sgd99

I was just thinking of getting into RL, made some progress with q-learning. perfect timing!

I love this: "Our singular focus means no distraction by management overhead or product cycles, and our business model means safety, security, and progress are all insulated from short-term commercial pressures."

SSI, here is "Safe SuperIntelligence Inc."

thank you!

can anyone explain how these visual tokens which are concatenated with the tokenizer outputs for the encoder are created?

Not self-hosted/local but Claude by Anthropic from what I've heard is really good but the API is not publicly available. It's apparently accessible via Poe (https://poe.com)

As for open models, HuggingFace has a nice leaderboard to see which ones are decent: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderb...

It's good that they chose to continue support for code-davinci-002 (https://twitter.com/sama/status/1638576434485825536?s=20) but it'd much better if they open-source it sooner or later as even OpenAI didn't expect that their model is being widely used.

or just pay for the photos instead?

I tried: https://codepen.io/shreydan/pen/ZEjRPwx

Pretty impressive! + It's open source + we can try this out on our own supported car or a simulator. CommaAI has come a long way

You can try Vim within VSCode with this extension: https://marketplace.visualstudio.com/items?itemName=vscodevi...

It's the best of both worlds

not everyone is happy about this change, pagination allowed people to keep track of previous results on different pages.