fixed the no attribute 'pipeline' bug

handler.py

@@ -1,32 +1,42 @@
-from typing import Dict, Any, List
-from transformers import AutoModelForCausalLM, AutoTokenizer, TextGenerationPipeline
+from typing import Dict, Any
 import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
 
 class EndpointHandler:
     def __init__(self, path=""):
-        # Load model and tokenizer
+        # Load tokenizer and model
         self.tokenizer = AutoTokenizer.from_pretrained(path)
-        self.model = AutoModelForCausalLM.from_pretrained(path, device_map="auto", torch_dtype=torch.float16)
+        self.model = AutoModelForCausalLM.from_pretrained(path)
+        
+        # Create a pipeline that the inference API expects
+        self.pipeline = pipeline(
+            "text-generation",
+            model=self.model,
+            tokenizer=self.tokenizer,
+            device=0 if torch.cuda.is_available() else -1
+        )
 
-    def __call__(self, data: Dict[str, Any]) -> List[Dict[str, str]]:
+    def __call__(self, data: Dict[str, Any]) -> Dict[str, Any]:
         prompt_input = data.get("inputs", "")
-        vibe = data.get("vibe", "Open to All Paths")  # Default fallback
-
-        # Construct Camino-aware prompt
-        full_prompt = (
+        vibe = data.get("vibe", "Open to All Paths")
+        
+        # Prepare prompt with Vela's persona
+        prompt = (
             f"#### Human (Vibe: {vibe}): {prompt_input.strip()}\n"
             f"#### Assistant (Vela - your Camino companion):"
         )
-
+        
         # Default generation params
         generation_args = data.get("parameters", {})
         generation_args.setdefault("max_new_tokens", 1024)
         generation_args.setdefault("temperature", 0.2)
         generation_args.setdefault("top_p", 0.95)
         generation_args.setdefault("do_sample", True)
-
-        # Generate response
-        outputs = self.pipeline(full_prompt, **generation_args)
-
-        # Return in correct format
-        return [{"generated_text": outputs[0]["generated_text"]}]
+        
+        # Use pipeline for generation
+        outputs = self.pipeline(
+            prompt,
+            **generation_args
+        )
+        
+        return outputs