| Tweetovi |
|
|
nfiedel
@nfiedel
|
28. sij |
|
Excited to share some of our research into open-domain dialog and natural language response generation (NLG)!
Paper: arxiv.org/abs/2001.09977
Huge thanks to a fantastic group of collaborators! @xpearhead @lmthang @quocleix and more. twitter.com/GoogleAI/statu…
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Google AI
@GoogleAI
|
18. pro |
|
Check out a novel approach to out-of-distribution detection, applied to a new benchmark dataset of genomic sequences, that enables a #MachineLearning model to better discriminate between anomalous data and that used in training. Learn all about it below ↓ goo.gle/2sCZz6O
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Adam Roberts
@ada_rob
|
13. pro |
|
I'm so excited about the program we've put together for Saturday's #NeurIPS2019 ML for Creativity and Design Workshop 3.0.
Aside from the amazing accepted talks and posters, we have a diverse set of invited speakers I want to highlight in this thread.
neurips2019creativity.github.io
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Adam Roberts
@ada_rob
|
9. pro |
|
As promised, we have made the Text-To-Text Transfer Transformer (T5) models much easier to fine-tune for new tasks, and we just released a Colab notebook where you can try it yourself on a free TPU!
👇
tiny.cc/t5-colab
(1/3) pic.twitter.com/LU5scmhbIs
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Arvind Neelakantan
@arvind_io
|
1. stu |
|
We explore a simple approach to task-oriented dialog. A single neural network consumes conversation history and external knowledge as input and generates the next turn text response along with the action (when necessary) as output. Paper: arxiv.org/pdf/1910.14613… 1/4 pic.twitter.com/BjptMZBHvm
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Colin Raffel
@colinraffel
|
24. lis |
|
New paper! We perform a systematic study of transfer learning for NLP using a unified text-to-text model, then push the limits to achieve SoTA on GLUE, SuperGLUE, CNN/DM, and SQuAD.
Paper: arxiv.org/abs/1910.10683
Code/models/data/etc: git.io/Je0cZ
Summary ⬇️ (1/14) pic.twitter.com/VP1nkkHefB
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Ian Simon
@iansimon
|
16. ruj |
|
We (@iansimon + @huangcza + @jesseengel + @fjord41 + @notwaldorf) have now released a Colab notebook where you can play around with some Transformer models for generating piano performances: magenta.tensorflow.org/piano-transfor…
And the models are trained on transcribed YouTube recordings!
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Arvind Neelakantan
@arvind_io
|
13. ruj |
|
We are excited to release Taskmaster-1, a new task-oriented dialog dataset. We explore two methods for data collection, two-person and self-dialogs. Surprisingly self-dialogs are an effective way to collect dialog.
Paper accepted to @emnlp2019 : arxiv.org/pdf/1909.05358… pic.twitter.com/gTj5ocfdrS
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Monica Dinculescu
@notwaldorf
|
17. srp |
|
There were *21.6M* harmonizations submitted by people to the Bach Doodle!🙀
We are open sourcing this dataset along with a set of fun interactive visualizations to explore the data, discover some familiar top hits and even regional childhood songs! 🎵
📝 magenta.tensorflow.org/bach-doodle-viz pic.twitter.com/8FrKcbQsOQ
|
||
|
|
||
|
|
nfiedel
@nfiedel
|
3. lip |
|
Thanks @dmontauk for the intro. Adding @unclepeddy who kindly offered to help out.
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Douglas Eck
@douglas_eck
|
8. svi |
|
Wayne from @theflaminglips Claire from @YACHT with @adarob and @jesseengel from @GoogleMagenta! pic.twitter.com/fDyIt2q9Ma
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Rachael Tatman
@rctatman
|
15. sij 2019. |
|
I promised this a while ago and I finally delivered; I wrote a gosh-darn magnum opus on what I don't like about BLEU & some alternatives you can use instead. medium.com/@rtatman/evalu…
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Anna Huang
@huangcza
|
13. pro 2018. |
|
Check out our new blog post on music transformer that generates minute-long pieces that I find myself listening to again! It also generates compelling harmonizations and realizes chord progressions.
Which is your favorite? Ours is the Jazz 🎷
👂g.co/magenta/music-… pic.twitter.com/KCs9d0BJdf
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
David Smith
@davidingolcar
|
28. stu 2018. |
|
We are the first human beings to see a Mars sunset. Its quite a thought. pic.twitter.com/cM5UJojiPy
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Nikhil Thorat
@nsthorat
|
13. stu 2018. |
|
TensorFlow.js now in Ableton Live via Magenta.js and MaxMSP + Node.
Achievement unlocked, congrats to the folks who made this possible! twitter.com/ada_rob/status…
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Jeff Dean
@JeffDean
|
10. stu 2018. |
|
Hard to believe that 3 years ago, about 20 of us gathered early one morning in a conference room on the 4th floor of Bldg. 1900 for the open-source launch of TensorFlow, watching live Google Analytics stats of tensorflow.org & Hacker News comments. We've come a long way! twitter.com/TensorFlow/sta…
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Matt Gardner
@nlpmattg
|
5. stu 2018. |
|
BLEU is Not Suitable for the Evaluation of Text Simplification, by Elior Sulem, Omri Abend, and Ari Rappoport. EMNLP 2018 short paper. arxiv.org/abs/1810.05995
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Jeremiah Harmsen
@JeremiahHarmsen
|
4. stu 2018. |
|
TensorFlow Serving handles *millions* of QPS inside Google - and it just got easier for everyone to use!
You can start with pre-built Docker images for GPU/CPU or compile an image optimised for your specific architecture - all with a single command.
Ready for blastoff! 🛠️🚀 twitter.com/TensorFlow/sta…
|
||
|
|
||
| nfiedel proslijedio/la je tweet | ||
|
Jeff Dean
@JeffDean
|
29. lis 2018. |
|
AI is a powerful tool for improving society. I'm excited about our AI for Social Good program, focusing on both research & empowering the ecosystem.
If you have ideas, submit to the AI Impact Challenge! We're funding the best ideas w/$25M. Learn more at ai.google/social-good
|
||
|
|
||
|
|
nfiedel
@nfiedel
|
18. kol 2018. |
|
Great analysis and potential for CA to continue innovating and leading the way to large scale, sustainable energy production. If you ever have a chance, I recommend watching Daniel’s talks. twitter.com/dan_kammen/sta…
|
||
|
|
||