import ctranslate2
import transformers
import datetime

for prompt in ["Hello, I am Joan and I am from Barcelona.", "Translate to Catalan: how are you today?", "Translate to Catalan:  It is a British public service broadcaster, Headquartered in London, and it is the world's oldest national broadcaster.", "Parafraseja en català: Avui serà un dia molt llarg perquè farem moltes coses junts."]:
    for model_name in [#"bloom-560m", 
        "bloomz-7b1-mt"]:

        generator = ctranslate2.Generator(model_name, compute_type="int8")
        tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)

        start_time = datetime.datetime.now()

        start_tokens = tokenizer.convert_ids_to_tokens(tokenizer.encode(prompt))
        results = generator.generate_batch([start_tokens], max_length=90)
        result = tokenizer.decode(results[0].sequences_ids[0])
        t = str(datetime.datetime.now() - start_time)
        print(f"Model: {model_name}")
        print(f"Prompt: {prompt}")    
        print(f"Result: {result}")
        print(f"Time: {t}")
        print("")

        del generator, tokenizer