Update app.py
Browse files
app.py
CHANGED
|
@@ -9,16 +9,29 @@ from indicai_projects.lite_tts import Lite_TTS
|
|
| 9 |
from indicai_projects.speech2text_translation import INDIC_SEAMLESS
|
| 10 |
from indicai_projects.speech2text_all import Indic_STT_ALL
|
| 11 |
from zipfile import ZipFile
|
| 12 |
-
|
| 13 |
device = Device("cuda" if cuda_is_available() else "cpu")
|
| 14 |
|
| 15 |
indic_tts_lang = getenv("indic_tts_lang","hi")
|
| 16 |
-
|
| 17 |
ZipFile(hf_hub_download("shethjenil/CONFORMER_INDIC_STT","conformer_onnx.zip"), 'r').extractall("conformer_onnx")
|
| 18 |
indic_stt_all_model = Indic_STT_ALL("conformer_onnx",device)
|
| 19 |
indic_tts = Indic_TTS(indic_tts_lang,device)
|
| 20 |
-
|
| 21 |
vits_tts = Lite_TTS(device)
|
| 22 |
indic_seamless_model = INDIC_SEAMLESS(device)
|
| 23 |
|
| 24 |
-
gr.TabbedInterface(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
from indicai_projects.speech2text_translation import INDIC_SEAMLESS
|
| 10 |
from indicai_projects.speech2text_all import Indic_STT_ALL
|
| 11 |
from zipfile import ZipFile
|
|
|
|
| 12 |
device = Device("cuda" if cuda_is_available() else "cpu")
|
| 13 |
|
| 14 |
indic_tts_lang = getenv("indic_tts_lang","hi")
|
|
|
|
| 15 |
ZipFile(hf_hub_download("shethjenil/CONFORMER_INDIC_STT","conformer_onnx.zip"), 'r').extractall("conformer_onnx")
|
| 16 |
indic_stt_all_model = Indic_STT_ALL("conformer_onnx",device)
|
| 17 |
indic_tts = Indic_TTS(indic_tts_lang,device)
|
| 18 |
+
sans_tts = SansTTS(device)
|
| 19 |
vits_tts = Lite_TTS(device)
|
| 20 |
indic_seamless_model = INDIC_SEAMLESS(device)
|
| 21 |
|
| 22 |
+
gr.TabbedInterface(
|
| 23 |
+
[
|
| 24 |
+
gr.Interface(indic_tts.predict,[gr.Textbox(label="Enter Text"),gr.Dropdown(indic_tts.speakers, label="speaker"),],gr.Audio(type="filepath", label="Speech")),
|
| 25 |
+
gr.Interface(sans_tts.predict,[gr.Textbox(value="उद्यमेन हि सिध्यन्ति कार्याणि न मनोरथैः"),gr.Dropdown(sans_tts.speakers,label='Speaker'),gr.Slider(0.5,2,1,step=0.1,label='Speaking Speed')],gr.Audio(label="Speech")),
|
| 26 |
+
gr.Interface(vits_tts.predict,[gr.Textbox(),gr.Dropdown(vits_tts.speakers,label='Speaker'),gr.Dropdown(vits_tts.styles,label='Style')],gr.Audio(label="Speech")),
|
| 27 |
+
gr.Interface(indic_seamless_model.predict,[gr.Audio(type="filepath"),gr.Dropdown(list(indic_seamless_model.lang_conf.keys()), label="Target Language"),],gr.Text(label="Translations"),title="Audio Translation",),
|
| 28 |
+
gr.Interface(indic_stt_all_model.predict,[gr.Audio(type="filepath"),gr.Dropdown(indic_stt_all_model.supported_langs,label='Language')],[gr.Text(label="CTC"),gr.Text(label="RNNT")]),
|
| 29 |
+
],
|
| 30 |
+
[
|
| 31 |
+
f"{indic_tts.full_name} TTS",
|
| 32 |
+
"Sanskrit TTS",
|
| 33 |
+
"Lite TTS With ",
|
| 34 |
+
"Audio Translation",
|
| 35 |
+
"All Indic Speech To Text",
|
| 36 |
+
],
|
| 37 |
+
).launch()
|