Abderrahmane Issam (@abdrahman_issam) 's Twitter Profile
Abderrahmane Issam

@abdrahman_issam

PhD student working on Neural Machine Translation @UM_DACS

ID: 1261376832580050947

calendar_today15-05-2020 19:24:26

61 Tweet

101 Followers

980 Following

Stephen Mayhew (@mayhewsw) 's Twitter Profile Photo

I’m excited to announce an ambitious project: ✨Universal NER ✨. Multilingual NLP is missing a gold standard multilingual NER dataset, and this project aims to fill that gap. This is a community effort: I need your help! See: universalner.org Or read more in the 🧵

Maarten Grootendorst (@maartengr) 's Twitter Profile Photo

Preview #1: multi-topic assignment on a token level Had a lot of fun developing this upcoming feature in BERTopic. It can be used to approximate the topic distributions in a document after having generated the topics. #machinelearning #datascience #nlp #naturallanguageprocessing

Preview #1: multi-topic assignment on a token level
Had a lot of fun developing this upcoming feature in BERTopic. It can be used to approximate the topic distributions in a document after having generated the topics. 
#machinelearning #datascience #nlp #naturallanguageprocessing
Calum Bird (@calumbirdo) 's Twitter Profile Photo

An optimal language model with 30B parameters can achieve better results than Davinci 175B, while being ~5x cheaper to host and ~2x cheaper to train [1]. I built a calculator so that you can avoid the mistake that cost OpenAI millions👇

An optimal language model with 30B parameters can achieve better results than Davinci 175B, while being ~5x cheaper to host and ~2x cheaper to train [1]. I built a calculator so that you can avoid the mistake that cost OpenAI millions👇
Dr. Khalil خليل ⵅⴰⵍⵉⵍ (@khalilmrini) 's Twitter Profile Photo

Abderrahmane was the lead author on our work on the first summarization model for Moroccan Darija. Can't wait to see what you will accomplish during your PhD! 👏👏👏

Guillaume Lample @ NeurIPS 2024 (@guillaumelample) 's Twitter Profile Photo

Today we release LLaMA, 4 foundation models ranging from 7B to 65B parameters. LLaMA-13B outperforms OPT and GPT-3 175B on most benchmarks. LLaMA-65B is competitive with Chinchilla 70B and PaLM 540B. The weights for all models are open and available at research.facebook.com/publications/l… 1/n

Today we release LLaMA, 4 foundation models ranging from 7B to 65B parameters.
LLaMA-13B outperforms OPT and GPT-3 175B on most benchmarks. LLaMA-65B is competitive with Chinchilla 70B and PaLM 540B.
The weights for all models are open and available at research.facebook.com/publications/l…
1/n
Jerry Spanakis 🟥 🦋 gerasimoss.bsky.social (@gerasimoss) 's Twitter Profile Photo

3/ Abderrahmane Issam (who just started) presenting his ideas for improving machine translation for non-native speakers. How knowledge distillation 🚰can be used and what can we learn from linguistic differences/similarities among languages? Also w/ Yusuf Can Semerci within VOXReality

3/ <a href="/abdrahman_issam/">Abderrahmane Issam</a> (who just started) presenting his ideas for improving machine translation for non-native speakers. How knowledge distillation 🚰can be used and what can we learn from linguistic differences/similarities among languages? Also w/ <a href="/YCSemerci/">Yusuf Can Semerci</a> within <a href="/VOXReality_eu/">VOXReality</a>
Jerry Spanakis 🟥 🦋 gerasimoss.bsky.social (@gerasimoss) 's Twitter Profile Photo

🙂 ✨ Check the latest post from Abderrahmane Issam detailing the latest developments in speech recognition (#ASR) and #Machinetranslation (#MT). 🙌 UM Department of Advanced Computing Sciences work in VOXReality will focus on how 🖼️ & 📝context can help in real-world #ASR and #MT esp. in a low-resource setup

Math&Maroc (@mathsmaroc) 's Twitter Profile Photo

Last week's first Generative AI hackathon in Morocco was a success thanks to 1337AI and Math&Maroc. Despite limited compute power and scarce data, ~75 hackers showed ingenuity and adaptability. 🧵of cool things that happened thinkai.ma #GenerativeAI #Hackathon

Oreva Ahia (@orevaahia) 's Twitter Profile Photo

Do you use LM APIs like ChatGPT in non-English languages? You might be overpaying and your in-context learning performance may also be negatively impacted. tinyurl.com/y6hakyc8 🧵⬇️

Do you use LM APIs like ChatGPT in non-English languages? You might be overpaying and your in-context learning performance may also be negatively impacted.
tinyurl.com/y6hakyc8 🧵⬇️
Jerry Spanakis 🟥 🦋 gerasimoss.bsky.social (@gerasimoss) 's Twitter Profile Photo

#VOXReality VOXReality 3rd GA finished in Amsterdam. On behalf of UM Department of Advanced Computing Sciences Yusuf Can Semerci Abderrahmane Issam Paweł Mąka shared updates on contextual machine translation & non-native ASR and together preparing 3 use cases (theater 🎭, virtual conference 🧑‍🏫, training agent 🧑‍🚒).

Jerry Spanakis 🟥 🦋 gerasimoss.bsky.social (@gerasimoss) 's Twitter Profile Photo

#acl2024nlp #iwslt2024 Abderrahmane Issam UM Department of Advanced Computing Sciences uses adapters 🧩 to balance ⚖️ latency & translation quality. Our method supports multiple latency levels in one model. Work part of VOXReality project, 🤝 with Yusuf Can Semerci, Jan Scholtes 📄: arxiv.org/abs/2407.13469 >>

Jerry Spanakis 🟥 🦋 gerasimoss.bsky.social (@gerasimoss) 's Twitter Profile Photo

👋 folks attending #ACL2024NLP! Wanna chat about #nlproc in 🧑‍⚖️,📱,MT, or 🌡️in Bangkok? HMU! Also, don’t miss Justus-Jonas Erker 🇪🇺 @ ACL2024 🇹🇭 (PS4, Tue 10:30) & Abderrahmane Issam (IWSLT, Fri 09:00). More info ⬇️

Dennis Soemers (@dennissoemers) 's Twitter Profile Photo

We have a new open position for a PhD student to work on Reinforcement Learning for Cyber Security! Check out vacancies.maastrichtuniversity.nl/job/Maastricht… for details, and feel free to contact me for any info as well. Application deadline: May 11 RTs much appreciated.

We have a new open position for a PhD student to work on Reinforcement Learning for Cyber Security! 

Check out vacancies.maastrichtuniversity.nl/job/Maastricht… for details, and feel free to contact me for any info as well.

Application deadline: May 11

RTs much appreciated.