You might like
Presented a session with Nebojsa Jojic on "Natural Language Processing: Azure + Research" at Harvard at the deep learning workshop computefest.seas.harvard.edu/workshops . Code for the tutorial is available here: github.com/erikamenezes/a… #nlp #research #deeplearning #azuremachinelearning
Finally got some of my CVPR 2019 notes together! Read more on the CVPR Meta-learning Workshop: link.medium.com/aQuB7DVU4X Happy Friday! #CVPR2019 #Metalearning
link.medium.com
CVPR 2019 — Meta-learning Workshop
This was my second time at CVPR, which was at Long Beach this year. With 9k attendees and sessions spread over 3 buildings, I got a chance…
CVPR 2019 Implementations: get all the papers with code in one place here: paperswithcode.com/conference/cvp…
XLNet, a new acronym to remember in #NLProc👇 Two key differences to BERT: - Learns with an objective which maximizes likelihood over all permutations of the factorization order. - Encodes with Transformer-XL which captures longer-term dependencies than vanilla Transformer.
XLNet: Generalized Autoregressive Pretraining for Language Understanding: outperforming BERT on 20 tasks (SQuAD, GLUE, sentiment analysis), while integrating ideas from Transformer-XL: arxiv: arxiv.org/abs/1906.08237 code + pretrained models: github.com/zihangdai/xlnet
Amazing growth @cvpr2019 this year — 5160 submitted papers, 1294 accepted papers, 9227 attendees. 56% growth and 26% growth acceleration over 2018. At this rate we’ll have 10.8 billion submitted papers by 2028
Blackhole enthusiasts at #CVPR2019, don't miss Katie Bauman at @WiCVworkshop @cvpr2019 Hyatt Regency A now!
This is a great post that highlights the connection between the building blocks used in Transformers and Capsule Networks. Definitely worth reading!
New project is live: gen.studio a collaboration with @metmuseum & @microsoft For each object at the Met, we can > find a corresponding object in GAN latent space > traverse latent space between objects to discover artworks that could have been made, but never were
This is a super cool resource: Papers With Code now includes 950+ ML tasks, 500+ evaluation tables (including SOTA results) and 8500+ papers with code. Probably the largest collection of NLP tasks I've seen including 140+ tasks and 100 datasets. paperswithcode.com/sota
Fantastic job by the @Azure DSVM team getting fastai and all the course notebooks pre-installed so that students can be up and running quickly and easily! I love how much easier it's getting for folks to get started with practical deep learning :) course-v3.fast.ai/start_azure.ht…
So far (2 days gone), this is my favorite #NeurIPS18 paper. I just love the paper that tells me why it works. It also showed that the popular previous belief "internal covariate shift" has little to do with batch norm's success. Ah! 😍 3-min video: youtu.be/ZOabsYbmBRM
Test of time award talk at #NeurIPS2018 concludes with saying the role of causality in AI needs to be investigated further. A part of what Doina Precup and I talked about at our mentoring table this morning.
Our PyTorch BERT is on pip! I took extra care to make it both easy to use and modular. Uses @ai2_allennlp file caching technique to download/cache/load Google's pretrained models Includes 6 PyTorch models with various architectures, tokenizer & optimizer 👉github.com/huggingface/py…
Chartify from Spotify: “Easier chart creation in Python for data scientists” labs.spotify.com/2018/11/15/int…
United States Trends
- 1. Sonny Gray 7,620 posts
- 2. Rush Hour 4 10.6K posts
- 3. Godzilla 21.2K posts
- 4. Red Sox 7,430 posts
- 5. Dick Fitts N/A
- 6. Chelsea 332K posts
- 7. Raising Arizona 1,011 posts
- 8. National Treasure 5,707 posts
- 9. 50 Cent 4,881 posts
- 10. Happy Thanksgiving 22K posts
- 11. Gone in 60 2,089 posts
- 12. Giolito N/A
- 13. Brett Ratner 3,286 posts
- 14. Muriel Bowser N/A
- 15. #yummymeets N/A
- 16. Clarke 6,607 posts
- 17. Valley Girl N/A
- 18. Ghost Rider 2,496 posts
- 19. Barca 120K posts
- 20. Lord of War 1,507 posts
Something went wrong.
Something went wrong.