from unsloth import FastLanguageModel
model, tokenizer = FastLanguageModel.from_pretrained(
  model_name = "llama-3-docker-command-lora",
  max_seq_length = max_seq_length,
  dtype = dtype,
  load_in_4bit = load_in_4bit,
)
FastLanguageModel.for_inference(model) # Enable native 2x faster inference

inputs = tokenizer(
[
    alpaca_prompt.format(
        "translate this sentence in docker command.", # instruction
        "Give me a list of all containers, indicating their status as well.", # input
        "", # output - leave this blank for generation!
    )
], return_tensors = "pt").to("cuda")

outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True)
tokenizer.batch_decode(outputs)