Entries for July 21, 2026

Apparently mosh does not allow bitmaps and herdr complicates things even more So my @pidotdev nyan cat context indicator has to be full unicode 🏳️‍🌈🏳️‍🌈🏳️‍🌈😺

My recent formulation is handy when it comes to claims like this, e.g. 50 tok/s on DGX spark with DS4 flash 🤨 Because if the formulation is correct, laws of physics do not permit > 28 tok/s in a single session, for weights comparable to antirez/deepseek-v4-gguf. Ignoring speculative decoding (rho = 1) Then I realized that the reported number is an aggregate of 4 sessions, so per single session it is like 12~14 tok/s (great result if it holds btw!) IMO people subjectively always assume single session speeds, so when a tok/s is claimed, we should all converge on a clear notation For example, 4x12.3 tok/s makes it very clear: 4 sessions, 12.3 tok/s each And when we want to emphasize aggregate, let's show them always together "4x12.3 tok/s = 50 tok/s aggregate" Another interesting result of the formulation: If you are willing to suffer 12 tok/s, then you can theoretically have up to 8 parallel sessions --- around 100 tok/s aggregate, for the same size of weights and KV cache You can find the specific hardware/model combo upper bounds in the hugging face space link below, and my formulation as well Let me know if you find any errors in my upper bound formulation

it’s not hard to have such long running tasks, they come up often in refactoring, dataset processing, etc. this is one of the longer ones. I think I used codex 4-5 resets in the last 3 weeks, token usage going vertical chatgpt pro is one of the highest value plans in the market