Issues

• Bishnupriya Manipuri language in Open-NLLB

  gordicaleksa/Open-NLLB#33 · Filter by author usangraha-source opened on Mar 29, 2026

• Is this project still actively maintained ?

  question
  gordicaleksa/Open-NLLB#32 · Filter by author thangld201 opened on May 20, 2025

• Preventing some words to be translated

  question
  gordicaleksa/Open-NLLB#31 · Filter by author ekmekovski opened on Oct 15, 2024

• Provide context for input and output

  enhancementhelp wanted
  gordicaleksa/Open-NLLB#29 · Filter by author joiemoie opened on Jan 1, 2024

• Standard Moroccan Tamazight is mislabeled

  documentation
  gordicaleksa/Open-NLLB#28 · Filter by author MedAymenF opened on Dec 31, 2023

• Creation of a small model file for a few languages

  question
  gordicaleksa/Open-NLLB#27 · Filter by author dlippold opened on Nov 4, 2023

• MinHash: benchmark memory, speed and accuracy with varying r and b

  enhancementgood first issue
  gordicaleksa/Open-NLLB#23 · Filter by author vienneraphael opened on Sep 29, 2023
  1

• [Future - outside current project scope] non-English LLMs (Serbian LLM, etc.)

  enhancement
  gordicaleksa/Open-NLLB#21 · Filter by author gordicaleksa opened on Sep 12, 2023

• [Future - outside current project scope] 7B lang-family-specific Open-NLLB checkpoint

  enhancement
  gordicaleksa/Open-NLLB#20 · Filter by author gordicaleksa opened on Sep 12, 2023

• [Data] Acquire additional high-quality (non-public) parallel corpora for HBS

  enhancement
  gordicaleksa/Open-NLLB#19 · Filter by author gordicaleksa opened on Sep 12, 2023
  

• [Modeling] Release a 615M English -> HBS Open-NLLB checkpoint

  enhancement
  gordicaleksa/Open-NLLB#18 · Filter by author gordicaleksa opened on Sep 12, 2023

• [Modeling] Release a 3.3B Open-NLLB checkpoint (~202 languages)

  enhancement
  gordicaleksa/Open-NLLB#17 · Filter by author gordicaleksa opened on Sep 12, 2023

• [Modeling] Release a 1.3B Slavic languages Open-NLLB checkpoint

  enhancement
  gordicaleksa/Open-NLLB#16 · Filter by author gordicaleksa opened on Sep 12, 2023

• [Modeling] Release a 615M HBS (Croatian, Bosnian, Serbian) Open-NLLB checkpoint

  enhancement
  gordicaleksa/Open-NLLB#15 · Filter by author gordicaleksa opened on Sep 12, 2023

• Get a compute grant

  question
  gordicaleksa/Open-NLLB#14 · Filter by author gordicaleksa opened on Sep 11, 2023
  

• Estimate the necessary compute and number of GPUs for Open-NLLB effort

  documentationhelp wanted
  gordicaleksa/Open-NLLB#13 · Filter by author gordicaleksa opened on Sep 11, 2023
  

• Understand how to do 4-stage curriculum learning from the paper

  documentationhelp wanted
  gordicaleksa/Open-NLLB#12 · Filter by author gordicaleksa opened on Sep 11, 2023

• Setup a pipeline for mined data (use Allen AI's OSS dataset replication)

  enhancementhelp wanted
  gordicaleksa/Open-NLLB#11 · Filter by author gordicaleksa opened on Sep 11, 2023

• Obtain high quality Serbian parallel corpus (currently 0 support in our public bi-text)

  enhancement
  gordicaleksa/Open-NLLB#10 · Filter by author gordicaleksa opened on Sep 11, 2023
  

• Choosing the LID model

  question
  gordicaleksa/Open-NLLB#8 · Filter by author vienneraphael opened on Sep 11, 2023

• LID model peak probabilities

  question
  gordicaleksa/Open-NLLB#7 · Filter by author vienneraphael opened on Sep 11, 2023

• Native language visualizations

  good first issuequestion
  gordicaleksa/Open-NLLB#6 · Filter by author vienneraphael opened on Sep 11, 2023
  2

• Hydra pickle issue in generate_multi.py

  bug
  gordicaleksa/Open-NLLB#3 · Filter by author vienneraphael opened on Sep 11, 2023

• Reduce peak memory when using FSDP on 2+ GPUs

  bughelp wantedquestion
  gordicaleksa/Open-NLLB#2 · Filter by author gordicaleksa opened on Sep 9, 2023