from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "MacLeanLuke/gemma-2b-tool-tuned"

model = AutoModelForCausalLM.from_pretrained(repo_id)
tokenizer = AutoTokenizer.from_pretrained(repo_id)

inputs = tokenizer("Hello, how are you?", return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0]))