greedy.zh.json

{
 "<h1>Greedy Sampling</h1>\n<p>Here we sample the most likely token from the distribution of logits.</p>\n<p>Here&#x27;s an <a href=\"experiment.html\">experiment</a> that uses these sampling techniques.</p>\n": "<h1>\u8d2a\u5a6a\u91c7\u6837</h1>\n<p>\u5728\u8fd9\u91cc\uff0c\u6211\u4eec\u4ece\u65e5\u5fd7\u5206\u5e03\u4e2d\u62bd\u53d6\u6700\u6709\u53ef\u80fd\u7684\u4ee4\u724c\u3002</p>\n<p>\u8fd9\u662f\u4e00\u4e2a\u4f7f\u7528\u8fd9\u4e9b\u91c7\u6837\u6280\u672f\u7684<a href=\"experiment.html\">\u5b9e\u9a8c</a>\u3002</p>\n",
 "<p> Sample the most likely token from the distribution of logits</p>\n": "<p>\u4ece\u65e5\u5fd7\u5206\u5e03\u4e2d\u62bd\u53d6\u6700\u6709\u53ef\u80fd\u7684\u4ee4\u724c</p>\n",
 "A PyTorch implementation of greedy sampling from language models.": "\u4ece\u8bed\u8a00\u6a21\u578b\u4e2d\u8fdb\u884c\u8d2a\u5a6a\u91c7\u6837\u7684 PyTorch \u5b9e\u73b0\u3002",
 "Greedy Sampling": "\u8d2a\u5a6a\u91c7\u6837"
}