以一定概率选择lstm输出，并不是简单地argmax

最新推荐文章于 2023-11-02 09:12:47 发布

xiewenbo

最新推荐文章于 2023-11-02 09:12:47 发布

阅读量1k

点赞数

分类专栏： RNN LSTM

本文链接：https://blog.csdn.net/xiewenbo/article/details/80367776

版权

LSTM 同时被 2 个专栏收录

12 篇文章 0 订阅

订阅专栏

RNN

8 篇文章 0 订阅

订阅专栏

def sample(preds, temperature=1.0):
    '''
    Generate some randomness with the given preds
    which is a list of numbers, if the temperature
    is very small, it will always pick the index
    with highest pred value
    '''
    preds = np.asarray(preds).astype('float64')
    preds = np.log(preds) / temperature
    exp_preds = np.exp(preds)
    preds = exp_preds / np.sum(exp_preds)
    probas = np.random.multinomial(1, preds, 1)
    return np.argmax(probas)