Add application file

app.py

@@ -0,0 +1,47 @@
+import torch
+import torchaudio
+from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
+import gradio as gr
+
+# Load model and processor from your fine-tuned directory
+MODEL_PATH = "C:\Users\Nam\My_First_ASR\imported_model"  # Update this if needed
+processor = Wav2Vec2Processor.from_pretrained(MODEL_PATH)
+model = Wav2Vec2ForCTC.from_pretrained(MODEL_PATH).eval()
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+
+# Define inference function
+def transcribe(audio):
+    if audio is None:
+        return "No audio provided."
+
+    sr, data = audio
+
+    # Convert to mono and resample to 16kHz if needed
+    waveform = torch.tensor(data).unsqueeze(0)
+    if sr != 16000:
+        resampler = torchaudio.transforms.Resample(orig_freq=sr, new_freq=16000)
+        waveform = resampler(waveform)
+    if waveform.shape[0] > 1:
+        waveform = waveform.mean(dim=0, keepdim=True)
+
+    # Inference
+    inputs = processor(waveform.squeeze().numpy(), sampling_rate=16000, return_tensors="pt", padding=True)
+    input_values = inputs.input_values.to(device)
+
+    with torch.no_grad():
+        logits = model(input_values).logits
+        predicted_ids = torch.argmax(logits, dim=-1)
+
+    transcription = processor.batch_decode(predicted_ids)[0]
+    return transcription.strip()
+
+# Gradio interface
+gr.Interface(
+    fn=transcribe,
+    inputs=gr.Audio(source="upload", type="numpy", label="Upload WAV/MP3 file"),
+    outputs=gr.Textbox(label="Transcription"),
+    title="🗣️ ASR Demo with Wav2Vec2",
+    description="Upload an audio file (WAV or MP3) and get the transcription using your fine-tuned model.",
+    live=False
+).launch()