Request-Code for counting objects using text prompt in an image and domain specific finetuning script

[AT38875]

Hi @ksugar @dhkim2810 @qiaoyu1002 @ChaoningZhang Can you also incorporate the code for counting objects in an image based on prompts. Also it will be great if you could publish the fine-tuning script for custom objects. Here is the example for counting objects. https://github.com/ylqi/Count-Anything/ It is using SAM model which takes roughly 7GB GPU memory for inference and around 15-20 minutes. Also throwing error for custom images.Can you incorporate those changes in mobile SAM for object counter as well. Thanks