Update app.py
Browse files
app.py
CHANGED
@@ -1,126 +1,83 @@
|
|
1 |
-
import streamlit as st
|
2 |
-
import yt_dlp
|
3 |
import os
|
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 |
-
st.
|
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 |
-
def download_video(youtube_url, format_id):
|
87 |
-
ydl_opts = {
|
88 |
-
'format': f"{format_id}+bestaudio/best",
|
89 |
-
'merge_output_format': 'mp4',
|
90 |
-
'outtmpl': 'downloaded_video.%(ext)s',
|
91 |
-
'nocache': True, # μΊμ λΉνμ±ν
|
92 |
-
'force_generic_extractor': True # νμ μλ‘μ΄ URL μ¬μ©
|
93 |
-
}
|
94 |
-
with yt_dlp.YoutubeDL(ydl_opts) as ydl:
|
95 |
-
ydl.download([youtube_url])
|
96 |
-
|
97 |
-
# "μΆμΆ" λ²νΌ μΆκ°
|
98 |
-
if st.button('μΆμΆ'):
|
99 |
-
format_id = st.session_state.selected_format_id
|
100 |
-
if format_id is None:
|
101 |
-
st.error('Selected format not found.')
|
102 |
-
else:
|
103 |
-
try:
|
104 |
-
# μΈμ
μνμ youtube_url μ¬μ©
|
105 |
-
download_video(st.session_state.youtube_url, format_id)
|
106 |
-
st.success('Video downloaded successfully!')
|
107 |
-
st.video('downloaded_video.mp4')
|
108 |
-
|
109 |
-
# λ€μ΄λ‘λλ νμΌ μ 곡 (optional)
|
110 |
-
with open('downloaded_video.mp4', 'rb') as file:
|
111 |
-
st.download_button(
|
112 |
-
label="Download Video File",
|
113 |
-
data=file,
|
114 |
-
file_name='downloaded_video.mp4',
|
115 |
-
mime='video/mp4'
|
116 |
-
)
|
117 |
-
|
118 |
-
# μΆμΆλ μμ μ 보 μΆλ ₯
|
119 |
-
st.write(f"**μ λͺ©:** {st.session_state.info_dict['title']}")
|
120 |
-
st.write(f"**μ±λλͺ
:** {st.session_state.info_dict['uploader']}")
|
121 |
-
st.write(f"**μ
λ‘λ λ μ§:** {st.session_state.info_dict['upload_date']}")
|
122 |
-
|
123 |
-
# λ€μ΄λ‘λ ν μΈμ
μν 리μ
|
124 |
-
reset_session_state()
|
125 |
-
except Exception as e:
|
126 |
-
st.error(f"An error occurred during download: {e}")
|
|
|
|
|
|
|
1 |
import os
|
2 |
+
import streamlit as st
|
3 |
+
import torch
|
4 |
+
from transformers import WhisperForConditionalGeneration, WhisperProcessor
|
5 |
+
import librosa
|
6 |
+
import moviepy.editor as mp
|
7 |
+
import srt
|
8 |
+
from datetime import timedelta
|
9 |
+
from tempfile import NamedTemporaryFile
|
10 |
+
|
11 |
+
# λͺ¨λΈ λ° νλ‘μΈμ λ‘λ
|
12 |
+
@st.cache_resource
|
13 |
+
def load_model():
|
14 |
+
model = WhisperForConditionalGeneration.from_pretrained("lcjln/AIME_Project_The_Final")
|
15 |
+
processor = WhisperProcessor.from_pretrained("lcjln/AIME_The_Final")
|
16 |
+
return model, processor
|
17 |
+
|
18 |
+
model, processor = load_model()
|
19 |
+
|
20 |
+
# μΉ μ ν리μΌμ΄μ
μΈν°νμ΄μ€
|
21 |
+
st.title("Whisper μλ§ μμ±κΈ°")
|
22 |
+
st.write("μμ νμΌμ μ
λ‘λνμ¬ μλ§μ μμ±νμΈμ.")
