Spaces:
Sleeping
Sleeping
Added time to output and default model value
Browse files
app.py
CHANGED
|
@@ -2,11 +2,15 @@ import gradio as gr
|
|
| 2 |
import torch
|
| 3 |
from transformers import pipeline
|
| 4 |
import numpy as np
|
|
|
|
| 5 |
|
| 6 |
pipe_base = pipeline("automatic-speech-recognition", model="aitor-medrano/lara-base-pushed")
|
| 7 |
pipe_small = pipeline("automatic-speech-recognition", model="aitor-medrano/whisper-small-lara")
|
| 8 |
|
| 9 |
-
def greet(modelo, grabacion):
|
|
|
|
|
|
|
|
|
|
| 10 |
sr, y = grabacion
|
| 11 |
# Pasamos el array de muestras a tipo NumPy de 32 bits
|
| 12 |
y = y.astype(np.float32)
|
|
@@ -17,14 +21,17 @@ def greet(modelo, grabacion):
|
|
| 17 |
else:
|
| 18 |
pipe = pipe_small
|
| 19 |
|
| 20 |
-
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
demo = gr.Interface(fn=greet,
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
demo.launch()
|
|
|
|
| 2 |
import torch
|
| 3 |
from transformers import pipeline
|
| 4 |
import numpy as np
|
| 5 |
+
import time
|
| 6 |
|
| 7 |
pipe_base = pipeline("automatic-speech-recognition", model="aitor-medrano/lara-base-pushed")
|
| 8 |
pipe_small = pipeline("automatic-speech-recognition", model="aitor-medrano/whisper-small-lara")
|
| 9 |
|
| 10 |
+
def greet(modelo="Base", grabacion):
|
| 11 |
+
|
| 12 |
+
inicio = time.time()
|
| 13 |
+
|
| 14 |
sr, y = grabacion
|
| 15 |
# Pasamos el array de muestras a tipo NumPy de 32 bits
|
| 16 |
y = y.astype(np.float32)
|
|
|
|
| 21 |
else:
|
| 22 |
pipe = pipe_small
|
| 23 |
|
| 24 |
+
result = modelo + ":" + pipe({"sampling_rate": sr, "raw": y})["text"]
|
| 25 |
+
fin = time.time()
|
| 26 |
+
|
| 27 |
+
return result, fin - inicio
|
| 28 |
|
| 29 |
demo = gr.Interface(fn=greet,
|
| 30 |
+
inputs=[
|
| 31 |
+
gr.Dropdown(
|
| 32 |
+
["base", "small"], label="Modelo", info="Modelos de Lara entrenados"
|
| 33 |
+
),
|
| 34 |
+
gr.Audio()
|
| 35 |
+
],
|
| 36 |
+
outputs=["text","number"])
|
| 37 |
demo.launch()
|