{"id":29680,"date":"2024-10-28T01:56:03","date_gmt":"2024-10-28T01:56:03","guid":{"rendered":"http:\/\/atmokpo.com\/w\/?p=29680"},"modified":"2024-11-26T06:51:45","modified_gmt":"2024-11-26T06:51:45","slug":"%ed%97%88%ea%b9%85%ed%8e%98%ec%9d%b4%ec%8a%a4-%ed%8a%b8%eb%a0%8c%ec%8a%a4%ed%8f%ac%eb%a8%b8-%ed%99%9c%ec%9a%a9%ea%b0%95%ec%a2%8c-%ec%9d%b8%ec%bd%94%eb%94%a9-%eb%b0%8f-%eb%94%94%ec%bd%94%eb%94%a9","status":"publish","type":"post","link":"https:\/\/atmokpo.com\/w\/29680\/","title":{"rendered":"\ud5c8\uae45\ud398\uc774\uc2a4 \ud2b8\ub80c\uc2a4\ud3ec\uba38 \ud65c\uc6a9\uac15\uc88c, \uc778\ucf54\ub529 \ubc0f \ub514\ucf54\ub529"},"content":{"rendered":"

<\/p>\n

\ub525\ub7ec\ub2dd \ubd84\uc57c\uc5d0\uc11c \uc790\uc5f0\uc5b4 \ucc98\ub9ac(NLP)\ub294 \ud2b9\ud788 \uc8fc\ubaa9\ubc1b\uace0 \uc788\ub294 \uc601\uc5ed \uc911 \ud558\ub098\uc785\ub2c8\ub2e4. 2018\ub144 \uacf5\uac1c\ub41c Hugging Face\uc758 Transformers \ub77c\uc774\ube0c\ub7ec\ub9ac<\/a>\ub294 NLP \ubaa8\ub378\uc744 \uc190\uc27d\uac8c \uc0ac\uc6a9\ud560 \uc218 \uc788\ub3c4\ub85d \ub3c4\uc640\uc8fc\ub294 \uac15\ub825\ud55c \ub3c4\uad6c\uc785\ub2c8\ub2e4. \ubcf8 \uac15\uc88c\uc5d0\uc11c\ub294 Hugging Face Transformers \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \ud65c\uc6a9\ud558\uc5ec \uc778\ucf54\ub529\uacfc \ub514\ucf54\ub529\uc744 \uc218\ud589\ud558\ub294 \ubc29\ubc95\uc5d0 \ub300\ud574 \uc54c\uc544\ubcf4\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n

1. Transformers \ub77c\uc774\ube0c\ub7ec\ub9ac \uc18c\uac1c<\/h2>\n

Transformers \ub77c\uc774\ube0c\ub7ec\ub9ac\ub294 BERT, GPT-2, T5\uc640 \uac19\uc740 \ub2e4\uc591\ud55c \uc2e0\uacbd\ub9dd \uc544\ud0a4\ud14d\ucc98\ub97c \uc9c0\uc6d0\ud569\ub2c8\ub2e4. \uc774 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \ud1b5\ud574 \ubcf5\uc7a1\ud55c NLP \ubaa8\ub378\uc744 \uc27d\uac8c \uad6c\ud604\ud560 \uc218 \uc788\uc73c\uba70, \uac1c\uc778\uc801\uc778 \uc5f0\uad6c\ub098 \uc0c1\uc5c5\uc801\uc778 \ud504\ub85c\uc81d\ud2b8 \ubaa8\ub450\uc5d0\uc11c \ud65c\uc6a9\ub429\ub2c8\ub2e4.<\/p>\n

