Spaces:
Running
Running
jhj0517
commited on
Commit
β’
b065a65
1
Parent(s):
1f8abba
refactoring to use data class
Browse files- app.py +29 -9
- modules/faster_whisper_inference.py +72 -170
app.py
CHANGED
@@ -8,6 +8,8 @@ from modules.nllb_inference import NLLBInference
|
|
8 |
from ui.htmls import *
|
9 |
from modules.youtube_manager import get_ytmetas
|
10 |
from modules.deepl_api import DeepLAPI
|
|
|
|
|
11 |
|
12 |
class App:
|
13 |
def __init__(self, args):
|
@@ -68,10 +70,16 @@ class App:
|
|
68 |
files_subtitles = gr.Files(label="Downloadable output file", scale=4, interactive=False)
|
69 |
btn_openfolder = gr.Button('π', scale=1)
|
70 |
|
71 |
-
params = [input_file,
|
72 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
73 |
btn_run.click(fn=self.whisper_inf.transcribe_file,
|
74 |
-
inputs=params +
|
75 |
outputs=[tb_indicator, files_subtitles])
|
76 |
btn_openfolder.click(fn=lambda: self.open_folder("outputs"), inputs=None, outputs=None)
|
77 |
dd_model.change(fn=self.on_change_models, inputs=[dd_model], outputs=[cb_translate])
|
@@ -108,10 +116,16 @@ class App:
|
|
108 |
files_subtitles = gr.Files(label="Downloadable output file", scale=4)
|
109 |
btn_openfolder = gr.Button('π', scale=1)
|
110 |
|
111 |
-
params = [tb_youtubelink,
|
112 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
btn_run.click(fn=self.whisper_inf.transcribe_youtube,
|
114 |
-
inputs=params +
|
115 |
outputs=[tb_indicator, files_subtitles])
|
116 |
tb_youtubelink.change(get_ytmetas, inputs=[tb_youtubelink],
|
117 |
outputs=[img_thumbnail, tb_title, tb_description])
|
@@ -141,10 +155,16 @@ class App:
|
|
141 |
files_subtitles = gr.Files(label="Downloadable output file", scale=4)
|
142 |
btn_openfolder = gr.Button('π', scale=1)
|
143 |
|
144 |
-
params = [mic_input,
|
145 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
146 |
btn_run.click(fn=self.whisper_inf.transcribe_mic,
|
147 |
-
inputs=params +
|
148 |
outputs=[tb_indicator, files_subtitles])
|
149 |
btn_openfolder.click(fn=lambda: self.open_folder("outputs"), inputs=None, outputs=None)
|
150 |
dd_model.change(fn=self.on_change_models, inputs=[dd_model], outputs=[cb_translate])
|
|
|
8 |
from ui.htmls import *
|
9 |
from modules.youtube_manager import get_ytmetas
|
10 |
from modules.deepl_api import DeepLAPI
|
11 |
+
from modules.whisper_data_class import *
|
12 |
+
|
13 |
|
14 |
class App:
|
15 |
def __init__(self, args):
|
|
|
70 |
files_subtitles = gr.Files(label="Downloadable output file", scale=4, interactive=False)
|
71 |
btn_openfolder = gr.Button('π', scale=1)
|
72 |
|
73 |
+
params = [input_file, dd_file_format, cb_timestamp]
|
74 |
+
whisper_params = WhisperGradioComponents(model_size=dd_model,
|
75 |
+
lang=dd_lang,
|
76 |
+
is_translate=cb_translate,
|
77 |
+
beam_size=nb_beam_size,
|
78 |
+
log_prob_threshold=nb_log_prob_threshold,
|
79 |
+
no_speech_threshold=nb_no_speech_threshold,
|
80 |
+
compute_type=dd_compute_type)
|
81 |
btn_run.click(fn=self.whisper_inf.transcribe_file,
|
82 |
+
inputs=params + whisper_params.to_list(),
|
83 |
outputs=[tb_indicator, files_subtitles])
|
84 |
btn_openfolder.click(fn=lambda: self.open_folder("outputs"), inputs=None, outputs=None)
|
85 |
dd_model.change(fn=self.on_change_models, inputs=[dd_model], outputs=[cb_translate])
|
|
|
116 |
files_subtitles = gr.Files(label="Downloadable output file", scale=4)
|
117 |
btn_openfolder = gr.Button('π', scale=1)
|
118 |
|
119 |
+
params = [tb_youtubelink, dd_file_format, cb_timestamp]
|
120 |
