Surprised to see our (@fadybaly) Arabic BERT model from 4 years ago as the TOP 10 most finetuned model on the @huggingface hub.
It now has ~9M total downloads, with ~600K monthly.
Thread/Paper: https://t.co/5CNnj2fUWd
Thanks to @huggingface Spaces and @streamlit, you can try our Arabic NLP models
The space includes:
- AraGPT2 Mega and co
- My rank 1 submission to the Arabic Sentiment Analysis competition
- QA with AraELECTRA
- Sarcasm Detection
- Our text preprocessor
https://t.co/odPfEvBvBz
The Researcher – Episode 5:
Using Arabic language in Artificial Intelligence.
How does a machine speak and how would you make it speak in Arabic? Do machines currently feel emotions?
Find out with #AUB Dr. Hazem Hajj, Associate Professor @MSFEA_AUB and Dr. @imadb15
We're proud to announce that we have 3 papers accepted to @WANLP_Ar, by our talented team members @tareknaous @reeeemahmoud @wissam_antoun@fadybaly @hh63_hopeful
The papers target Arabic conversational agents and language models (details below ↓)
#wanlp2021#ArabicNLP 1/6
Arabic poetry generation using (fine-tuning) the newly released GPT2 model by AUB Mindlab. The base model was used (memory restrictions). Underlined text was given. #nlp#deeplearning#arabic
Happy New Year to the Arabic NLP community. An awsome and powerful set of Arabic language models are just released by @wissam_antoun@fadybaly at @aubmindlab
Isn't 2021 great already? 👏👏👏
New Year, New Arabic Models 🥳
@aubmindlab,@fadybaly,and I are happy to release 10 new Arabic Transformer Models on @huggingface🤗
- AraGPT2: Mega (1.46B), Large(792M), Medium(370M), Base (135M)
- AraELECTRA: Base (135M)
- AraBERTv2,v0.2: Large (371M),Base (136M)
#ArabicNLP
1/5
It's pretty awesome how dancing makes robots less intimidating. Looking forward to seeing more nontrivial Machine Learning on these robots. Credit: Boston Dynamics.
Introducing TensorSensor: a lib that clarifies exceptions by augmenting messages + visualizing Python code to indicate shape of tensor variables; works with Tensorflow, PyTorch, Numpy, and higher-level libraries like Keras and fastai. Article: https://t.co/lqZgWOu4FC
New Adapter by @mapmeld:
Adapter for AraBERT (aubmindlab/bert-base-arabert) trained to classify Arabic by dialect, trained for 3 epochs on samples from University of British Columbia and John Hopkins University.
Check it out at https://t.co/TBMQ1ZZ5Iw
IT'S FINALLY OUT!
After months of work, I'm happy to announce that the first video in a series on visualizing deep learning is out. It's on visualizing the structure of a neural network.
RTs appreciated :)
https://t.co/oXuOPuN0AV
The material for my "Stance Detection" tutorial at @icwsm with @AldayelAbeer and @Walid_Magdy is available at: https://t.co/GqyacRlsEP
We plan to record the tutorial and will send out the link when it becomes available.
#icwsm2020#stanceDetection