File size: 12,143 Bytes
17c6d62
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
<!--Copyright 2020 The HuggingFace Team. All rights reserved.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
specific language governing permissions and limitations under the License.

โš ๏ธ Note that this file is in Markdown but contain specific syntax for our doc-builder (similar to MDX) that may not be
rendered properly in your Markdown viewer.

-->

# BART[[bart]]

<div class="flex flex-wrap space-x-1">
<a href="https://huggingface.co/models?filter=bart">
<img alt="Models" src="https://img.shields.io/badge/All_model_pages-bart-blueviolet">
</a>
<a href="https://huggingface.co/spaces/docs-demos/bart-large-mnli">
<img alt="Spaces" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue">
</a>
</div>

## ๊ฐœ์š” [[overview]]

Bart ๋ชจ๋ธ์€ 2019๋…„ 10์›” 29์ผ Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov, Luke Zettlemoyer๊ฐ€ ๋ฐœํ‘œํ•œ [BART: ์ž์—ฐ์–ด ์ƒ์„ฑ, ๋ฒˆ์—ญ, ์ดํ•ด๋ฅผ ์œ„ํ•œ ์žก์Œ ์ œ๊ฑฐ seq2seq ์‚ฌ์ „ ํ›ˆ๋ จ](https://arxiv.org/abs/1910.13461)์ด๋ผ๋Š” ๋…ผ๋ฌธ์—์„œ ์†Œ๊ฐœ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.

๋…ผ๋ฌธ์˜ ์ดˆ๋ก์— ๋”ฐ๋ฅด๋ฉด,

- Bart๋Š” ์–‘๋ฐฉํ–ฅ ์ธ์ฝ”๋”(BERT์™€ ์œ ์‚ฌ)์™€ ์™ผ์ชฝ์—์„œ ์˜ค๋ฅธ์ชฝ์œผ๋กœ ๋””์ฝ”๋”ฉํ•˜๋Š” ๋””์ฝ”๋”(GPT์™€ ์œ ์‚ฌ)๋ฅผ ์‚ฌ์šฉํ•˜๋Š” ํ‘œ์ค€ seq2seq/๊ธฐ๊ณ„ ๋ฒˆ์—ญ ์•„ํ‚คํ…์ฒ˜๋ฅผ ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค.
- ์‚ฌ์ „ ํ›ˆ๋ จ ์ž‘์—…์€ ์›๋ž˜ ๋ฌธ์žฅ์˜ ์ˆœ์„œ๋ฅผ ๋ฌด์ž‘์œ„๋กœ ์„ž๊ณ , ํ…์ŠคํŠธ์˜ ์ผ๋ถ€ ๊ตฌ๊ฐ„์„ ๋‹จ์ผ ๋งˆ์Šคํฌ ํ† ํฐ์œผ๋กœ ๋Œ€์ฒดํ•˜๋Š” ์ƒˆ๋กœ์šด ์ธํ•„๋ง(in-filling) ๋ฐฉ์‹์„ ํฌํ•จํ•ฉ๋‹ˆ๋‹ค.
- BART๋Š” ํŠนํžˆ ํ…์ŠคํŠธ ์ƒ์„ฑ์„ ์œ„ํ•œ ๋ฏธ์„ธ ์กฐ์ •์— ํšจ๊ณผ์ ์ด์ง€๋งŒ ์ดํ•ด ์ž‘์—…์—๋„ ์ž˜ ์ž‘๋™ํ•ฉ๋‹ˆ๋‹ค. GLUE์™€ SQuAD์—์„œ ๋น„์Šทํ•œ ํ›ˆ๋ จ ๋ฆฌ์†Œ์Šค๋กœ RoBERTa์˜ ์„ฑ๋Šฅ๊ณผ ์ผ์น˜ํ•˜๋ฉฐ, ์ถ”์ƒ์  ๋Œ€ํ™”, ์งˆ์˜์‘๋‹ต, ์š”์•ฝ ์ž‘์—… ๋“ฑ์—์„œ ์ตœ๋Œ€ 6 ROUGE ์ ์ˆ˜์˜ ํ–ฅ์ƒ์„ ๋ณด์ด๋ฉฐ ์ƒˆ๋กœ์šด ์ตœ๊ณ  ์„ฑ๋Šฅ์„ ๋‹ฌ์„ฑํ–ˆ์Šต๋‹ˆ๋‹ค.

