build1 publisher
A 4-bit Gemma 4 26B on one L4 trails TypeSafe's Jev by 2.1 points overall
A pre-registered run reads Gemma's label logits on one 24 GB GPU and scores it against the hosted API on the same 3,880 records, where it is level on yes/no questions and 4.5 points behind on multiple choice. One temperature fitted on 50 labels closes the calibration gap.
Publishers:dev.to
Reality
- Evidence70
- Adoption20
- Hype gap−12
- Incentives35
- Confidence55