Soul Player C64 – A real transformer running on a 1 MHz Commodore 64 (opens in new tab)

Since it seems to just produce broken and nonsensical sentences (at least based on the one example given) I'm not sure if it does work at this scale.

Anyway, as written this passage doesn't really make a whole lot of sense (the point is that it produces broken sentences?), and given that it was almost certainly written by an AI, it demonstrates that the architecture doesn't work especially well at any scale (I kid, I kid).

Transformers don't scale down very well, in my experience - I used to train local models all the time as new ones were released, as I recall transformers were the first ones I couldn't get better results out of with my limited training data and GPU.

It (v3) mostly only says hello and bye, but I guess for 25k parameters you can't complain. (I think the rather exuberant copy is probably the product of Claude et al.)

If it turns out that LLM-like models can produce genuinely useful outputs on something as constrained as a Commodore 64—or even more convincingly, if someone manages to train a capable model within the limits of hardware from that era—it would suggest we may have left a lot of progress on the table. Not just in terms of efficiency, but in how we framed the problem space for decades.

(Or maybe not, if it doesn't perform better than random, I haven't actually tried it out yet. Some more examples would have been nice!)

I wonder how far you could push this while still staying period correct, e.g. by adding a REU (RAM Expansion Unit), or even a GeoRAM (basically a REU on steroids).

SuperCPU would also be an option, but for me it's always blurring the line of "what is a C64" a bit too much, and it likely just makes it faster anyway.

Brings back memories

I'm not sure what the venn diagram of knowledge to understand what that sentence is suggesting looks like, it's probably more crowded in the intersection than one might think.