Spaces:
Build error
Build error
Commit
·
d0e4d54
1
Parent(s):
16a0bfc
Delete app.py
Browse files
app.py
DELETED
@@ -1,154 +0,0 @@
|
|
1 |
-
# -*- coding: utf-8 -*-
|
2 |
-
"""
|
3 |
-
Created on Tue Nov 22 09:54:41 2022
|
4 |
-
|
5 |
-
@author: luol2
|
6 |
-
"""
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
import streamlit as st
|
11 |
-
import argparse
|
12 |
-
from src.nn_model import bioTag_CNN,bioTag_BERT,bioTag_Bioformer
|
13 |
-
from src.dic_ner import dic_ont
|
14 |
-
from src.tagging_text import bioTag
|
15 |
-
import os
|
16 |
-
import time
|
17 |
-
import json
|
18 |
-
import sys
|
19 |
-
|
20 |
-
|
21 |
-
st.set_page_config(
|
22 |
-
page_title="PhenoTagger",
|
23 |
-
page_icon=":shark:",
|
24 |
-
# layout="wide",
|
25 |
-
initial_sidebar_state="expanded",
|
26 |
-
menu_items={
|
27 |
-
'Get Help': 'https://www.extremelycoolapp.com/help',
|
28 |
-
'Report a bug': "https://www.extremelycoolapp.com/bug",
|
29 |
-
'About': "# This is a header. This is an *extremely* cool app!"
|
30 |
-
}
|
31 |
-
)
|
32 |
-
st.title('PhenoTagger Demo')
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
# with st.spinner('Model is being loaded..'):
|
41 |
-
|
42 |
-
# print('load model done!')
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
with st.form(key="my_form"):
|
48 |
-
|
49 |
-
@st.cache(allow_output_mutation=True)
|
50 |
-
def load_model():
|
51 |
-
ontfiles={'dic_file':'./dict_new/noabb_lemma.dic',
|
52 |
-
'word_hpo_file':'./dict_new/word_id_map.json',
|
53 |
-
'hpo_word_file':'./dict_new/id_word_map.json'}
|
54 |
-
|
55 |
-
# if para_set['model_type']=='cnn':
|
56 |
-
# vocabfiles={'w2vfile':'../vocab/bio_embedding_intrinsic.d200',
|
57 |
-
# 'charfile':'../vocab/char.vocab',
|
58 |
-
# 'labelfile':'../dict_new/lable.vocab',
|
59 |
-
# 'posfile':'../vocab/pos.vocab'}
|
60 |
-
# modelfile='../models/cnn_p5n5_b128_95_hponew1.h5'
|
61 |
-
|
62 |
-
# elif para_set['model_type']=='bioformer':
|
63 |
-
vocabfiles={'labelfile':'./dict_new/lable.vocab',
|
64 |
-
'config_path':'./vocab/bioformer-cased-v1.0/bert_config.json',
|
65 |
-
'checkpoint_path':'./vocab/bioformer-cased-v1.0/bioformer-cased-v1.0-model.ckpt-2000000',
|
66 |
-
'vocab_path':'./vocab/bioformer-cased-v1.0/vocab.txt'}
|
67 |
-
modelfile='./vocab/bioformer_p5n5_b64_1e-5_95_hponew3.h5'