+
whisper_params = WhisperGradioComponents(model_size=dd_model,
|
121 |
+
lang=dd_lang,
|
122 |
+
is_translate=cb_translate,
|
123 |
+
beam_size=nb_beam_size,
|
124 |
+
log_prob_threshold=nb_log_prob_threshold,
|
125 |
+
no_speech_threshold=nb_no_speech_threshold,
|
126 |
+
compute_type=dd_compute_type)
|
127 |
btn_run.click(fn=self.whisper_inf.transcribe_youtube,
|
128 |
+
inputs=params + whisper_params.to_list(),
|
129 |
outputs=[tb_indicator, files_subtitles])
|
130 |
tb_youtubelink.change(get_ytmetas, inputs=[tb_youtubelink],
|
131 |
outputs=[img_thumbnail, tb_title, tb_description])
|
|
|
155 |
files_subtitles = gr.Files(label="Downloadable output file", scale=4)
|
156 |
btn_openfolder = gr.Button('π', scale=1)
|
157 |
|
158 |
+
params = [mic_input, dd_file_format]
|
159 |
+
whisper_params = WhisperGradioComponents(model_size=dd_model,
|
160 |
+
lang=dd_lang,
|
161 |
+
is_translate=cb_translate,
|
162 |
+
beam_size=nb_beam_size,
|
163 |
+
log_prob_threshold=nb_log_prob_threshold,
|
164 |
+
no_speech_threshold=nb_no_speech_threshold,
|
165 |
+
compute_type=dd_compute_type)
|
166 |
btn_run.click(fn=self.whisper_inf.transcribe_mic,
|
167 |
+
inputs=params + whisper_params.to_list(),
|
168 |
outputs=[tb_indicator, files_subtitles])
|
169 |
btn_openfolder.click(fn=lambda: self.open_folder("outputs"), inputs=None, outputs=None)
|
170 |
dd_model.change(fn=self.on_change_models, inputs=[dd_model], outputs=[cb_translate])
|
modules/faster_whisper_inference.py
CHANGED
@@ -1,10 +1,9 @@
|
|
1 |
import os
|
2 |
|
3 |
-
import tqdm
|
4 |
import time
|
5 |
import numpy as np
|
6 |
from typing import BinaryIO, Union, Tuple, List
|
7 |
-
from datetime import datetime
|
8 |
|
9 |
import faster_whisper
|
10 |
import ctranslate2
|
@@ -15,6 +14,7 @@ import gradio as gr
|
|
15 |
from .base_interface import BaseInterface
|
16 |
from modules.subtitle_manager import get_srt, get_vtt, get_txt, write_file, safe_filename
|
17 |
from modules.youtube_manager import get_ytdata, get_ytaudio
|
|
|
18 |
|
19 |
|
20 |
class FasterWhisperInference(BaseInterface):
|
@@ -26,22 +26,17 @@ class FasterWhisperInference(BaseInterface):
|
|
26 |
self.available_langs = sorted(list(whisper.tokenizer.LANGUAGES.values()))
|
27 |
self.translatable_models = ["large", "large-v1", "large-v2", "large-v3"]
|
28 |
self.device = "cuda" if torch.cuda.is_available() else "cpu"
|
29 |
-
self.available_compute_types = ctranslate2.get_supported_compute_types(
|
|
|
30 |
self.current_compute_type = "float16" if self.device == "cuda" else "float32"
|
31 |
self.default_beam_size = 1
|
32 |
|
33 |
def transcribe_file(self,
|
34 |
fileobjs: list,
|
35 |
-
model_size: str,
|
36 |
-
lang: str,
|
37 |
file_format: str,
|
38 |
-
istranslate: bool,
|
39 |
add_timestamp: bool,
|
40 |
-
|
41 |
-
|
42 |
-
no_speech_threshold: float,
|
43 |
-
compute_type: str,
|
44 |
-
progress=gr.Progress()
|
45 |
) -> list:
|
46 |
"""
|
47 |
Write subtitle file from Files
|
@@ -50,31 +45,14 @@ class FasterWhisperInference(BaseInterface):
|
|
50 |
----------
|
51 |
fileobjs: list
|
52 |
List of files to transcribe from gr.Files()
|
53 |
-
model_size: str
|
54 |
-
Whisper model size from gr.Dropdown()
|
55 |
-
lang: str
|
56 |
-
Source language of the file to transcribe from gr.Dropdown()
|
57 |
file_format: str
|
58 |
-
File format to write from gr.Dropdown(). Supported format: [SRT, WebVTT, txt]
|
59 |
-
istranslate: bool
|
60 |
-
Boolean value from gr.Checkbox() that determines whether to translate to English.
