Update endpoints.py

endpoints.py

@@ -16,10 +16,7 @@ import requests
 def LLM(llm_name, length):
     print(llm_name)
     tokenizer = AutoTokenizer.from_pretrained(llm_name)
-    model = AutoModelForCausalLM.from_pretrained(llm_name,
-                                                 trust_remote_code=True, 
-                                                 device_map="auto",
-                                                 load_in_8bit=True)
+    model = AutoModelForCausalLM.from_pretrained(llm_name)                                                
     pipe = pipeline("text-generation",
                     model=model,
                     tokenizer=tokenizer,
@@ -31,7 +28,7 @@ def LLM(llm_name, length):
     return pipe
 
 
-pipe = LLM("WizardLM/WizardCoder-Python-7B-V1.0",4000)
+pipe = LLM("Salesforce/codegen-16B-nl",4000)
 # tokenizer = AutoTokenizer.from_pretrained("WizardLM/WizardCoder-1B-V1.0")
 # base_model = AutoModelForCausalLM.from_pretrained("WizardLM/WizardCoder-1B-V1.0")
 # Mistral 7B