|
--- |
|
license: mit |
|
language: |
|
- en |
|
--- |
|
# Visually Grounded embeddings for Fast-text and GloVe |
|
|
|
This repository contains multiple visually grounded word embedding models. |
|
All of these embeddings have been effectively infused with visual information from images.<br> |
|
They have been proven to show stronger correlations (compared to textual embeddings) |
|
to human judgments on various word similarities and relatedness benchmarks. |
|
|
|
# Usage |
|
|
|
All of the models are encoded in [gensim](https://pypi.org/project/gensim/) format. |
|
Loading the model: |
|
```python |
|
import gensim |
|
|
|
model_g = gensim.models.KeyedVectors.load_word2vec_format('path_to_embeddings' , binary=True) |
|
|
|
#retrieve the most similar words |
|
print(model_g.most_similar('together',topn=10)) |
|
|
|
[('togther', 0.6425853967666626), ('togehter', 0.6374243497848511), ('togeather', 0.6196791529655457), |
|
('togather', 0.5998020172119141), ('togheter', 0.5819681882858276),('toghether', 0.5738174319267273), |
|
('2gether', 0.5187329053878784), ('togethor', 0.501663088798523), ('gether', 0.49128714203834534), |
|
('toegther', 0.48457157611846924)] |
|
|
|
print(model_g.most_similar('sad',topn=10)) |
|
|
|
[('saddening', 0.6763913631439209), ('depressing', 0.6676110029220581), ('saddened', 0.6352651715278625), |
|
('sorrowful', 0.6336953043937683), ('heartbreaking', 0.6180269122123718), ('heartbroken', 0.6099187135696411), |
|
('tragic', 0.6039361953735352), ('pathetic', 0.5848405361175537), ('Sad', 0.5826965570449829), |
|
('mournful', 0.5742306709289551)] |
|
|
|
#find the outlier word |
|
print(model_g.doesnt_match(['fire', 'water', 'land', 'sea', 'air', 'car'])) |
|
|
|
car |
|
``` |
|
|
|
where 'path_to_embeddings' is the path to the embeddings you intend to use. |
|
|
|
# Which embeddings to use |
|
Under the **Files and Versions** tab, you can see the list of 4 available embeddings. |
|
|
|
The following embedding files are from the paper [Learning Zero-Shot Multifaceted Visually Grounded Word Embeddings via Multi-Task Training](https://aclanthology.org/2021.conll-1.12/): |
|
- v_glove_1024d_1.0 |
|
- v_fasttext_1024d_1.0 |
|
|
|
The following embedding files are from the paper [Language with Vision: a Study on Grounded Word and Sentence Embeddings](https://arxiv.org/pdf/2206.08823.pdf): |
|
|
|
- v_glove_1024d_2.0 |
|
- v_glove_300_d_2.0 |
|
|
|
All of them come with 1024-dimensional word vectors except v_glove_300_d_2.0 which contains 300-dimensional word vectors. |
|
|
