4by4by4

Does anyone have rough numbers on how this compares to vLLM in terms of price performance? Maybe with something like Dynamo for disaggregated serving and KV-aware routing.

I'm curious if it's a modest increase in performance or a significant one.

I'd be curious to see the raw test results.

I have a suspicion that most models will miss the `database_migration` Checkpoint 2 test that includes a `default_value` because it could be interpreted as either a JSON-literal or a SQL-expression.

There might be other tests as well that are prone to failure for reasons other than the reasons cited in the paper.

I think a cool experiment would be to adjust the order of the features implemented (e.g. checkpoint 3 then 2 then 5 then 4) where dependencies allow it. Then one could account for some checkpoints being more difficult than others.

> So the same effort you would expend to get out of those activities [such as one year of smoking] on account of their risk, the same effort you should be willing to expend to get a full-body mri

That's the conclusion of the article, which I didn't immediately understand from the title. I read "earns" as a negative reward, not a positive one.

This seems similar to fzf with a nice default experience (quick access to fuzzy search files, shell history, git log) and a subset of features.

I use fzf to hack together custom TUIs (a calculator, clipboard search, quick LLM chat). The killer feature for me is `--bind` to bind keys and events to custom actions.

That would be nice and that’s why we first tried MinIO.

But with MinIO and erasure coding a single PUT results in more IOPS and we saw lower performance.

Also, expanding MinIO must be done in increments of your original buildout which is annoying. So if you start with 4 servers and 500TB, they recommend you expand by adding another 4 servers with 500TB at least.

Not the only reason, but we have a distributed workload so HTTP is a better protocol than NFS for our use case.

Setup is a little obscure, but the developer is responsive on Slack and GH.

We are only a couple months in and haven’t had to add to our cluster yet, storing about 250TB, so it’s still early for us. Promising so far and hardware has already paid for itself.

We tested both SeaweedFS and Min.io for cheaply (HDD) storing > 100TB of audio data.

Seaweed had much better performance for our use case.

Using LazyVim got me over the hump to switch to neovim from vscode. My previous attempts failed when it took too long to get an IDE-compatible setup. I suppose I am the type of person that wants and IDE/editor that lives in the terminal!

As a teenager I had a chess lesson with Ben Finegold and I practiced this puzzle for weeks on a physical board.

He also mentioned this riddle off hand which I had a ton of fun solving. https://fivethirtyeight.com/features/how-about-a-nice-game-o...

Hi HN! This game is similar to typing tests (e.g. typings.gg) but for speech. Hope you enjoy!

I’m curious what services you interact with that have five nines availability. Most services that I interact with, AWS APIs as an example, top out at an SLA of four nines for availability.

I have a four month old and was inspired to buy one outfit with magnets because of this tweet. It works fairly well, but is no better than a single zipper outfit and is much more expensive.

I’ve thought about this, too. Sadly, times only started being recorded in the last 20 years and last time I tried I couldn’t find a large dataset with the times included.

Chess.com did do one study and found a large percentage of mistakes occurs in moves 36-40 because in some time controls additional time is added at move 40.

True, it’s works differently though. And Stockfish does not compute win/draw/loss probabilities as part of its eval. It converts cp to wdl using an exponential fit based stockfish vs stockfish games. So draw percentage in Leela is a lot more interesting and useful.

In my opinion, neural network engines like Alpha Zero or Leela Zero do a much better job of assessing how difficult a position is to hold. They also report the evaluation in a completely different manner, win/draw/loss probability as opposed to centipawn loss.

For example, in yesterday’s game Stockfish was often giving a drawn evaluation (0.00) where Leela Chess gave a win probability of 30%+. I was posting about this during the game.

https://twitter.com/nik_king_/status/1466794534214504454?s=2...

I agree with others that your salary is low. Another important question is: how much would you like to make?