создал файл App
Browse files
app.py
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from transformers import CLIPProcessor, CLIPModel
|
| 3 |
+
from PIL import Image
|
| 4 |
+
import torch
|
| 5 |
+
|
| 6 |
+
# Загружаем модель и препроцессор
|
| 7 |
+
model = CLIPModel.from_pretrained("openai/clip-vit-large-patch14")
|
| 8 |
+
processor = CLIPProcessor.from_pretrained("openai/clip-vit-large-patch14")
|
| 9 |
+
|
| 10 |
+
# Основная функция для классификации
|
| 11 |
+
def classify(image, labels):
|
| 12 |
+
# labels — строка вида "Titanic, Avatar, Interstellar"
|
| 13 |
+
candidate_labels = [label.strip() for label in labels.split(",") if label.strip()]
|
| 14 |
+
inputs = processor(text=candidate_labels, images=image, return_tensors="pt", padding=True)
|
| 15 |
+
|
| 16 |
+
with torch.no_grad():
|
| 17 |
+
outputs = model(**inputs)
|
| 18 |
+
logits_per_image = outputs.logits_per_image
|
| 19 |
+
probs = logits_per_image.softmax(dim=1)
|
| 20 |
+
|
| 21 |
+
results = {label: round(float(p) * 100, 2) for label, p in zip(candidate_labels, probs[0])}
|
| 22 |
+
return results
|
| 23 |
+
|
| 24 |
+
# Интерфейс Gradio
|
| 25 |
+
iface = gr.Interface(
|
| 26 |
+
fn=classify,
|
| 27 |
+
inputs=[
|
| 28 |
+
gr.Image(type="pil", label="Загрузите постер фильма"),
|
| 29 |
+
gr.Textbox(label="Введите названия фильмов через запятую")
|
| 30 |
+
],
|
| 31 |
+
outputs="label",
|
| 32 |
+
title="Zero-Shot CLIP Movie Classifier",
|
| 33 |
+
description="Сравни изображение с предложенными названиями фильмов"
|
| 34 |
+
)
|
| 35 |
+
|
| 36 |
+
iface.launch()
|