Commit
·
2441c62
1
Parent(s):
b188778
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,67 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
language:
|
4 |
+
- en
|
5 |
---
|
6 |
+
|
7 |
+
# rrivera1849/LUAR-CRUD
|
8 |
+
|
9 |
+
Author Style Representations using [LUAR](https://aclanthology.org/2021.emnlp-main.70.pdf).
|
10 |
+
|
11 |
+
The LUAR training and evaluation repository can be found [here](https://github.com/llnl/luar).
|
12 |
+
|
13 |
+
This model was trained on the Reddit Million User Dataset (MUD) found [here](https://aclanthology.org/2021.naacl-main.415.pdf).
|
14 |
+
|
15 |
+
## Usage
|
16 |
+
|
17 |
+
```python
|
18 |
+
from transformers import AutoModel, AutoTokenizer
|
19 |
+
|
20 |
+
tokenizer = AutoTokenizer.from_pretrained("rrivera1849/LUAR-CRUD")
|
21 |
+
model = AutoModel.from_pretrained("rrivera1849/LUAR-CRUD")
|
22 |
+
|
23 |
+
# we embed `episodes`, a colletion of documents presumed to come from an author
|
24 |
+
# NOTE: make sure that `episode_length` consistent across `episode`
|
25 |
+
batch_size = 3
|
26 |
+
episode_length = 16
|
27 |
+
text = [
|
28 |
+
["Foo"] * episode_length,
|
29 |
+
["Bar"] * episode_length,
|
30 |
+
["Zoo"] * episode_length,
|
31 |
+
]
|
32 |
+
text = [j for i in text for j in i]
|
33 |
+
tokenized_text = tokenizer(
|
34 |
+
text,
|
35 |
+
max_length=32,
|
36 |
+
padding="max_length",
|
37 |
+
truncation=True,
|
38 |
+
return_tensors="pt"
|
39 |
+
)
|
40 |
+
# inputs size: (batch_size, episode_length, max_token_length)
|
41 |
+
tokenized_text["input_ids"] = tokenized_text["input_ids"].reshape(batch_size, episode_length, -1)
|
42 |
+
tokenized_text["attention_mask"] = tokenized_text["attention_mask"].reshape(batch_size, episode_length, -1)
|
43 |
+
print(tokenized_text["input_ids"].size()) # torch.Size([3, 16, 32])
|
44 |
+
print(tokenized_text["attention_mask"].size()) # torch.Size([3, 16, 32])
|
45 |
+
|
46 |
+
out = model(**tokenized_text)
|
47 |
+
print(out.size()) # torch.Size([3, 512])
|
48 |
+
```
|
49 |
+
|
50 |
+
## Citing & Authors
|
51 |
+
|
52 |
+
If you find this model helpful, feel free to cite our [publication](https://aclanthology.org/2021.emnlp-main.70.pdf).
|
53 |
+
|
54 |
+
```
|
55 |
+
@inproceedings{uar-emnlp2021,
|
56 |
+
author = {Rafael A. Rivera Soto and Olivia Miano and Juanita Ordonez and Barry Chen and Aleem Khan and Marcus Bishop and Nicholas Andrews},
|
57 |
+
title = {Learning Universal Authorship Representations},
|
58 |
+
booktitle = {EMNLP},
|
59 |
+
year = {2021},
|
60 |
+
}
|
61 |
+
```
|
62 |
+
|
63 |
+
## License
|
64 |
+
|
65 |
+
LUAR is distributed under the terms of the Apache License (Version 2.0).
|
66 |
+
|
67 |
+
All new contributions must be made under the Apache-2.0 licenses.
|