Fix handler to accept HF Endpoints inputs format

handler.py

@@ -71,8 +71,15 @@ class EndpointHandler:
         """
         Expect `data` dict like:
         {
-          "description": "... voice description ...",
+          "inputs": {
+            "text": "... text to speak ...",
+            "description": "... voice description ..."
+          }
+        }
+        Or simplified:
+        {
           "text": "... text to speak ...",
+          "description": "... voice description ...",
           "generation_args": { optional dict for text generation params }
         }
         Returns dict with base64 audio:
@@ -81,8 +88,24 @@ class EndpointHandler:
           "sampling_rate": 24000
         }
         """
-        description = data.get("description", "")
-        text = data.get("text", "")
+        # Handle HF Endpoints format with "inputs" key
+        if "inputs" in data:
+            inputs = data["inputs"]
+            # Handle both dict and string inputs
+            if isinstance(inputs, dict):
+                description = inputs.get("description", "")
+                text = inputs.get("text", "")
+            elif isinstance(inputs, str):
+                text = inputs
+                description = ""
+            else:
+                text = str(inputs)
+                description = ""
+        else:
+            # Handle direct format
+            description = data.get("description", "")
+            text = data.get("text", "")
+        
         if not text:
             return {"error": "No text provided."}
         prompt = description + "\n" + text
@@ -121,7 +144,11 @@ class EndpointHandler:
 
         inputs = self.tokenizer(prompt, return_tensors="pt").to(self.device)
         # generate token ids with default or custom params
-        gen_args = data.get("generation_args", {})
+        # Get generation_args from either top level or from inputs
+        if "inputs" in data and isinstance(data["inputs"], dict):
+            gen_args = data["inputs"].get("generation_args", {})
+        else:
+            gen_args = data.get("generation_args", {})
         outputs = self.model.generate(**inputs, **gen_args)
         token_ids = outputs[0]