์ด ๋ชจ๋ธ์€ [sshleifer](https://huggingface.co/sshleifer)์— ์˜ํ•ด ๊ธฐ์—ฌ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์ €์ž์˜ ์ฝ”๋“œ๋Š” [์ด๊ณณ](https://github.com/pytorch/fairseq/tree/master/examples/bart)์—์„œ ํ™•์ธํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

## ์‚ฌ์šฉ ํŒ:[[usage-tips]]

- BART๋Š” ์ ˆ๋Œ€ ์œ„์น˜ ์ž„๋ฒ ๋”ฉ์„ ์‚ฌ์šฉํ•˜๋Š” ๋ชจ๋ธ์ด๋ฏ€๋กœ ์ผ๋ฐ˜์ ์œผ๋กœ ์ž…๋ ฅ์„ ์™ผ์ชฝ๋ณด๋‹ค๋Š” ์˜ค๋ฅธ์ชฝ์— ํŒจ๋”ฉํ•˜๋Š” ๊ฒƒ์ด ์ข‹์Šต๋‹ˆ๋‹ค.
- ์ธ์ฝ”๋”์™€ ๋””์ฝ”๋”๊ฐ€ ์žˆ๋Š” seq2seq ๋ชจ๋ธ์ž…๋‹ˆ๋‹ค. ์ธ์ฝ”๋”์—๋Š” ์†์ƒ๋œ ํ† ํฐ์ด(corrupted tokens) ์ž…๋ ฅ๋˜๊ณ , ๋””์ฝ”๋”์—๋Š” ์›๋ž˜ ํ† ํฐ์ด ์ž…๋ ฅ๋ฉ๋‹ˆ๋‹ค(๋‹จ, ์ผ๋ฐ˜์ ์ธ ํŠธ๋žœ์Šคํฌ๋จธ ๋””์ฝ”๋”์ฒ˜๋Ÿผ ๋ฏธ๋ž˜ ๋‹จ์–ด๋ฅผ ์ˆจ๊ธฐ๋Š” ๋งˆ์Šคํฌ๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค). ์‚ฌ์ „ ํ›ˆ๋ จ ์ž‘์—…์—์„œ ์ธ์ฝ”๋”์— ์ ์šฉ๋˜๋Š” ๋ณ€ํ™˜๋“ค์˜ ๊ตฌ์„ฑ์€ ๋‹ค์Œ๊ณผ ๊ฐ™์Šต๋‹ˆ๋‹ค:

์‚ฌ์ „ ํ›ˆ๋ จ ์ž‘์—…์—์„œ ์ธ์ฝ”๋”์— ์ ์šฉ๋˜๋Š” ๋ณ€ํ™˜๋“ค์˜ ๊ตฌ์„ฑ์€ ๋‹ค์Œ๊ณผ ๊ฐ™์Šต๋‹ˆ๋‹ค:

  * ๋ฌด์ž‘์œ„ ํ† ํฐ ๋งˆ์Šคํ‚น (BERT ์ฒ˜๋Ÿผ)
  * ๋ฌด์ž‘์œ„ ํ† ํฐ ์‚ญ์ œ
  * k๊ฐœ ํ† ํฐ์˜ ๋ฒ”์œ„๋ฅผ ๋‹จ์ผ ๋งˆ์Šคํฌ ํ† ํฐ์œผ๋กœ ๋งˆ์Šคํ‚น (0๊ฐœ ํ† ํฐ์˜ ๋ฒ”์œ„๋Š” ๋งˆ์Šคํฌ ํ† ํฐ์˜ ์‚ฝ์ž…์„ ์˜๋ฏธ)
  * ๋ฌธ์žฅ ์ˆœ์„œ ๋’ค์„ž๊ธฐ
  * ํŠน์ • ํ† ํฐ์—์„œ ์‹œ์ž‘ํ•˜๋„๋ก ๋ฌธ์„œ ํšŒ์ „

