Spaces:
Sleeping
Sleeping
Upload folder using huggingface_hub
Browse files
.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
README.md
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
---
|
2 |
title: EwMeetingNotes
|
3 |
-
emoji:
|
4 |
-
colorFrom:
|
5 |
-
colorTo:
|
6 |
sdk: gradio
|
7 |
sdk_version: 4.12.0
|
8 |
app_file: app.py
|
|
|
1 |
---
|
2 |
title: EwMeetingNotes
|
3 |
+
emoji: 🏢
|
4 |
+
colorFrom: red
|
5 |
+
colorTo: blue
|
6 |
sdk: gradio
|
7 |
sdk_version: 4.12.0
|
8 |
app_file: app.py
|
app.py
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import whisper
|
3 |
+
# Use a pipeline as a high-level helper
|
4 |
+
from transformers import pipeline
|
5 |
+
from datetime import timedelta
|
6 |
+
|
7 |
+
model_size = "medium"
|
8 |
+
model = whisper.load_model(model_size)
|
9 |
+
summarizer_pipe = pipeline('summarization', model="sshleifer/distilbart-cnn-12-6")
|
10 |
+
|
11 |
+
def format_seconds(seconds):
|
12 |
+
# Create a timedelta object with the given seconds
|
13 |
+
delta = timedelta(seconds=seconds)
|
14 |
+
|
15 |
+
# Format the timedelta as a string in the desired format
|
16 |
+
formatted_time = str(delta)
|
17 |
+
|
18 |
+
# Extract hours, minutes, and seconds
|
19 |
+
hours, remainder = divmod(delta.seconds, 3600)
|
20 |
+
minutes, seconds = divmod(remainder, 60)
|
21 |
+
|
22 |
+
# Format milliseconds with three decimal places
|
23 |
+
# milliseconds = int(delta.microseconds / 1000)
|
24 |
+
|
25 |
+
# Create the final formatted string
|
26 |
+
formatted_string = "{:02}:{:02}:{:02}".format(hours, minutes, seconds)
|
27 |
+
|
28 |
+
return formatted_string
|
29 |
+
|
30 |
+
|
31 |
+
def process_meeting_video(mp3_path):
|
32 |
+
transcription = model.transcribe(mp3_path, word_timestamps=True)
|
33 |
+
transcript_text = ""
|
34 |
+
for segment in transcription["segments"]:
|
35 |
+
start = segment['start']
|
36 |
+
end = segment['end']
|
37 |
+
text = segment['text']
|
38 |
+
if start == 0.0000:
|
39 |
+
startTime = "00:00:00"
|
40 |
+
else:
|
41 |
+
startTime = format_seconds(start)
|
42 |
+
endTime = format_seconds(end)
|
43 |
+
transcript_text += (f"[{startTime}->{endTime}] {text}\n")
|
44 |
+
|
45 |
+
summary_output = summarizer_pipe(transcription["text"])[0]['summary_text']
|
46 |
+
return transcript_text,summary_output
|
47 |
+
|
48 |
+
summary_box = gr.Textbox(label="Summary", lines=10)
|
49 |
+
transcription_box = gr.Textbox(label="Transcript", lines=30)
|
50 |
+
demo = gr.Interface(process_meeting_video,
|
51 |
+
inputs=gr.File(),
|
52 |
+
outputs=[transcription_box, summary_box],
|
53 |
+
)
|
54 |
+
|
55 |
+
if __name__ == "__main__":
|
56 |
+
demo.launch()
|
requirements.txt
ADDED
@@ -0,0 +1,222 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
absl-py==2.0.0
|
2 |
+
accelerate==0.24.1
|
3 |
+
aiofiles==23.2.1
|
4 |
+
aiohttp==3.8.6
|
5 |
+
aiosignal==1.3.1
|
6 |
+
altair==5.1.2
|
7 |
+
annotated-types==0.6.0
|
8 |
+
anyio==3.7.1
|
9 |
+
appnope==0.1.3
|
10 |
+
argon2-cffi==23.1.0
|
11 |
+
argon2-cffi-bindings==21.2.0
|
12 |
+
arrow==1.3.0
