lackonendes/PAW-27B-X3-GGUF

Add follow link

Move X3.1 notice to the top

Link to PAW-27B-X3.1

Upload README.md with huggingface_hub

model card: rename IFB/HumanEval/MBPP column headers, drop % from labels

model card: label speed column PP512/TG128

model card: combine PP512/TG128 into one speed column

model card: merge IFB-L/IFB-S into one loose/strict column

model card: show MMLU/IFB/HumanEval/MBPP/GSM8K as percentages

model card: move size into artifact column, split bpw column

model card: drop csweep/reasonfull/v6/v12d/v7 rows from eval table

model card: stack base/plus scores on separate lines in HumanEval & MBPP

model card: merge HumanEval/MBPP base+plus into one column each

model card: revert gguf frontmatter (grid is filename-driven; keep honest names)

model card: add gguf file grid (x3 1.5-3.5bit + DFlash2 drafter)

add model card

add Qwen3.8-27B-DFlash2-Q2_K.gguf

add PAW-27B-X3-1.5bit.gguf

add PAW-27B-X3-2bit.gguf

add PAW-27B-X3-2.5bit.gguf

add PAW-27B-X3-3bit.gguf

add PAW-27B-X3-3.5bit.gguf

initial commit