|
a |
|
b/QueryExtraction/output.txt |
|
|
1 |
Extraction: |
|
|
2 |
SpaCy TextRank: |
|
|
3 |
Captured 4 |
|
|
4 |
longer sequences |
|
|
5 |
long input sentences |
|
|
6 |
high bleu score |
|
|
7 |
the complete input sequence |
|
|
8 |
vector |
|
|
9 |
the input sequence |
|
|
10 |
information |
|
|
11 |
long dependencies |
|
|
12 |
a candidate translation |
|
|
13 |
simpler words |
|
|
14 |
the entire sequence |
|
|
15 |
fixed length |
|
|
16 |
valuable parts |
|
|
17 |
parallel processing |
|
|
18 |
one or more |
|
|
19 |
reference translations |
|
|
20 |
Seq2Seq Models |
|
|
21 |
some input words |
|
|
22 |
encodes |
|
|
23 |
the input |
|
|
24 |
sequence |
|
|
25 |
hidden state |
|
|
26 |
Vaswani et al |
|
|
27 |
a context vector |
|
|
28 |
this context vector |
|
|
29 |
this fixed-length context vector design |
|
|
30 |
the |
|
|
31 |
complete sequence |
|
|
32 |
an output word |
|
|
33 |
image captioning |
|
|
34 |
various problems |
|
|
35 |
Attention |
|
|
36 |
attention |
|
|
37 |
------------------------------ |
|
|
38 |
Gensim TextRank: |
|
|
39 |
Captured 4 |
|
|
40 |
learning |
|
|
41 |
learn |
|
|
42 |
attention |
|
|
43 |
words |
|
|
44 |
word |
|
|
45 |
encoder |
|
|
46 |
encodes |
|
|
47 |
translation |
|
|
48 |
translated |
|
|
49 |
translations |
|
|
50 |
translate |
|
|
51 |
decoder |
|
|
52 |
paper |
|
|
53 |
input sequence |
|
|
54 |
sequences |
|
|
55 |
vector |
|
|
56 |
bleu |
|
|
57 |
sentence |
|
|
58 |
sentences |
|
|
59 |
unit |
|
|
60 |
ideas |
|
|
61 |
idea |
|
|
62 |
parts |
|
|
63 |
models |
|
|
64 |
model |
|
|
65 |
like image |
|
|
66 |
length |
|
|
67 |
designed |
|
|
68 |
design |
|
|
69 |
long |
|
|
70 |
context |
|
|
71 |
works |
|
|
72 |
work |
|
|
73 |
processes |
|
|
74 |
processed |
|
|
75 |
processing |
|
|
76 |
hidden |
|
|
77 |
evaluation |
|
|
78 |
information |
|
|
79 |
------------------------------ |
|
|
80 |
Rake: |
|
|
81 |
Captured 1 |
|
|
82 |
2014 paper “ neural machine translation |
|
|
83 |
neural machine translation using seq2seq models |
|
|
84 |
various problems like image captioning |
|
|
85 |
“ thought vector ”) |
|
|
86 |
every encoder hidden state |
|
|
87 |
vaswani et al ., |
|
|
88 |
decoder unit works well |
|
|
89 |
high bleu score ). |
|
|
90 |
length context vector design |
|
|
91 |
retain longer sequences |
|
|
92 |
bilingual evaluation understudy |
|
|
93 |
decoder unit fails |
|
|
94 |
whole long sentence |
|
|
95 |
famous paper attention |
|
|
96 |
deep learning community |
|
|
97 |
deep learning arena |
|
|
98 |
long input sentences |
|
|
99 |
complete input sequence |
|
|
100 |
candidate translation |
|
|
101 |
et al |
|
|
102 |
seq2seq model |
|
|
103 |
context vector |
|
|
104 |
paper laid |
|
|
105 |
complete sequence |
|
|
106 |
long dependencies |
|
|
107 |
jointly learning |
|
|
108 |
shorter sentences |
|
|
109 |
fixed length |
|
|
110 |
input sequence |
|
|
111 |
decoder architecture |
|
|
112 |
------------------------------ |
|
|
113 |
Rakun: |
|
|
114 |
01-Mar-21 20:38:03 - Initiated a keyword detector instance. |
|
|
115 |
01-Mar-21 20:38:03 - Number of nodes reduced from 128 to 119 |
|
|
116 |
Captured 5 |
|
|
117 |
attention |
|
|
118 |
