dev(narugo): add percentile value

app.py

@@ -1,7 +1,10 @@
 import os
+from functools import lru_cache
 
 import gradio as gr
-from huggingface_hub import HfFileSystem
+import numpy as np
+import pandas as pd
+from huggingface_hub import HfFileSystem, hf_hub_download
 from imgutils.generic import classify_predict_score
 from natsort import natsorted
 
@@ -17,14 +20,32 @@ _MODELS = natsorted([
 LABELS = ["worst", "low", "normal", "good", "great", "best", "masterpiece"]
 
 
+@lru_cache()
+def _get_mark_table(model):
+    df = pd.read_csv(hf_hub_download(
+        repo_id=_REPOSITORY,
+        repo_type='model',
+        filename=f'{model}/samples.csv',
+    ))
+    df = df.sort_values(['score'])
+    df['cnt'] = list(range(len(df)))
+    df['final_score'] = df['cnt'] / len(df)
+
+    x = np.concatenate([[0.0], df['score'], [6.0]])
+    y = np.concatenate([[0.0], df['final_score'], [1.0]])
+    return x, y
+
+
 def _fn_predict(image, model):
     scores = classify_predict_score(
         image=image,
         repo_id=_REPOSITORY,
         model_name=model,
     )
-    final_score = sum(i * scores[label] for i, label in enumerate(LABELS))
-    return final_score, scores
+    weighted_mean = sum(i * scores[label] for i, label in enumerate(LABELS))
+    x, y = _get_mark_table(model)
+    percentile = y[np.searchsorted(x, np.clip(weighted_mean, a_min=0.0, a_max=6.0))]
+    return weighted_mean, percentile, scores
 
 
 if __name__ == '__main__':
@@ -36,13 +57,17 @@ if __name__ == '__main__':
                 gr_submit = gr.Button(value='Submit', variant='primary')
 
             with gr.Column():
-                gr_score = gr.Text(label='Aesthetic Score (0~6)', value='')
-                gr_output = gr.Label(label='Aesthetic Classes')
+                with gr.Row():
+                    gr_score = gr.Text(label='Aesthetic Score (0~6)', value='')
+                    gr_percentile = gr.Text(label='Percentile (0.0-1.0)', value='')
+
+                with gr.Row():
+                    gr_output = gr.Label(label='Aesthetic Classes')
 
             gr_submit.click(
                 _fn_predict,
                 inputs=[gr_input_image, gr_model],
-                outputs=[gr_score, gr_output],
+                outputs=[gr_score, gr_percentile, gr_output],
             )
 
     demo.queue(os.cpu_count()).launch()