update pretrained name
Browse files
README.md
CHANGED
@@ -36,7 +36,8 @@ Notice, we removed the 'B-','I-' etc from data label.🗡
|
|
36 |
```python
|
37 |
from transformers import pipeline
|
38 |
|
39 |
-
|
|
|
40 |
ner("Your text", aggregation_strategy="first")
|
41 |
```
|
42 |
And here is to make your output more consecutive ⭐️
|
@@ -44,7 +45,7 @@ And here is to make your output more consecutive ⭐️
|
|
44 |
```python
|
45 |
import pandas as pd
|
46 |
from transformers import AutoTokenizer
|
47 |
-
tokenizer = AutoTokenizer.from_pretrained(
|
48 |
|
49 |
def clean_output(outputs):
|
50 |
results = []
|
@@ -96,3 +97,7 @@ def entity_table(pipeline, **pipeline_kw):
|
|
96 |
# will return a dataframe
|
97 |
entity_table(ner)(YOUR_VERY_CONTENTFUL_TEXT)
|
98 |
```
|
|
|
|
|
|
|
|
|
|
36 |
```python
|
37 |
from transformers import pipeline
|
38 |
|
39 |
+
PRETRAINED = "raynardj/ner-gene-dna-rna-jnlpba-pubmed"
|
40 |
+
ner = pipeline(task="ner",model=PRETRAINED, tokenizer=PRETRAINED)
|
41 |
ner("Your text", aggregation_strategy="first")
|
42 |
```
|
43 |
And here is to make your output more consecutive ⭐️
|
|
|
45 |
```python
|
46 |
import pandas as pd
|
47 |
from transformers import AutoTokenizer
|
48 |
+
tokenizer = AutoTokenizer.from_pretrained(PRETRAINED)
|
49 |
|
50 |
def clean_output(outputs):
|
51 |
results = []
|
|
|
97 |
# will return a dataframe
|
98 |
entity_table(ner)(YOUR_VERY_CONTENTFUL_TEXT)
|
99 |
```
|
100 |
+
|
101 |
+
> check our NER model on
|
102 |
+
* [gene and gene products](/raynardj/ner-gene-dna-rna-jnlpba-pubmed)
|
103 |
+
* [chemical substance](/raynardj/ner-chemical-bionlp-bc5cdr-pubmed).
|