d2l.Vocab(sentences, min_freq=5, reserved_tokens=[‘＜pad＞‘, ‘＜mask＞‘, ‘＜cls＞‘, ‘＜sep＞‘]) 参数讲解

d2l.Vocab(sentences, min_freq=2, reserved_tokens=['<pad>', '<mask>', '<cls>', '<sep>'])
sentences：源句子，比如说 sentences赋值为下边的五个句子 注意第一句我手动添加了两个'<unk>',第二句添加了'<pad>', '<mask>', '<cls>', '<sep>',作为第三个参数的测试
[
	['<unk>', '<unk>', 'the', 'ottoman', 'turkish', 'empire', 'entered', 'the', 'first', 'world', 'war', 'on', 'the', 'side', 'of', 'the', 'central', 'powers', 'on', '31', 'october', '1914'], 
	['<pad>', '<mask>', '<cls>', '<sep>', 'the', 'stalemate', 'of', 'trench', 'warfare', 'on', 'the', 'western', 'front', 'convinced', 'the', 'british', 'imperial', 'war', 'cabinet', 'that', 'an', 'attack', 'on', 'the', 'central', 'powers', 'elsewhere', ',', 'particularly', 'turkey', ',', 'could', 'be', 'the', 'best', 'way', 'of', 'winning', 'the', 'war'], 
	['from', 'february', '1915', 'this', 'took', 'the', 'form', 'of', 'naval', 'operations', 'aimed', 'at', 'forcing', 'a', 'passage', 'through', 'the', 'dardanelles', ',', 'but', 'after', 'several', 'setbacks', 'it', 'was', 'decided', 'that', 'a', 'land', 'campaign', 'was', 'also', 'necessary'],
	['to', 'that', 'end', ',', 'the', 'mediterranean', 'expeditionary', 'force', 'was', 'formed', 'under', 'the', 'command', 'of', 'general', 'ian', 'hamilton'],
	['three', 'amphibious', 'landings', 'were', 'planned', 'to', 'secure', 'the', 'gallipoli', 'peninsula', ',', 'which', 'would', 'allow', 'the', 'navy', 'to', 'attack', 'the', 'turkish', 'capital', 'constantinople', ',', 'in', 'the', 'hope', 'that', 'would', 'convince', 'the', 'turks', 'to', 'ask', 'for', 'an', 'armistice', '.']
]
min_freq：把出现次数少于2次的低频率词元视为相同的未知词元即视为'<unk>'
reserved_tokens：如果遇到reserved_tokens中的词也作为token保留下来

一个完整的例子：