ThibWeb

for me it’s more an environmental consideration than financial? Passivhaus Standard for houses is about 50kWh/m²/year, so 30kWh is getting close to 10% of total

(OP) yes all energy / carbon footprint numbers are from Neuralwatt. ty for your work! Switched from Standard to Pro last week, it’s great

ah my bad, I’ve switched the visual to A index by cost per task. Note the blog post’s plot is filtered further than AA’s, based on what models are available with inference providers we can actually work with (Europe only). Vibe-coded filtering is here (very rough): https://pareto-eco-sum.netlify.app/

Might do more of that next time! Inference was with TensorX and Neuralwatt.

The high usage was due to omp in vibe mode overnight, probably working way too hard through things. The high cost, yes we’d rather pay extra to work with providers that provide other benefits than just lowest cost possible (open models, no training, EU DC, etc)

No those numbers are GPU only. See https://cleerdash.sustainableaigroup.com/ for a fuller model

For what it’s worth the energy numbers I get from Neuralwatt are within the ballpark of what is available elsewhere like https://cleerdash.sustainableaigroup.com/. What’s your source that there is profit / capex in there? Their published methodology seems pretty transparent

I agree but it does worry me how fast my usage is increasing. Two months ago I was using 10x less tokens and probably not much more than 5kWh on inference. This month about 30kWh on inference. If it becomes more affordable, is there going to be another jump? Not quite sure

Yep, I think next month will be on that. It feels slightly better from a few days of use, and in our WIP benchmarking it scores way higher

I have a hard time justifying GLM 5.3 these days. It’s slightly better than Flash but rarely enough to justify the much steeper price. We chose to use usage-based billing only so are very sensitive to model price.

Yes, all from Neuralwatt, GPU energy use only. Makes models’ “efficiency” much more visible than tokens.

Hmm I might need to rephrase. Initial challenge was to use GLM 5.3 Flash and I was on the non-Flash version for the whole vibe coded build. Just wasn’t paying attention and I didn’t realize that one session was a quarter of the month’s spend and 150% of the budget (big price difference between models)

It was a bit of a silly challenge, wasn’t sure how workable, learned a lot in the process about what actually drives usage / costs, and how to keep both under control

done, added three options to the blog post! ty for the feeback

Good point, i’ll fix that. Instructions are here: https://github.com/wagtail/bakerydemo#api-demo

Let us know how it goes! This is very new for us so feedback is key

Data on the number of releases comes from ecosyste.ms which was very convenient to do this while comparing with other packages outside PyPI. For what it’s worth, Rails is at 510 on there, which I believe is all releases since 2004. And Laravel is at 247 but there’s 2-3 years missing

it’s my 1st attempt at reporting on CI downloads specifically. Interpreting this is more of an art than a science, I’d love to hear if others have ideas on what to do with this data!

for me the surprise is the pace? I’d expect people to be more set in their tools that it takes longer than a few months for a new tool, no matter how good, to become the majority use one. Though perhaps people adopt new tools more easily in CI where install times matter more

Between this and the relative color syntax, looks like much cleverer theming options are on the way, yay for more control for end users. For anyone discovering APCA, aside from it giving better results, I’ve also found it super interesting to read about the logic behind the algorithm. The Easy Intro mentioned in the article (https://git.apcacontrast.com/documentation/APCAeasyIntro) is well worth the time and it has all the relevant follow-up resources.