Fairseq Bug Fix Last updated on Jan 8, 2026 Go to Project Site GitHub Repo Open Graph (OG) Image Fairseq Bug Fix LASER Word Embeddings Deep Learning Lydia Nishimwe AI Research Scientist | PhD AI Research Scientist studying how large models behave, fail, and generalise in the real world. Related Your Fairseq-trained model might have more embedding parameters than it should. Robust Neural Machine Translation of User-Generated Content Making Sentence Embeddings Robust to User-Generated Content RoLASER UGC Translation Eval Challenges