Spaces:
Sleeping
Sleeping
Aboubacar OUATTARA - kaira
commited on
Commit
·
4273f74
1
Parent(s):
589123e
add examples files
Browse files
app.py
CHANGED
@@ -1,7 +1,11 @@
|
|
|
|
|
|
1 |
import spaces
|
2 |
import torch
|
3 |
from transformers import pipeline
|
4 |
import gradio as gr
|
|
|
|
|
5 |
from bambara_utils import BambaraWhisperTokenizer
|
6 |
|
7 |
# Determine the appropriate device (GPU or CPU)
|
@@ -31,15 +35,34 @@ def transcribe(audio):
|
|
31 |
text = pipe(audio)["text"]
|
32 |
return text
|
33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
|
35 |
def main():
|
|
|
|
|
|
|
36 |
# Setup Gradio interface
|
37 |
iface = gr.Interface(
|
38 |
fn=transcribe,
|
39 |
-
inputs=gr.Audio(type="filepath"),
|
40 |
outputs="text",
|
41 |
title="Bambara Automatic Speech Recognition",
|
42 |
-
description="Realtime demo for Bambara speech recognition based on a fine-tuning of the Whisper model."
|
|
|
43 |
)
|
44 |
|
45 |
# Launch the interface
|
|
|
1 |
+
import os
|
2 |
+
|
3 |
import spaces
|
4 |
import torch
|
5 |
from transformers import pipeline
|
6 |
import gradio as gr
|
7 |
+
# Please note that the below import will override whisper LANGUAGES to add bambara
|
8 |
+
# this is not the best way to do it but at least it works. for more info check the bambara_utils code
|
9 |
from bambara_utils import BambaraWhisperTokenizer
|
10 |
|
11 |
# Determine the appropriate device (GPU or CPU)
|
|
|
35 |
text = pipe(audio)["text"]
|
36 |
return text
|
37 |
|
38 |
+
def get_wav_files(directory):
|
39 |
+
"""
|
40 |
+
Returns a list of absolute paths to all .wav files in the specified directory.
|
41 |
+
|
42 |
+
Args:
|
43 |
+
directory (str): The directory to search for .wav files.
|
44 |
+
|
45 |
+
Returns:
|
46 |
+
list: A list of absolute paths to the .wav files.
|
47 |
+
"""
|
48 |
+
# List all files in the directory
|
49 |
+
files = os.listdir(directory)
|
50 |
+
# Filter for .wav files and create absolute paths
|
51 |
+
wav_files = [os.path.abspath(os.path.join(directory, file)) for file in files if file.endswith('.wav')]
|
52 |
+
return wav_files
|
53 |
|
54 |
def main():
|
55 |
+
# Get a list of all .wav files in the examples directory
|
56 |
+
example_files = get_wav_files("./examples")
|
57 |
+
|
58 |
# Setup Gradio interface
|
59 |
iface = gr.Interface(
|
60 |
fn=transcribe,
|
61 |
+
inputs=gr.Audio(type="filepath", value=example_files[0]),
|
62 |
outputs="text",
|
63 |
title="Bambara Automatic Speech Recognition",
|
64 |
+
description="Realtime demo for Bambara speech recognition based on a fine-tuning of the Whisper model.",
|
65 |
+
examples=example_files,
|
66 |
)
|
67 |
|
68 |
# Launch the interface
|