Minor import fix

[jayant07-yb]

If you run the exampler command provided in the readme $ python Inference.py --model_path ./weights/FastSAM.pt --img_path ./images/dogs.jpg --text_prompt "the yellow dog" then you will be getting this error:

0: 576x1024 21 objects, 309.2ms
Speed: 13.9ms preprocess, 309.2ms inference, 201.5ms postprocess per image at shape (1, 3, 1024, 1024)
Traceback (most recent call last):
  File "/home/janand/robot_hand_control/src/wrist/FastSegExp/FastSAM/Inference.py", line 122, in <module>
    main(args)
  File "/home/janand/robot_hand_control/src/wrist/FastSegExp/FastSAM/Inference.py", line 98, in main
    ann = prompt_process.text_prompt(text=args.text_prompt)
  File "/home/janand/robot_hand_control/src/wrist/FastSegExp/FastSAM/fastsam/prompt.py", line 445, in text_prompt
    clip_model, preprocess = clip.load('ViT-B/32', device=self.device)
NameError: name 'clip' is not defined

It seems like the import clip was missing from fastsam/prompt.py

[Armadillooutfitter]

fastsam/prompt.py uses CLIP but never imports it, so clip.load() crashes when --text_prompt is used.
1️⃣ Open this file
fastsam/prompt.py
2️⃣ Add ONE import at the top
fastsam/prompt.py

3️⃣ Install CLIP
pip install git+https://github.com/openai/CLIP.git

4️⃣ Run the example again
python Inference.py
--model_path ./weights/FastSAM.pt
--img_path ./images/dogs.jpg
--text_prompt "the yellow dog"

✅ This alone fixes the crash.