|
61 |
-
It's Whisper's feature to translate speech from another language directly into English end-to-end.
|
62 |
add_timestamp: bool
|
63 |
-
Boolean value from gr.Checkbox() that determines whether to add a timestamp at the end of the filename.
|
64 |
-
beam_size: int
|
65 |
-
Int value from gr.Number() that is used for decoding option.
|
66 |
-
log_prob_threshold: float
|
67 |
-
float value from gr.Number(). If the average log probability over sampled tokens is
|
68 |
-
below this value, treat as failed.
|
69 |
-
no_speech_threshold: float
|
70 |
-
float value from gr.Number(). If the no_speech probability is higher than this value AND
|
71 |
-
the average log probability over sampled tokens is below `log_prob_threshold`,
|
72 |
-
consider the segment as silent.
|
73 |
-
compute_type: str
|
74 |
-
compute type from gr.Dropdown().
|
75 |
-
see more info : https://opennmt.net/CTranslate2/quantization.html
|
76 |
progress: gr.Progress
|
77 |
Indicator to show progress directly in gradio.
|
|
|
|
|
78 |
|
79 |
Returns
|
80 |
----------
|
@@ -83,18 +61,12 @@ class FasterWhisperInference(BaseInterface):
|
|
83 |
Files to return to gr.Files()
|
84 |
"""
|
85 |
try:
|
86 |
-
self.update_model_if_needed(model_size=model_size, compute_type=compute_type, progress=progress)
|
87 |
-
|
88 |
files_info = {}
|
89 |
for fileobj in fileobjs:
|
90 |
transcribed_segments, time_for_task = self.transcribe(
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
beam_size=beam_size,
|
95 |
-
log_prob_threshold=log_prob_threshold,
|
96 |
-
no_speech_threshold=no_speech_threshold,
|
97 |
-
progress=progress
|
98 |
)
|
99 |
|
100 |
file_name, file_ext = os.path.splitext(os.path.basename(fileobj.name))
|
@@ -105,7 +77,7 @@ class FasterWhisperInference(BaseInterface):
|
|
105 |
add_timestamp=add_timestamp,
|
106 |
file_format=file_format
|
107 |
)
|
108 |
-
files_info[file_name] = {"subtitle": subtitle, "time_for_task": time_for_task, "path":
|
109 |
|
110 |
total_result = ''
|
111 |
total_time = 0
|
@@ -115,10 +87,10 @@ class FasterWhisperInference(BaseInterface):
|
|
115 |
total_result += f'{info["subtitle"]}'
|
116 |
total_time += info["time_for_task"]
|
117 |
|
118 |
-
|
119 |
-
|
120 |
|
121 |
-
return [
|
122 |
|
123 |
except Exception as e:
|
124 |
print(f"Error transcribing file on line {e}")
|
@@ -128,50 +100,27 @@ class FasterWhisperInference(BaseInterface):
|
|
128 |
self.remove_input_files([fileobj.name for fileobj in fileobjs])
|
129 |
|
130 |
def transcribe_youtube(self,
|
131 |
-
|
132 |
-
model_size: str,
|
133 |
-
lang: str,
|
134 |
file_format: str,
|
135 |
-
istranslate: bool,
|
136 |
add_timestamp: bool,
|
137 |
-
|
138 |
-
|
139 |
-
no_speech_threshold: float,
|
140 |
-
compute_type: str,
|
141 |
-
progress=gr.Progress()
|
142 |
) -> list:
|
143 |
"""
|
144 |
Write subtitle file from Youtube
|
145 |
|
146 |
Parameters
|
147 |
----------
|
148 |
-
|
149 |
-
|
150 |
-
model_size: str
|
151 |
-
Whisper model size from gr.Dropdown()
|
152 |
-
lang: str
|
153 |
-
Source language of the file to transcribe from gr.Dropdown()
|
154 |
file_format: str
|
155 |
-
File format to write from gr.Dropdown(). Supported format: [SRT, WebVTT, txt]
|
156 |
-
istranslate: bool
|
157 |
-
Boolean value from gr.Checkbox() that determines whether to translate to English.
