small change

pages/3_Oxford_Vases_Disentanglement.py

@@ -56,7 +56,7 @@ with input_col_1:
         concept_ids = st.multiselect('Concepts:', tuple(labels), max_selections=2, default=['AMPHORA', 'CHALICE'])
 
         st.write('**Choose a latent space to disentangle**')
-        space_id = st.selectbox('Space:', tuple(['Z', 'W']))
+        space_id = st.selectbox('Space:', tuple(['Z', 'W']), default='W')
 
         choose_text_button = st.form_submit_button('Choose the defined concept and space to disentangle')
           
@@ -90,7 +90,7 @@ smoothgrad_col_1, smoothgrad_col_2, smoothgrad_col_3, smoothgrad_col_4, smoothgr
 
 # ---------------------------- DISPLAY COL 1 ROW 1 ------------------------------
 with output_col_1:
-    separation_vector, number_important_features, imp_nodes, performance = get_separation_space(concept_ids, annotations, ann_df, latent_space=st.session_state.space_id)
+    separation_vector, number_important_features, imp_nodes, performance = get_separation_space(concept_ids, annotations, ann_df, latent_space=st.session_state.space_id, samples=200)
     # st.write(f'Class ID {input_id} - {input_label}: {pred_prob*100:.3f}% confidence')
     st.write('Concept vector', separation_vector)
     header_col_1.write(f'Concept {st.session_state.concept_ids} - Space {st.session_state.space_id} - Number of relevant nodes: {number_important_features} - Val classification performance: {performance}')# - Nodes {",".join(list(imp_nodes))}')
@@ -108,7 +108,7 @@ with input_col_2:
         random_id = st.form_submit_button('Generate a random image')
 
         if random_id:
-            image_id = random.randint(0, 50000)
+            image_id = random.randint(0, 20000)
             st.session_state.image_id = image_id
             chosen_image_id_input.number_input('Image ID:', format='%d', step=1, value=st.session_state.image_id)
             
@@ -143,13 +143,14 @@ else:
     original_image_vec = annotations['w_vectors'][st.session_state.image_id]
 
 img = generate_original_image(original_image_vec, model, latent_space=st.session_state.space_id)
+top_pred = ann_df.iloc[st.session_state.image_id].idxmax()
 # input_image = original_image_dict['image']
 # input_label = original_image_dict['label']
 # input_id = original_image_dict['id']
 
 with smoothgrad_col_3:
     st.image(img)
-    smooth_head_3.write(f'Base image')
+    smooth_head_3.write(f'Base image, predicted as {top_pred}')
 
 
 images, lambdas = regenerate_images(model, original_image_vec, separation_vector, min_epsilon=-(int(epsilon)), max_epsilon=int(epsilon), latent_space=st.session_state.space_id, layers=layers)

pages/4_todo.py

@@ -1,56 +0,0 @@
-import streamlit as st
-import pandas as pd
-
-from backend.utils import load_dataset, use_container_width_percentage
-
-st.set_page_config(layout='wide')
-
-st.title('ImageNet-1k')
-st.markdown('This page shows the summary of 50,000 images in the validation set of [ImageNet-1k](https://huggingface.co/datasets/imagenet-1k)')
-
-# SCREEN_WIDTH, SCREEN_HEIGHT = 2560, 1664
-
-with st.spinner("Loading dataset..."):
-    dataset_dict = {}
-    for data_index in range(5):
-        dataset_dict[data_index] = load_dataset(data_index)
-        
-imagenet_df = pd.read_csv('./data/ImageNet_metadata.csv')
-
-class_labels = imagenet_df.ClassLabel.unique().tolist()
-class_labels.sort()
-selected_classes = st.multiselect('Class filter: ', options=['All'] + class_labels)
-if not ('All' in selected_classes or len(selected_classes) == 0):
-    imagenet_df = imagenet_df[imagenet_df['ClassLabel'].isin(selected_classes)]
-# st.write(class_labels)
-
-col1, col2 = st.columns([2, 1])
-with col1:
-    st.dataframe(imagenet_df)
-    use_container_width_percentage(100)
-
-with col2:
-    st.text_area('Type anything here to copy later :)')
-    image = None
-    with st.form("display image"):
-        img_index = st.text_input('Image ID to display')
-    
-        submitted = st.form_submit_button('Display this image')
-        error_container = st.empty()
-
-        if submitted:
-            try:
-                img_index = int(img_index)
-                if img_index > 50000-1 or img_index < 0:
-                    error_container.error('The Image ID must be in range from 0 to 49999', icon="🚫")
-                else:
-                    image = dataset_dict[img_index//10_000][img_index%10_000]['image']
-                    class_label = dataset_dict[img_index//10_000][img_index%10_000]['label']
-                    class_id = dataset_dict[img_index//10_000][img_index%10_000]['id']
-            except ValueError:
-                error_container.error('Please enter an integer number for Image ID', icon = "🚫")
-            
-    if image != None:
-        st.image(image)
-        st.write('**Class label:** ', class_label)
-        st.write('\n**Class id:** ', str(class_id))