Another benchmark has hit the quant war
retained text
If you have not been following [](https://huggingface.co/ISTA-DASLab)[ISTA-DASLab](https://huggingface.co/ISTA-DASLab) the creators of the gsq-rco quants have received some criticism from [unsloth](https://www.reddit.com/r/LocalLLaMA/comments/1xm102d/unsloth_vs_gsqrco_512k_context_deepseekv41_gguf/) regarding the inabilty to reproduce the benchmark numbers that [ISTA-DASLab](https://huggingface.co/ISTA-DASLab) have quoted. The methodology used to produce their benchmark numbers has not been posted and their has been difficulty reproducing their numbers with unsloth finding in there testing unsloth quants outperforming the gsq-rco quants. The numbers in this post here are from another creator the creator of [basalt](https://github.com/jesdga95/basalt/) (a inference engine) previously his qwen-flash-next quants (his inference engine uses specially packed quants based on a existing model similar to [ninfer](https://github.com/Neroued/ninfer) ) used ISTA-DASLab's quants as the base but after doing some testing he replaced them with similar sized unsloth equivalents the performance numbers are in the screenshots
document fetched 2026-11-08 00:34 · canonical https://reddit.com/r/LocalLLaMA/comments/1xpxfob/another_benchmark_has_hit_the_quant_war