Update app.py
Browse files
app.py
CHANGED
@@ -1,15 +1,13 @@
|
|
1 |
import gradio as gr
|
2 |
import spaces
|
3 |
-
import torch
|
4 |
-
from accelerate import init_empty_weights
|
5 |
import random
|
6 |
import json
|
|
|
|
|
7 |
from difflib import SequenceMatcher
|
8 |
from jiwer import wer
|
9 |
import torchaudio
|
10 |
from transformers import pipeline
|
11 |
-
import os
|
12 |
-
import string
|
13 |
|
14 |
# Load metadata
|
15 |
with open("common_voice_en_validated_249_hf_ready.json") as f:
|
@@ -20,19 +18,7 @@ ages = sorted(set(entry["age"] for entry in data))
|
|
20 |
genders = sorted(set(entry["gender"] for entry in data))
|
21 |
accents = sorted(set(entry["accent"] for entry in data))
|
22 |
|
23 |
-
#
|
24 |
-
pipe_whisper_tiny = pipeline("automatic-speech-recognition", model="openai/whisper-tiny")
|
25 |
-
pipe_whisper_tiny_en = pipeline("automatic-speech-recognition", model="openai/whisper-tiny.en")
|
26 |
-
pipe_whisper_base = pipeline("automatic-speech-recognition", model="openai/whisper-base")
|
27 |
-
pipe_whisper_base_en = pipeline("automatic-speech-recognition", model="openai/whisper-base.en")
|
28 |
-
pipe_whisper_medium = pipeline("automatic-speech-recognition", model="openai/whisper-medium")
|
29 |
-
pipe_whisper_medium_en = pipeline("automatic-speech-recognition", model="openai/whisper-medium.en")
|
30 |
-
pipe_distil_whisper_large = pipeline("automatic-speech-recognition", model="distil-whisper/distil-large-v3.5")
|
31 |
-
pipe_wav2vec2_base_960h = pipeline("automatic-speech-recognition", model="facebook/wav2vec2-base-960h")
|
32 |
-
pipe_wav2vec2_large_960h = pipeline("automatic-speech-recognition", model="facebook/wav2vec2-large-960h")
|
33 |
-
pipe_hubert_large_ls960_ft = pipeline("automatic-speech-recognition", model="facebook/hubert-large-ls960-ft")
|
34 |
-
|
35 |
-
# Functions
|
36 |
def convert_to_wav(file_path):
|
37 |
wav_path = file_path.replace(".mp3", ".wav")
|
38 |
if not os.path.exists(wav_path):
|
@@ -41,10 +27,6 @@ def convert_to_wav(file_path):
|
|
41 |
torchaudio.save(wav_path, waveform, sample_rate)
|
42 |
return wav_path
|
43 |
|
44 |
-
def transcribe(pipe, file_path):
|
45 |
-
result = pipe(file_path)
|
46 |
-
return result["text"].strip().lower()
|
47 |
-
|
48 |
def highlight_differences(ref, hyp):
|
49 |
sm = SequenceMatcher(None, ref.split(), hyp.split())
|
50 |
result = []
|
@@ -74,7 +56,7 @@ def generate_audio(age, gender, accent):
|
|
74 |
wav_file_path = convert_to_wav(file_path)
|
75 |
return wav_file_path, wav_file_path
|
76 |
|
77 |
-
# Transcribe & Compare
|
78 |
@spaces.GPU
|
79 |
def transcribe_audio(file_path):
|
80 |
if not file_path:
|
@@ -89,29 +71,33 @@ def transcribe_audio(file_path):
|
|
89 |
if not gold:
|
90 |
return "Reference not found.", "", "", "", "", "", ""
|
91 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
92 |
outputs = {}
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
"facebook/wav2vec2-large-960h": pipe_wav2vec2_large_960h,
|
103 |
-
"facebook/hubert-large-ls960-ft": pipe_hubert_large_ls960_ft,
|
104 |
-
}
|
105 |
-
|
106 |
-
for name, model in models.items():
|
107 |
-
text = transcribe(model, file_path)
|
108 |
-
clean = normalize(text)
|
109 |
-
wer_score = wer(gold, clean)
|
110 |
-
outputs[name] = f"<b>{name} (WER: {wer_score:.2f}):</b><br>{highlight_differences(gold, clean)}"
|
111 |
|
112 |
return (gold, *outputs.values())
|
113 |
|
114 |
-
# Gradio
|
115 |
with gr.Blocks() as demo:
|
116 |
gr.Markdown("# Comparing ASR Models on Diverse English Speech Samples")
|
117 |
gr.Markdown("""
|
@@ -119,7 +105,7 @@ with gr.Blocks() as demo:
|
|
119 |
Users can select age, gender, and accent to generate diverse English audio samples.
