[Tepix]

Kimi K2.6 does not run well on 256GB.

[zozbot234]

Have you tried it? It would be slow for sure, but the main limitation AIUI would actually be storing the context in RAM - models like Kimi and GLM have high demands there which limit your ability to get meaningful aggregate throughput via large batches.

[Tepix]

No need to try really. 1100b weights with 256GB RAM that‘s less than 1.8 bits per weight if you want a little bit of context.

How is that supposed to give good results?

[girvo]

True, I might be thinking of some of the communities four-Spark clusters for it; it’s already int4 right?

[Tepix]

Yeah, the default quants are 595GB. Even four Sparks would require a quant lower than 4bit