Skip to content

Commit e74ad23

Browse files
authored
Add chunked maze data links
1 parent 6a1cd84 commit e74ad23

File tree

1 file changed

+7
-1
lines changed

1 file changed

+7
-1
lines changed

spirl/data/README.md

+7-1
Original file line numberDiff line numberDiff line change
@@ -16,9 +16,15 @@ gdown https://drive.google.com/uc?id=1VobNYJQw_Uwax0kbFG7KOXTgv6ja2s1M
1616
1717
# Download Office Dataset
1818
gdown https://drive.google.com/uc?id=1FOE1kiU71nB-3KCDuxGqlAqRQbKmSk80
19+
20+
# Download Maze Dataset in smaller chunks (~3GB each)
21+
gdown https://drive.google.com/uc?id=15fjt8QMC6xMpqTMftLF1RCWAe3jvklGb
22+
gdown https://drive.google.com/uc?id=1aTz94EJYPU5A-h-EV5CxCIfd8io2e-EH
23+
gdown https://drive.google.com/uc?id=1EiahoGgGiS7ol-Xx-DHIiEroZSeG9jEr
24+
gdown https://drive.google.com/uc?id=1y5VafZN_95tHStEHKsoQvLrvw2y0pmRQ
1925
```
2026

21-
Finally, unzip the downloaded zip files, update the `data_dir` in your config file to the unzipped directory with the dataset you want to use and then you are ready to start training!
27+
Finally, unzip the downloaded zip files, update the `data_dir` in your config file to the unzipped directory with the dataset you want to use and then you are ready to start training! For the chunked maze data, make sure to copy all unzipped contents into a single folder called `maze`.
2228

2329
## Re-Generating Datasets
2430

0 commit comments

Comments
 (0)