Hervé Bredin commited on
Commit
ebc74bd
0 Parent(s):

feat: initial import

Browse files
.github/workflows/check_size.yaml ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ name: Check file size
2
+
3
+ on:
4
+ pull_request:
5
+ branches: [main]
6
+
7
+ # to run this workflow manually from the Actions tab
8
+ workflow_dispatch:
9
+
10
+ jobs:
11
+ sync-to-hub:
12
+ runs-on: ubuntu-latest
13
+ steps:
14
+ - name: Check large files
15
+ uses: ActionsDesk/[email protected]
16
+ with:
17
+ filesizelimit: 10485760 # = 10MB, so we can sync to HF spaces
.github/workflows/sync_to_hub.yaml ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ name: Sync to Hugging Face hub
2
+
3
+ on:
4
+ push:
5
+ branches: [main]
6
+
7
+ # to run this workflow manually from the Actions tab
8
+ workflow_dispatch:
9
+
10
+ jobs:
11
+ sync-to-hub:
12
+ runs-on: ubuntu-latest
13
+ steps:
14
+ - uses: actions/checkout@v2
15
+ with:
16
+ fetch-depth: 0
17
+ - name: Push to hub
18
+ env:
19
+ HF_TOKEN: ${{ secrets.HF_TOKEN }}
20
+ run: git push https://hbredin:[email protected]/spaces/pyannote/pretrained-pipelines main --force
README.md ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ title: Pretrained pipelines
3
+ emoji: 🎹
4
+ colorFrom: red
5
+ colorTo: red
6
+ sdk: streamlit
7
+ app_file: app.py
8
+ pinned: true
9
+ ---
10
+
11
+ # Configuration
12
+
13
+ `title`: _string_
14
+ Display title for the Space
15
+
16
+ `emoji`: _string_
17
+ Space emoji (emoji-only character allowed)
18
+
19
+ `colorFrom`: _string_
20
+ Color for Thumbnail gradient (red, yellow, green, blue, indigo, purple, pink, gray)
21
+
22
+ `colorTo`: _string_
23
+ Color for Thumbnail gradient (red, yellow, green, blue, indigo, purple, pink, gray)
24
+
25
+ `sdk`: _string_
26
+ Can be either `gradio` or `streamlit`
27
+
28
+ `sdk_version` : _string_
29
+ Only applicable for `streamlit` SDK.
30
+ See [doc](https://hf.co/docs/hub/spaces) for more info on supported versions.
31
+
32
+ `app_file`: _string_
33
+ Path to your main application file (which contains either `gradio` or `streamlit` Python code).
34
+ Path is relative to the root of the repository.
35
+
36
+ `pinned`: _boolean_
37
+ Whether the Space stays on top of your list.
app.py ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from huggingface_hub import HfApi
2
+ import matplotlib.pyplot as plt
3
+ import streamlit as st
4
+ from pyannote.audio import Pipeline
5
+ from pyannote.audio import Audio
6
+ from pyannote.core import notebook, Segment
7
+ import io
8
+ import base64
9
+
10
+ from matplotlib.backends.backend_agg import RendererAgg
11
+
12
+ _lock = RendererAgg.lock
13
+
14
+ PYANNOTE_LOGO = "https://avatars.githubusercontent.com/u/7559051?s=400&v=4"
15
+ EXCERPT = 30.0
16
+
17
+ st.set_page_config(
18
+ page_title="pyannote.audio pretrained pipelines",
19
+ page_icon=PYANNOTE_LOGO)
20
+
21
+ st.sidebar.image(PYANNOTE_LOGO)
22
+
23
+ st.markdown(
24
+ f"""
25
+ # 🎹 Pretrained pipelines
26
+
27
+ Upload an audio file and the first {EXCERPT:g} seconds will be processed automatically.
28
+ """
29
+ )
30
+
31
+ PIPELINES = [p.modelId for p in HfApi().list_models(filter="pyannote-audio-pipeline") if p.modelId.startswith("pyannote/")]
32
+
33
+ audio = Audio(sample_rate=16000, mono=True)
34
+
35
+ selected_pipeline = st.selectbox("", PIPELINES, index=0)
36
+
37
+ with st.spinner('Loading pipeline...'):
38
+ pipeline = Pipeline.from_pretrained(selected_pipeline)
39
+
40
+ uploaded_file = st.file_uploader("")
41
+ if uploaded_file is not None:
42
+
43
+ try:
44
+ duration = audio.get_duration(uploaded_file)
45
+ except RuntimeError as e:
46
+ st.error(e)
47
+ st.stop()
48
+ waveform, sample_rate = audio.crop(uploaded_file, Segment(0, min(duration, EXCERPT)))
49
+ file = {"waveform": waveform, "sample_rate": sample_rate, "uri": uploaded_file.name}
50
+
51
+ with st.spinner('Running pipeline...'):
52
+ output = pipeline(file)
53
+
54
+ with _lock:
55
+
56
+ notebook.reset()
57
+ notebook.crop = Segment(0, min(duration, EXCERPT))
58
+
59
+ fig, ax = plt.subplots(nrows=1, ncols=1)
60
+ fig.set_figwidth(12)
61
+ fig.set_figheight(2.0)
62
+ notebook.plot_annotation(output, ax=ax, time=True, legend=True)
63
+
64
+ plt.tight_layout()
65
+ st.pyplot(fig=fig, clear_figure=True)
66
+ plt.close(fig)
67
+
68
+ with io.StringIO() as fp:
69
+ output.write_rttm(fp)
70
+ content = fp.getvalue()
71
+
72
+ b64 = base64.b64encode(content.encode()).decode()
73
+ href = f'<a download="{output.uri}.rttm" href="data:file/text;base64,{b64}">Download as RTTM</a>'
74
+ st.markdown(href, unsafe_allow_html=True)
75
+
76
+
77
+ st.sidebar.markdown(
78
+ """
79
+ -------------------
80
+
81
+ To use these pipelines on more and longer files on your own (GPU, hence much faster) servers, check the [documentation](https://github.com/pyannote/pyannote-audio).
82
+
83
+ For [technical questions](https://github.com/pyannote/pyannote-audio/discussions) and [bug reports](https://github.com/pyannote/pyannote-audio/issues), please check [pyannote.audio](https://github.com/pyannote/pyannote-audio) Github repository.
84
+
85
+ For commercial enquiries and scientific consulting, please contact [me](mailto:[email protected]).
86
+ """
87
+ )
packages.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ libsndfile1
requirements.txt ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ git+https://github.com/pyannote/pyannote-audio.git@develop#egg=pyannote-audio
2
+ speechbrain == 0.5.10
3
+ matplotlib == 3.3.3
4
+