joelniklaus
commited on
Commit
·
012296c
1
Parent(s):
4832d46
improved dataset card
Browse files
README.md
CHANGED
@@ -105,12 +105,13 @@ text_type is one of the following:
|
|
105 |
- legislation
|
106 |
- other
|
107 |
|
108 |
-
|
109 |
-
|
|
|
110 |
from datasets import load_dataset
|
111 |
|
112 |
-
config = 'en_contracts'
|
113 |
-
dataset = load_dataset('joelito/Multi_Legal_Pile', config, split='
|
114 |
```
|
115 |
|
116 |
'config' is a combination of language and text_type, e.g. 'en_contracts' or 'de_caselaw'.
|
@@ -179,7 +180,7 @@ we disregard many US resources that are either very specialized (e.g., tax rulin
|
|
179 |
outdated/historical (e.g., founding letters),
|
180 |
very small (less than 20MB),
|
181 |
not legal language in the strict sense (conversations),
|
182 |
-
or the information overlaps with other sources (study materials).
|
183 |
If you are interested in a US-based (US-biased) model, refer to the "Pile of (US) Law" by Henderson et al. (2022).
|
184 |
Analyses are put into the "other" category because in mc4_legal we also likely have similar text.
|
185 |
|
|
|
105 |
- legislation
|
106 |
- other
|
107 |
|
108 |
+
|
109 |
+
Use the dataset like this:
|
110 |
+
```python
|
111 |
from datasets import load_dataset
|
112 |
|
113 |
+
config = 'en_contracts' # {language}_{text_type}
|
114 |
+
dataset = load_dataset('joelito/Multi_Legal_Pile', config, split='train', streaming=True)
|
115 |
```
|
116 |
|
117 |
'config' is a combination of language and text_type, e.g. 'en_contracts' or 'de_caselaw'.
|
|
|
180 |
outdated/historical (e.g., founding letters),
|
181 |
very small (less than 20MB),
|
182 |
not legal language in the strict sense (conversations),
|
183 |
+
or the information overlaps with other sources (e.g. study materials or eurlex).
|
184 |
If you are interested in a US-based (US-biased) model, refer to the "Pile of (US) Law" by Henderson et al. (2022).
|
185 |
Analyses are put into the "other" category because in mc4_legal we also likely have similar text.
|
186 |
|