Ochieng Millicent
@iam_OchiengMMachine Learning Research Scientist at Microsoft Research Africa
Similar User
@siminyu_kat
@alienelf
@Melissahei
@KreutzerJulia
@bonadossou
@davlanade
@ChrisEmezue
@m_0_a
@danqi_chen
@Kolokodess
@GhanaNLP
@UCRiverside
@alex_conneau
@vukosi
@KaliTessera
Led by: @davlanade, @JessicaOjo__ , @Israel_Abebe , Jian Yun Zhuang, @alabi_jesujoba, @zodiacJRH, @iam_OchiengM, @sarahookr, @AndieBukula, En-Shiun Annie Lee, Chiamaka Chukwuneke, @happybuzaaba1, @banQomania, @jonathanmukiibi, @SalomonKabongo, @yuehgoh, Lolwethu Ndolela...
Special thanks to @iam_OchiengM and the Microsoft AI for Good team for inviting me!
Benchmark is now on @huggingface collection huggingface.co/collections/ma…
A big thank you to @OpenAI @CohereForAI and @Oracle for the compute/API credits. Also, we thank @haileysch__ and @lintangsutawika for their assistance with the EleutherAI lm-eval tool.
This is a great collaboration with many authors @JessicaOjo__ @Israel_Abebe , @alabi_jesujoba @zodiacJRH @iam_OchiengM @sarahookr @AndieBukula @happybuzaaba1 @banQomania @jonathanmukiibi @SalomonKabongo @yuehgoh @RoowyM @Shmuhammadd @NanaYaaSally @SambSokhar
In few-shot evaluation, LLaMa 3 70B significantly benefits from few-shot examples for AfriMMLU and AfriXNLI but did not for AfriMGSM since it's only able to reason effectively on maths in English. GPT-4o consistently improves in performance with additional few-shot examples.
GPT-4o is the best across all tasks for native African languages, however, the performance is worse for eng & fra, where GPT-4-Turbo is more than +9.0 better. Aya-101 is the best open-model, but in the translate-test setting LLaMa 3 70B is better since it's more English-centric
We provide zero/few-shot evaluation of the performance of 14 LLMs (open weight models and closed models) in two settings leveraging lm-eval: 1) in-language and 2) translate-test (where test sets were automatically translated to English using NLLB-200-3B).
We cover 18 languages in our benchmark, 16 native African languages (amh, ewe, hau, ibo, kin, lin, lug, orm, sna, sot, swa, twi, wol, xho, yor, & zul) , and two European languages (eng and fra) translated from MGSM, MMLU and XNLI datasets.
I'm very happy to share our new benchmark, IrokoBench--a human translated benchmark dataset for 16 African languages covering: - NLI (AfriXNLI) - Maths reasoning (AfriMGSM) - Multi-choice QA (AfriMMLU) arxiv.org/abs/2406.03368 Project funded by @LacunaFund with @MasakhaneNLP
For everyone wondering about the #emnlp2024 cfp - it is out now with an ARR June deadline (15th June) #nlproc
Thrilled to have presented on "Multilingual Evaluation of Generative AI Models" at #EMNLP2023 in Singapore. 🌏 Meeting @kalikabali, Sunayana Sitaram, and @monojitchou in person was a highlight! Thanks, Jacki O'Neil for the support! Inspired by the journey. 🚀 #NLP
This work would not have been possible without an amazing set of collaborators: @rishanky @VarunGumma23 @divyanshuaggarw @ashutoshbsathe @iam_OchiengM, Ishaan Watts, maxamed Axmed, Prachi Jain, @kalikabali and Sunayana Sitaram. Read the paper here: arxiv.org/abs/2311.07463 (6/n)
Microsoft Africa Research Institute is on a mission to explore how small businesses in Nigeria 🇳🇬 are harnessing generative AI. The team will be engaging with businesses, to dive into their daily operations. Would you like to participate? Here's more: aka.ms/AiResearchNige…
Excited to be part of the Deep Learning Indaba @DeepIndaba at Accra, Ghana. I will be talking at a panel on Efficient multilingual NLP for African languages in the age of LLMs. Thanks to the organizingers @MasakhaneNLP and @iam_OchiengM for inviting me.
⭐️Exciting news! We're gearing up for Deep Learning Indaba 2023! @DeepIndaba 🚀 Mark your calendars for our "Efficiency in Africa NLP" Workshop, scheduled on Friday, September 8, 2023, in Accra, Ghana. This workshop is an ideal fit for learners at any level.
The next version of our benchmarking effort, Mega V2, across 16 datasets, 71 languages and a larger set of models, including GPT3.5Turbo and GPT4.We also discuss the directions for future research on multilingual NLP based on their performance. #NLProc lnkd.in/g9zFZA2i
🎉 Exciting news! We are thrilled to announce Melissa Densmore as our General Chair for #COMPASS2023 conference in Cape Town from 16-19 August 2023! Melissa is an Associate Professor in Computer Science at UCT, leading the HCI Lab and Hasso Plattner Institute Research School.
COMPASS 2023 Submission! Missed the deadline? Don't worry! You can still present at COMPASS2023 by submitting a short paper, poster, demo, workshop or panel directly to the conference. Check out the details at compass.acm.org/late-submissio…. #COMPASS23 #LateSubmissions #CallforPapers
United States Trends
- 1. Luka 43,3 B posts
- 2. Mavs 20,1 B posts
- 3. Shai 15,9 B posts
- 4. Jamie Foxx 22,2 B posts
- 5. Kyrie 12,8 B posts
- 6. Greece 45,5 B posts
- 7. Greece 45,5 B posts
- 8. #SnoopDoggXboxChainGiveaway N/A
- 9. Taina 2.376 posts
- 10. Randy Moss 3.899 posts
- 11. WNBA 33,2 B posts
- 12. #SkeletonCrew 7.060 posts
- 13. Thunder 91,6 B posts
- 14. Jake Burger 2.015 posts
- 15. Infowars 18,2 B posts
- 16. Caruso 4.075 posts
- 17. Mitch 63,1 B posts
- 18. SantaMoney 2.539 posts
- 19. Chet 5.231 posts
- 20. Ortiz 25,2 B posts
Who to follow
-
Kathleen S.
@siminyu_kat -
Jade Abbott
@alienelf -
Melissa Heikkilä
@Melissahei -
Julia Kreutzer
@KreutzerJulia -
Bonaventure F. P. Dossou
@bonadossou -
David Ifeoluwa Adelani 🇳🇬
@davlanade -
Chris Chinenye Emezue
@ChrisEmezue -
Mohamed Ahmed
@m_0_a -
Danqi Chen
@danqi_chen -
Ada Nduka Oyom
@Kolokodess -
Ghana Natural Language Processing (NLP)
@GhanaNLP -
UC Riverside
@UCRiverside -
Alexis Conneau @ NeurIPS
@alex_conneau -
Vukosi Marivate 🇿🇦🚀📊
@vukosi -
Kale-ab Tessera
@KaliTessera
Something went wrong.
Something went wrong.