From: Evaluating the construct validity of text embeddings with application to survey questions
Model | Name | Dimension | File size |
---|---|---|---|
fastText | cc.en.300.bin | 300 | 2.44 GB |
GloVe | glove.840B.300d | 300 | 2.03 GB |
BERT | BERT-base-uncased | 768 | 420 MB |
BERT | BERT-large-uncased | 1024 | 1.25 GB |
Sentence-BERT | All-DistilRoBERTa-V1 | 768 | 292 MB |
Sentence-BERT | All-MPNet-base-V2 | 768 | 418 MB |
USE | USE-V4 | 512 | 916 MB |