cfontes
/

qwen-dflash2-spark

Text Generation

speculative-decoding

Model card Files Files and versions

qwen-dflash2-spark / bench

7.04 kB

Ctrl+K

Ctrl+K

1 contributor

History: 1 commit

cfontes's picture

Qwen3.8-27B DFlash2 on 2x DGX Spark: one-stop stack (135 tok/s C1, lm_head BF16 fix)

5be2256 verified 3 days ago

edit_bench.py

4.4 kB
Qwen3.8-27B DFlash2 on 2x DGX Spark: one-stop stack (135 tok/s C1, lm_head BF16 fix) 3 days ago
prod_tps.py

1.33 kB
Qwen3.8-27B DFlash2 on 2x DGX Spark: one-stop stack (135 tok/s C1, lm_head BF16 fix) 3 days ago
quality_capture.py

1.31 kB
Qwen3.8-27B DFlash2 on 2x DGX Spark: one-stop stack (135 tok/s C1, lm_head BF16 fix) 3 days ago