Comments (2)
Hi @davidberenstein1957, saw the capabailities of sense2vec library and how it will work much better than some of the pretrained glove word2vec models.
My question was is there are a way we can add support for more state of the art word vector embeddings like sentence transformers,BERT etc.?
from concise-concepts.
@prakhar251998 thanks for the suggestion, but sadly this wouldn´t be possible. The concise-concepts
library works based on a find_most_similar
search within pre-defined embeddings based on tokens
present in the embedding model. For word2vec-like models, these tokens
are pre-defined/indexed and have a stand-alone semantical meaning like apple
being used in a similar context as pear
. For transformer-based models, the index
is mostly limited to a sub-word/character level and therefore doesn´t allow for a find_most_similar
operation.
I you would like to use these kinds of embeddings, you could potentially create a semantic-search knowledge base with KNN/ANN and embeddings based on the descriptions of the potential entities, but maybe this costs too much effort.
from concise-concepts.
Related Issues (20)
- error: missing ), unterminated subpattern at position x HOT 2
- Example fail while using GPUs HOT 2
- Python latest package 0.6.2 failing. Error in Conceptualizer.py.Results Deterioration HOT 8
- Unable to load local custom gensim model HOT 2
- duplicate logging regarding missing entires in embedding model HOT 1
- matching_patterns.json HOT 2
- multi token patterns HOT 12
- OSError on while adding concise_concepts to spacy nlp pipeline HOT 1
- add spaczz fuzzymatcher option to concise-concepts
- Custom models showing different confidences even 0 in case of mixed casing text HOT 2
- Loading transformer based models and handling phrases HOT 2
- consider generative LLM prompt based word expansion
- Question: How to use (external) transformer-based embeddings? HOT 3
- Model Sensitivity HOT 1
- Including Entities in concise concepts HOT 1
- determine fuzzyness with character distance `fuzzy=0` -> `fuzzy=n`
- Handling of Multiple Words HOT 5
- Lemmatization need for LEMMA patterns HOT 2
- json array too large HOT 2
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 concise-concepts.