Spaces:
Running
Running
junzhaosun
commited on
Commit
•
3bb42a7
1
Parent(s):
6e4a518
Add application file
Browse files- app.py +54 -0
- examples/henry5.mp3 +0 -0
- examples/hmm_i_dont_know.wav +0 -0
- examples/see_in_eyes.wav +0 -0
- examples/yearn_for_time.mp3 +0 -0
- examples/zhiqi.wav +0 -0
- packages.txt +0 -0
- requirements.txt +0 -0
app.py
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/usr/local/bin/python3
|
2 |
+
#-*- coding:utf-8 -*-
|
3 |
+
import gradio as gr
|
4 |
+
|
5 |
+
title = "OpenAI Whisper Large v2"
|
6 |
+
|
7 |
+
description = """
|
8 |
+
本例用于演示 <b>openai/whisper-large-v2</b> 模型的语音识别(ASR)能力。目前没有对模型做微调,基于原始模型开发。 Whisper原始模型主要支持英语语音的识别。英语的效果最好,中文语音识别后只会输出汉语拼音。
|
9 |
+
|
10 |
+
<b>更多的信息请参考:</b> <a href="https://huggingface.co/openai/whisper-large-v2">openai/whisper-large-v2</a>。
|
11 |
+
|
12 |
+
<b>使用方法:</b> 上传一个音频文件或直接在页面中录制音频。音频会在传递到模型之前转换为单声道并重新采样为16 kHz。
|
13 |
+
"""
|
14 |
+
|
15 |
+
article = """
|
16 |
+
<div style='margin:20px auto;'>
|
17 |
+
|
18 |
+
<p>
|
19 |
+
参考:
|
20 |
+
<a href="https://huggingface.co/openai/whisper-large-v2">OpenAI Whisper Large v2</a> |
|
21 |
+
<a href="https://github.com/innev">Innev GitHub</a>
|
22 |
+
</p>
|
23 |
+
|
24 |
+
<p>音频案例:<p>
|
25 |
+
<ul>
|
26 |
+
<li>"春日阳光普照大地,正是踏春好时节" 来源: 知琪(Zhiqi)
|
27 |
+
<li>"Hmm, I don't know" 来源: <a href="https://freesound.org/people/InspectorJ/sounds/519189/">InspectorJ</a> (CC BY 4.0 license)
|
28 |
+
<li>"Henry V" excerpt 来源: <a href="https://freesound.org/people/acclivity/sounds/24096/">acclivity</a> (CC BY-NC 4.0 license)
|
29 |
+
<li>"You can see it in the eyes" 来源: <a href="https://freesound.org/people/JoyOhJoy/sounds/165348/">JoyOhJoy</a> (CC0 license)
|
30 |
+
<li>"We yearn for time" 来源: <a href="https://freesound.org/people/Sample_Me/sounds/610529/">Sample_Me</a> (CC0 license)
|
31 |
+
</ul>
|
32 |
+
</div>
|
33 |
+
"""
|
34 |
+
|
35 |
+
examples = [
|
36 |
+
["examples/zhiqi.wav", None],
|
37 |
+
["examples/hmm_i_dont_know.wav", None],
|
38 |
+
["examples/henry5.mp3", None],
|
39 |
+
["examples/yearn_for_time.mp3", None],
|
40 |
+
["examples/see_in_eyes.wav", None],
|
41 |
+
]
|
42 |
+
|
43 |
+
gr.load(
|
44 |
+
"models/openai/whisper-large-v2",
|
45 |
+
inputs=[
|
46 |
+
gr.Audio(label="上传语音", source="upload", type="numpy"),
|
47 |
+
gr.Audio(label="录制语音", source="microphone", type="numpy"),
|
48 |
+
],
|
49 |
+
outputs=gr.Text(label="识别出的文字"),
|
50 |
+
title=title,
|
51 |
+
description=description,
|
52 |
+
article=article,
|
53 |
+
examples=examples
|
54 |
+
).launch()
|
examples/henry5.mp3
ADDED
Binary file (375 kB). View file
|
|
examples/hmm_i_dont_know.wav
ADDED
Binary file (203 kB). View file
|
|
examples/see_in_eyes.wav
ADDED
Binary file (65.2 kB). View file
|
|
examples/yearn_for_time.mp3
ADDED
Binary file (56.3 kB). View file
|
|
examples/zhiqi.wav
ADDED
Binary file (322 kB). View file
|
|
packages.txt
ADDED
File without changes
|
requirements.txt
ADDED
File without changes
|