This card is a 272 page report. So now we are redefining names :)

Does the model card fit in the model's context :)

well it will saturate your 5h limit window at least

Yes totally agree it’s regurgitating crazy expansive text like book author who needs to publish 10 books a day

I notice recently they push way too many tokens to explain things than they use to. I have not measured it, but working with it every day I can tell.

Working on my own agentic system. It’s been so much fun and learning rust amd svelte along the way. Thanks for the inspiration hacker-news community.

Totally agree lmstudio headless server on a remote machine but control models from your laptop is an amazing workflow. But Gemma 4 was not a good model atleast in my trials “find me the largest text file in all of the current sub folders” it went on a loopy tool call for ever even with Q8

Streisand effect. I didn’t care about the book but I will have listened to this book by this weekend.

Yes same experience. Goes into loop mode where is sends same command again and again, till we kill it. This was Q_8 version on lmstudio

After spending some time on how Claude code (leaked) tools were written, it makes sense why we constantly hit limits, for all the amazing llm capabilities, CC does not have edit tool and always read before write ( this makes sense) but I expected some surgical precision magic software, seems like other agents like master open code and pi are way better than CC in taken usage

I tried Claude code with zed and zed will eat memory like crazy (128GB RAM) after long sessions. I gave up on zed after that happening for a month

If we are living in an era where software release everyday is the norm, no amount of testing is enough to claim stability. Roll the dice everyday with these beautiful stochastic imitators.