Add front end
Browse files
app.py
CHANGED
@@ -1,3 +1,33 @@
|
|
1 |
import gradio as gr
|
|
|
|
|
2 |
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import gradio as gr
|
2 |
+
from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
|
3 |
+
import torch
|
4 |
|
5 |
+
# Load the model and processor
|
6 |
+
processor = Wav2Vec2Processor.from_pretrained("jonatasgrosman/wav2vec2-large-xlsr-53-chinese-zh-cn")
|
7 |
+
model = Wav2Vec2ForCTC.from_pretrained("jonatasgrosman/wav2vec2-large-xlsr-53-chinese-zh-cn")
|
8 |
+
|
9 |
+
# Function to transcribe the audio
|
10 |
+
def transcribe_audio(audio):
|
11 |
+
input_values = processor(audio["array"], sampling_rate=audio["sampling_rate"], return_tensors="pt").input_values
|
12 |
+
|
13 |
+
# Inference
|
14 |
+
with torch.no_grad():
|
15 |
+
logits = model(input_values).logits
|
16 |
+
|
17 |
+
# Decode the transcription
|
18 |
+
predicted_ids = torch.argmax(logits, dim=-1)
|
19 |
+
transcription = processor.batch_decode(predicted_ids)
|
20 |
+
|
21 |
+
return transcription[0] # Since we're only handling one audio file
|
22 |
+
|
23 |
+
# Set up the Gradio interface
|
24 |
+
interface = gr.Interface(
|
25 |
+
fn=transcribe_audio,
|
26 |
+
inputs=gr.Audio(source="microphone", type="filepath"), # Accept audio files
|
27 |
+
outputs="text",
|
28 |
+
title="Chinese Audio Transcription",
|
29 |
+
description="Upload or record an audio file to transcribe it into Chinese."
|
30 |
+
)
|
31 |
+
|
32 |
+
# Launch the interface
|
33 |
+
interface.launch()
|