andrewsleader
commited on
Commit
•
dbc0542
1
Parent(s):
3f25e97
Update README.md
Browse files
README.md
CHANGED
@@ -18,45 +18,8 @@ Overview of the Temporal Graph Benchmark (TGB) pipeline:
|
|
18 |
- Novel TG models can be easily evaluated on TGB datasets via reproducible and realistic evaluation protocols.
|
19 |
- TGB provides public and online leaderboards to track recent developments in temporal graph learning domain.
|
20 |
|
21 |
-
![TGB dataloading and evaluation pipeline](imgs/pipeline.png)
|
22 |
|
23 |
-
**To submit to [TGB leaderboard](https://tgb.complexdatalab.com/), please fill in this [google form](https://forms.gle/SEsXvN1QHo9tSFwx9)**
|
24 |
|
25 |
-
**See all version differences and update notes [here](https://tgb.complexdatalab.com/docs/update/)**
|
26 |
-
|
27 |
-
### Announcements
|
28 |
-
|
29 |
-
**Excited to announce TGX, a companion package for analyzing temporal graphs in WSDM 2024 Demo Track**
|
30 |
-
|
31 |
-
TGX supports all TGB datasets and provides numerous temporal graph visualization plots and statistics out of the box. See our paper: [Temporal Graph Analysis with TGX](https://arxiv.org/abs/2402.03651) and [TGX website](https://complexdata-mila.github.io/TGX/).
|
32 |
-
|
33 |
-
**Excited to announce that TGB has been accepted to NeurIPS 2023 Datasets and Benchmarks Track**
|
34 |
-
|
35 |
-
Thanks to everyone for your help in improving TGB! we will continue to improve TGB based on your feedback and suggestions.
|
36 |
-
|
37 |
-
**Please update to version `0.9.2`**
|
38 |
-
|
39 |
-
#### version `0.9.2`
|
40 |
-
|
41 |
-
Update the fix for `tgbl-flight` where now the unix timestamps are provided directly in the dataset. If you had issues with `tgbl-flight`, please remove `TGB/tgb/datasets/tgbl_flight`and redownload the dataset for a clean install
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
#### version `0.9.1`
|
46 |
-
|
47 |
-
Fixed an issue for `tgbl-flight` where the timestamp conversion is incorrect due to time zone differences. If you had issues with `tgbl-flight` before, please update your package.
|
48 |
-
|
49 |
-
|
50 |
-
#### version `0.9.0`
|
51 |
-
|
52 |
-
Added the large `tgbn-token` dataset with 72 million edges to the `nodeproppred` dataset.
|
53 |
-
|
54 |
-
Fixed errors in `tgbl-coin` and `tgbl-flight` where a small set of edges are not sorted chronologically. Please update your dataset version for them to version 2 (will be prompted in terminal).
|
55 |
-
|
56 |
-
|
57 |
-
### Pip Install
|
58 |
-
|
59 |
-
You can install TGB via [pip](https://pypi.org/project/py-tgb/). **Requires python >= 3.9**
|
60 |
```
|
61 |
pip install py-tgb
|
62 |
```
|
|
|
18 |
- Novel TG models can be easily evaluated on TGB datasets via reproducible and realistic evaluation protocols.
|
19 |
- TGB provides public and online leaderboards to track recent developments in temporal graph learning domain.
|
20 |
|
|
|
21 |
|
|
|
22 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
```
|
24 |
pip install py-tgb
|
25 |
```
|