tourist_on_road

Can a employee sponsored greencard holder move jobs within 3 months after getting greencard approved granted the new role is pretty similar to one before? will there be any issues in terms of naturalization/reapproval.?

Nice post. Instead of using inpainting for model explainability, why not use attention maps to understand where the model is focusing to get a particular prediction?

Super impressive to see how OpenAI managed to bring the project from research to production (something usable for creatives). This is non trivial since the usecase involves filtering NSFW content, reducing bias in generated images. Kudos to the entire team.

Looks like the site went down due. you can access the article here -- https://archive.ph/QQ5iV

Good point. The fact there is no inductive bias inherent to transformers makes it difficult to train a decent model on small datasets from scratch. However, there are recent research directions that try to address this problem [1].

Also baking in some sort of domain specific inductive bias into model architecture itself can address this problem as well [2].

[1]: Escaping the Big Data Paradigm with Compact Transformers: https://arxiv.org/abs/2104.05704

[2]: CvT: Introducing Convolutions to Vision Transformers: https://arxiv.org/abs/2103.15808

Transformers gained popularity due to the scalable nature of the architecture and how well it can be parallelized on existing GPU/XLA hardware. Modeling is always conditioned on the hardware available at hand. Transfomers lack inductive bias which make it generic building blocks unlike CNN/RNN like models and by injecting inductive bias like positional encoding, it can be well translated to various domains.

Congratulations to the Streamlit team!

Personally as a machine learning practitioner, Streamlit was a god send tool to quickly setup demos without having to fiddle with front end code. I hope they continue to support the open source ecosystem even after the acquisition.

Something to note -- Indiana U doesn't have engineering stream unlike Purdue. Only thing close to engineering is computer science and Intelligent systems engineering.

Great work! Looks like you are using HNSWLIB. From what I understand HNSW graph based approach can be memory intensive compared PQ code based approach. FAISS has support for both HNSW and PQ codes. Any plans on extending your work to support PQ code based index in future?

I sense greed on part of coinbase. If coinbase doesn't want to have 4 year vest schedule which makes up a significant portion of salary, Why don't they follow what netflix is doing and just give out salary as full base salary instead of any RSUs.

..without a fan

In this regard. I like the google approach for not asking questions that are publicly available on internet.

These are basic computer science fundamentals and critical thinking capabilities that essential for performing the job atleast for software development roles.

Isn't it what computer science is all about. You won't necessarily see the exact problem. But having the ability to reduce the problem at hand to an efficient algorithm that you might have "regurgitated".

But it might come handy when implementing - route optimization problems - Building database query optimization engine

You can argue that not everyone is implementing these from scratch every day. But it could be argued that given an opportunity, an engineer should have the skills and ability to build these systems.

Well, the entire business started with his father Dhiru bhai ambani. It just happens to operate in a different way compared to the organizations we know in the west.

The irony

How does tiledb compares to something similar like milvus

From the looks of it. It is from uber ATG not from AI labs. I believe those are two different orgs. Someone from uber can clarify

I read somewhere they realized that just giving a thumbs up or down on a title is much reliable than having star recommendation system