๐โต๏ธ ๐ต๐ปโโ๏ธ๐จ๐ผโ๐ป๐ฅฝ๐โท๐บ๐ & @atomic_bits
Joined May 2008
- Tweets421
- Following2.4K
- Followers160
- Likes52.4K
Peter Rigole retweeted
Stop the executions in Iran.
The Islamic Republic has executed Erfan Kiani, and the world remains silent.
Executions are becoming routine in Iran, and every time we turn a blind eye, we enable the next one.
We must not ignore this.
#ErfanKiani ๐
Peter Rigole retweeted
gpt-oss @openai is here! ๐ฅ
> two MoEs with 21B/3.6B and 117B/5.1B total/active params, efficient reasoning models ๐คฏ
> use & fine-tune with transformers & TRL ๐ ๏ธ
> inference powered by @huggingface Inference Providers ๐ซก
> apache 2.0 license ๐
Peter Rigole retweeted
In Europe, too many leaders are used to speak and say nothing. Conversations are not real because they have to be polite first instead of having content.
This is why they canโt deal with what @JDVance said, simply because it is the truth.
The truth is that the US Vice President made people angry in Munich by simply telling the truth, by engaging in a real conversation instead of saying what the elites want to hear.
Indeed, European leaders are ignoring the will of their people, ignoring religious freedoms and not acting to halt illegal migration or protect the European values. That is not fake. Europe isnโt investing in itself. But indeed it is uncomfortable. You canโt change anything if you skip the uncomfortable facts and uncomfortable talks.
When certain leaders stop being angry when facing the truth perhaps they will be ready for the change that we all need.
Perhaps European leaders should have just replied honestly. No change without honest conversations.
Honest conversations! Isnโt this exactly what the Munich Security Conference is for?
Assita Kanko
Peter Rigole retweeted
so the local LLM action begin.
Apple is leading the pack with the MLX work that allows you to run LLMโs locally.
everything points towards private personal LLMโs at the WWDC in Juneโฆ
Peter Rigole retweeted
#PyTorch goes to the Linux Foundation. This is massive for open source software! ๐๐ฅ๐
โAs of August 2022, PyTorch was one of the five-fastest growing open source software communities in the world alongside the Linux kernel and Kubernetes.โ linuxfoundation.org/blog/welโฆ
Peter Rigole retweeted
I am developing and teaching a new course on 3D Vision at Berkeley this semester: CS294: Geometry and Learning for 3D Vision. It will be finished in about two weeks. I will make all the lecture slides available. It is time to close the loop between reconstruction and recognition.
Peter Rigole retweeted
Sneak peek van een testopstelling voor het Surv-AI-llance project waar we met heel veel trots aan meewerken! Meer weten? Lees het volledige project hier: surv-ai-llance.com.
Peter Rigole retweeted
Welcome to the future. Cop pulls over driverless car (because no lights?) Then Cruise goes on the lamb. (via instagram.com/tv/Cb1q-vggFn9โฆ)
Peter Rigole retweeted
Hereโs our new computer vision system achieving state of the art results in image segmentation, without needing any labeled training data. This new model was trained on random, unlabeled data, but quickly achieved state-of-the-art results. Itโs awesome.
Peter Rigole retweeted
Reconstructing the 3D world from 2D video requires monocular depth estimation and video panoptic segmentation, but they are typically considered separately. Today we present ViP-DeepLab, which does both at once while achieving state-of-the-art performanceโgoo.gle/2S9BVM0
Peter Rigole retweeted
A new video dataset from @facebookai designed to evaluate the fairness of computer vision and audio systems
As part of our ongoing efforts to help surface fairness issues in #AI systems, weโve open sourced a new data set of 45,186 videos to help evaluate fairness in #computervision and audio models across age, gender, apparent skin tone, & ambient lighting. ai.facebook.com/blog/sheddinโฆ
Peter Rigole retweeted
If you have an interest in continual learning, check out this incredible collective project from @ContinualAI:
github.com/ContinualAI/avalaโฆ
A PyTorch based, end-to-end library implementing a number of benchmarks, models, metrics, to setup your project and simplify comparisons.
Peter Rigole retweeted
Self supervision is the future of computer vision. Get started and reproduce results with VISSL and @PyTorch
Introducing VISSL (vissl.ai) - a library for reproducible, SOTA self-supervised learning for computer vision! Over 10 methods implemented, 60 pre-trained models, 15 benchmarks, and counting.
Peter Rigole retweeted
The DALL-E results look really cool, especially image generation. It's basically an extension of GPT-3 to text and images. openai.com/blog/dall-e/
Peter Rigole retweeted
Finding the natural parts of an object and their intrinsic coordinate frames without supervision is a crucial step in learning to parse images into part-whole hierarchies. If we start with point clouds, we can do it!
๐ข๐ข๐ข ๐๐ฎ๐ง ๐๐ญ ๐๐ฅ. "๐๐๐ง๐จ๐ง๐ข๐๐๐ฅ ๐๐๐ฉ๐ฌ๐ฎ๐ฅ๐๐ฌ: ๐๐ง๐ฌ๐ฎ๐ฉ๐๐ซ๐ฏ๐ข๐ฌ๐๐ ๐๐๐ฉ๐ฌ๐ฎ๐ฅ๐๐ฌ ๐ข๐ง ๐๐๐ง๐จ๐ง๐ข๐๐๐ฅ ๐๐จ๐ฌ๐"
A network design that realizes the concept of "๐ฆ๐๐ง๐ญ๐๐ฅ ๐ฉ๐ข๐๐ญ๐ฎ๐ซ๐" for unsupervised 3D deep learning.
canonical-capsules.github.io
Peter Rigole retweeted
This is sci-fi to reality! An epic #Starship #SN8 testflight ๐คฏ we have lost our minds @SpaceX @elonmusk @Erdayastronaut ๐๏ธ๐ท cosmicperspective.com/starshโฆ