	scheme: W4A16
	targets: Linear
	ignore: [lm_head, 're:visual.*']