leon-se
/

SmolVLM-Instruct-W4A16-G128

Image-Text-to-Text

compressed-tensors

Model card Files Files and versions Community

SmolVLM-Instruct-W4A16-G128 / recipe.yaml

leon-se's picture

Upload folder using huggingface_hub

fc7c18f verified about 2 months ago

history blame contribute delete

224 Bytes

	DEFAULT_stage:
	DEFAULT_modifiers:
	GPTQModifier:
	sequential_targets: [LlamaDecoderLayer]
	scheme: W4A16
	targets: Linear
	ignore: ['re:.lm_head', 're:model.vision_model.', 're:model.connector.*']