Deploying DeepSeek on 96 H100 GPUs

Wow, please edit the title to include Open-source !

Why? Open source isn't in the original title

Also “open source” I feel covers for “open weights” which is not the same thing.

These open models are just commercial binary distributions made available at zero cost with intention to cripple opportunities for Western LLM providers to capitalize on investments.

These are more like really gorgeous corporate swags than FOSS.

> intention to cripple opportunities for Western LLM providers to capitalize on investments.

Western LLM providers release open weight models too (e.g. Mistral).

> cripple opportunities for Western LLM

Good! If they're not open, they're creating more lock-in. And on top of that, they're using information they don't own to do so and then renting it back to us.

I 100% support it as a consumer :P but IMO we do have to be aware that it's just a happy coincidence.

Do you realize how silly you sound ?

"Linux was a Finish conspiracy to cripple hard-working US operating systems makers. It's not really open because I don't understand it."

Leave nationalism to those who want to use you as mere pawns in their imaginary Chess game.

I don't care as long as it doesn't sound sillier than to call them open source. They're literally binaries. Often lossy compressed, even.

Open weights is the equivalent of open source here. DeepSeek is open weight.

If you have some reason to believe a different definition should be used, please provide it. Because there is no source code here.

What does “open source” even mean when there is no source code?

There is a source, it would be the training data. There is also kind of the training code.

Almost absolutely no one releases their training data.