Models that fit on common consumer hardware(8GiB VRAM). ~8-12B
ConicCat
ConicCat
AI & ML interests
Efficient (and thereby accessible) machine learning through distillation, test time scaling, PEFT, etc.
Recent Activity
new activity
about 2 hours ago
BasedBase/Qwen3-Coder-30B-A3B-Instruct-480B-Distill-V2:Can we get benchmarks to compare to the official Qwen3 Coder 30b a3b?
updated
a dataset
about 2 hours ago
ConicCat/BasedBaseEvals
Organizations
None yet