Spaces:
Running
Running
File size: 7,488 Bytes
63ab978 67cc6b1 40f2b57 91b9b83 63ab978 91b9b83 2ff01cb 91b9b83 67cc6b1 63ab978 67cc6b1 5845bfc f9abd83 63ab978 5845bfc 63ab978 91b9b83 63ab978 91b9b83 6074f61 a70c074 6074f61 a70c074 6074f61 67cc6b1 6074f61 67cc6b1 63ab978 6074f61 63ab978 6074f61 21c25c6 6074f61 a70c074 91b9b83 6074f61 21c25c6 91b9b83 21c25c6 91b9b83 21c25c6 91b9b83 21c25c6 6074f61 21c25c6 a70c074 21c25c6 40f2b57 21c25c6 40f2b57 21c25c6 6074f61 67cc6b1 6074f61 67cc6b1 91b9b83 5845bfc 91b9b83 63ab978 91b9b83 6074f61 a70c074 63ab978 6074f61 63ab978 6074f61 63ab978 6074f61 63ab978 6074f61 63ab978 6074f61 a70c074 63ab978 6074f61 91b9b83 6074f61 67cc6b1 6074f61 67cc6b1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 |
import whisper
from .base_interface import BaseInterface
from modules.subtitle_manager import get_srt, get_vtt, write_file, safe_filename
from modules.youtube_manager import get_ytdata, get_ytaudio
import gradio as gr
import os
from datetime import datetime
DEFAULT_MODEL_SIZE = "large-v2"
class WhisperInference(BaseInterface):
def __init__(self):
super().__init__()
self.current_model_size = None
self.model = None
self.available_models = whisper.available_models()
self.available_langs = sorted(list(whisper.tokenizer.LANGUAGES.values()))
def transcribe_file(self, fileobjs,
model_size, lang, subformat, istranslate,
progress=gr.Progress()):
def progress_callback(progress_value):
progress(progress_value, desc="Transcribing..")
try:
if model_size != self.current_model_size or self.model is None:
progress(0, desc="Initializing Model..")
self.current_model_size = model_size
self.model = whisper.load_model(name=model_size, download_root=os.path.join("models", "Whisper"))
if lang == "Automatic Detection":
lang = None
progress(0, desc="Loading Audio..")
files_info = {}
for fileobj in fileobjs:
audio = whisper.load_audio(fileobj.name)
translatable_model = ["large", "large-v1", "large-v2"]
if istranslate and self.current_model_size in translatable_model:
result = self.model.transcribe(audio=audio, language=lang, verbose=False, task="translate",
progress_callback=progress_callback)
else:
result = self.model.transcribe(audio=audio, language=lang, verbose=False,
progress_callback=progress_callback)
progress(1, desc="Completed!")
file_name, file_ext = os.path.splitext(os.path.basename(fileobj.orig_name))
file_name = file_name[:-9]
file_name = safe_filename(file_name)
timestamp = datetime.now().strftime("%m%d%H%M%S")
output_path = os.path.join("outputs", f"{file_name}-{timestamp}")
if subformat == "SRT":
subtitle = get_srt(result["segments"])
write_file(subtitle, f"{output_path}.srt")
elif subformat == "WebVTT":
subtitle = get_vtt(result["segments"])
write_file(subtitle, f"{output_path}.vtt")
files_info[file_name] = subtitle
total_result = ''
for file_name, subtitle in files_info.items():
total_result += '------------------------------------\n'
total_result += f'{file_name}\n\n'
total_result += f'{subtitle}'
return f"Done! Subtitle is in the outputs folder.\n\n{total_result}"
except Exception as e:
return f"Error: {str(e)}"
finally:
self.release_cuda_memory()
self.remove_input_files([fileobj.name for fileobj in fileobjs])
def transcribe_youtube(self, youtubelink,
model_size, lang, subformat, istranslate,
progress=gr.Progress()):
def progress_callback(progress_value):
progress(progress_value, desc="Transcribing..")
try:
if model_size != self.current_model_size or self.model is None:
progress(0, desc="Initializing Model..")
self.current_model_size = model_size
self.model = whisper.load_model(name=model_size, download_root=os.path.join("models", "Whisper"))
if lang == "Automatic Detection":
lang = None
progress(0, desc="Loading Audio from Youtube..")
yt = get_ytdata(youtubelink)
audio = whisper.load_audio(get_ytaudio(yt))
translatable_model = ["large", "large-v1", "large-v2"]
if istranslate and self.current_model_size in translatable_model:
result = self.model.transcribe(audio=audio, language=lang, verbose=False, task="translate",
progress_callback=progress_callback)
else:
result = self.model.transcribe(audio=audio, language=lang, verbose=False,
progress_callback=progress_callback)
progress(1, desc="Completed!")
file_name = safe_filename(yt.title)
timestamp = datetime.now().strftime("%m%d%H%M%S")
output_path = os.path.join("outputs", f"{file_name}-{timestamp}")
if subformat == "SRT":
subtitle = get_srt(result["segments"])
write_file(subtitle, f"{output_path}.srt")
elif subformat == "WebVTT":
subtitle = get_vtt(result["segments"])
write_file(subtitle, f"{output_path}.vtt")
return f"Done! Subtitle file is in the outputs folder.\n\n{subtitle}"
except Exception as e:
return f"Error: {str(e)}"
finally:
yt = get_ytdata(youtubelink)
file_path = get_ytaudio(yt)
self.release_cuda_memory()
self.remove_input_files([file_path])
def transcribe_mic(self, micaudio,
model_size, lang, subformat, istranslate,
progress=gr.Progress()):
def progress_callback(progress_value):
progress(progress_value, desc="Transcribing..")
try:
if model_size != self.current_model_size or self.model is None:
progress(0, desc="Initializing Model..")
self.current_model_size = model_size
self.model = whisper.load_model(name=model_size, download_root=os.path.join("models", "Whisper"))
if lang == "Automatic Detection":
lang = None
progress(0, desc="Loading Audio..")
translatable_model = ["large", "large-v1", "large-v2"]
if istranslate and self.current_model_size in translatable_model:
result = self.model.transcribe(audio=micaudio, language=lang, verbose=False, task="translate",
progress_callback=progress_callback)
else:
result = self.model.transcribe(audio=micaudio, language=lang, verbose=False,
progress_callback=progress_callback)
progress(1, desc="Completed!")
timestamp = datetime.now().strftime("%m%d%H%M%S")
output_path = os.path.join("outputs", f"{file_name}-{timestamp}")
if subformat == "SRT":
subtitle = get_srt(result["segments"])
write_file(subtitle, f"{output_path}.srt")
elif subformat == "WebVTT":
subtitle = get_vtt(result["segments"])
write_file(subtitle, f"{output_path}.vtt")
return f"Done! Subtitle file is in the outputs folder.\n\n{subtitle}"
except Exception as e:
return f"Error: {str(e)}"
finally:
self.release_cuda_memory()
self.remove_input_files([micaudio])
|