Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,48 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
tags:
|
6 |
+
- conversation
|
7 |
+
- dialogue
|
8 |
+
- commonsense
|
9 |
---
|
10 |
+
|
11 |
+
# Model Card for ConvoSenseGenerator
|
12 |
+
|
13 |
+
ConvoSenseGenerator is a generative model that produces commonsense inferences for dialogue contexts, covering 10 common social commonsense types such as emotional reactions, motivations, causes, subsequent events, and more! It is trained on the large-scale dataset, ConvoSense, that is collected synthetically using ChatGPT3.5. ConvoSenseGenerator produces inferences that humans judge to achieve high reasonability, high rates of novel information for the corresponding dialogue contexts, and high degree of detail, outperforming models trained on previous datasets that are human-written.
|
14 |
+
|
15 |
+
## Model Description
|
16 |
+
- **Repository:** [Code](https://github.com/emorynlp/convosense)
|
17 |
+
- **Paper:** [ConvoSense: Overcoming Monotonous Commonsense Inferences for Conversational AI](https://github.com/emorynlp/ConvoSense/blob/main/paper.pdf)
|
18 |
+
- **Point of Contact:** [Sarah E. Finch](mailto:sfillwo@emory.edu)
|
19 |
+
|
20 |
+
## Model Training
|
21 |
+
|
22 |
+
ConvoSenseGenerator is trained on our recent dataset: 🥤[ConvoSense](https://huggingface.co/datasets/allenai/soda).
|
23 |
+
The backbone model of ConvoSenseGenerator is [T5-3b](https://huggingface.co/t5-3b).
|
24 |
+
|
25 |
+
### How to use
|
26 |
+
|
27 |
+
Below is a simple code snippet to get ConvoSenseGenerator running :)
|
28 |
+
|
29 |
+
```python
|
30 |
+
import torch
|
31 |
+
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
32 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
33 |
+
tokenizer = AutoTokenizer.from_pretrained("sefinch/ConvoSenseGenerator")
|
34 |
+
model = AutoModelForSeq2SeqLM.from_pretrained("sefinch/ConvoSenseGenerator").to(device)
|
35 |
+
...
|
36 |
+
```
|
37 |
+
|
38 |
+
### Citation
|
39 |
+
|
40 |
+
Please cite our work if you find the resources in this repository useful:
|
41 |
+
```
|
42 |
+
@article{convosense_finch:24,
|
43 |
+
author = {Finch, Sarah E. and Choi, Jinho D.},
|
44 |
+
title = "{ConvoSense: Overcoming Monotonous Commonsense Inferences for Conversational AI}",
|
45 |
+
journal = {Transactions of the Association for Computational Linguistics},
|
46 |
+
year = {2024}
|
47 |
+
}
|
48 |
+
```
|