Comments (4)
The LSTM-CRF model predicts a sequence of tags for the entire sequence. As a result, there is not a real notion of entity score, but only sequence score (and the model returns the sequence with the best score). However, you can do something like taking the average of the LSTM probability scores in your entity, and this should give you a good proxy for a confidence score. For instance, if you have
"Barack Obama" in your sentence, and that the model tags these two words as "B_PER" and "E_PER" then you can report the average (or the product) of P(B_PER|Barack) and P(E_PER|Obama) given by the model.
from tagger.
I understand what you are saying above,
I am quite new to Theano. I am sorry if I am wrong
While calling the forward function in the f_eval alpha variable would return the probability, but when I make return_best_sequence as False, the code breaks and gives the following error-
File "tagger.py", line 49, in classify_ner
y_preds = np.array(f_eval(*input))[1:-1]
IndexError: too many indices for array
from tagger.
What you want to look at is probably the tag probability scores:
Line 278 in c735605
from tagger.
Thanks a lot , it worked ..
from tagger.
Related Issues (20)
- utils.py issue line 303 list index out of range HOT 8
- Why my program goes into infinite loop? HOT 2
- Cuda and Theano Version HOT 4
- Can you
- Can my Chinese data be used in this program?(character-level) HOT 3
- Which tokenizer did you use? HOT 11
- Script for training embeddings HOT 9
- IOError "No such file or directory: './evaluation/temp/eval.1181043.scores " HOT 3
- SGD x Adam HOT 2
- Are you planning to release models in German, Spanish and Dutch as well? HOT 1
- Data size and decoding time
- Pretrained word embedding HOT 5
- Confusion about lable conversion. HOT 2
- Inconsistent conversion for IOBES to IOB HOT 1
- Token level or Entity level? HOT 2
- Equations of LSTM
- transition scores HOT 1
- How to set the parameters of a small dataset? HOT 1
- . HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from tagger.