Create app.py

app.py

@@ -0,0 +1,35 @@
+import gradio as gr
+import torchaudio
+import torch
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+
+# Load model and processor
+processor = Wav2Vec2Processor.from_pretrained("Mustafaa4a/ASR-Somali")
+model = Wav2Vec2ForCTC.from_pretrained("Mustafaa4a/ASR-Somali")
+
+def transcribe(audio):
+    waveform, sample_rate = torchaudio.load(audio)
+
+    if sample_rate != 16000:
+        resampler = torchaudio.transforms.Resample(orig_freq=sample_rate, new_freq=16000)
+        waveform = resampler(waveform)
+
+    inputs = processor(waveform.squeeze(), sampling_rate=16000, return_tensors="pt")
+    with torch.no_grad():
+        logits = model(**inputs).logits
+
+    predicted_ids = torch.argmax(logits, dim=-1)
+    transcription = processor.decode(predicted_ids[0])
+    return transcription
+
+# Gradio Interface setup
+interface = gr.Interface(
+    fn=transcribe,
+    inputs=gr.Audio(type="filepath", label="Upload Somali Audio (.wav)"),
+    outputs=gr.Textbox(label="Transcription"),
+    title="Somali-speech_to_text",
+    description="Upload a Somali speech audio file (mono WAV, 16kHz) and get the text transcription."
+)
+
+# Launch the Gradio app and make it publicly available by using 'share=True'
+interface.launch()  # Don't use share=True in Hugging Face Spaces