Add initial custom handler

handler.py

@@ -0,0 +1,28 @@
+from typing import Dict, List, Any
+from transformers import AutoProcessor, MusicgenForConditionalGeneration
+import torch
+
+class EndpointHandler:
+    def __init__(self, path=""):
+        self.processor = AutoProcessor.from_pretrained(path)
+        self.model = MusicgenForConditionalGeneration.from_pretrained(path)
+    
+    def __call__(self, data: Dict[str, Any]) -> Dict[str, str]:
+        text_input = data.pop("inputs", data)
+        parameters = data.pop("parameters", None)
+
+        inputs = self.processor(
+            text = [text_input],
+            return_tensors="pt",
+            padding=True).to("cuda")
+        
+        if parameters is not None:
+            with torch.autocast("cuda"):
+                outputs = self.model.generate(**inputs, **parameters)
+        else:
+            with torch.autocast("cuda"):
+                outputs = self.model.generate(**inputs)
+        
+        prediction = outputs[0].cpu().numpy().tolist()
+
+        return [{"generated_audio": prediction}]