|
68 |
-
# else:
|
69 |
-
# print('Model type is wrong, please select cnn or bioformer.')
|
70 |
-
# sys.exit()
|
71 |
-
|
72 |
-
|
73 |
-
biotag_dic=dic_ont(ontfiles)
|
74 |
-
|
75 |
-
# if para_set['model_type']=='cnn':
|
76 |
-
# nn_model=bioTag_CNN(vocabfiles)
|
77 |
-
# nn_model.load_model(modelfile)
|
78 |
-
# elif para_set['model_type']=='bioformer':
|
79 |
-
nn_model=bioTag_Bioformer(vocabfiles)
|
80 |
-
session=nn_model.load_model(modelfile)
|
81 |
-
test_tag='1232'
|
82 |
-
return nn_model,biotag_dic,test_tag,session
|
83 |
-
|
84 |
-
|
85 |
-
#hyper-parameter
|
86 |
-
st.sidebar.header("Hyperparameter Settings")
|
87 |
-
sbform = st.sidebar.form("Hyper-paramiters")
|
88 |
-
# para_model=sbform.selectbox('Model', ['cnn', 'bioformer'])
|
89 |
-
para_overlap=sbform.selectbox('Return overlapping concepts', ['True', 'False'])
|
90 |
-
para_abbr=sbform.selectbox('Identify abbreviations', ['True', 'False'])
|
91 |
-
para_threshold = sbform.slider('Threshold:', min_value=0.5, max_value=0.95, value=0.95, step=0.05)
|
92 |
-
sbform.form_submit_button("Setting")
|
93 |
-
|
94 |
-
st.write('parameters:', para_overlap,para_abbr,para_threshold)
|
95 |
-
nn_model,biotag_dic,test_tag,session=load_model()
|
96 |
-
|
97 |
-
|
98 |
-
input_text = st.text_area(
|
99 |
-
"Paste your text below (max 500 words)",
|
100 |
-
height=510,
|
101 |
-
)
|
102 |
-
|
103 |
-
MAX_WORDS = 500
|
104 |
-
import re
|
105 |
-
res = len(re.findall(r"\w+", input_text))
|
106 |
-
if res > MAX_WORDS:
|
107 |
-
st.warning(
|
108 |
-
"⚠️ Your text contains "
|
109 |
-
+ str(res)
|
110 |
-
+ " words."
|
111 |
-
+ " Only the first 500 words will be reviewed. Stay tuned as increased allowance is coming! 😊"
|
112 |
-
)
|
113 |
-
|
114 |
-
input_text = input_text[:MAX_WORDS]
|
115 |
-
|
116 |
-
submit_button = st.form_submit_button(label="✨ Get me the data!")
|
117 |
-
|
118 |
-
if para_overlap=='True':
|
119 |
-
para_overlap=True
|
120 |
-
else:
|
121 |
-
para_overlap=False
|
122 |
-
if para_abbr=='True':
|
123 |
-
para_abbr=True
|
124 |
-
else:
|
125 |
-
para_abbr=False
|
126 |
-
para_set={
|
127 |
-
#model_type':para_model, # cnn or bioformer
|
128 |
-
'onlyLongest':para_overlap, # False: return overlap concepts, True only longgest
|
129 |
-
'abbrRecog':para_abbr,# False: don't identify abbr, True: identify abbr
|
130 |
-
'ML_Threshold':para_threshold,# the Threshold of deep learning model
|
131 |
-
}
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
if not submit_button:
|
136 |
-
st.stop()
|
137 |
-
|
138 |
-
|
139 |
-
st.markdown(f"""**Results:**\n""")
|
140 |
-
# print('dic...........:',biotag_dic.keys())
|
141 |
-
print('........:',test_tag)
|
142 |
-
print('........!!!!!!:',input_text)
|
143 |
-
print('...input:',input_text)
|
144 |
-
tag_result=bioTag(session,input_text,biotag_dic,nn_model,onlyLongest=para_set['onlyLongest'], abbrRecog=para_set['abbrRecog'],Threshold=para_set['ML_Threshold'])
|
145 |
-
for ele in tag_result:
|
146 |
-
start = ele[0]
|
147 |
-
last = ele[1]
|
148 |
-
mention = input_text[int(ele[0]):int(ele[1])]
|
149 |
-
type='Phenotype'
|
150 |
-
id=ele[2]
|
151 |
-
score=ele[3]
|
152 |
-
output=start+"\t"+last+"\t"+mention+"\t"+id+"\n"
|
153 |
-
st.info(output)
|
154 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|