Update app.py

app.py

@@ -1,25 +1,21 @@
+from huggingface_hub import from_pretrained_fastai
 import gradio as gr
-from fastai.vision.all import load_learner, PILImage
+from fastai.vision.all import *
 
-# 1. Carga directa de los pesos en un learner preconfigurado
-learn = load_learner('export.pkl') if Path('export.pkl').exists() \
-        else load_learner('model.pkl')  # por si lo quieres probar
 
-# 2. O, si usas la opción 2 con .save():
-# from fastai.vision.all import cnn_learner, resnet34, imagenet_stats
-# learn = cnn_learner(dls=None, arch=resnet34, metrics=[], cbs=[])
-# learn.load('tobacco_weights')
 
-def predict_image(img):
-    # img ya es PILImage si usas type="pil"
-    pred, idx, probs = learn.predict(PILImage.create(img))
-    return {str(pred): float(probs[idx])}
+# repo_id = "YOUR_USERNAME/YOUR_LEARNER_NAME"
+repo_id = "alramil/Practica1"
 
-iface = gr.Interface(
-    fn=predict_image,
-    inputs=gr.Image(type="pil", label="Sube una imagen"),
-    outputs=gr.Label(num_top_classes=3, label="Predicciones"),
-)
+learner = from_pretrained_fastai(repo_id)
+labels = learner.dls.vocab
 
-if __name__ == "__main__":
-    iface.launch()
+# Definimos una función que se encarga de llevar a cabo las predicciones
+def predict(img):
+    #img = PILImage.create(img)
+    pred,pred_idx,probs = learner.predict(img)
+    return {labels[i]: float(probs[i]) for i in range(len(labels))}
+    
+# Creamos la interfaz y la lanzamos. 
+gr.Interface(fn=predict, inputs=gr.inputs.Image(shape=(128, 128)), outputs=gr.outputs.Label(num_top_classes=3),examples=['fry.jpg','leela.jpg']).launch(share=False)
+