## ๊ตฌํ˜„ ๋…ธํŠธ[[implementation-notes]]

- Bart๋Š” ์‹œํ€€์Šค ๋ถ„๋ฅ˜์— `token_type_ids`๋ฅผ ์‚ฌ์šฉํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ์ ์ ˆํ•˜๊ฒŒ ๋‚˜๋ˆ„๊ธฐ ์œ„ํ•ด์„œ [`BartTokenizer`]๋‚˜
  [`~BartTokenizer.encode`]๋ฅผ ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค.
- [`BartModel`]์˜ ์ •๋ฐฉํ–ฅ ์ „๋‹ฌ์€ `decoder_input_ids`๊ฐ€ ์ „๋‹ฌ๋˜์ง€ ์•Š์œผ๋ฉด `decoder_input_ids`๋ฅผ ์ž๋™์œผ๋กœ ์ƒ์„ฑํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ด๋Š” ๋‹ค๋ฅธ ์ผ๋ถ€ ๋ชจ๋ธ๋ง API์™€ ๋‹ค๋ฅธ ์ ์ž…๋‹ˆ๋‹ค. ์ด ๊ธฐ๋Šฅ์˜ ์ผ๋ฐ˜์ ์ธ ์‚ฌ์šฉ ์‚ฌ๋ก€๋Š” ๋งˆ์Šคํฌ ์ฑ„์šฐ๊ธฐ(mask filling)์ž…๋‹ˆ๋‹ค.
-  ๋ชจ๋ธ ์˜ˆ์ธก์€ `forced_bos_token_id=0`์ผ ๋•Œ ๊ธฐ์กด ๊ตฌํ˜„๊ณผ ๋™์ผํ•˜๊ฒŒ ์ž‘๋™ํ•˜๋„๋ก ์˜๋„๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ, [`fairseq.encode`]์— ์ „๋‹ฌํ•˜๋Š” ๋ฌธ์ž์—ด์ด ๊ณต๋ฐฑ์œผ๋กœ ์‹œ์ž‘ํ•  ๋•Œ๋งŒ ์ด ๊ธฐ๋Šฅ์ด ์ž‘๋™ํ•ฉ๋‹ˆ๋‹ค.
- [`~generation.GenerationMixin.generate`]๋Š” ์š”์•ฝ๊ณผ ๊ฐ™์€ ์กฐ๊ฑด๋ถ€ ์ƒ์„ฑ ์ž‘์—…์— ์‚ฌ์šฉ๋˜์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ์ž์„ธํ•œ ๋‚ด์šฉ์€ ํ•ด๋‹น ๋ฌธ์„œ์˜ ์˜ˆ์ œ๋ฅผ ์ฐธ์กฐํ•˜์„ธ์š”.
- *facebook/bart-large-cnn* ๊ฐ€์ค‘์น˜๋ฅผ ๋กœ๋“œํ•˜๋Š” ๋ชจ๋ธ์€ `mask_token_id`๊ฐ€ ์—†๊ฑฐ๋‚˜, ๋งˆ์Šคํฌ ์ฑ„์šฐ๊ธฐ ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค.

## ๋งˆ์Šคํฌ ์ฑ„์šฐ๊ธฐ[[mask-filling]]

`facebook/bart-base`์™€ `facebook/bart-large` ์ฒดํฌํฌ์ธํŠธ๋Š” ๋ฉ€ํ‹ฐ ํ† ํฐ ๋งˆ์Šคํฌ๋ฅผ ์ฑ„์šฐ๋Š”๋ฐ ์‚ฌ์šฉ๋  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

```python
from transformers import BartForConditionalGeneration, BartTokenizer

model = BartForConditionalGeneration.from_pretrained("facebook/bart-large", forced_bos_token_id=0)
tok = BartTokenizer.from_pretrained("facebook/bart-large")
example_english_phrase = "UN Chief Says There Is No <mask> in Syria"
batch = tok(example_english_phrase, return_tensors="pt")
generated_ids = model.generate(batch["input_ids"])
assert tok.batch_decode(generated_ids, skip_special_tokens=True) == [
    "UN Chief Says There Is No Plan to Stop Chemical Weapons in Syria"
]
```

## ์ž๋ฃŒ[[resources]]

BART๋ฅผ ์‹œ์ž‘ํ•˜๋Š” ๋ฐ ๋„์›€์ด ๋˜๋Š” Hugging Face์™€ community ์ž๋ฃŒ ๋ชฉ๋ก(๐ŸŒŽ๋กœ ํ‘œ์‹œ๋จ) ์ž…๋‹ˆ๋‹ค. ์—ฌ๊ธฐ์— ํฌํ•จ๋  ์ž๋ฃŒ๋ฅผ ์ œ์ถœํ•˜๊ณ  ์‹ถ์œผ์‹œ๋‹ค๋ฉด PR(Pull Request)๋ฅผ ์—ด์–ด์ฃผ์„ธ์š”. ๋ฆฌ๋ทฐ ํ•ด๋“œ๋ฆฌ๊ฒ ์Šต๋‹ˆ๋‹ค! ์ž๋ฃŒ๋Š” ๊ธฐ์กด ์ž๋ฃŒ๋ฅผ ๋ณต์ œํ•˜๋Š” ๋Œ€์‹  ์ƒˆ๋กœ์šด ๋‚ด์šฉ์„ ๋‹ด๊ณ  ์žˆ์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.

