Create app.py

app.py

@@ -0,0 +1,59 @@
+import gradio as gr
+from transformers import pipeline
+import torch
+import librosa
+import soundfile
+
+SAMPLE_RATE = 16000
+
+pipe = pipeline(model="birgermoell/whisper-small-sv-bm")  # change to "your-username/the-name-you-picked"
+
+
+def process_audio_file(file):
+    data, sr = librosa.load(file)
+
+    if sr != SAMPLE_RATE:
+        data = librosa.resample(data, sr, SAMPLE_RATE)
+
+    # monochannel
+    data = librosa.to_mono(data)
+    return data
+
+
+def transcribe(Microphone, File_Upload):
+    warn_output = ""
+    if (Microphone is not None) and (File_Upload is not None):
+        warn_output = "WARNING: You've uploaded an audio file and used the microphone. " \
+                      "The recorded file from the microphone will be used and the uploaded audio will be discarded.\n"
+        file = Microphone
+
+    elif (Microphone is None) and (File_Upload is None):
+        return "ERROR: You have to either use the microphone or upload an audio file"
+
+    elif Microphone is not None:
+        file = Microphone
+    else:
+        file = File_Upload
+
+    audio_data = process_audio_file(file)
+    
+    text = pipe(audio_data)["text"]
+
+    return warn_output + text
+
+
+iface = gr.Interface(
+    fn=transcribe,
+    inputs=[
+        gr.inputs.Audio(source="microphone", type='filepath', optional=True),
+        gr.inputs.Audio(source="upload", type='filepath', optional=True),
+    ],
+    outputs="text",
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Small SV",
+    description="Demo for Swedish speech recognition using the [Whisper Small SV BM checkpoint](https://huggingface.co/birgermoell/whisper-small-sv-bm).",
+    allow_flagging='never',
+)
+
+iface.launch(enable_queue=True)