|
158 |
-
It's Whisper's feature to translate speech from another language directly into English end-to-end.
|
159 |
add_timestamp: bool
|
160 |
Boolean value from gr.Checkbox() that determines whether to add a timestamp at the end of the filename.
|
161 |
-
beam_size: int
|
162 |
-
Int value from gr.Number() that is used for decoding option.
|
163 |
-
log_prob_threshold: float
|
164 |
-
float value from gr.Number(). If the average log probability over sampled tokens is
|
165 |
-
below this value, treat as failed.
|
166 |
-
no_speech_threshold: float
|
167 |
-
float value from gr.Number(). If the no_speech probability is higher than this value AND
|
168 |
-
the average log probability over sampled tokens is below `log_prob_threshold`,
|
169 |
-
consider the segment as silent.
|
170 |
-
compute_type: str
|
171 |
-
compute type from gr.Dropdown().
|
172 |
-
see more info : https://opennmt.net/CTranslate2/quantization.html
|
173 |
progress: gr.Progress
|
174 |
Indicator to show progress directly in gradio.
|
|
|
|
|
175 |
|
176 |
Returns
|
177 |
----------
|
@@ -180,20 +129,14 @@ class FasterWhisperInference(BaseInterface):
|
|
180 |
Files to return to gr.Files()
|
181 |
"""
|
182 |
try:
|
183 |
-
self.update_model_if_needed(model_size=model_size, compute_type=compute_type, progress=progress)
|
184 |
-
|
185 |
progress(0, desc="Loading Audio from Youtube..")
|
186 |
-
yt = get_ytdata(
|
187 |
audio = get_ytaudio(yt)
|
188 |
|
189 |
transcribed_segments, time_for_task = self.transcribe(
|
190 |
-
audio
|
191 |
-
|
192 |
-
|
193 |
-
beam_size=beam_size,
|
194 |
-
log_prob_threshold=log_prob_threshold,
|
195 |
-
no_speech_threshold=no_speech_threshold,
|
196 |
-
progress=progress
|
197 |
)
|
198 |
|
199 |
progress(1, desc="Completed!")
|
@@ -214,7 +157,7 @@ class FasterWhisperInference(BaseInterface):
|
|
214 |
finally:
|
215 |
try:
|
216 |
if 'yt' not in locals():
|
217 |
-
yt = get_ytdata(
|
218 |
file_path = get_ytaudio(yt)
|
219 |
else:
|
220 |
file_path = get_ytaudio(yt)
|
@@ -225,47 +168,24 @@ class FasterWhisperInference(BaseInterface):
|
|
225 |
pass
|
226 |
|
227 |
def transcribe_mic(self,
|
228 |
-
|
229 |
-
model_size: str,
|
230 |
-
lang: str,
|
231 |
file_format: str,
|
232 |
-
|
233 |
-
|
234 |
-
log_prob_threshold: float,
|
235 |
-
no_speech_threshold: float,
|
236 |
-
compute_type: str,
|
237 |
-
progress=gr.Progress()
|
238 |
) -> list:
|
239 |
"""
|
240 |
Write subtitle file from microphone
|
241 |
|
242 |
Parameters
|
243 |
----------
|
244 |
-
|
245 |
Audio file path from gr.Microphone()
|
246 |
-
model_size: str
|
247 |
-
Whisper model size from gr.Dropdown()
|
248 |
-
lang: str
|
249 |
-
Source language of the file to transcribe from gr.Dropdown()
|
250 |
file_format: str
|
251 |
-
File format to write from gr.Dropdown(). Supported format: [SRT, WebVTT, txt]
|
252 |
-
istranslate: bool
|
253 |
-
Boolean value from gr.Checkbox() that determines whether to translate to English.
