Update handler.py

handler.py

@@ -1,5 +1,4 @@
 # Adapted from https://huggingface.co/nvidia/NVLM-D-72B#inference
-
 import math
 from typing import Any, Dict, List
 
@@ -152,15 +151,24 @@ class EndpointHandler:
         )
 
     def __call__(self, data: Dict[str, Any]) -> Dict[str, List[Any]]:
-        if "instances" not in data:
+        if "instances" in data:
+            logger.warning("Using `instances` instead of `inputs` is deprecated.")
+            data["inputs"] = data.pop("instances")
+
+        if "inputs" not in data:
             raise ValueError(
-                "The request body must contain a key 'instances' with a list of instances."
+                "The request body must contain a key 'inputs' with a list of inputs."
             )
-
+        
         logger.debug(f"Received incoming request with {data=}")
 
         predictions = []
-        for input in data["instances"]:
+        for input in data["inputs"]:
+            if "prompt" not in input:
+                raise ValueError(
+                    "The request input body must contain a key 'prompt' with the prompt to use."
+                )
+
             generation_config = input.get("generation_config", dict(max_new_tokens=1024, do_sample=False))
 
             if "image_url" not in input:
@@ -186,4 +194,4 @@ class EndpointHandler:
                 )
 
             predictions.append(response)
-        return {"predictions": predictions}
+        return {"predictions": predictions}