Upload PyTorch INT4 quantized FastVLM-1.5B (weights + config + eval)

Upload evaluation results

initial commit