Update handler.py

handler.py

@@ -1,8 +1,50 @@
+# import torch
+# from typing import  Dict, List, Any
+# from transformers import AutoTokenizer, AutoModelForCausalLM
+
+# class EndpointHandler:
+#     def __init__(self, path: str = ""):
+
+#         self.tokenizer = AutoTokenizer.from_pretrained(path, padding_side = "left")
+#         self.model = AutoModelForCausalLM.from_pretrained(path, device_map = "auto", torch_dtype=torch.float16)
+
+#     def __call__(self, data: Dict[str, Any]) -> Dict[str, Any]:
+#         """
+#         Args:
+#             data (:obj:):
+#                 includes the input data and the parameters for the inference.
+#         Return:
+#             A :obj:`list`:. The list contains the answer and scores of the inference inputs
+#         """
+
+#         # process input
+#         inputs_dict = data.pop("inputs", data)
+#         parameters = data.pop("parameters", {})
+        
+#         prompts = [f"<human>: {prompt}\n<bot>:" for prompt in inputs_dict]
+
+#         self.tokenizer.pad_token = self.tokenizer.eos_token
+
+#         inputs = self.tokenizer(prompts, truncation=True, max_length=2048-512,
+#                                 return_tensors='pt', padding=True).to(self.model.device)
+#         input_length = inputs.input_ids.shape[1]
+
+#         if parameters.get("deterministic", False):
+#             torch.manual_seed(42)
+
+#         outputs = self.model.generate(
+#             **inputs, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.7, top_k=50
+#          )
+        
+#         output_strs = self.tokenizer.batch_decode(outputs[:, input_length:], skip_special_tokens=True)
+        
+#         return {"generated_text": output_strs}
+
 import torch
 from typing import  Dict, List, Any
 from transformers import AutoTokenizer, AutoModelForCausalLM
 
-class EndpointHandler:
+class EndpointHandler():
     def __init__(self, path: str = ""):
 
         self.tokenizer = AutoTokenizer.from_pretrained(path, padding_side = "left")
@@ -18,10 +60,10 @@ class EndpointHandler:
         """
 
         # process input
-        inputs_dict = data.pop("inputs", data)
+        inputs_list = data.pop("inputs", data)
         parameters = data.pop("parameters", {})
         
-        prompts = [f"<human>: {prompt}\n<bot>:" for prompt in inputs_dict]
+        prompts = [f"<human>: {prompt}\n<bot>:" for prompt in inputs_list]
 
         self.tokenizer.pad_token = self.tokenizer.eos_token