Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
chenpangpang
transformers
Commits
036c2c6b
Unverified
Commit
036c2c6b
authored
Jun 01, 2020
by
Mehrdad Farahani
Committed by
GitHub
Jun 01, 2020
Browse files
Update HooshvareLab/bert-base-parsbert-uncased (#4687)
mBERT results added regarding NER datasets!
parent
74872c19
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
16 additions
and
16 deletions
+16
-16
model_cards/HooshvareLab/bert-base-parsbert-uncased/README.md
...l_cards/HooshvareLab/bert-base-parsbert-uncased/README.md
+16
-16
No files found.
model_cards/HooshvareLab/bert-base-parsbert-uncased/README.md
View file @
036c2c6b
...
...
@@ -28,29 +28,29 @@ The following table summarizes the F1 score obtained by ParsBERT as compared to
### Sentiment Analysis (SA) task
| Dataset | ParsBERT |
Multilingual
BERT | DeepSentiPers |
|:--------------------------:|:---------:|:-----
------------
:|:-------------:|
| Digikala User Comments | 81.74
*
|
80.74
| - |
| SnappFood User Comments | 88.12
*
|
87.87
| - |
| SentiPers (Multi Class) | 71.11
*
|
-
| 69.33 |
| SentiPers (Binary Class) | 92.13
*
|
-
| 91.98 |
| Dataset | ParsBERT |
m
BERT | DeepSentiPers |
|:--------------------------:|:---------:|:-----:|:-------------:|
| Digikala User Comments | 81.74
*
|
80.74
| - |
| SnappFood User Comments | 88.12
*
|
87.87
| - |
| SentiPers (Multi Class) | 71.11
*
|
-
| 69.33 |
| SentiPers (Binary Class) | 92.13
*
|
-
| 91.98 |
### Text Classification (TC) task
| Dataset | ParsBERT |
Multilingual
BERT |
|:-----------------:|:--------:|:-----
------------
:|
| Digikala Magazine | 93.59
*
|
90.72
|
| Persian News | 97.19
*
|
95.79
|
| Dataset | ParsBERT |
m
BERT |
|:-----------------:|:--------:|:-----:|
| Digikala Magazine | 93.59
*
|
90.72
|
| Persian News | 97.19
*
|
95.79
|
### Named Entity Recognition (NER) task
| Dataset | ParsBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|:-------:|:--------:|:----------:|:--------------:|:----------:|:----------------:|:------------:|
| PEYMA | 9
8.79
*
| - | 90.59 | - | 84.00 | - |
| ARMAN | 9
3.10
*
| 89.9 | 84.03 | 86.55 | - | 77.45 |
| Dataset | ParsBERT |
mBERT |
MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|:-------:|:--------:|:--------
:|:--------
--:|:--------------:|:----------:|:----------------:|:------------:|
| PEYMA | 9
3.10
*
| 86.64
| - | 90.59 | - | 84.00 | - |
| ARMAN | 9
8.79
*
| 95.89
| 89.9 | 84.03 | 86.55 | - | 77.45 |
**If you tested ParsBERT on a public dataset and you want to add your results to the table above, open a pull request or contact us. Also make sure to have your code available online so we can add it as a reference**
...
...
@@ -66,10 +66,10 @@ config = AutoConfig.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
tokenizer
=
AutoTokenizer
.
from_pretrained
(
"HooshvareLab/bert-base-parsbert-uncased"
)
model
=
AutoModel
.
from_pretrained
(
"HooshvareLab/bert-base-parsbert-uncased"
)
text
=
"ما در هوشواره معتقدیم با انتقال صحیح دانش و آگاهی، همه
ی
افراد میتوانند از ابزارهای هوشمند استفاده کنند. شعار ما هوش مصنوعی برای همه است."
text
=
"ما در هوشواره معتقدیم با انتقال صحیح دانش و آگاهی، همه افراد میتوانند از ابزارهای هوشمند استفاده کنند. شعار ما هوش مصنوعی برای همه است."
tokenizer
.
tokenize
(
text
)
>>>
[
'ما'
,
'در'
,
'هوش'
,
'##واره'
,
'معتقدیم'
,
'با'
,
'انتقال'
,
'صحیح'
,
'دانش'
,
'و'
,
'اگاهی'
,
'،'
,
'همه
ی
'
,
'افراد'
,
'میتوانند'
,
'از'
,
'ابزارهای'
,
'هوشمند'
,
'استفاده'
,
'کنند'
,
'.'
,
'شعار'
,
'ما'
,
'هوش'
,
'مصنوعی'
,
'برای'
,
'همه'
,
'است'
,
'.'
]
>>>
[
'ما'
,
'در'
,
'هوش'
,
'##واره'
,
'معتقدیم'
,
'با'
,
'انتقال'
,
'صحیح'
,
'دانش'
,
'و'
,
'اگاهی'
,
'،'
,
'همه'
,
'افراد'
,
'میتوانند'
,
'از'
,
'ابزارهای'
,
'هوشمند'
,
'استفاده'
,
'کنند'
,
'.'
,
'شعار'
,
'ما'
,
'هوش'
,
'مصنوعی'
,
'برای'
,
'همه'
,
'است'
,
'.'
]
```
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment