fixed bug for llama-2 auth token handling

app_modules/llm_loader.py

@@ -356,20 +356,20 @@ class LLMLoader:
                         model = MODEL_NAME_OR_PATH
 
                     pipe = pipeline(
-                            task,
-                            model=model,
-                            tokenizer=tokenizer,
-                            streamer=self.streamer,
-                            return_full_text=return_full_text,  # langchain expects the full text
-                            device=hf_pipeline_device_type,
-                            torch_dtype=torch_dtype,
-                            max_new_tokens=2048,
-                            trust_remote_code=True,
-                            temperature=temperature,
-                            top_p=0.95,
-                            top_k=0,  # select from top 0 tokens (because zero, relies on top_p)
-                            repetition_penalty=1.115,
-                        )
+                        task,
+                        model=model,
+                        tokenizer=tokenizer,
+                        streamer=self.streamer,
+                        return_full_text=return_full_text,  # langchain expects the full text
+                        device=hf_pipeline_device_type,
+                        torch_dtype=torch_dtype,
+                        max_new_tokens=2048,
+                        trust_remote_code=True,
+                        temperature=temperature,
+                        top_p=0.95,
+                        top_k=0,  # select from top 0 tokens (because zero, relies on top_p)
+                        repetition_penalty=1.115,
+                    )
 
                 self.llm = HuggingFacePipeline(pipeline=pipe, callbacks=callbacks)
             elif self.llm_model_type == "mosaicml":