|
254 |
-
It's Whisper's feature to translate speech from another language directly into English end-to-end.
|
255 |
-
beam_size: int
|
256 |
-
Int value from gr.Number() that is used for decoding option.
|
257 |
-
log_prob_threshold: float
|
258 |
-
float value from gr.Number(). If the average log probability over sampled tokens is
|
259 |
-
below this value, treat as failed.
|
260 |
-
no_speech_threshold: float
|
261 |
-
float value from gr.Number(). If the no_speech probability is higher than this value AND
|
262 |
-
the average log probability over sampled tokens is below `log_prob_threshold`,
|
263 |
-
compute_type: str
|
264 |
-
compute type from gr.Dropdown().
|
265 |
-
see more info : https://opennmt.net/CTranslate2/quantization.html
|
266 |
-
consider the segment as silent.
|
267 |
progress: gr.Progress
|
268 |
Indicator to show progress directly in gradio.
|
|
|
|
|
269 |
|
270 |
Returns
|
271 |
----------
|
@@ -274,18 +194,11 @@ class FasterWhisperInference(BaseInterface):
|
|
274 |
Files to return to gr.Files()
|
275 |
"""
|
276 |
try:
|
277 |
-
self.update_model_if_needed(model_size=model_size, compute_type=compute_type, progress=progress)
|
278 |
-
|
279 |
progress(0, desc="Loading Audio..")
|
280 |
-
|
281 |
transcribed_segments, time_for_task = self.transcribe(
|
282 |
-
|
283 |
-
|
284 |
-
|
285 |
-
beam_size=beam_size,
|
286 |
-
log_prob_threshold=log_prob_threshold,
|
287 |
-
no_speech_threshold=no_speech_threshold,
|
288 |
-
progress=progress
|
289 |
)
|
290 |
progress(1, desc="Completed!")
|
291 |
|
@@ -302,16 +215,12 @@ class FasterWhisperInference(BaseInterface):
|
|
302 |
print(f"Error transcribing file on line {e}")
|
303 |
finally:
|
304 |
self.release_cuda_memory()
|
305 |
-
self.remove_input_files([
|
306 |
|
307 |
def transcribe(self,
|
308 |
audio: Union[str, BinaryIO, np.ndarray],
|
309 |
-
|
310 |
-
|
311 |
-
beam_size: int,
|
312 |
-
log_prob_threshold: float,
|
313 |
-
no_speech_threshold: float,
|
314 |
-
progress: gr.Progress
|
315 |
) -> Tuple[List[dict], float]:
|
316 |
"""
|
317 |
transcribe method for faster-whisper.
|
@@ -320,22 +229,10 @@ class FasterWhisperInference(BaseInterface):
|
|
320 |
----------
|
321 |
audio: Union[str, BinaryIO, np.ndarray]
|
322 |
Audio path or file binary or Audio numpy array
|
323 |
-
lang: str
|
324 |
-
Source language of the file to transcribe from gr.Dropdown()
|
325 |
-
istranslate: bool
|
326 |
-
Boolean value from gr.Checkbox() that determines whether to translate to English.
|
327 |
-
It's Whisper's feature to translate speech from another language directly into English end-to-end.
|
328 |
-
beam_size: int
|
329 |
-
Int value from gr.Number() that is used for decoding option.
|
330 |
-
log_prob_threshold: float
|
331 |
-
float value from gr.Number(). If the average log probability over sampled tokens is
|
332 |
-
below this value, treat as failed.
|
333 |
-
no_speech_threshold: float
|
334 |
-
float value from gr.Number(). If the no_speech probability is higher than this value AND
|
335 |
-
the average log probability over sampled tokens is below `log_prob_threshold`,
|
336 |
-
consider the segment as silent.
|
337 |
progress: gr.Progress
|
338 |
Indicator to show progress directly in gradio.
