INNER CODE UNIT · Python

complete

sign-language-translator/sign-language-translator · sign_language_translator/cli.py:301

def complete(
    inputs,
    model_code,
    end_token,
    max_length,
    beam_width,
    model_weight,
    selection_strategy,
    join,
):
    """
    Complete a sequence using Language Models.

    Predicts next tokens in the given sequence and completes it using the specified model and beam search.
    You can also use multiple models and also pass along model selection weights for each model (using --model-weights).
    Currently you can use the following model-codes:\n
    1. urdu-mixed-ngram (token | start: "<")\n
    2. ur-supported-gpt (token | start: "<")\n

View source record →