qwen38_27bq40 / benchmarks

602 kB

Ctrl+K

1 contributor

History: 6 commits

HelloSun

Add DFlash2 CPU speed benchmarks (n_max 3/5/7)

a57bd00 verified 3 days ago

dflash
Add DFlash2 CPU speed benchmarks (n_max 3/5/7) 3 days ago
iq3_xxs
Add UD-IQ3_XXS: VNNI IQ3_XXS/IQ3_S kernels, benchmarks, PPL, model card 3 days ago
mtp
MTP draft length sweep k=3..6: decode, prefill, acceptance (k=3 best) 3 days ago
ops_v3
Q4_0: single-thread SCALE fix, SIMD RMS_NORM+MUL fusion, fusion on/off experiment (this CPU only) 3 days ago
bench_run.sh

454 Bytes
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
bin_q4_new_r1.csv

1.53 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
bin_q4_new_r2.csv

1.53 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
bin_q4_new_r3.csv

1.53 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
bin_stock_r1.csv

1.53 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
bin_stock_r2.csv

1.53 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
bin_stock_r3.csv

1.53 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
ppl_patched.log

2.37 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
ppl_stock.log

2.37 kB
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
prefetch_sweep_tg64.log

238 Bytes
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago
summary.csv

411 Bytes
Qwen3.8-27B Q4_0 CPU optimization: patch, one-click build, benchmarks, model card 3 days ago