decider:0.8b
decider · Mapika · 0.75B · ONNX Runtime · Apache-2.0
- Score
- 47.8
- 95 % interval 45.7–50.0
- Rank
- 19
- of 35 scored models
- Public benchmark
- 66.9
- macro accuracy, %
- typed-decisions
- 50.5
- accuracy, %
By task
Chance-corrected accuracy on each of the 14 tasks: 0 is guessing, 100 is every answer right.
Accuracy and calibration
| Benchmark | Accuracy | Choice | Noul | Score | ECE | Brier | Rejected |
|---|---|---|---|---|---|---|---|
| Public benchmark | 66.9(65.4–68.1) | 74.5 | 71.9 | 45.9 | 0.061 | 0.421 | 0 / 3,880 |
| typed-decisions | 50.5(48.2–53.0) | 46.5 | 55.2 | 50.1 | 0.172 | 0.648 | 0 / 400 requests |
Accuracy in %, with its 95 % bootstrap interval. A rejected request (the model can’t take the question, or the input is too long for it) counts as wrong.
Speed and memory
| Device | p50 | p90 | Load | VRAM | RAM peak | Precision |
|---|---|---|---|---|---|---|
| RTX 5090 CUDA | 189 ms | 191 ms | 17.4 s | 5.0 GB | 6.9 GB | F32 |
| Threadripper 3960X CPU | 3.11 s | 3.11 s | 15.1 s | – | 6.6 GB | F32 |
Five-question requests over HTTP, one at a time, each a different short message. VRAM: the rise of the GPU’s used memory while the model is loaded. RAM peak: the runner process’s peak resident memory.
Parity with the authors’ code
Its parity check is in the model’s family documentation in the repository (docs/families/); the compiled parity file does not list it yet.
Reproduce
The tools are in convert/ollaya_convert and need only Python’s standard library. Run them against a local Ollaya 0.13.0 with the same model digest.
ollaya pull decider:0.8b
# typed-decisions; export the test split once with
# uv run python -m ollaya_convert.bench_typed export --out typed.jsonl
python -m ollaya_convert.bench_typed run --model decider:0.8b \
--url http://127.0.0.1:11435 --data typed.jsonl --out typed/decider-0.8b
# Bespoke Labs' public benchmark; build its 13 subsets first
NIMBLE_SRC=nimble python -m ollaya_convert.bench_public run --model decider:0.8b \
--url http://127.0.0.1:11435 --data public --out public/decider-0.8b
# speed and memory
python -m ollaya_convert.bench_latency --models decider:0.8b --memory \
--machine my-machine --device cuda --out latency.json