Spaces:
Running
Running
Update transcription.py
Browse files- transcription.py +21 -25
transcription.py
CHANGED
@@ -2,7 +2,7 @@ import os
|
|
2 |
from faster_whisper import WhisperModel
|
3 |
|
4 |
class TranscriptionMaker():
|
5 |
-
|
6 |
def __init__(self,output_dir=os.path.abspath("/tmp/data/transcriptions")):
|
7 |
self.model = WhisperModel("base", device="cpu")
|
8 |
self.output_dir = output_dir
|
@@ -13,35 +13,31 @@ class TranscriptionMaker():
|
|
13 |
print(f"Error creating directory {self.output_dir}: {e}")
|
14 |
raise
|
15 |
|
16 |
-
|
17 |
-
def create_transcription(self,
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
|
|
24 |
|
|
|
25 |
for segment in segments:
|
26 |
results.append({
|
27 |
"start": segment.start,
|
28 |
"end": segment.end,
|
29 |
"text": segment.text
|
30 |
})
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
print(f"Error writing transcription file: {e}")
|
40 |
-
raise
|
41 |
-
return output_file
|
42 |
-
except FileNotFoundError as e:
|
43 |
-
print(f"Error: {e}")
|
44 |
raise
|
45 |
-
|
46 |
-
|
47 |
-
raise
|
|
|
2 |
from faster_whisper import WhisperModel
|
3 |
|
4 |
class TranscriptionMaker():
|
5 |
+
#書き起こしファイルを吐き出すディレクトリを指定
|
6 |
def __init__(self,output_dir=os.path.abspath("/tmp/data/transcriptions")):
|
7 |
self.model = WhisperModel("base", device="cpu")
|
8 |
self.output_dir = output_dir
|
|
|
13 |
print(f"Error creating directory {self.output_dir}: {e}")
|
14 |
raise
|
15 |
|
16 |
+
#音声ファイルのディレクトリを受け取り、書き起こしファイルを作成する
|
17 |
+
def create_transcription(self,audio_directory):
|
18 |
+
results = []
|
19 |
+
#ディレクトリ内のファイルを全て取得
|
20 |
+
audio_files = os.listdir(audio_directory)
|
21 |
+
for audio_file in audio_files:
|
22 |
+
if os.path.splitext(audio_file)[-1].lower() != '.wav':
|
23 |
+
continue
|
24 |
+
audio_path = os.path.join(audio_directory, audio_file)
|
25 |
|
26 |
+
segments,info = list(self.model.transcribe(audio_path))
|
27 |
for segment in segments:
|
28 |
results.append({
|
29 |
"start": segment.start,
|
30 |
"end": segment.end,
|
31 |
"text": segment.text
|
32 |
})
|
33 |
+
#ファイルの書き込み。ファイル名は"読み込みディレクトリ名_transcription.txt"
|
34 |
+
output_file=os.path.join(self.output_dir,os.path.basename(audio_directory)+"_transcription.txt")
|
35 |
+
try:
|
36 |
+
with open(output_file,"w",encoding="utf-8") as f:
|
37 |
+
for result in results:
|
38 |
+
f.write(f"{result['text']}\n")
|
39 |
+
except OSError as e:
|
40 |
+
print(f"Error writing transcription file: {e}")
|
|
|
|
|
|
|
|
|
|
|
41 |
raise
|
42 |
+
return output_file
|
43 |
+
|
|