Update app.py

app.py

@@ -5,7 +5,7 @@ from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
 
 token = os.environ["HUGGINGFACEHUB_API_TOKEN"]
 
-model_id = 'Deci-early-access/DeciLM-7B-instruct-early'
+model_id = 'Deci/DeciLM-7B-instruct'
 
 SYSTEM_PROMPT_TEMPLATE = """### System: You are an AI assistant that follows instruction extremely well. Help as much as you can.
 ### User:
@@ -18,7 +18,7 @@ SYSTEM_PROMPT_TEMPLATE = """### System: You are an AI assistant that follows ins
 DESCRIPTION = """
 # <p style="text-align: center; color: #292b47;"> 🤖 <span style='color: #3264ff;'>DeciLM-7B-Instruct:</span> A Fast Instruction-Tuned Model💨 </p>
 <span style='color: #292b47;'>Welcome to <a href="https://huggingface.co/Deci/DeciLM-7B-instruct" style="color: #3264ff;">DeciLM-7B-Instruct</a>! DeciLM-7B-Instruct is a 7B parameter instruction-tuned language model and released under the Apache 2.0 license. It's an instruction-tuned model, not a chat-tuned model;  you should prompt the model with an instruction that describes a task, and the model will respond appropriately to complete the task.</span>
-<p><span style='color: #292b47;'>Learn more about the base model <a href="" style="color: #3264ff;">DeciLM-7B.</a></span></p>
+<p><span style='color: #292b47;'>Learn more about the base model <a href="https://huggingface.co/Deci/DeciLM-7B" style="color: #3264ff;">DeciLM-7B.</a></span></p>
 """
 
 bnb_config = BitsAndBytesConfig(
@@ -33,10 +33,7 @@ if torch.cuda.is_available():
     model = AutoModelForCausalLM.from_pretrained(model_id,
                                                  device_map="auto",
                                                  trust_remote_code=True,
-                                                 quantization_config=bnb_config,
-                                                 # use_flash_attention_2=True, #DeciLM doesn't use flash_attention_2
-                                                #  this token will be deleted
-                                                 token=token
+                                                 quantization_config=bnb_config
                                                  )
 else:
     model = None