Dmitry Chaplinsky
commited on
Commit
·
c2e7c9e
1
Parent(s):
72f6b63
Yet another try
Browse files- pipeline.py +5 -10
pipeline.py
CHANGED
@@ -1,13 +1,14 @@
|
|
1 |
from typing import Dict, List, Any
|
2 |
from nemo.collections.nlp.models import PunctuationCapitalizationModel
|
3 |
|
4 |
-
|
|
|
5 |
def __init__(self, path=""):
|
6 |
# IMPLEMENT_THIS
|
7 |
# Preload all the elements you are going to need at inference.
|
8 |
# For instance your model, processors, tokenizer that might be needed.
|
9 |
# This function is only called once, so do all the heavy processing I/O here"""
|
10 |
-
self.model = PunctuationCapitalizationModel.
|
11 |
|
12 |
def __call__(self, inputs: str) -> List[Dict[str, Any]]:
|
13 |
"""
|
@@ -31,14 +32,8 @@ class PreTrainedPipeline():
|
|
31 |
offset = 0
|
32 |
for tok, lab in zip(tokens, labels):
|
33 |
if lab != "OO":
|
34 |
-
res.append({
|
35 |
-
"entity_group": lab,
|
36 |
-
"word": tok,
|
37 |
-
"start": offset,
|
38 |
-
"end": offset + len(tok),
|
39 |
-
"score": 1
|
40 |
-
})
|
41 |
|
42 |
offset += len(tok) + 1
|
43 |
|
44 |
-
return res
|
|
|
1 |
from typing import Dict, List, Any
|
2 |
from nemo.collections.nlp.models import PunctuationCapitalizationModel
|
3 |
|
4 |
+
|
5 |
+
class PreTrainedPipeline:
|
6 |
def __init__(self, path=""):
|
7 |
# IMPLEMENT_THIS
|
8 |
# Preload all the elements you are going to need at inference.
|
9 |
# For instance your model, processors, tokenizer that might be needed.
|
10 |
# This function is only called once, so do all the heavy processing I/O here"""
|
11 |
+
self.model = PunctuationCapitalizationModel.restore_from("punctuation_uk_bert.nemo")
|
12 |
|
13 |
def __call__(self, inputs: str) -> List[Dict[str, Any]]:
|
14 |
"""
|
|
|
32 |
offset = 0
|
33 |
for tok, lab in zip(tokens, labels):
|
34 |
if lab != "OO":
|
35 |
+
res.append({"entity_group": lab, "word": tok, "start": offset, "end": offset + len(tok), "score": 0.99})
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
|
37 |
offset += len(tok) + 1
|
38 |
|
39 |
+
return res
|