Show HN: We quantized Qwen3.6-35B-A3B to 2-bit: 12.3 GB, 225 tok/s on a 4090 (huggingface.co) 3 points by yzh 1mo ago ↗ HN
[–] riknos314 1mo ago ↗ I'm missing an accuracy vs bits graph here.Cool to see this running but if it's at 10%+ loss vs 16bit that's less cool.Hard to understand that story from what's published here.
[–] magicalhippo 1mo ago ↗ How does it compare against Unsloth's UD-Q2_K_XL which has the same size? Ie, why pick yours over theirs?
3 comments
[ 1.6 ms ] story [ 7.5 ms ] threadCool to see this running but if it's at 10%+ loss vs 16bit that's less cool.
Hard to understand that story from what's published here.