model = XDecoderModel.from_pretrained("xdecoder-10.5") model.eval().cuda()
image = Image.open("street_scene.jpg")
prompt = "The pedestrian wearing a red jacket" xdecoder 10.5
You might be thinking of a video/audio decoder (e.g., XviD, ffdshow, LAV Filters) — but none use version 10.5 with “xdecoder”. model = XDecoderModel