|
|
|
|
|
339 |
|
340 |
Returns
|
341 |
----------
|
@@ -346,18 +243,24 @@ class FasterWhisperInference(BaseInterface):
|
|
346 |
"""
|
347 |
start_time = time.time()
|
348 |
|
349 |
-
|
|
|
|
|
|
|
|
|
|
|
350 |
lang = None
|
351 |
else:
|
352 |
language_code_dict = {value: key for key, value in whisper.tokenizer.LANGUAGES.items()}
|
353 |
-
lang = language_code_dict[lang]
|
|
|
354 |
segments, info = self.model.transcribe(
|
355 |
audio=audio,
|
356 |
language=lang,
|
357 |
-
task="translate" if
|
358 |
-
beam_size=beam_size,
|
359 |
-
log_prob_threshold=log_prob_threshold,
|
360 |
-
no_speech_threshold=no_speech_threshold,
|
361 |
)
|
362 |
progress(0, desc="Loading audio..")
|
363 |
|
@@ -373,24 +276,23 @@ class FasterWhisperInference(BaseInterface):
|
|
373 |
elapsed_time = time.time() - start_time
|
374 |
return segments_result, elapsed_time
|
375 |
|
376 |
-
def
|
377 |
-
|
378 |
-
|
379 |
-
|
380 |
-
|
381 |
"""
|
382 |
-
|
383 |
"""
|
384 |
-
|
385 |
-
|
386 |
-
|
387 |
-
|
388 |
-
self.
|
389 |
-
|
390 |
-
|
391 |
-
|
392 |
-
|
393 |
-
)
|
394 |
|
395 |
@staticmethod
|
396 |
def generate_and_write_file(file_name: str,
|
|
|
1 |
import os
|
2 |
|
|
|
3 |
import time
|
4 |
import numpy as np
|
5 |
from typing import BinaryIO, Union, Tuple, List
|
6 |
+
from datetime import datetime
|
7 |
|
8 |
import faster_whisper
|
9 |
import ctranslate2
|
|
|
14 |
from .base_interface import BaseInterface
|
15 |
from modules.subtitle_manager import get_srt, get_vtt, get_txt, write_file, safe_filename
|
16 |
from modules.youtube_manager import get_ytdata, get_ytaudio
|
17 |
+
from modules.whisper_data_class import *
|
18 |
|
19 |
|
20 |
class FasterWhisperInference(BaseInterface):
|
|
|
26 |
self.available_langs = sorted(list(whisper.tokenizer.LANGUAGES.values()))
|
27 |
self.translatable_models = ["large", "large-v1", "large-v2", "large-v3"]
|
28 |
self.device = "cuda" if torch.cuda.is_available() else "cpu"
|
29 |
+
self.available_compute_types = ctranslate2.get_supported_compute_types(
|
30 |
+
"cuda") if self.device == "cuda" else ctranslate2.get_supported_compute_types("cpu")
|
31 |
self.current_compute_type = "float16" if self.device == "cuda" else "float32"
|
32 |
self.default_beam_size = 1
|
33 |
|
34 |
def transcribe_file(self,
|
35 |
fileobjs: list,
|
|
|
|
|
36 |
file_format: str,
|
|
|
37 |
add_timestamp: bool,
|
38 |
+
progress=gr.Progress(),
|
39 |
+
*whisper_params,
|
|
|
|
|
|
|
40 |
) -> list:
|
41 |
"""
|
42 |
Write subtitle file from Files
|
|
|
45 |
----------
|
46 |
fileobjs: list
|
47 |
List of files to transcribe from gr.Files()
|
|
|
|
|
|
|
|
|
48 |
file_format: str
|
49 |
+
Subtitle File format to write from gr.Dropdown(). Supported format: [SRT, WebVTT, txt]
|
|
|
|
|
|
|
50 |
add_timestamp: bool
|
51 |
+
Boolean value from gr.Checkbox() that determines whether to add a timestamp at the end of the subtitle filename.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
52 |
progress: gr.Progress
|
53 |
Indicator to show progress directly in gradio.
|
54 |
+
*whisper_params: tuple
|
55 |
+
Gradio components related to Whisper. see whisper_data_class.py for details.