sequence |
|
|
119 |
input sequence attention |
|
|
120 |
model |
|
|
121 |
paper |
|
|
122 |
using |
|
|
123 |
encoder |
|
|
124 |
seq2seq |
|
|
125 |
encoder-decoder |
|
|
126 |
vector |
|
|
127 |
information |
|
|
128 |
architecture |
|
|
129 |
learning |
|
|
130 |
decoder |
|
|
131 |
ideas |
|
|
132 |
prominent |
|
|
133 |
translation |
|
|
134 |
hence |
|
|
135 |
processes |
|
|
136 |
previous |
|
|
137 |
extension |
|
|
138 |
natural |
|
|
139 |
reads |
|
|
140 |
translate |
|
|
141 |
align |
|
|
142 |
candidate |
|
|
143 |
comparing |
|
|
144 |
score |
|
|
145 |
instead |
|
|
146 |
concentrated |
|
|
147 |
------------------------------ |
|
|
148 |
Yake: |
|
|
149 |
Captured 5 |
|
|
150 |
neural machine translation |
|
|
151 |
input sequence |
|
|
152 |
deep learning community |
|
|
153 |
context vector |
|
|
154 |
sequence |
|
|
155 |
input |
|
|
156 |
neural machine |
|
|
157 |
machine translation |
|
|
158 |
attention |
|
|
159 |
model |
|
|
160 |
complete input sequence |
|
|
161 |
learning community |
|
|
162 |
deep learning |
|
|
163 |
context |
|
|
164 |
vector |
|
|
165 |
bilingual evaluation understudy |
|
|
166 |
encoder-decoder |
|
|
167 |
translation |
|
|
168 |
learning |
|
|
169 |
prominent ideas |
|
|
170 |
fixed-length context vector |
|
|
171 |
context vector design |
|
|
172 |
long input sentences |
|
|
173 |
machine |
|
|
174 |
encoder-decoder unit |
|
|
175 |
neural |
|
|
176 |
encoder-decoder model |
|
|
177 |
words |
|
|
178 |
encoder |
|
|
179 |
complete sequence |
|
|
180 |
------------------------------ |
|
|
181 |
KeyBERT: |
|
|
182 |
01-Mar-21 20:38:03 - Load pretrained SentenceTransformer: distilbert-base-nli-mean-tokens |
|
|
183 |
01-Mar-21 20:38:03 - Did not find folder distilbert-base-nli-mean-tokens |
|
|
184 |
01-Mar-21 20:38:03 - Try to download model from server: https://sbert.net/models/distilbert-base-nli-mean-tokens.zip |
|
|
185 |
01-Mar-21 20:38:03 - Load SentenceTransformer from folder: /Users/irene/.cache/torch/sentence_transformers/sbert.net_models_distilbert-base-nli-mean-tokens |
|
|
186 |
01-Mar-21 20:38:04 - Use pytorch device: cpu |
|
|
187 |
Batches: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 9.71it/s] |
|
|
188 |
Batches: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 11/11 [00:01<00:00, 6.87it/s] |
|
|
189 |
Captured 1 |
|
|
190 |
deep learning |
|
|
191 |
neural machine |
|
|
192 |
encoder decoder |
|
|
193 |
learning arena |
|
|
194 |
ideas deep |
|
|
195 |
context neural |
|
|
196 |
decoder architecture |
|
|
197 |
machine translation |
|
|
198 |
transformers revolutionized |
|
|
199 |
graph encoder |
|
|
200 |
memorize long |
|
|
201 |
architecture encoder |
|
|
202 |
learning community |
|
|
203 |
decoder initialized |
|
|
204 |
prominent ideas |
|
|
205 |
reflected graph |
|
|
206 |
revolutionized deep |
|
|
207 |
translations graph |
|
|
208 |
valuable parts |
|
|
209 |
connecting encoder |
|
|
210 |
famous paper |
|
|
211 |
focus valuable |
|
|
212 |
candidate translation |
|
|
213 |
paper neural |
|
|
214 |
composed encoder |
|
|
215 |
arena concept |
|
|
216 |
helps model |
|
|
217 |
output critical |
|
|
218 |
memorize |
|
|
219 |
foundation famous |