Update app.py

app.py

@@ -81,12 +81,15 @@ def extract_frames(video_in, interval=24, output_format='.jpg'):
 
     return frames
 
-from transformers import AutoModelForCausalLM, CodeGenTokenizerFast as Tokenizer
+from transformers import AutoModelForCausalLM, AutoTokenizer
 from PIL import Image
 
-cap_model_id = "vikhyatk/moondream1"
-cap_model = AutoModelForCausalLM.from_pretrained(cap_model_id, trust_remote_code=True)
-cap_tokenizer = Tokenizer.from_pretrained(cap_model_id)
+model_id = "vikhyatk/moondream2"
+revision = "2024-03-06"
+model = AutoModelForCausalLM.from_pretrained(
+    model_id, trust_remote_code=True, revision=revision
+)
+tokenizer = AutoTokenizer.from_pretrained(model_id, revision=revision)
 
 def process_image(image_in):
     '''
@@ -100,9 +103,9 @@ def process_image(image_in):
     print(result)
     return result
     '''
-    image = Image.open(image_in)
-    enc_image = cap_model.encode_image(image)
-    result = cap_model.answer_question(enc_image, "Describe precisely the image in one sentence.", cap_tokenizer)
+    image = Image.open('<IMAGE_PATH>')
+    enc_image = model.encode_image(image)
+    result = model.answer_question(enc_image, "Describe the image in one sentence.", tokenizer)
     print(result)
     return result