holylovenia
commited on
Commit
•
d917635
1
Parent(s):
9e806a6
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -11,9 +11,9 @@ tags:
|
|
11 |
- self-supervised-pretraining
|
12 |
---
|
13 |
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
|
18 |
|
19 |
## Languages
|
@@ -23,25 +23,25 @@ ind, sun, jav
|
|
23 |
## Supported Tasks
|
24 |
|
25 |
Self Supervised Pretraining
|
26 |
-
|
27 |
## Dataset Usage
|
28 |
### Using `datasets` library
|
29 |
```
|
30 |
-
|
31 |
-
|
32 |
```
|
33 |
### Using `seacrowd` library
|
34 |
```import seacrowd as sc
|
35 |
# Load the dataset using the default config
|
36 |
-
|
37 |
# Check all available subsets (config names) of the dataset
|
38 |
-
|
39 |
# Load the dataset using a specific config
|
40 |
-
|
41 |
```
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
|
46 |
## Dataset Homepage
|
47 |
|
|
|
11 |
- self-supervised-pretraining
|
12 |
---
|
13 |
|
14 |
+
Indo4B-Plus is an extension of Indo4B, a large-scale Indonesian self-supervised pre-training corpus.
|
15 |
+
Indo4B-Plus extend Indo4B by adding two low-resource Indonesian local languages to the corpus, i.e., Sundanese and Javanese.
|
16 |
+
Indo4B-Plus adds 82,582,025 words (∼2.07%) of Sundanese sentences and 331,041,877 words (∼8.29%) of Javanese
|
17 |
|
18 |
|
19 |
## Languages
|
|
|
23 |
## Supported Tasks
|
24 |
|
25 |
Self Supervised Pretraining
|
26 |
+
|
27 |
## Dataset Usage
|
28 |
### Using `datasets` library
|
29 |
```
|
30 |
+
from datasets import load_dataset
|
31 |
+
dset = datasets.load_dataset("SEACrowd/indo4b_plus", trust_remote_code=True)
|
32 |
```
|
33 |
### Using `seacrowd` library
|
34 |
```import seacrowd as sc
|
35 |
# Load the dataset using the default config
|
36 |
+
dset = sc.load_dataset("indo4b_plus", schema="seacrowd")
|
37 |
# Check all available subsets (config names) of the dataset
|
38 |
+
print(sc.available_config_names("indo4b_plus"))
|
39 |
# Load the dataset using a specific config
|
40 |
+
dset = sc.load_dataset_by_config_name(config_name="<config_name>")
|
41 |
```
|
42 |
+
|
43 |
+
More details on how to load the `seacrowd` library can be found [here](https://github.com/SEACrowd/seacrowd-datahub?tab=readme-ov-file#how-to-use).
|
44 |
+
|
45 |
|
46 |
## Dataset Homepage
|
47 |
|