|
13 |
+
asttokens==2.4.1
|
14 |
+
async-lru==2.0.4
|
15 |
+
async-timeout==4.0.3
|
16 |
+
attrs==23.1.0
|
17 |
+
audioread==3.0.1
|
18 |
+
Babel==2.13.1
|
19 |
+
backcall==0.2.0
|
20 |
+
beautifulsoup4==4.12.2
|
21 |
+
bleach==6.1.0
|
22 |
+
blis==0.7.11
|
23 |
+
cachetools==5.3.2
|
24 |
+
catalogue==2.0.10
|
25 |
+
certifi==2023.7.22
|
26 |
+
cffi==1.16.0
|
27 |
+
charset-normalizer==3.3.0
|
28 |
+
click==8.1.7
|
29 |
+
cloudpathlib==0.16.0
|
30 |
+
colorama==0.4.6
|
31 |
+
comm==0.1.4
|
32 |
+
confection==0.1.4
|
33 |
+
contourpy==1.1.1
|
34 |
+
cycler==0.12.1
|
35 |
+
cymem==2.0.8
|
36 |
+
datasets==2.14.6
|
37 |
+
debugpy==1.8.0
|
38 |
+
decorator==5.1.1
|
39 |
+
deepmultilingualpunctuation==1.0.1
|
40 |
+
defusedxml==0.7.1
|
41 |
+
dill==0.3.7
|
42 |
+
distro==1.9.0
|
43 |
+
en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.7.1/en_core_web_sm-3.7.1-py3-none-any.whl#sha256=86cc141f63942d4b2c5fcee06630fd6f904788d2f0ab005cce45aadb8fb73889
|
44 |
+
evaluate==0.4.1
|
45 |
+
executing==2.0.0
|
46 |
+
fastapi==0.104.1
|
47 |
+
fastjsonschema==2.18.1
|
48 |
+
ffmpy==0.3.1
|
49 |
+
filelock==3.12.4
|
50 |
+
fonttools==4.43.1
|
51 |
+
fqdn==1.5.1
|
52 |
+
frozenlist==1.4.0
|
53 |
+
fsspec==2023.9.2
|
54 |
+
google-auth==2.23.4
|
55 |
+
google-auth-oauthlib==1.1.0
|
56 |
+
gradio==4.0.2
|
57 |
+
gradio_client==0.7.0
|
58 |
+
grpcio==1.59.2
|
59 |
+
h11==0.14.0
|
60 |
+
httpcore==0.18.0
|
61 |
+
httpx==0.25.0
|
62 |
+
huggingface-hub==0.19.4
|
63 |
+
idna==3.4
|
64 |
+
importlib-resources==6.1.0
|
65 |
+
ipykernel==6.26.0
|
66 |
+
ipython==8.16.1
|
67 |
+
ipywidgets==8.1.1
|
68 |
+
isoduration==20.11.0
|
69 |
+
jedi==0.19.1
|
70 |
+
Jinja2==3.1.2
|
71 |
+
jiwer==3.0.3
|
72 |
+
joblib==1.3.2
|
73 |
+
json5==0.9.14
|
74 |
+
jsonpointer==2.4
|
75 |
+
jsonschema==4.19.1
|
76 |
+
jsonschema-specifications==2023.7.1
|
77 |
+
jupyter-events==0.8.0
|
78 |
+
jupyter-lsp==2.2.0
|
79 |
+
jupyter_client==8.5.0
|
80 |
+
jupyter_core==5.4.0
|
81 |
+
jupyter_server==2.9.1
|
82 |
+
jupyter_server_terminals==0.4.4
|
83 |
+
jupyterlab==4.0.7
|
84 |
+
jupyterlab-pygments==0.2.2
|
85 |
+
jupyterlab-widgets==3.0.9
|
86 |
+
jupyterlab_server==2.25.0
|
87 |
+
kiwisolver==1.4.5
|
88 |
+
langcodes==3.3.0
|
89 |
+
lazy_loader==0.3
|
90 |
+
librosa==0.10.1
|
91 |
+
llvmlite==0.41.0
|
92 |
+
lxml==4.9.3
|
93 |
+
Markdown==3.5.1
|
94 |
+
markdown-it-py==3.0.0
|
95 |
+
MarkupSafe==2.1.3
|
96 |
+
matplotlib==3.8.1
|
97 |
+
matplotlib-inline==0.1.6
|
98 |
+
mdurl==0.1.2
|
99 |
+
mistune==3.0.2
|
100 |
+
more-itertools==10.1.0
|
101 |
+
mpmath==1.3.0
|
102 |
+
msgpack==1.0.7
|
103 |
+
multidict==6.0.4
|
104 |
+
multiprocess==0.70.15
|
105 |
+
murmurhash==1.0.10
|
106 |
+
nbclient==0.8.0
|
107 |
+
nbconvert==7.9.2
|
108 |
+
nbformat==5.9.2
|
109 |
+
nest-asyncio==1.5.8
|
110 |
+
networkx==3.1
|
111 |
+
nltk==3.8.1
|
112 |
+
notebook_shim==0.2.3
|
113 |
+
numba==0.58.0
|
114 |
+
numpy==1.25.2
|
115 |
+
oauthlib==3.2.2
|
116 |
+
openai==1.6.1
|
117 |
+
openai-whisper @ git+https://github.com/openai/whisper.git@e58f28804528831904c3b6f2c0e473f346223433
