Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import WhisperTokenizer
|
2 |
+
|
3 |
+
tokenizer = WhisperTokenizer.from_pretrained("openai/whisper-small", language="marathi", task="transcribe")
|
4 |
+
|
5 |
+
from transformers import pipeline
|
6 |
+
import gradio as gr
|
7 |
+
import torch
|
8 |
+
|
9 |
+
pipe = pipeline(model="thak123/whisper-small-gom",
|
10 |
+
task="automatic-speech-recognition", tokenizer= tokenizer) # change to "your-username/the-name-you-picked"
|
11 |
+
|
12 |
+
pipe.model.config.forced_decoder_ids = (
|
13 |
+
pipe.tokenizer.get_decoder_prompt_ids(
|
14 |
+
language="marathi", task="transcribe"
|
15 |
+
)
|
16 |
+
)
|
17 |
+
|
18 |
+
def transcribe(audio):
|
19 |
+
text = pipe(audio)["text"]
|
20 |
+
return text
|
21 |
+
|
22 |
+
iface = gr.Interface(
|
23 |
+
fn=transcribe,
|
24 |
+
inputs=gr.Audio(source="microphone", type="filepath"),
|
25 |
+
outputs="text",
|
26 |
+
title="Whisper Small Konkani",
|
27 |
+
description="Realtime demo for Konkani speech recognition using a fine-tuned Whisper small model.",
|
28 |
+
)
|
29 |
+
|
30 |
+
|
31 |
+
iface.launch()
|