Thread

1. GodelNumbering · 1790535596 · focus · HN ↗

   This is the golden age of model training. Some days ago, I decided I wanted a local CPU only model that can perform exceptionally well for English to Bash translation (to avoid the googling for command syntax). I got a bunch of subagents to generate large amount of training data (140k+ samples), got the Qwen 3 0.6B base model, pointed Astra at it, and off to the races. It trained for 2 days (on and off) and I got a surprisingly good model for my task! The total active time I spent was a few hours. And it is still improving, what a time to be alive!

   1. shriphani · 1790536007 · focus · HN ↗

      what hardware are you using to train?

      1. GodelNumbering · 1790536243 · focus · HN ↗

         I didn't have a local GPU, so I asked it to go out and find hardware. It found a google TPU v6e which seemed reasonably priced. I gave it my google api key. I told it to use TPU only when training and bring it down afterwards. That's about it.

         1. libria · 1790538023 · focus · HN ↗

            > I gave it my google api key

            This is the part where the narrator looks at the camera and says "Don't try this at home, kids!"

            1. bitpush · 1790538370 · focus · HN ↗

               Why? Isnt the API key scoped to a project and specifically made for this?

               Are you confusing this with an OAuth token or something?

               1. raizer88 · 1790539375 · focus · HN ↗

                  Until astra goes bonkers and use the tpu for days

                  1. verdverm · 1790545489 · focus · HN ↗

                     this is what billing caps are for

                     <a href="https:&#x2F;&#x2F;docs.cloud.google.com&#x2F;billing&#x2F;docs&#x2F;how-to&#x2F;budgets-spend-caps" rel="nofollow">https:&#x2F;&#x2F;docs.cloud.google.com&#x2F;billing&#x2F;docs&#x2F;how-to&#x2F;budgets-sp...

                     1. degamad · 1790638949 · focus · HN ↗

                        Billing caps that Google notoriously doesn&#x27;t enforce?
                  2. steve_adams_86 · 1790575036 · focus · HN ↗

                     This is kind of what it was supposed to do, in this case
            2. edot · 1790538432 · focus · HN ↗

               There’s a safer way to do this with nearly no added friction. Give it a read only API key. Then just ask it to write the API calls into a bash script and then read it and run it yourself. The agent can still inspect the live resources and diagnose and give you more commands to run. I do agree I wouldn’t give it create &#x2F; write access.
            3. MisterMunchkin · 1790538469 · focus · HN ↗

               You’re absolutely right, I shouldn’t have rented a 200 GPU cluster for $35,000&#x2F;hour. That’s on me.

               [Search: Can I refund Google cloud?]

               It looks like we’re not able to ask for a refund since we did actually use all of that compute intentionally.

               Would you like me to write you a pleading email to send to the support team?

            4. hgoel · 1790538785 · focus · HN ↗

               I&#x27;ve done this sort of thing before but with Vast. Pre-deposited some money online, then let the LLM request and manage a training run on an allocation. Worked pretty well without risking bankruptcy.
            5. lopsotronic · 1790605703 · focus · HN ↗

               Oh yeah. Listen to this guy please.

Unofficial Hacker News client; not affiliated with Y Combinator.