gemma-4-E4B-it-GMCE (GMCoder Efficiency)
google/gemma-4-E4B-it fine-tuned for GMCoder Efficiency, which makes answers shorter and more direct. Files: GGUF (Q8_0, Q6_K, Q5_K_M, Q4_K_M, BF16) and BF16 safetensors.
What it does (measured)
Our own unseen "Form B" exams, default chat mode, 60k token cap, single run per model, graded blind by one grader. Scores are approximate; a few points is noise.
| Tokens (3 exams) | Ultimate (/25) | Historical (/100, approx) | |
|---|---|---|---|
| gemma-4-E4B-it | 19.6k | 19 | ~72 |
| gemma-4-E4B-it-GMCE | 12.1k (-38%) | 20 | ~68 |
Per exam: Ultimate -27% tokens, Historical -42%, Everyday -41%. Everyday was spot-checked, not fully graded; it looked about even.
Caveats
- Only the default mode was tested (our local server ignored Gemma's thinking switch).
- Benchmarks are our own and small. Do not read the numbers as leaderboard results.