|
23 |
+
|
24 |
+
# μμ νμΌ μ
λ‘λ
|
25 |
+
uploaded_file = st.file_uploader("μ¬κΈ°μ μμ νμΌμ λλκ·Έ μ€ λλ‘ νμΈμ", type=["mp4", "mkv", "mov"])
|
26 |
+
|
27 |
+
if uploaded_file is not None:
|
28 |
+
# μ§νλ° μ΄κΈ°ν
|
29 |
+
progress_bar = st.progress(0)
|
30 |
+
|
31 |
+
# μ
λ‘λλ μμ νμΌμμ μ€λμ€ μΆμΆ
|
32 |
+
st.write("μμμ μ²λ¦¬νλ μ€μ
λλ€...")
|
33 |
+
progress_bar.progress(10)
|
34 |
+
|
35 |
+
with NamedTemporaryFile(suffix=".mp4") as temp_video_file:
|
36 |
+
temp_video_file.write(uploaded_file.read())
|
37 |
+
video = mp.VideoFileClip(temp_video_file.name)
|
38 |
+
audio_path = temp_video_file.name.replace(".mp4", ".wav")
|
39 |
+
video.audio.write_audiofile(audio_path, codec='pcm_s16le')
|
40 |
+
|
41 |
+
progress_bar.progress(30)
|
42 |
+
|
43 |
+
# μ€λμ€ νμΌ λ‘λ λ° μ²λ¦¬
|
44 |
+
st.write("μ€λμ€ νμΌμ μ²λ¦¬νλ μ€μ
λλ€...")
|
45 |
+
audio, sr = librosa.load(audio_path, sr=16000)
|
46 |
+
|
47 |
+
progress_bar.progress(50)
|
48 |
+
|
49 |
+
# Whisper λͺ¨λΈλ‘ λ³ν
|
50 |
+
st.write("λͺ¨λΈμ ν΅ν΄ μλ§μ μμ±νλ μ€μ
λλ€...")
|
51 |
+
inputs = processor(audio, return_tensors="pt", sampling_rate=16000)
|
52 |
+
with torch.no_grad():
|
53 |
+
predicted_ids = model.generate(inputs["input_features"], max_length=2048)
|
54 |
+
|
55 |
+
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)[0].strip()
|
56 |
+
|
57 |
+
progress_bar.progress(80)
|
58 |
+
|
59 |
+
# μλ§μ SRT νμΌλ‘ μ μ₯
|
60 |
+
st.write("SRT νμΌμ μμ±νλ μ€μ
λλ€...")
|
61 |
+
lines = transcription.split(". ")
|
62 |
+
subs = []
|
63 |
+
step = len(audio) / sr / len(lines)
|
64 |
+
start_time = 0.0
|
65 |
+
|
66 |
+
for i, line in enumerate(lines):
|
67 |
+
end_time = start_time + step
|
68 |
+
subs.append(srt.Subtitle(index=i+1, start=timedelta(seconds=start_time), end=timedelta(seconds=end_time), content=line))
|
69 |
+
start_time = end_time
|
70 |
+
|
71 |
+
srt_content = srt.compose(subs)
|
72 |
+
|
73 |
+
srt_file_path = audio_path.replace(".wav", ".srt")
|
74 |
+
with open(srt_file_path, "w", encoding="utf-8") as f:
|
75 |
+
f.write(srt_content)
|
76 |
+
|
77 |
+
progress_bar.progress(100)
|
78 |
+
|
79 |
+
st.success("SRT νμΌμ΄ μ±κ³΅μ μΌλ‘ μμ±λμμ΅λλ€!")
|
80 |
+
|
81 |
+
# SRT νμΌ λ€μ΄λ‘λ λ²νΌ
|
82 |
+
with open(srt_file_path, "rb") as srt_file:
|
83 |
+
st.download_button(label="SRT νμΌ λ€μ΄λ‘λ", data=srt_file, file_name=os.path.basename(srt_file_path), mime="text/srt")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|