from transformers import AutoModelForCausalLM, AutoTokenizer, TextGenerationPipeline
model_path = 'fiveflow/ATOMM'
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(model_path, 
                                                  device_map="auto",
                                                #   load_in_4bit=True,
                                                  low_cpu_mem_usage=True)

pipe = TextGenerationPipeline(model = model, tokenizer = tokenizer)