feat: initial import

.github/workflows/check_size.yaml

@@ -0,0 +1,17 @@
+name: Check file size
+
+on:
+  pull_request:
+    branches: [main]
+
+  # to run this workflow manually from the Actions tab
+  workflow_dispatch:
+
+jobs:
+  sync-to-hub:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check large files
+        uses: ActionsDesk/[email protected]
+        with:
+          filesizelimit: 10485760 # = 10MB, so we can sync to HF spaces

.github/workflows/sync_to_hub.yaml

@@ -0,0 +1,20 @@
+name: Sync to Hugging Face hub
+
+on:
+  push:
+    branches: [main]
+
+  # to run this workflow manually from the Actions tab
+  workflow_dispatch:
+
+jobs:
+  sync-to-hub:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          fetch-depth: 0
+      - name: Push to hub
+        env:
+          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        run: git push https://hbredin:[email protected]/spaces/pyannote/pretrained-pipelines main --force

README.md

@@ -0,0 +1,37 @@
+---
+title: Pretrained pipelines
+emoji: 🎹
+colorFrom: red
+colorTo: red
+sdk: streamlit
+app_file: app.py
+pinned: true
+---
+
+# Configuration
+
+`title`: _string_  
+Display title for the Space
+
+`emoji`: _string_  
+Space emoji (emoji-only character allowed)
+
+`colorFrom`: _string_  
+Color for Thumbnail gradient (red, yellow, green, blue, indigo, purple, pink, gray)
+
+`colorTo`: _string_  
+Color for Thumbnail gradient (red, yellow, green, blue, indigo, purple, pink, gray)
+
+`sdk`: _string_  
+Can be either `gradio` or `streamlit`
+
+`sdk_version` : _string_  
+Only applicable for `streamlit` SDK.  
+See [doc](https://hf.co/docs/hub/spaces) for more info on supported versions.
+
+`app_file`: _string_  
+Path to your main application file (which contains either `gradio` or `streamlit` Python code).  
+Path is relative to the root of the repository.
+
+`pinned`: _boolean_  
+Whether the Space stays on top of your list.

app.py

@@ -0,0 +1,87 @@
+from huggingface_hub import HfApi
+import matplotlib.pyplot as plt
+import streamlit as st
+from pyannote.audio import Pipeline
+from pyannote.audio import Audio
+from pyannote.core import notebook, Segment
+import io
+import base64
+
+from matplotlib.backends.backend_agg import RendererAgg
+
+_lock = RendererAgg.lock
+
+PYANNOTE_LOGO = "https://avatars.githubusercontent.com/u/7559051?s=400&v=4"
+EXCERPT = 30.0
+
+st.set_page_config(
+    page_title="pyannote.audio pretrained pipelines",
+    page_icon=PYANNOTE_LOGO)
+
+st.sidebar.image(PYANNOTE_LOGO)
+
+st.markdown(
+    f"""
+# 🎹 Pretrained pipelines
+
+Upload an audio file and the first {EXCERPT:g} seconds will be processed automatically.
+"""
+)
+
+PIPELINES = [p.modelId for p in HfApi().list_models(filter="pyannote-audio-pipeline") if p.modelId.startswith("pyannote/")]
+
+audio = Audio(sample_rate=16000, mono=True)
+
+selected_pipeline = st.selectbox("", PIPELINES, index=0)
+
+with st.spinner('Loading pipeline...'):
+    pipeline = Pipeline.from_pretrained(selected_pipeline)
+
+uploaded_file = st.file_uploader("")
+if uploaded_file is not None:
+
+    try:
+        duration = audio.get_duration(uploaded_file)
+    except RuntimeError as e:
+        st.error(e)
+        st.stop()
+    waveform, sample_rate = audio.crop(uploaded_file, Segment(0, min(duration, EXCERPT)))
+    file = {"waveform": waveform, "sample_rate": sample_rate, "uri": uploaded_file.name}
+
+    with st.spinner('Running pipeline...'):
+        output = pipeline(file)
+
+    with _lock:
+
+        notebook.reset()
+        notebook.crop = Segment(0, min(duration, EXCERPT))
+
+        fig, ax = plt.subplots(nrows=1, ncols=1)
+        fig.set_figwidth(12)
+        fig.set_figheight(2.0)
+        notebook.plot_annotation(output, ax=ax, time=True, legend=True)
+
+        plt.tight_layout()
+        st.pyplot(fig=fig, clear_figure=True)
+        plt.close(fig)
+
+    with io.StringIO() as fp:
+        output.write_rttm(fp)
+        content = fp.getvalue()
+
+        b64 = base64.b64encode(content.encode()).decode()
+        href = f'<a download="{output.uri}.rttm" href="data:file/text;base64,{b64}">Download as RTTM</a>'
+        st.markdown(href, unsafe_allow_html=True)
+
+
+st.sidebar.markdown(
+    """
+-------------------
+
+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).  
+
+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.
+
+For commercial enquiries and scientific consulting, please contact [me](mailto:[email protected]).
+"""
+)

packages.txt

@@ -0,0 +1 @@
+libsndfile1

requirements.txt

@@ -0,0 +1,4 @@
+git+https://github.com/pyannote/pyannote-audio.git@develop#egg=pyannote-audio
+speechbrain == 0.5.10
+matplotlib == 3.3.3
+