feat: add whisper speech-to-text model

Add OpenAI's Whisper large-v3 model for automatic speech recognition.
Key changes:
- Import required libraries (transformers, torch, datasets)
- Load Whisper model and processor
- Create pipeline for speech-to-text transcription
- Update Gradio interface to accept audio input and return text output

app.py

@@ -1,14 +1,43 @@
 import spaces
 import gradio as gr
 # Use a pipeline as a high-level helper
-from transformers import pipeline
+import torch
+from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
+from datasets import load_dataset
 
-pipe = pipeline("automatic-speech-recognition", model="openai/whisper-large-v3")
 
-@spaces.GPU
 def transcribe_audio(audio):
-    return pipe(audio)
+    device = "cuda:0" if torch.cuda.is_available() else "cpu"
+    torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
 
-demo = gr.Interface(fn=transcribe_audio, inputs=gr.Audio(sources="upload", type="filepath"), outputs="text")
+    model_id = "openai/whisper-large-v3"
+
+    model = AutoModelForSpeechSeq2Seq.from_pretrained(
+        model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
+    )
+    model.to(device)
+
+    processor = AutoProcessor.from_pretrained(model_id)
+
+    pipe = pipeline(
+        "automatic-speech-recognition",
+        model=model,
+        tokenizer=processor.tokenizer,
+        feature_extractor=processor.feature_extractor,
+        max_new_tokens=128,
+        chunk_length_s=25,
+        batch_size=16,
+        torch_dtype=torch_dtype,
+        device=device,
+    )
+
+    result = pipe(audio)
+    return result["text"]
+
+
+demo = gr.Interface(fn=transcribe_audio,
+                    inputs=gr.Audio(sources="upload", type="filepath"),
+                    outputs="text")
 demo.launch()
 
+