Update utils/target_classifier.py
Browse files- utils/target_classifier.py +99 -99
utils/target_classifier.py
CHANGED
@@ -1,127 +1,127 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
|
17 |
-
|
18 |
|
19 |
-
|
20 |
-
|
21 |
|
22 |
-
|
23 |
|
24 |
-
#
|
25 |
-
|
26 |
|
27 |
-
|
28 |
|
29 |
-
#
|
30 |
-
|
31 |
|
32 |
-
#
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
|
38 |
-
#
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
|
44 |
-
#
|
45 |
-
|
46 |
|
47 |
-
|
48 |
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
|
72 |
-
|
73 |
|
74 |
-
|
75 |
-
|
76 |
-
|
77 |
|
78 |
-
|
79 |
|
80 |
|
81 |
-
|
82 |
-
|
83 |
-
|
84 |
-
|
85 |
-
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
|
105 |
-
|
106 |
|
107 |
-
|
108 |
|
109 |
-
|
110 |
|
111 |
-
|
112 |
|
113 |
-
#
|
114 |
-
|
115 |
|
116 |
-
#
|
117 |
-
|
118 |
|
119 |
-
#
|
120 |
-
|
121 |
|
122 |
|
123 |
-
#
|
124 |
-
#
|
125 |
# # classifier_model = st.session_state['target_classifier']
|
126 |
|
127 |
# # results = classifier_model(list(haystack_doc.text))
|
@@ -137,4 +137,4 @@
|
|
137 |
# # df.index += 1
|
138 |
# # # df['Label_def'] = df['Target Label'].apply(lambda i: _lab_dict[i])
|
139 |
|
140 |
-
|
|
|
1 |
+
from typing import List, Tuple
|
2 |
+
from typing_extensions import Literal
|
3 |
+
import logging
|
4 |
+
import pandas as pd
|
5 |
+
from pandas import DataFrame, Series
|
6 |
+
from utils.config import getconfig
|
7 |
+
from utils.preprocessing import processingpipeline
|
8 |
+
import streamlit as st
|
9 |
+
from transformers import pipeline
|
10 |
|
11 |
+
## Labels dictionary ###
|
12 |
+
_lab_dict = {
|
13 |
+
'0':'NO',
|
14 |
+
'1':'YES',
|
15 |
+
}
|
16 |
|
17 |
+
def get_target_labels(preds):
|
18 |
|
19 |
+
"""
|
20 |
+
Function that takes the numerical predictions as an input and returns a list of the labels.
|
21 |
|
22 |
+
"""
|
23 |
|
24 |
+
# Get label names
|
25 |
+
preds_list = preds.tolist()
|
26 |
|
27 |
+
predictions_names=[]
|
28 |
|
29 |
+
# loop through each prediction
|
30 |
+
for ele in preds_list:
|
31 |
|
32 |
+
# see if there is a value 1 and retrieve index
|
33 |
+
try:
|
34 |
+
index_of_one = ele.index(1)
|
35 |
+
except ValueError:
|
36 |
+
index_of_one = "NA"
|
37 |
|
38 |
+
# Retrieve the name of the label (if no prediction made = NA)
|
39 |
+
if index_of_one != "NA":
|
40 |
+
name = label_dict[index_of_one]
|
41 |
+
else:
|
42 |
+
name = "Other"
|
43 |
|
44 |
+
# Append name to list
|
45 |
+
predictions_names.append(name)
|
46 |
|
47 |
+
return predictions_names
|
48 |
|
49 |
+
@st.cache_resource
|
50 |
+
def load_targetClassifier(config_file:str = None, classifier_name:str = None):
|
51 |
+
"""
|
52 |
+
loads the document classifier using haystack, where the name/path of model
|
53 |
+
in HF-hub as string is used to fetch the model object.Either configfile or
|
54 |
+
model should be passed.
|
55 |
+
1. https://docs.haystack.deepset.ai/reference/document-classifier-api
|
56 |
+
2. https://docs.haystack.deepset.ai/docs/document_classifier
|
57 |
+
Params
|
58 |
+
--------
|
59 |
+
config_file: config file path from which to read the model name
|
60 |
+
classifier_name: if modelname is passed, it takes a priority if not \
|
61 |
+
found then will look for configfile, else raise error.
|
62 |
+
Return: document classifier model
|
63 |
+
"""
|
64 |
+
if not classifier_name:
|
65 |
+
if not config_file:
|
66 |
+
logging.warning("Pass either model name or config file")
|
67 |
+
return
|
68 |
+
else:
|
69 |
+
config = getconfig(config_file)
|
70 |
+
classifier_name = config.get('target','MODEL')
|
71 |
|
72 |
+
logging.info("Loading classifier")
|
73 |
|
74 |
+
doc_classifier = pipeline("text-classification",
|
75 |
+
model=classifier_name,
|
76 |
+
top_k =1)
|
77 |
|
78 |
+
return doc_classifier
|
79 |
|
80 |
|
81 |
+
@st.cache_data
|
82 |
+
def target_classification(haystack_doc:pd.DataFrame,
|
83 |
+
threshold:float = 0.5,
|
84 |
+
classifier_model:pipeline= None
|
85 |
+
)->Tuple[DataFrame,Series]:
|
86 |
+
"""
|
87 |
+
Text-Classification on the list of texts provided. Classifier provides the
|
88 |
+
most appropriate label for each text. There labels indicate whether the paragraph
|
89 |
+
references a specific action, target or measure in the paragraph.
|
90 |
+
---------
|
91 |
+
haystack_doc: List of haystack Documents. The output of Preprocessing Pipeline
|
92 |
+
contains the list of paragraphs in different format,here the list of
|
93 |
+
Haystack Documents is used.
|
94 |
+
threshold: threshold value for the model to keep the results from classifier
|
95 |
+
classifiermodel: you can pass the classifier model directly,which takes priority
|
96 |
+
however if not then looks for model in streamlit session.
|
97 |
+
In case of streamlit avoid passing the model directly.
|
98 |
+
Returns
|
99 |
+
----------
|
100 |
+
df: Dataframe with two columns['SDG:int', 'text']
|
101 |
+
x: Series object with the unique SDG covered in the document uploaded and
|
102 |
+
the number of times it is covered/discussed/count_of_paragraphs.
|
103 |
+
"""
|
104 |
|
105 |
+
logging.info("Working on target/action identification")
|
106 |
|
107 |
+
haystack_doc['Target Label'] = 'NA'
|
108 |
|
109 |
+
if not classifier_model:
|
110 |
|
111 |
+
classifier_model = st.session_state['target_classifier']
|
112 |
|
113 |
+
# Get predictions
|
114 |
+
predictions = classifier_model(list(haystack_doc.text))
|
115 |
|
116 |
+
# Get labels for predictions
|
117 |
+
pred_labels = getlabels(predictions)
|
118 |
|
119 |
+
# Save labels
|
120 |
+
haystack_doc['Target Label'] = pred_labels
|
121 |
|
122 |
|
123 |
+
# logging.info("Working on action/target extraction")
|
124 |
+
# if not classifier_model:
|
125 |
# # classifier_model = st.session_state['target_classifier']
|
126 |
|
127 |
# # results = classifier_model(list(haystack_doc.text))
|
|
|
137 |
# # df.index += 1
|
138 |
# # # df['Label_def'] = df['Target Label'].apply(lambda i: _lab_dict[i])
|
139 |
|
140 |
+
return haystack_doc
|