Spaces:
Runtime error
Runtime error
Commit
路
fdc5234
1
Parent(s):
6d7326c
Subiendo
Browse files
app.py
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
|
3 |
+
from utils import carga_modelo, genera
|
4 |
+
|
5 |
+
##P谩gina principal
|
6 |
+
st.title('Generador de mariposas')
|
7 |
+
st.write('Este es un generador de mariposas creado con Huggan y Streamlit')
|
8 |
+
|
9 |
+
##Barra lateral
|
10 |
+
st.sidebar.subheader('隆Esta mariposa no existe! 驴Puedes creerlo?')
|
11 |
+
st.sidebar.image('assets/logo.png', width = 200)
|
12 |
+
st.sidebar.caption('Demo creado en vivo')
|
13 |
+
|
14 |
+
##Carga del modelo
|
15 |
+
|
16 |
+
repo_id = 'ceyda/butterfly_cropped_uniq1K_512'
|
17 |
+
modelo_gan = carga_modelo(repo_id)
|
18 |
+
|
19 |
+
##Generaci贸n de im谩genes de mariposas
|
20 |
+
n_mariposas = 4
|
21 |
+
|
22 |
+
def corre():
|
23 |
+
with st.spinner('Cargando modelo...'):
|
24 |
+
ims = genera(modelo_gan, n_mariposas):
|
25 |
+
st,session_state['ims'] = ims
|
26 |
+
|
27 |
+
if 'ims' not in st.session_state:
|
28 |
+
st.session_state['ims'] = None
|
29 |
+
corre()
|
30 |
+
|
31 |
+
ims = st.session_state['ims']
|
32 |
+
|
33 |
+
corre_boton = st.button('Generar mariposas',
|
34 |
+
on_click = corre,
|
35 |
+
help = 'Estamos en vuelo, abrocha tu cintur贸n'
|
36 |
+
)
|
37 |
+
|
38 |
+
if ims is not None:
|
39 |
+
cols = st.columns(n_mariposas)
|
40 |
+
for j, im in enumerate(ims):
|
41 |
+
i = j & n_mariposas
|
42 |
+
cols[i].image(im, use_column_width=True)
|
utils.py
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import numpy as np
|
2 |
+
import torch
|
3 |
+
import huggan.pytorch.lightweight_gan import LightweightGan
|
4 |
+
|
5 |
+
def carga_modelo(model_name = 'ceyda/butterfly_cropped_uniq1K_512', model_version = None)
|
6 |
+
gan = LightweightGan(model_name, version = model_version)
|
7 |
+
gan.eval()
|
8 |
+
return gan
|
9 |
+
|
10 |
+
def genera(gan, batch_size = 1)
|
11 |
+
with torch.no_grad() #no queremos entrenar el modelo
|
12 |
+
ims = gan.G(torch.randn(batch_size, gan.latent_dim)).clamp(0.0,1.0) * 255 #generamos im谩genes y las aplastamos entre 0 y 1
|
13 |
+
ims = ims.permute(0,2,3,1).detach().cpu().numpy().astype(np.uint8) #las pasamos a numpy
|
14 |
+
return ims
|
15 |
+
|