2Vasabi
/

qwen3-vl-imba

compressed-tensors

Model card Files Files and versions

qwen3-vl-imba / quantization_info.json

2Vasabi's picture

Upload folder using huggingface_hub

dc07f24 verified 10 months ago

175 Bytes

	{
	"model": "Qwen/Qwen3-VL-4B-Instruct",
	"quantization_method": "SmoothQuant W8A8",
	"num_calibration_samples": 512,
	"max_seq_length": 4096,
	"smoothquant_alpha": 0.3
	}