Spaces:
Runtime error
Runtime error
rashid101b
commited on
Commit
•
154da93
1
Parent(s):
8b3ca86
Update app.py
Browse files
app.py
CHANGED
@@ -5,12 +5,12 @@ import torch
|
|
5 |
import gradio as gr
|
6 |
from transformers import Wav2Vec2Tokenizer, Wav2Vec2ForCTC
|
7 |
nltk.download("punkt")
|
|
|
8 |
#Loading the pre-trained model and the tokenizer
|
9 |
model_name = "facebook/wav2vec2-base-960h"
|
10 |
tokenizer = Wav2Vec2Tokenizer.from_pretrained(model_name)
|
11 |
model = Wav2Vec2ForCTC.from_pretrained(model_name)
|
12 |
def load_data(input_file):
|
13 |
-
|
14 |
#reading the file
|
15 |
speech, sample_rate = librosa.load(input_file)
|
16 |
#make it 1-D
|
@@ -23,7 +23,8 @@ def load_data(input_file):
|
|
23 |
def correct_casing(input_sentence):
|
24 |
sentences = nltk.sent_tokenize(input_sentence)
|
25 |
return (' '.join([s.replace(s[0],s[0].capitalize(),1) for s in sentences]))
|
26 |
-
|
|
|
27 |
speech = load_data(input_file)
|
28 |
#Tokenize
|
29 |
input_values = tokenizer(speech, return_tensors="pt").input_values
|
@@ -37,9 +38,9 @@ def load_data(input_file):
|
|
37 |
transcription = correct_casing(transcription.lower())
|
38 |
return transcription
|
39 |
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
|
|
5 |
import gradio as gr
|
6 |
from transformers import Wav2Vec2Tokenizer, Wav2Vec2ForCTC
|
7 |
nltk.download("punkt")
|
8 |
+
|
9 |
#Loading the pre-trained model and the tokenizer
|
10 |
model_name = "facebook/wav2vec2-base-960h"
|
11 |
tokenizer = Wav2Vec2Tokenizer.from_pretrained(model_name)
|
12 |
model = Wav2Vec2ForCTC.from_pretrained(model_name)
|
13 |
def load_data(input_file):
|
|
|
14 |
#reading the file
|
15 |
speech, sample_rate = librosa.load(input_file)
|
16 |
#make it 1-D
|
|
|
23 |
def correct_casing(input_sentence):
|
24 |
sentences = nltk.sent_tokenize(input_sentence)
|
25 |
return (' '.join([s.replace(s[0],s[0].capitalize(),1) for s in sentences]))
|
26 |
+
|
27 |
+
def asr_transcript(input_file):
|
28 |
speech = load_data(input_file)
|
29 |
#Tokenize
|
30 |
input_values = tokenizer(speech, return_tensors="pt").input_values
|
|
|
38 |
transcription = correct_casing(transcription.lower())
|
39 |
return transcription
|
40 |
|
41 |
+
gr.Interface(asr_transcript,
|
42 |
+
inputs = gr.inputs.Audio(source="microphone", type="filepath", optional=True, label="Speaker"),
|
43 |
+
outputs = gr.outputs.Textbox(label="Output Text"),
|
44 |
+
title="ASR using Wav2Vec 2.0",
|
45 |
+
description = "This application displays transcribed text for given audio input",
|
46 |
+
examples = [["Test_File1.wav"], ["Test_File2.wav"], ["Test_File3.wav"]], theme="grass").launch()
|