patrickvonplaten commited on
Commit
f946462
1 Parent(s): 5ca1e5b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -1
README.md CHANGED
@@ -14,11 +14,15 @@ This however only loads the first of >100 files.
14
  We now need to add **all** other files to this repo. This can be done as follows:
15
 
16
  1) Clone this repo (assuming `git lfs` is installed): `git clone https://huggingface.co/datasets/flax-community/german_common_crawl`
 
17
  2) For each file:
18
  `https://opendata.iisys.de/systemintegration/Datasets/CommonCrawl/head/de_head_0000_2016-18.tar.gz` - `https://opendata.iisys.de/systemintegration/Datasets/CommonCrawl/middle/de_middle_0009_2019-47.tar.gz`
19
 
20
  run the command `./convert_file.sh <file_name>` This command will download the file via `wget`, filter out all text that is below a threshold as explained here: https://opendata.iisys.de/systemintegration/Datasets/CommonCrawl/middle/de_middle_0009_2019-47.tar.gz and then converts the file into the correct format.
 
21
  3) Upload the file to this repo:
22
  `git add . && git commit -m "add file x" && git push
23
 
24
- Ideally this can be done in a loop on a computer that has enough CPU memory (Note that if this is done on a TPU VM, make sure to disable the TPU via `export JAX_PLATFORM_NAME=cpu`.
 
 
 
14
  We now need to add **all** other files to this repo. This can be done as follows:
15
 
16
  1) Clone this repo (assuming `git lfs` is installed): `git clone https://huggingface.co/datasets/flax-community/german_common_crawl`
17
+
18
  2) For each file:
19
  `https://opendata.iisys.de/systemintegration/Datasets/CommonCrawl/head/de_head_0000_2016-18.tar.gz` - `https://opendata.iisys.de/systemintegration/Datasets/CommonCrawl/middle/de_middle_0009_2019-47.tar.gz`
20
 
21
  run the command `./convert_file.sh <file_name>` This command will download the file via `wget`, filter out all text that is below a threshold as explained here: https://opendata.iisys.de/systemintegration/Datasets/CommonCrawl/middle/de_middle_0009_2019-47.tar.gz and then converts the file into the correct format.
22
+
23
  3) Upload the file to this repo:
24
  `git add . && git commit -m "add file x" && git push
25
 
26
+ Ideally this can be done in a loop on a computer that has enough CPU memory (Note that if this is done on a TPU VM, make sure to disable the TPU via `export JAX_PLATFORM_NAME=cpu`.
27
+
28
+ Also some description and file names have to be added correctly to the dataset.py script