|
120 |
The models are evaluated on their ability to transcribe those samples.
|
121 |
Data is sourced from 249 validated entries in the Common Voice English Delta Segment 21.0 release.
|
122 |
-
|
123 |
|
124 |
with gr.Row():
|
125 |
age = gr.Dropdown(choices=ages, label="Age")
|
|
|
1 |
import gradio as gr
|
2 |
import spaces
|
|
|
|
|
3 |
import random
|
4 |
import json
|
5 |
+
import os
|
6 |
+
import string
|
7 |
from difflib import SequenceMatcher
|
8 |
from jiwer import wer
|
9 |
import torchaudio
|
10 |
from transformers import pipeline
|
|
|
|
|
11 |
|
12 |
# Load metadata
|
13 |
with open("common_voice_en_validated_249_hf_ready.json") as f:
|
|
|
18 |
genders = sorted(set(entry["gender"] for entry in data))
|
19 |
accents = sorted(set(entry["accent"] for entry in data))
|
20 |
|
21 |
+
# Utility functions
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
def convert_to_wav(file_path):
|
23 |
wav_path = file_path.replace(".mp3", ".wav")
|
24 |
if not os.path.exists(wav_path):
|
|
|
27 |
torchaudio.save(wav_path, waveform, sample_rate)
|
28 |
return wav_path
|
29 |
|
|
|
|
|
|
|
|
|
30 |
def highlight_differences(ref, hyp):
|
31 |
sm = SequenceMatcher(None, ref.split(), hyp.split())
|
32 |
result = []
|
|
|
56 |
wav_file_path = convert_to_wav(file_path)
|
57 |
return wav_file_path, wav_file_path
|
58 |
|
59 |
+
# Transcribe & Compare (GPU Decorated)
|
60 |
@spaces.GPU
|
61 |
def transcribe_audio(file_path):
|
62 |
if not file_path:
|
|
|
71 |
if not gold:
|
72 |
return "Reference not found.", "", "", "", "", "", ""
|
73 |
|
74 |
+
model_ids = [
|
75 |
+
"openai/whisper-tiny",
|
76 |
+
"openai/whisper-tiny.en",
|
77 |
+
"openai/whisper-base",
|
78 |
+
"openai/whisper-base.en",
|
79 |
+
"openai/whisper-medium",
|
80 |
+
"openai/whisper-medium.en",
|
81 |
+
"distil-whisper/distil-large-v3.5",
|
82 |
+
"facebook/wav2vec2-base-960h",
|
83 |
+
"facebook/wav2vec2-large-960h",
|
84 |
+
"facebook/hubert-large-ls960-ft",
|
85 |
+
]
|
86 |
+
|
87 |
outputs = {}
|
88 |
+
for model_id in model_ids:
|
89 |
+
try:
|
90 |
+
pipe = pipeline("automatic-speech-recognition", model=model_id)
|
91 |
+
text = pipe(file_path)["text"].strip().lower()
|
92 |
+
clean = normalize(text)
|
93 |
+
wer_score = wer(gold, clean)
|
94 |
+
outputs[model_id] = f"<b>{model_id} (WER: {wer_score:.2f}):</b><br>{highlight_differences(gold, clean)}"
|
95 |
+
except Exception as e:
|
96 |
+
outputs[model_id] = f"<b>{model_id}:</b><br><span style='color:red'>Error: {str(e)}</span>"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
97 |
|
98 |
return (gold, *outputs.values())
|
99 |
|
100 |
+
# Gradio UI
|
101 |
with gr.Blocks() as demo:
|
102 |
gr.Markdown("# Comparing ASR Models on Diverse English Speech Samples")
|
103 |
gr.Markdown("""
|
|
|
105 |
Users can select age, gender, and accent to generate diverse English audio samples.
|
106 |
The models are evaluated on their ability to transcribe those samples.
|
107 |
Data is sourced from 249 validated entries in the Common Voice English Delta Segment 21.0 release.
|
108 |
+
""")
|
109 |
|
110 |
with gr.Row():
|
111 |
age = gr.Dropdown(choices=ages, label="Age")
|