added upload file from device

app.py

@@ -1,3 +1,4 @@
+import io
 import streamlit as st
 import requests
 import numpy as np
@@ -11,14 +12,8 @@ def get_model():
 
 caption_model = get_model()
 
-img_url = st.text_input(label='Enter Image URL')
-
-if (img_url != "") and (img_url != None):
-    img = Image.open(requests.get(img_url, stream=True).raw)
-    img = img.convert('RGB')
-    st.image(img)
 
-    img.save('tmp.jpg')
+def predict():
     pred_caption = generate_caption('tmp.jpg', caption_model)
 
     st.write('Predicted Captions:')
@@ -27,3 +22,25 @@ if (img_url != "") and (img_url != None):
     for _ in range(4):
         pred_caption = generate_caption('tmp.jpg', caption_model, add_noise=True)
         st.write(pred_caption)
+
+
+img_url = st.text_input(label='Enter Image URL')
+
+if (img_url != "") and (img_url != None):
+    img = Image.open(requests.get(img_url, stream=True).raw)
+    img = img.convert('RGB')
+    st.image(img)
+    img.save('tmp.jpg')
+    st.image(img)
+    predict()
+
+
+img = st.file_uploader(label='Upload Image', type=['jpg', 'png'])
+
+if img != None:
+    img = img.read()
+    img = Image.open(io.BytesIO(img))
+    img = img.convert('RGB')
+    img.save('tmp.jpg')
+    st.image(img)
+    predict()

model.py

@@ -21,12 +21,13 @@ tokenizer = tf.keras.layers.TextVectorization(
     standardize=None,
     output_sequence_length=MAX_LENGTH,
     vocabulary=vocab
-    )
+)
 
 idx2word = tf.keras.layers.StringLookup(
     mask_token="",
     vocabulary=tokenizer.get_vocabulary(),
-    invert=True)
+    invert=True
+)
 
 
 # MODEL