Decision ·
Qwen3-1.7B = main 8 GB engine; 4B = optional "more complete" tier
From the record
- Context
- decided by: author
- Why
- Pixel 8a: 1.7B 13-16 tok/s, answers 13-16 s; 4B 3.6-6 tok/s, 29-46 s, one LMK kill
- What followed
- adopted; confirmed at 19:07 grill
- Reversible
- yes
- Status
- kept
- Alternatives considered
- keep 4B default
- Qwen2.5-1.5B (faster but weaker)
Linked facts · 2
- Qwen3-1.7B = main 8 GB engine; 4B = optional "more complete" tierQwen3-1.7B = main 8 GB engine; 4B = optional "more complete" tieradopted; confirmed at 19:07 grillDate2026-10-20 ~11:50Kinda record, not a measurement (set, status, judge, n, seeds: not recorded)Caveatwho decided: author; why: Pixel 8a: 1.7B 13-16 tok/s, answers 13-16 s; 4B 3.6-6 tok/s, 29-46 s, one LMK kill; alternatives: keep 4B default; Qwen2.5-1.5B (faster but weaker); reversible: yesSourcenot in the public repositories yetOpen the data file
- Qwen3-1.7B becomes the main 8 GB engineQwen3-1.7B becomes the main 8 GB engine4B becomes optional tier (19:07)Date2026-10-20Kinda record, not a measurement (set, status, judge, n, seeds: not recorded)Caveatwho: author; why: Pixel 8a: 1.7B 13-16 tok/s vs 4B 3.6-6 tok/s, 4B low-memory kill; reversible: yesSourcenot in the public repositories yetOpen the data file