GUI-Owl-32B-W4A16-G128

GTPQ W4A16 quantized GUI-Owl-32B via llm-compressor, using flickr30k dataset.

Memory Usage

Type GUI-Owl-32B GUI-Owl-32B-W4A16-G128
Memory Size 63 GB 19.2 GB

Inference

Run with:

vllm serve JessieWei/GUI-Owl-32B-W4A16-G128

Detailed parameter tuning refer to GUI-Owl-32B.

Downloads last month
1
Safetensors
Model size
6B params
Tensor type
BF16
·
I64
·
I32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for JessieWei/GUI-Owl-32B-W4A16-G128

Quantized
(28)
this model

Dataset used to train JessieWei/GUI-Owl-32B-W4A16-G128