|
118 |
+
orjson==3.9.10
|
119 |
+
overrides==7.4.0
|
120 |
+
packaging==23.2
|
121 |
+
pandas==2.1.2
|
122 |
+
pandocfilters==1.5.0
|
123 |
+
parso==0.8.3
|
124 |
+
pexpect==4.8.0
|
125 |
+
pickleshare==0.7.5
|
126 |
+
Pillow==10.1.0
|
127 |
+
platformdirs==3.11.0
|
128 |
+
pooch==1.8.0
|
129 |
+
portalocker==2.8.2
|
130 |
+
preshed==3.0.9
|
131 |
+
prometheus-client==0.17.1
|
132 |
+
prompt-toolkit==3.0.39
|
133 |
+
protobuf==4.23.4
|
134 |
+
psutil==5.9.6
|
135 |
+
ptyprocess==0.7.0
|
136 |
+
pure-eval==0.2.2
|
137 |
+
pyarrow==14.0.0
|
138 |
+
pyasn1==0.5.0
|
139 |
+
pyasn1-modules==0.3.0
|
140 |
+
pycparser==2.21
|
141 |
+
pydantic==2.4.2
|
142 |
+
pydantic_core==2.10.1
|
143 |
+
pydub==0.25.1
|
144 |
+
Pygments==2.16.1
|
145 |
+
pyparsing==3.1.1
|
146 |
+
pysubs2==1.6.1
|
147 |
+
python-dateutil==2.8.2
|
148 |
+
python-json-logger==2.0.7
|
149 |
+
python-multipart==0.0.6
|
150 |
+
pytube==15.0.0
|
151 |
+
pytz==2023.3.post1
|
152 |
+
PyYAML==6.0.1
|
153 |
+
pyzmq==25.1.1
|
154 |
+
rapidfuzz==3.5.1
|
155 |
+
referencing==0.30.2
|
156 |
+
regex==2023.10.3
|
157 |
+
requests==2.31.0
|
158 |
+
requests-oauthlib==1.3.1
|
159 |
+
responses==0.18.0
|
160 |
+
rfc3339-validator==0.1.4
|
161 |
+
rfc3986-validator==0.1.1
|
162 |
+
rich==13.6.0
|
163 |
+
rpds-py==0.10.6
|
164 |
+
rsa==4.9
|
165 |
+
sacrebleu==2.4.0
|
166 |
+
safetensors==0.4.0
|
167 |
+
scikit-learn==1.3.2
|
168 |
+
scipy==1.11.3
|
169 |
+
semantic-version==2.10.0
|
170 |
+
Send2Trash==1.8.2
|
171 |
+
sentence-transformers==2.2.2
|
172 |
+
sentencepiece==0.1.99
|
173 |
+
shellingham==1.5.4
|
174 |
+
six==1.16.0
|
175 |
+
smart-open==6.4.0
|
176 |
+
sniffio==1.3.0
|
177 |
+
soundfile==0.12.1
|
178 |
+
soupsieve==2.5
|
179 |
+
soxr==0.3.7
|
180 |
+
spacy==3.7.2
|
181 |
+
spacy-legacy==3.0.12
|
182 |
+
spacy-loggers==1.0.5
|
183 |
+
srsly==2.4.8
|
184 |
+
stack-data==0.6.3
|
185 |
+
starlette==0.27.0
|
186 |
+
sympy==1.12
|
187 |
+
tabulate==0.9.0
|
188 |
+
tensorboard==2.15.0
|
189 |
+
tensorboard-data-server==0.7.2
|
190 |
+
terminado==0.17.1
|
191 |
+
thinc==8.2.2
|
192 |
+
threadpoolctl==3.2.0
|
193 |
+
tiktoken==0.3.3
|
194 |
+
tinycss2==1.2.1
|
195 |
+
tokenizers==0.15.0
|
196 |
+
tomlkit==0.12.0
|
197 |
+
toolz==0.12.0
|
198 |
+
torch==2.1.2
|
199 |
+
torchaudio==2.1.2
|
200 |
+
torchvision==0.16.2
|
201 |
+
tornado==6.3.3
|
202 |
+
tqdm==4.66.1
|
203 |
+
traitlets==5.12.0
|
204 |
+
transformers @ git+https://github.com/huggingface/transformers@35551f9a0f66a22de4971b4a51b3c172d3b87f95
|
205 |
+
typer==0.9.0
|
206 |
+
types-python-dateutil==2.8.19.14
|
207 |
+
typing_extensions==4.8.0
|
208 |
+
tzdata==2023.3
|
209 |
+
uri-template==1.3.0
|
210 |
+
urllib3==2.0.6
|
211 |
+
uvicorn==0.23.2
|
212 |
+
wasabi==1.1.2
|
213 |
+
wcwidth==0.2.8
|
214 |
+
weasel==0.3.4
|
215 |
+
webcolors==1.13
|
216 |
+
webencodings==0.5.1
|
217 |
+
websocket-client==1.6.4
|
218 |
+
websockets==11.0.3
|
219 |
+
Werkzeug==3.0.1
|
220 |
+
widgetsnbextension==4.0.9
|
221 |
+
xxhash==3.4.1
|
222 |
+
yarl==1.9.2
|