nimool commited on
Commit
645c5d6
·
1 Parent(s): 5a8783b

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +10 -9
app.py CHANGED
@@ -4,7 +4,7 @@ from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
4
  import gradio as gr
5
  import sox
6
  import subprocess
7
-
8
 
9
  def read_file_and_process(wav_file):
10
  filename = wav_file.split('.')[0]
@@ -36,10 +36,12 @@ def parse(wav_file):
36
  logits = model(**input_values).logits
37
  return parse_transcription(logits)
38
 
39
- model_id = "facebook/wav2vec2-base-100h"
 
40
  processor = Wav2Vec2Processor.from_pretrained(model_id)
41
  model = Wav2Vec2ForCTC.from_pretrained(model_id)
42
 
 
43
  input_ = gr.Audio(source="microphone", type="filepath")
44
  txtbox = gr.Textbox(
45
  label="persian text output:",
@@ -53,12 +55,11 @@ txtbox = gr.Textbox(
53
 
54
  title = "Speech-to-Text (persian)"
55
  description = "Upload a prsian audio, and let AI do the hard work of transcribing."
56
- article = "<p style='text-align: center'><a href='https://arxiv.org/abs/2104.06678'>Large-Scale Self- and Semi-Supervised Learning for Speech Translation</a></p>"
 
57
 
58
 
59
- while output != "خروج":
60
- demo = gr.Interface(fn=parse, inputs = input_, outputs=txtbox, title=title, description=description, article = article,
61
- streaming=True, interactive=True,
62
- analytics_enabled=False, show_tips=False, enable_queue=True)
63
- demo.launch(share=True)
64
- output = outpts
 
4
  import gradio as gr
5
  import sox
6
  import subprocess
7
+ from fuzzywuzzy import fuzz
8
 
9
  def read_file_and_process(wav_file):
10
  filename = wav_file.split('.')[0]
 
36
  logits = model(**input_values).logits
37
  return parse_transcription(logits)
38
 
39
+
40
+ model_id = "jonatasgrosman/wav2vec2-large-xlsr-53-persian"
41
  processor = Wav2Vec2Processor.from_pretrained(model_id)
42
  model = Wav2Vec2ForCTC.from_pretrained(model_id)
43
 
44
+
45
  input_ = gr.Audio(source="microphone", type="filepath")
46
  txtbox = gr.Textbox(
47
  label="persian text output:",
 
55
 
56
  title = "Speech-to-Text (persian)"
57
  description = "Upload a prsian audio, and let AI do the hard work of transcribing."
58
+ article = "<p style='text-align: center'><a href='https://github.com/nimaprgrmr'>Large-Scale Self- and Semi-Supervised Learning for Speech Translation</a></p>"
59
+
60
 
61
 
62
+ demo = gr.Interface(fn=parse, inputs = input_, outputs=txtbox, title=title, description=description, article = article,
63
+ streaming=True, interactive=True,
64
+ analytics_enabled=False, show_tips=False, enable_queue=True)
65
+ demo.launch(share=True)