Add local files for interface

.gitattributes

@@ -10,10 +10,12 @@
 *.lfs.* filter=lfs diff=lfs merge=lfs -text
 *.mlmodel filter=lfs diff=lfs merge=lfs -text
 *.model filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs, merge=lfs -text
 *.msgpack filter=lfs diff=lfs merge=lfs -text
 *.npy filter=lfs diff=lfs merge=lfs -text
 *.npz filter=lfs diff=lfs merge=lfs -text
 *.onnx filter=lfs diff=lfs merge=lfs -text
+*.opus filter=lfs diff=lfs, merge=lfs -text
 *.ot filter=lfs diff=lfs merge=lfs -text
 *.parquet filter=lfs diff=lfs merge=lfs -text
 *.pb filter=lfs diff=lfs merge=lfs -text
@@ -29,6 +31,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.tflite filter=lfs diff=lfs merge=lfs -text
 *.tgz filter=lfs diff=lfs merge=lfs -text
 *.wasm filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs, merge=lfs -text
 *.xz filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,6 +1,6 @@
 ---
 title: Audiobot
-emoji: 📉
+emoji: 🎵
 colorFrom: blue
 colorTo: yellow
 sdk: gradio

app.py

@@ -0,0 +1,62 @@
+from pathlib import Path
+
+import gradio as gr
+import librosa
+from transformers import (AutoFeatureExtractor,
+                          AutoModelForAudioClassification, pipeline)
+
+feature_extractor = AutoFeatureExtractor.from_pretrained(
+    "sanchit-gandhi/whisper-medium-fleurs-lang-id",
+    do_normalize=True
+)
+SR = 16_000
+
+label2id = {
+    'blues': '0',
+    'classical': '1',
+    'country': '2',
+    'disco': '3',
+    'hiphop': '4',
+    'jazz': '5',
+    'metal': '6',
+    'pop': '7',
+    'reggae': '8',
+    'rock': '9'
+}
+num_labels = len(label2id)
+
+path = "best_Whisper-Small_model_92"
+model = AutoModelForAudioClassification.from_pretrained("allispaul/whisper-small-gtzan")
+classifier = pipeline("audio-classification", model=model, feature_extractor=feature_extractor)
+
+def predict(audio):
+    if isinstance(audio, str) or isinstance(audio, Path):
+        audio, sr = librosa.load(audio, sr=16_000)
+    else:
+        sr, audio = audio
+    preds = classifier({"array": audio, "sampling_rate": sr})
+    probs = {k: 0.0 for k in label2id}
+    for pred in preds:
+        probs[pred["label"]] = pred["score"]
+    return probs
+
+title = "🤖 🎵 Audiobot 🎹 ⚡"
+description = "<h2>♯ The superpowered music genre classifier ♭</h2>"
+article = """
+<p>🤖 This model is a version of <a href="https://huggingface.co/openai/whisper-medium">Whisper Medium</a>,
+fine-tuned on the <a href="https://huggingface.co/datasets/marsyas/gtzan">GTZAN</a> dataset.</p>
+<p>🎹 It recognizes 10 genres: blues, classical, country, hip-hop, jazz, metal, pop, reggae, and rock.</p>
+<p>🎵 Upload a song or click one of the examples to try it out!</p>
+<p>⚡ Part of a project for the <a href="https://www.erdosinstitute.org/">Erdős Institute</a> Deep Learning bootcamp,
+by Dylan Bates, Muhammed Cifci, Aycan Katitas, Johann Thiel, Soheyl Anbouhi, and Paul VanKoughnett.</p>
+"""
+iface = gr.Interface(
+    fn=predict,
+    inputs=gr.Audio(type="filepath"),
+    outputs=gr.Label(num_top_classes=3),
+    title=title,
+    description=description,
+    article=article,
+    examples=["examples/country.00008.wav", "examples/hiphop.00057.wav", "examples/gamblersblues.opus"]
+)
+iface.launch()

examples/country.00008.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbfd35f47e913737b62baab8dc75d4e362be1e2b1a63533b1850ff1746652a7d
+size 1327524

examples/gamblersblues.opus

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7dc17b447eeb37cb5d4aaa3c2014d58a9f8434d057ee8c62789cdb0ac23672cf
+size 2260647

examples/hiphop.00057.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1152e98c15f222c972f4f6426fa53586775097c6e599c4165c7faf3d20c0db60
+size 1323632