File size: 2,317 Bytes
e108204
 
97178a3
 
 
 
 
 
d1b1f9b
44b2a96
97178a3
 
 
 
 
 
e108204
97178a3
 
 
 
334e795
 
 
97178a3
 
 
 
 
 
 
 
192c00a
97178a3
 
192c00a
97178a3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
226beba
97178a3
 
 
 
 
 
 
ffc95d8
97178a3
 
f5fd2e6
97178a3
 
 
 
 
 
 
 
 
deb5584
192c00a
 
 
 
 
 
 
 
deb5584
97178a3
 
44b2a96
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
---
license: apache-2.0
datasets:
- asset
- wi_locness
- GEM/wiki_auto_asset_turk
- discofuse
- zaemyung/IteraTeR_plus
- jfleg
- grammarly/coedit
language:
- en
metrics:
- sari
- bleu
- accuracy
---
# Model Card for CoEdIT-xxl

This model was obtained by fine-tuning the corresponding google/flan-t5-xxl model on the CoEdIT dataset.

**Paper:** CoEdIT: ext Editing by Task-Specific Instruction Tuning

**Authors:** Vipul Raheja, Dhruv Kumar, Ryan Koo, Dongyeop Kang

## Model Details

### Model Description

- **Language(s) (NLP)**: English
- **Finetuned from model:** google/flan-t5-xxl

### Model Sources

- **Repository:** https://github.com/vipulraheja/coedit
- **Paper:** https://arxiv.org/abs/2305.09857

## How to use
We make available the models presented in our paper. 

<table>
  <tr>
    <th>Model</th>
    <th>Number of parameters</th>
  </tr>
  <tr>
    <td>CoEdIT-large</td>
    <td>770M</td>
  </tr>
  <tr>
    <td>CoEdIT-xl</td>
    <td>3B</td>
  </tr>
  <tr>
    <td>CoEdIT-xxl</td>
    <td>11B</td>
  </tr>  
</table>


## Uses

## Text Revision Task
Given an edit instruction and an original text, our model can generate the edited version of the text.<br>

![task_specs](https://huggingface.co/grammarly/coedit-xl/resolve/main/task_examples.png)

## Usage
```python
from transformers import AutoTokenizer, T5ForConditionalGeneration

tokenizer = AutoTokenizer.from_pretrained("grammarly/coedit-xxl")
model = T5ForConditionalGeneration.from_pretrained("grammarly/coedit-xxl")
input_text = 'Fix grammatical errors in this sentence: When I grow up, I start to understand what he said is quite right.'
input_ids = tokenizer(input_text, return_tensors="pt").input_ids
outputs = model.generate(input_ids, max_length=256)
edited_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
```


#### Software
https://github.com/vipulraheja/coedit

## Citation

**BibTeX:**
```
@article{raheja2023coedit,
      title={CoEdIT: Text Editing by Task-Specific Instruction Tuning}, 
      author={Vipul Raheja and Dhruv Kumar and Ryan Koo and Dongyeop Kang},
      year={2023},
      eprint={2305.09857},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```

**APA:**
Raheja, V., Kumar, D., Koo, R., & Kang, D. (2023). CoEdIT: Text Editing by Task-Specific Instruction Tuning. ArXiv. /abs/2305.09857