<PipelineTag pipeline="summarization"/>

- [๋ถ„์‚ฐํ˜• ํ•™์Šต: ๐Ÿค— Transformers์™€ Amazon SageMaker๋ฅผ ์ด์šฉํ•˜์—ฌ ์š”์•ฝํ•˜๊ธฐ ์œ„ํ•œ BART/T5 ํ•™์Šต](https://huggingface.co/blog/sagemaker-distributed-training-seq2seq)์— ๋Œ€ํ•œ ๋ธ”๋กœ๊ทธ ํฌ์ŠคํŠธ.
- [blurr๋ฅผ ์ด์šฉํ•˜์—ฌ fastai๋กœ ์š”์•ฝํ•˜๊ธฐ ์œ„ํ•œ BART๋ฅผ ๋ฏธ์„ธ ์กฐ์ •](https://colab.research.google.com/github/ohmeow/ohmeow_website/blob/master/posts/2021-05-25-mbart-sequence-classification-with-blurr.ipynb)ํ•˜๋Š” ๋ฐฉ๋ฒ•์— ๋Œ€ํ•œ ๋…ธํŠธ๋ถ. ๐ŸŒŽ
- [Trainer ํด๋ž˜์Šค๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ๋‘ ๊ฐ€์ง€ ์–ธ์–ด๋กœ ์š”์•ฝํ•˜๊ธฐ ์œ„ํ•œ BART ๋ฏธ์„ธ ์กฐ์ •](https://colab.research.google.com/github/elsanns/xai-nlp-notebooks/blob/master/fine_tune_bart_summarization_two_langs.ipynb)ํ•˜๋Š” ๋ฐฉ๋ฒ•์— ๋Œ€ํ•œ ๋…ธํŠธ๋ถ. ๐ŸŒŽ
- ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/pytorch/summarization)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/summarization.ipynb)์—์„œ๋Š”  [`BartForConditionalGeneration`]์ด ์ง€์›๋ฉ๋‹ˆ๋‹ค.
- [`TFBartForConditionalGeneration`]๋Š” ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/summarization)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/summarization-tf.ipynb)์—์„œ ์ง€์›๋ฉ๋‹ˆ๋‹ค.
- ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ์—์„œ๋Š”](https://github.com/huggingface/transformers/tree/main/examples/flax/summarization)[`FlaxBartForConditionalGeneration`]์ด ์ง€์›๋ฉ๋‹ˆ๋‹ค.
- Hugging Face `datasets` ๊ฐ์ฒด๋ฅผ ํ™œ์šฉํ•˜์—ฌ [`BartForConditionalGeneration`]์„ ํ•™์Šต์‹œํ‚ค๋Š” ๋ฐฉ๋ฒ•์˜ ์˜ˆ๋Š” ์ด [ํฌ๋Ÿผ ํ† ๋ก ](https://discuss.huggingface.co/t/train-bart-for-conditional-generation-e-g-summarization/1904)์—์„œ ์ฐพ์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
- ๐Ÿค— Hugging Face ์ฝ”์Šค์˜ [์š”์•ฝ](https://huggingface.co/course/chapter7/5?fw=pt#summarization)์žฅ.
- [์š”์•ฝ ์ž‘์—… ๊ฐ€์ด๋“œ](../tasks/summarization)

<PipelineTag pipeline="fill-mask"/>

- [`BartForConditionalGeneration`]๋Š” ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#robertabertdistilbert-and-masked-language-modeling)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb)์„ ์ฐธ๊ณ ํ•˜์„ธ์š”.
- [`TFBartForConditionalGeneration`]๋Š” ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_mlmpy)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb)์„ ์ฐธ๊ณ ํ•˜์„ธ์š”.
- [`FlaxBartForConditionalGeneration`]๋Š” ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/flax/language-modeling#masked-language-modeling)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/masked_language_modeling_flax.ipynb)์„ ์ฐธ๊ณ  ํ•˜์„ธ์š”.
- ๐Ÿค— Hugging Face ์ฝ”์Šค์˜ [๋งˆ์Šคํฌ ์–ธ์–ด ๋ชจ๋ธ๋ง](https://huggingface.co/course/chapter7/3?fw=pt) ์ฑ•ํ„ฐ.
- [๋งˆ์Šคํฌ ์–ธ์–ด ๋ชจ๋ธ๋ง ์ž‘์—… ๊ฐ€์ด๋“œ](../tasks/masked_language_modeling)