1.1 \uc124\uce58<\/h3>\n

Transformers \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \uc124\uce58\ud558\uae30 \uc704\ud574\uc11c\ub294 pip\ub97c \uc774\uc6a9\ud569\ub2c8\ub2e4. \ub2e4\uc74c \uba85\ub839\uc5b4\ub97c \uc2e4\ud589\ud574\uc8fc\uc138\uc694.<\/p>\n

pip install transformers<\/code><\/pre>\n
2. \ud14d\uc2a4\ud2b8 \uc778\ucf54\ub529<\/h2>\n
\uc778\ucf54\ub529\uc740 \ud14d\uc2a4\ud2b8 \ub370\uc774\ud130\ub97c \ubaa8\ub378\uc774 \uc774\ud574\ud560 \uc218 \uc788\ub294 \ud615\uc2dd\uc73c\ub85c \ubcc0\ud658\ud558\ub294 \uacfc\uc815\uc785\ub2c8\ub2e4. Transformers \ub77c\uc774\ube0c\ub7ec\ub9ac\uc5d0\uc11c\ub294 \ud1a0\ud06c\ub098\uc774\uc800\ub97c \uc0ac\uc6a9\ud558\uc5ec \ud14d\uc2a4\ud2b8\ub97c \uc778\ucf54\ub529\ud569\ub2c8\ub2e4. \ub2e4\uc74c\uc740 BERT \ubaa8\ub378\uc758 \ud1a0\ud06c\ub098\uc774\uc800\ub97c \uc0ac\uc6a9\ud558\uc5ec \ud14d\uc2a4\ud2b8\ub97c \uc778\ucf54\ub529\ud558\ub294 \uc608\uc81c\uc785\ub2c8\ub2e4.<\/p>\n
2.1 BERT \ud1a0\ud06c\ub098\uc774\uc800 \uc608\uc81c<\/h3>\n
\uc544\ub798 \ucf54\ub4dc\ub294 BERT \ubaa8\ub378\uc758 \uae30\ubcf8 \ud1a0\ud06c\ub098\uc774\uc800\ub97c \uc0ac\uc6a9\ud558\uc5ec \uc785\ub825 \ubb38\uc7a5\uc744 \uc778\ucf54\ub529\ud558\ub294 \uacfc\uc815\uc744 \ubcf4\uc5ec\uc90d\ub2c8\ub2e4.<\/p>\n
from transformers import BertTokenizer\n\n# BERT \ubaa8\ub378\uc758 \ud1a0\ud06c\ub098\uc774\uc800 \ucd08\uae30\ud654\ntokenizer = BertTokenizer.from_pretrained('bert-base-uncased')\n\n# \uc778\ucf54\ub529\ud560 \ud14d\uc2a4\ud2b8\ntext = \"Hello, how are you?\"\n\n# \ud14d\uc2a4\ud2b8 \uc778\ucf54\ub529\nencoded_input = tokenizer(text, return_tensors='pt')\n\n# \uacb0\uacfc \ucd9c\ub825\nprint(encoded_input)<\/code><\/pre>\n
\uc704 \ucf54\ub4dc\uc5d0\uc11c BertTokenizer.from_pretrained()<\/code> \uba54\uc11c\ub4dc\ub97c \uc0ac\uc6a9\ud558\uc5ec \uc0ac\uc804\ud559\uc2b5\ub41c BERT \ud1a0\ud06c\ub098\uc774\uc800\ub97c \ub85c\ub4dc\ud569\ub2c8\ub2e4. \uadf8\ub9ac\uace0 tokenizer()<\/code> \uba54\uc11c\ub4dc\ub97c \ud1b5\ud574 \uc785\ub825 \ubb38\uc7a5\uc744 \uc778\ucf54\ub529\ud569\ub2c8\ub2e4. return_tensors='pt'<\/code>\ub294 \ud150\uc11c\ud50c\ub85c\uc6b0 \ub300\uc2e0 \ud30c\uc774\ud1a0\uce58 \ud150\uc11c\ub97c \ubc18\ud658\ud558\ub3c4\ub85d \ud569\ub2c8\ub2e4.<\/p>\n
2.2 \uc778\ucf54\ub529 \uacb0\uacfc \uc124\uba85<\/h3>\n
\uc778\ucf54\ub529 \uacb0\uacfc\ub294 \ub2e4\uc74c\uacfc \uac19\uc740 \uad6c\uc870\ub97c \uac00\uc9d1\ub2c8\ub2e4:<\/p>\n