LASER

Making Sentence Embeddings Robust to User-Generated Content
[Microsoft Seminar] Extended presentation of the LREC-COLING 2024 paper of the same title.
May 29, 2024 2:30 PM — 3:30 PM Online
Lydia Nishimwe
Making Sentence Embeddings Robust to User-Generated Content
Making Sentence Embeddings Robust to User-Generated Content
Lydia Nishimwe, Benoît Sagot, Rachel Bawden
Making Sentence Embeddings Robust to User-Generated Content
Your Fairseq-trained model might have more embedding parameters than it should.
How a bug in reading SentencePiece vocabulary files causes some Fairseq-trained models to have up to 3k extra parameters in the embedding layer.
Lydia Nishimwe, posted on Mar 16, 2024
Last updated on Mar 22, 2025
Your Fairseq-trained model might have more embedding parameters than it should.
Fairseq Bug Fix
A bug in reading SentencePiece vocabulary files causes models to have 3k extra params in the embedding layer.
Fairseq Bug Fix
RoLASER
Making LASER sentence embeddings robust to user-generated content via Knowledge Distillation and Data Augmentation.
RoLASER