<PipelineTag pipeline="translation"/>

- [Seq2SeqTrainer๋ฅผ ์ด์šฉํ•˜์—ฌ ์ธ๋„์–ด๋ฅผ ์˜์–ด๋กœ ๋ฒˆ์—ญํ•˜๋Š” mBART๋ฅผ ๋ฏธ์„ธ ์กฐ์ •](https://colab.research.google.com/github/vasudevgupta7/huggingface-tutorials/blob/main/translation_training.ipynb)ํ•˜๋Š” ๋ฐฉ๋ฒ•์— ๋Œ€ํ•œ ๊ฐ€์ด๋“œ. ๐ŸŒŽ
- [`BartForConditionalGeneration`]๋Š” ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/pytorch/translation)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/translation.ipynb)์„ ์ฐธ๊ณ ํ•˜์„ธ์š”.
- [`TFBartForConditionalGeneration`]๋Š” ์ด [์˜ˆ์ œ ์Šคํฌ๋ฆฝํŠธ](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/translation)์™€ [๋…ธํŠธ๋ถ](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/translation-tf.ipynb)์„ ์ฐธ๊ณ  ํ•˜์„ธ์š”.
- [๋ฒˆ์—ญ ์ž‘์—… ๊ฐ€์ด๋“œ](../tasks/translation)

์ถ”๊ฐ€์ ์œผ๋กœ ๋ณผ ๊ฒƒ๋“ค:
- [ํ…์ŠคํŠธ ๋ถ„๋ฅ˜ ์ž‘์—… ๊ฐ€์ด๋“œ](../tasks/sequence_classification)
- [์งˆ๋ฌธ ๋‹ต๋ณ€ ์ž‘์—… ๊ฐ€์ด๋“œ](../tasks/question_answering)
- [์ธ๊ณผ์  ์–ธ์–ด ๋ชจ๋ธ๋ง ์ž‘์—… ๊ฐ€์ด๋“œ](../tasks/language_modeling)
- ์ด [๋…ผ๋ฌธ](https://arxiv.org/abs/2010.13002)์€ [์ฆ๋ฅ˜๋œ ์ฒดํฌํฌ์ธํŠธ](https://huggingface.co/models?search=distilbart)์— ๋Œ€ํ•ด ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค.

## BartConfig[[transformers.BartConfig]]

[[autodoc]] BartConfig
    - all

## BartTokenizer[[transformers.BartTokenizer]]

[[autodoc]] BartTokenizer
    - all

## BartTokenizerFast[[transformers.BartTokenizerFast]]

[[autodoc]] BartTokenizerFast
    - all


<frameworkcontent>
<pt>

## BartModel[[transformers.BartModel]]

[[autodoc]] BartModel
    - forward

## BartForConditionalGeneration[[transformers.BartForConditionalGeneration]]

[[autodoc]] BartForConditionalGeneration
    - forward

## BartForSequenceClassification[[transformers.BartForSequenceClassification]]

[[autodoc]] BartForSequenceClassification
    - forward

## BartForQuestionAnswering[[transformers.BartForQuestionAnswering]]

[[autodoc]] BartForQuestionAnswering
    - forward

## BartForCausalLM[[transformers.BartForCausalLM]]

[[autodoc]] BartForCausalLM
    - forward

</pt>
<tf>

## TFBartModel[[transformers.TFBartModel]]

[[autodoc]] TFBartModel
    - call

## TFBartForConditionalGeneration[[transformers.TFBartForConditionalGeneration]]

[[autodoc]] TFBartForConditionalGeneration
    - call

## TFBartForSequenceClassification[[transformers.TFBartForSequenceClassification]]

[[autodoc]] TFBartForSequenceClassification
    - call

</tf>
<jax>

## FlaxBartModel[[transformers.FlaxBartModel]]

[[autodoc]] FlaxBartModel
    - __call__
    - encode
    - decode

## FlaxBartForConditionalGeneration[[transformers.FlaxBartForConditionalGeneration]]

[[autodoc]] FlaxBartForConditionalGeneration
    - __call__
    - encode
    - decode

## FlaxBartForSequenceClassification[[transformers.FlaxBartForSequenceClassification]]

[[autodoc]] FlaxBartForSequenceClassification
    - __call__
    - encode
    - decode

## FlaxBartForQuestionAnswering[[transformers.FlaxBartForQuestionAnswering]]

[[autodoc]] FlaxBartForQuestionAnswering
    - __call__
    - encode
    - decode

## FlaxBartForCausalLM[[transformers.FlaxBartForCausalLM]]

[[autodoc]] FlaxBartForCausalLM
    - __call__
</jax>
</frameworkcontent>