import gradio as gr
from fastai.vision.all import *
import skimage

learn = load_learner('export.pkl')

def from_csv(x):
  try:
    pattern = '\/([A-Za-z\d]+.jpg)'

    match = re.findall(pattern,str(x))
    # print(match)
    x = df.loc[df['image'] == match[0]]
    # print(x)
    y = x['emotion'].item() # #y=x['label'].values[0] 
    return str(y)
  except:
    # print('check these files')
    # print(x)
    return 0 
  
labels = learn.dls.vocab
def predict(img):
    img = PILImage.create(img)
    pred,pred_idx,probs = learn.predict(img)
    return {labels[i]: float(probs[i]) for i in range(len(labels))}

title = "EMOTIONAL DAMAGE"
description = "A EMOTION classifier trained  with fastai. Created as a demo for Gradio and HuggingFace Spaces."
article="<p style='text-align: center'><a href='https://tmabraham.github.io/blog/gradio_hf_spaces_tutorial' target='_blank'>Blog post</a></p>"
examples = ['happylee.jpg','damu.jpeg','angry.jpeg','kili.jpg']
interpretation='default'
enable_queue=True

gr.Interface(fn=predict,inputs=gr.inputs.Image(shape=(512, 512)),outputs=gr.outputs.Label(num_top_classes=5),title=title,description=description,article=article,examples=examples,interpretation=interpretation,enable_queue=enable_queue).launch()