bsaleh03/Qwen3.8-27B-ASCII-Condensed
244.9k
Update README.md
Update README.md
Add measured throughput; correct context comparison to unpruned baseline
Remove plain IQ4_XS build; superseded by UD-IQ4_XS
Add model card
Add UD-IQ4_XS build (vocab 248320 -> 129006, 135K context on 16GB)
Add upstream Apache 2.0 LICENSE (Qwen3.8-27B)
Add Qwen3.8-27B-ASCII-Condensed IQ4_XS (vocab 248320 -> 129006)
initial commit
