Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,45 +1,25 @@
|
|
1 |
import gradio as gr
|
2 |
import tempfile
|
3 |
-
import os
|
4 |
-
from openai import OpenAI
|
5 |
from gtts import gTTS
|
6 |
from gtts.lang import tts_langs
|
7 |
|
8 |
-
# Initialize OpenAI client
|
9 |
-
client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))
|
10 |
-
|
11 |
# Get available languages for Google TTS
|
12 |
google_langs = tts_langs()
|
13 |
|
14 |
-
# Define voice options for
|
15 |
google_voice_options = {
|
16 |
-
"en": ["
|
17 |
-
|
18 |
-
"
|
19 |
-
"
|
20 |
-
"
|
21 |
-
"
|
22 |
-
"
|
23 |
-
"
|
24 |
-
"
|
|
|
25 |
}
|
26 |
|
27 |
-
def openai_tts(text, voice, model):
|
28 |
-
try:
|
29 |
-
response = client.audio.speech.create(
|
30 |
-
model=model,
|
31 |
-
voice=voice,
|
32 |
-
input=text
|
33 |
-
)
|
34 |
-
|
35 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as temp_audio:
|
36 |
-
temp_audio.write(response.content)
|
37 |
-
temp_audio_path = temp_audio.name
|
38 |
-
|
39 |
-
return temp_audio_path, f"Speech generated with OpenAI TTS using {voice} voice and {model} model"
|
40 |
-
except Exception as e:
|
41 |
-
return None, f"Error in OpenAI TTS speech generation: {str(e)}"
|
42 |
-
|
43 |
def google_tts(text, lang, tld):
|
44 |
try:
|
45 |
tts = gTTS(text=text, lang=lang, tld=tld, slow=False)
|
@@ -54,43 +34,30 @@ def google_tts(text, lang, tld):
|
|
54 |
|
55 |
# Gradio interface
|
56 |
with gr.Blocks() as iface:
|
57 |
-
gr.Markdown("#
|
58 |
|
59 |
text_input = gr.Textbox(label="Enter text for speech generation")
|
60 |
-
tts_method = gr.Radio(["OpenAI TTS", "Google TTS"], label="TTS Method", value="OpenAI TTS")
|
61 |
-
|
62 |
-
with gr.Group() as openai_options:
|
63 |
-
openai_voice_input = gr.Dropdown(["alloy", "echo", "fable", "onyx", "nova", "shimmer"], label="Select Voice", value="nova")
|
64 |
-
openai_model_input = gr.Dropdown(["tts-1", "tts-1-hd"], label="Select Model", value="tts-1")
|
65 |
|
66 |
with gr.Group() as google_options:
|
67 |
google_lang_input = gr.Dropdown(list(google_langs.keys()), label="Select Language", value="en")
|
68 |
-
google_voice_input = gr.Dropdown(google_voice_options["en"], label="Select Voice Variant", value="
|
69 |
|
70 |
speech_button = gr.Button("Generate Speech")
|
71 |
speech_output = gr.Audio(label="Generated Speech")
|
72 |
speech_message = gr.Textbox(label="Message")
|
73 |
|
74 |
-
def generate_speech(text,
|
75 |
-
|
76 |
-
|
77 |
-
|
78 |
-
return google_tts(text, google_lang, google_voice)
|
79 |
-
|
80 |
-
def update_visible_options(method):
|
81 |
-
if method == "OpenAI TTS":
|
82 |
-
return gr.Group(visible=True), gr.Group(visible=False)
|
83 |
-
else:
|
84 |
-
return gr.Group(visible=False), gr.Group(visible=True)
|
85 |
|
86 |
def update_google_voice_options(lang):
|
87 |
-
return gr.Dropdown(choices=google_voice_options.get(lang, ["com"]), value="
|
88 |
|
89 |
speech_button.click(generate_speech,
|
90 |
-
inputs=[text_input,
|
91 |
outputs=[speech_output, speech_message])
|
92 |
|
93 |
-
tts_method.change(update_visible_options, inputs=[tts_method], outputs=[openai_options, google_options])
|
94 |
google_lang_input.change(update_google_voice_options, inputs=[google_lang_input], outputs=[google_voice_input])
|
95 |
|
96 |
-
iface.launch()
|
|
|
1 |
import gradio as gr
|
2 |
import tempfile
|
|
|
|
|
3 |
from gtts import gTTS
|
4 |
from gtts.lang import tts_langs
|
5 |
|
|
|
|
|
|
|
6 |
# Get available languages for Google TTS
|
7 |
google_langs = tts_langs()
|
8 |
|
9 |
+
# Define voice options with descriptive names for each language
|
10 |
google_voice_options = {
|
11 |
+
"en": [("Australia", "com.au"), ("Canada", "ca"), ("United Kingdom", "co.uk"),
|
12 |
+
("United States", "com"), ("India", "co.in"), ("Ireland", "ie"), ("South Africa", "co.za")],
|
13 |
+
"es": [("Spain", "com"), ("Latin America", "es"), ("Mexico", "com.mx")],
|
14 |
+
"fr": [("France", "fr"), ("Canada", "ca")],
|
15 |
+
"de": [("Germany", "de"), ("Austria", "at")],
|
16 |
+
"it": [("Italy", "it"), ("United States", "com")],
|
17 |
+
"ja": [("Japan", "jp")],
|
18 |
+
"ko": [("Korea", "com")],
|
19 |
+
"pt": [("Brazil", "com.br"), ("Portugal", "pt")],
|
20 |
+
"zh": [("China", "cn"), ("Taiwan", "tw")]
|
21 |
}
|
22 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
def google_tts(text, lang, tld):
|
24 |
try:
|
25 |
tts = gTTS(text=text, lang=lang, tld=tld, slow=False)
|
|
|
34 |
|
35 |
# Gradio interface
|
36 |
with gr.Blocks() as iface:
|
37 |
+
gr.Markdown("# Google TTS Tool")
|
38 |
|
39 |
text_input = gr.Textbox(label="Enter text for speech generation")
|
|
|
|
|
|
|
|
|
|
|
40 |
|
41 |
with gr.Group() as google_options:
|
42 |
google_lang_input = gr.Dropdown(list(google_langs.keys()), label="Select Language", value="en")
|
43 |
+
google_voice_input = gr.Dropdown([x[0] for x in google_voice_options["en"]], label="Select Voice Variant", value="United States")
|
44 |
|
45 |
speech_button = gr.Button("Generate Speech")
|
46 |
speech_output = gr.Audio(label="Generated Speech")
|
47 |
speech_message = gr.Textbox(label="Message")
|
48 |
|
49 |
+
def generate_speech(text, google_lang, google_voice_desc):
|
50 |
+
# Find the tld (country code) based on the description
|
51 |
+
google_voice_tld = dict(google_voice_options[google_lang])[google_voice_desc]
|
52 |
+
return google_tts(text, google_lang, google_voice_tld)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
|
54 |
def update_google_voice_options(lang):
|
55 |
+
return gr.Dropdown(choices=[x[0] for x in google_voice_options.get(lang, [("United States", "com")])], value="United States")
|
56 |
|
57 |
speech_button.click(generate_speech,
|
58 |
+
inputs=[text_input, google_lang_input, google_voice_input],
|
59 |
outputs=[speech_output, speech_message])
|
60 |
|
|
|
61 |
google_lang_input.change(update_google_voice_options, inputs=[google_lang_input], outputs=[google_voice_input])
|
62 |
|
63 |
+
iface.launch()
|