|
56 |
|
57 |
Returns
|
58 |
----------
|
|
|
61 |
Files to return to gr.Files()
|
62 |
"""
|
63 |
try:
|
|
|
|
|
64 |
files_info = {}
|
65 |
for fileobj in fileobjs:
|
66 |
transcribed_segments, time_for_task = self.transcribe(
|
67 |
+
fileobj.name,
|
68 |
+
progress,
|
69 |
+
*whisper_params,
|
|
|
|
|
|
|
|
|
70 |
)
|
71 |
|
72 |
file_name, file_ext = os.path.splitext(os.path.basename(fileobj.name))
|
|
|
77 |
add_timestamp=add_timestamp,
|
78 |
file_format=file_format
|
79 |
)
|
80 |
+
files_info[file_name] = {"subtitle": subtitle, "time_for_task": time_for_task, "path": file_path}
|
81 |
|
82 |
total_result = ''
|
83 |
total_time = 0
|
|
|
87 |
total_result += f'{info["subtitle"]}'
|
88 |
total_time += info["time_for_task"]
|
89 |
|
90 |
+
result_str = f"Done in {self.format_time(total_time)}! Subtitle is in the outputs folder.\n\n{total_result}"
|
91 |
+
result_file_path = [info['path'] for info in files_info.values()]
|
92 |
|
93 |
+
return [result_str, result_file_path]
|
94 |
|
95 |
except Exception as e:
|
96 |
print(f"Error transcribing file on line {e}")
|
|
|
100 |
self.remove_input_files([fileobj.name for fileobj in fileobjs])
|
101 |
|
102 |
def transcribe_youtube(self,
|
103 |
+
youtube_link: str,
|
|
|
|
|
104 |
file_format: str,
|
|
|
105 |
add_timestamp: bool,
|
106 |
+
progress=gr.Progress(),
|
107 |
+
*whisper_params,
|
|
|
|
|
|
|
108 |
) -> list:
|
109 |
"""
|
110 |
Write subtitle file from Youtube
|
111 |
|
112 |
Parameters
|
113 |
----------
|
114 |
+
youtube_link: str
|
115 |
+
URL of the Youtube video to transcribe from gr.Textbox()
|
|
|
|
|
|
|
|
|
116 |
file_format: str
|
117 |
+
Subtitle File format to write from gr.Dropdown(). Supported format: [SRT, WebVTT, txt]
|
|
|
|
|
|
|
118 |
add_timestamp: bool
|
119 |
Boolean value from gr.Checkbox() that determines whether to add a timestamp at the end of the filename.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
120 |
progress: gr.Progress
|
121 |
Indicator to show progress directly in gradio.
|
122 |
+
*whisper_params: tuple
|
123 |
+
Gradio components related to Whisper. see whisper_data_class.py for details.
|
124 |
|
125 |
Returns
|
126 |
----------
|
|
|
129 |
Files to return to gr.Files()
|
130 |
"""
|
131 |
try:
|
|
|
|
|
132 |
progress(0, desc="Loading Audio from Youtube..")
|
133 |
+
yt = get_ytdata(youtube_link)
|
134 |
audio = get_ytaudio(yt)
|
135 |
|
136 |
transcribed_segments, time_for_task = self.transcribe(
|
137 |
+
audio,
|
138 |
+
progress,
|
139 |
+
*whisper_params,
|
|
|
|
|
|
|
|
|
140 |
)
|
141 |
|
142 |
progress(1, desc="Completed!")
|
|
|
157 |
finally:
|
158 |
try:
|
159 |
if 'yt' not in locals():
|
160 |
+
yt = get_ytdata(youtube_link)
|
161 |
file_path = get_ytaudio(yt)
|
162 |
else:
|
163 |
file_path = get_ytaudio(yt)
|
|
|
168 |
pass
|
169 |
|
170 |
def transcribe_mic(self,
|
171 |
+
mic_audio: str,
|
|
|
|
|
172 |
file_format: str,
|
173 |
+
progress=gr.Progress(),
|
174 |
+
*whisper_params,
|
|
|
|
|
|
|
|
|
175 |
) -> list:
|
176 |
"""
|
177 |
Write subtitle file from microphone
|
178 |
|
179 |
Parameters
|
180 |
----------
|
181 |
+
mic_audio: str
|
182 |
Audio file path from gr.Microphone()
|
|
|
|
|
|
|
|
|
183 |
file_format: str
|
184 |
+
Subtitle File format to write from gr.Dropdown(). Supported format: [SRT, WebVTT, txt]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
185 |
progress: gr.Progress
|
186 |
Indicator to show progress directly in gradio.
