Update README.md
Browse files
README.md
CHANGED
@@ -16,13 +16,51 @@ This model was created as part of the OLM project, which has the goal of continu
|
|
16 |
This is important because we want our models to know about events like COVID or
|
17 |
a presidential election right after they happen.
|
18 |
|
19 |
-
## Intended uses
|
20 |
-
|
21 |
-
You can use the raw model for
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
|
27 |
## Dataset
|
28 |
|
|
|
16 |
This is important because we want our models to know about events like COVID or
|
17 |
a presidential election right after they happen.
|
18 |
|
19 |
+
## Intended uses & limitations
|
20 |
+
|
21 |
+
You can use the raw model for masked language modeling, but it's mostly intended to
|
22 |
+
be fine-tuned on a downstream task, such as sequence classification, token classification or question answering.
|
23 |
+
|
24 |
+
### How to use
|
25 |
+
|
26 |
+
You can use this model directly with a pipeline for masked language modeling:
|
27 |
+
|
28 |
+
```python
|
29 |
+
>>> from transformers import pipeline
|
30 |
+
>>> unmasker = pipeline('fill-mask', model='olm/olm-roberta-base-oct-2022')
|
31 |
+
>>> unmasker("Hello I'm a <mask> model.")
|
32 |
+
[{'score': 0.10601336508989334,
|
33 |
+
'token': 2450,
|
34 |
+
'token_str': ' role',
|
35 |
+
'sequence': "Hello I'm a role model."},
|
36 |
+
{'score': 0.05792810395359993,
|
37 |
+
'token': 2677,
|
38 |
+
'token_str': ' former',
|
39 |
+
'sequence': "Hello I'm a former model."},
|
40 |
+
{'score': 0.057744599878787994,
|
41 |
+
'token': 1968,
|
42 |
+
'token_str': ' professional',
|
43 |
+
'sequence': "Hello I'm a professional model."},
|
44 |
+
{'score': 0.029099510982632637,
|
45 |
+
'token': 932,
|
46 |
+
'token_str': ' business',
|
47 |
+
'sequence': "Hello I'm a business model."},
|
48 |
+
{'score': 0.024220379069447517,
|
49 |
+
'token': 1840,
|
50 |
+
'token_str': ' young',
|
51 |
+
'sequence': "Hello I'm a young model."}]
|
52 |
+
```
|
53 |
+
|
54 |
+
Here is how to use this model to get the features of a given text in PyTorch:
|
55 |
+
|
56 |
+
```python
|
57 |
+
from transformers import AutoTokenizer, RobertaModel
|
58 |
+
tokenizer = AutoTokenizer.from_pretrained('olm/olm-roberta-base-oct-2022')
|
59 |
+
model = RobertaModel.from_pretrained("olm/olm-roberta-base-oct-2022")
|
60 |
+
text = "Replace me by any text you'd like."
|
61 |
+
encoded_input = tokenizer(text, return_tensors='pt')
|
62 |
+
output = model(**encoded_input)
|
63 |
+
```
|
64 |
|
65 |
## Dataset
|
66 |
|