[ML News] New ImageNet SOTA | Uber's H3 hexagonal coordinate system | New text-image-pair dataset

Subscribers:
285,000
Published on ● Video Link: https://www.youtube.com/watch?v=DkojaN7_f4E



Duration: 14:13
16,799 views
0


#truthfulqa #efficientnet #laion400M

Your regularly irregular updates on what's happening in the Machine Learning world.

OUTLINE:
0:00 - Intro
0:20 - TruthfulQA benchmark shines new light on GPT-3
2:00 - LAION-400M image-text-pair dataset
4:10 - GoogleAI's EfficientNetV2 and CoAtNet
6:15 - Uber's H3: A hexagonal coordinate system
7:40 - AWS NeurIPS 2021 DeepRacer Challenge
8:15 - Helpful Libraries
9:20 - State of PyTorch in September 2021
10:05 - Physics-Based Deep Learning Book
10:35 - Music-conditioned 3D dance generation
11:40 - Stallman's take on legal issues with Codex
12:20 - Tensorflow DirectML on AMD GPUs
13:00 - Schmidhuber Blog: Turing Oversold

ERRATA:
Uber's H3 is actually not new, but from 2018

References:
TruthfulQA - A benchmark assessing truthfulness of language models
https://owainevans.github.io/pdfs/truthfulQA_lin_evans.pdf

LAION-400M image-text-pair dataset
https://laion.ai/laion-400-open-dataset/
https://laion.ai/#top
https://gogetfunding.com/help-us-build-the-worlds-largest-open-billion-scale-image-text-dataset-perfect-for-training-dall-e-clip-other-multimodal-models/
https://rom1504.github.io/clip-retrieval/?back=https%3A%2F%2Fsplunk.vra.ro&index=laion_400m_128G&query=yellow+train

GooleAI releases EfficientNetV2 and CoAtNet
https://ai.googleblog.com/2021/09/toward-fast-and-accurate-neural.html

Uber's H3 hexagonal coordinate systems
https://eng.uber.com/h3/?utm_source=pocket_mylist

NeurIPS 2021 DeepRacer Challenge
https://www.aicrowd.com/challenges/neurips-2021-aws-deepracer-ai-driving-olympics-challenge?utm_source=pocket_mylist
https://aws.amazon.com/deepracer/
https://gitlab.aicrowd.com/deepracer/neurips-2021-aws-deepracer-starter-kit/-/tree/master/deepracer-gym

Helpful Libraries
https://github.com/rom1504/img2dataset
https://github.com/facebookresearch/vissl?utm_source=pocket_mylist
https://github.com/pyg-team/pytorch_geometric
https://aws.amazon.com/blogs/machine-learning/announcing-the-amazon-s3-plugin-for-pytorch/

State of PyTorch in September 2021
https://dev-discuss.pytorch.org/t/state-of-pytorch-core-september-2021-edition/332

Physics-Based Deep Learning Book
http://physicsbaseddeeplearning.org/intro.html
https://arxiv.org/pdf/2109.05237.pdf

Music Conditioned 3D dance generation
https://ai.googleblog.com/2021/09/music-conditioned-3d-dance-generation.html

Richard Stallman on Codex legal issues
https://news.slashdot.org/story/21/09/18/0432224/richard-stallman-shares-his-concerns-about-githubs-copilot----and-about-github

Tensorflow DirectML on AMD
https://wccftech.com/amd-microsoft-bring-tensorflow-directml-to-life-4x-improvement-with-rdna-2-gpus/

Schmidhuber: Turing Oversold
https://people.idsia.ch//~juergen/turing-oversold.html

Links:
TabNine Code Completion (Referral): http://bit.ly/tabnine-yannick
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
Discord: https://discord.gg/4H8xxDF
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Parler: https://parler.com/profile/YannicKilcher
LinkedIn: https://www.linkedin.com/in/ykilcher
BiliBili: https://space.bilibili.com/1824646584

If you want to support me, the best thing to do is to share out the content :)

If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
SubscribeStar: https://www.subscribestar.com/yannickilcher
Patreon: https://www.patreon.com/yannickilcher
Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq
Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2
Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m
Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n




Other Videos By Yannic Kilcher


2021-10-27I went to an AI Art Festival in Geneva (AiiA Festival Trip Report)
2021-10-24Symbolic Knowledge Distillation: from General Language Models to Commonsense Models (Explained)
2021-10-21I took a Swiss train and it was awesome! Train Seat Review - SBB InterCity 1 - Geneva to St. Gallen
2021-10-20[ML News] Microsoft trains 530B model | ConvMixer model fits into single tweet | DeepMind profitable
2021-10-07[ML News] DeepMind does Nowcasting | The Guardian's shady reporting | AI finishes Beethoven's 10th
2021-10-06Grokking: Generalization beyond Overfitting on small algorithmic datasets (Paper Explained)
2021-10-02How far can we scale up? Deep Learning's Diminishing Returns (Article Review)
2021-09-29[ML News] Plagiarism Case w/ Plot Twist | CLIP for video surveillance | OpenAI summarizes books
2021-09-27Inconsistency in Conference Peer Review: Revisiting the 2014 NeurIPS Experiment (Paper Explained)
2021-09-26100K Subs AMA (Ask Me Anything)
2021-09-24[ML News] New ImageNet SOTA | Uber's H3 hexagonal coordinate system | New text-image-pair dataset
2021-09-21Does GPT-3 lie? - Misinformation and fear-mongering around the TruthfulQA dataset
2021-09-20Topographic VAEs learn Equivariant Capsules (Machine Learning Research Paper Explained)
2021-09-16[ML News] Roomba Avoids Poop | Textless NLP | TikTok Algorithm Secrets | New Schmidhuber Blog
2021-09-14Celebrating 100k Subscribers! (w/ Channel Statistics)
2021-09-10[ML News] AI predicts race from X-Ray | Google kills HealthStreams | Boosting Search with MuZero
2021-09-06∞-former: Infinite Memory Transformer (aka Infty-Former / Infinity-Former, Research Paper Explained)
2021-09-03[ML News] Blind Chess AI Competition | Graph NNs for traffic | AI gift suggestions
2021-09-02ALiBi - Train Short, Test Long: Attention with linear biases enables input length extrapolation
2021-08-27[ML News] Stanford HAI coins Foundation Models & High-profile case of plagiarism uncovered
2021-08-26Fastformer: Additive Attention Can Be All You Need (Machine Learning Research Paper Explained)



Tags:
deep learning
machine learning
arxiv
explained
neural networks
ai
artificial intelligence
paper
mlnews
laion
schmidhuber
coatnet
efficientnetv2
truthfulqa
gpt-3
pyg
deepracer
turing