|
187 |
+
*whisper_params: tuple
|
188 |
+
Gradio components related to Whisper. see whisper_data_class.py for details.
|
189 |
|
190 |
Returns
|
191 |
----------
|
|
|
194 |
Files to return to gr.Files()
|
195 |
"""
|
196 |
try:
|
|
|
|
|
197 |
progress(0, desc="Loading Audio..")
|
|
|
198 |
transcribed_segments, time_for_task = self.transcribe(
|
199 |
+
mic_audio,
|
200 |
+
progress,
|
201 |
+
*whisper_params,
|
|
|
|
|
|
|
|
|
202 |
)
|
203 |
progress(1, desc="Completed!")
|
204 |
|
|
|
215 |
print(f"Error transcribing file on line {e}")
|
216 |
finally:
|
217 |
self.release_cuda_memory()
|
218 |
+
self.remove_input_files([mic_audio])
|
219 |
|
220 |
def transcribe(self,
|
221 |
audio: Union[str, BinaryIO, np.ndarray],
|
222 |
+
progress: gr.Progress,
|
223 |
+
*whisper_params,
|
|
|
|
|
|
|
|
|
224 |
) -> Tuple[List[dict], float]:
|
225 |
"""
|
226 |
transcribe method for faster-whisper.
|
|
|
229 |
----------
|
230 |
audio: Union[str, BinaryIO, np.ndarray]
|
231 |
Audio path or file binary or Audio numpy array
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
232 |
progress: gr.Progress
|
233 |
Indicator to show progress directly in gradio.
|
234 |
+
*whisper_params: tuple
|
235 |
+
Gradio components related to Whisper. see whisper_data_class.py for details.
|
236 |
|
237 |
Returns
|
238 |
----------
|
|
|
243 |
"""
|
244 |
start_time = time.time()
|
245 |
|
246 |
+
params = WhisperGradioComponents.to_values(*whisper_params)
|
247 |
+
|
248 |
+
if params.model_size != self.current_model_size or self.model is None or self.current_compute_type != params.compute_type:
|
249 |
+
self.update_model(params.model_size, params.compute_type, progress)
|
250 |
+
|
251 |
+
if params.lang == "Automatic Detection":
|
252 |
lang = None
|
253 |
else:
|
254 |
language_code_dict = {value: key for key, value in whisper.tokenizer.LANGUAGES.items()}
|
255 |
+
lang = language_code_dict[params.lang]
|
256 |
+
|
257 |
segments, info = self.model.transcribe(
|
258 |
audio=audio,
|
259 |
language=lang,
|
260 |
+
task="translate" if params.is_translate and self.current_model_size in self.translatable_models else "transcribe",
|
261 |
+
beam_size=params.beam_size,
|
262 |
+
log_prob_threshold=params.log_prob_threshold,
|
263 |
+
no_speech_threshold=params.no_speech_threshold,
|
264 |
)
|
265 |
progress(0, desc="Loading audio..")
|
266 |
|
|
|
276 |
elapsed_time = time.time() - start_time
|
277 |
return segments_result, elapsed_time
|
278 |
|
279 |
+
def update_model(self,
|
280 |
+
model_size: str,
|
281 |
+
compute_type: str,
|
282 |
+
progress: gr.Progress
|
283 |
+
):
|
284 |
"""
|
285 |
+
update current model setting
|
286 |
"""
|
287 |
+
progress(0, desc="Initializing Model..")
|
288 |
+
self.current_model_size = model_size
|
289 |
+
self.current_compute_type = compute_type
|
290 |
+
self.model = faster_whisper.WhisperModel(
|
291 |
+
device=self.device,
|
292 |
+
model_size_or_path=model_size,
|
293 |
+
download_root=os.path.join("models", "Whisper", "faster-whisper"),
|
294 |
+
compute_type=self.current_compute_type
|
295 |
+
)
|
|
|
296 |
|
297 |
@staticmethod
|
298 |
def generate_and_write_file(file_name: str,
|