decima:base-cleanrejects 6 %
Decima · A. M. Madani · 321M · ONNX Runtime · Apache-2.0
- Score
- 23.5
- 95 % interval 21.3–25.6
- Rank
- 53
- of 66 scored models
- Public benchmark
- 49.2
- macro accuracy, %
- typed-decisions
- 47.4
- accuracy, %
By task
Chance-corrected accuracy on each of the 14 tasks: 0 is guessing, 100 is every answer right.
Workflows
typed-decisions
23.0
Classification
MASSIVE en
64.9
MASSIVE de
56.4
Civil Comments
46.0
Aegis 2
0.0
Understanding
MultiNLI
71.9
VitaminC
39.4
PAWS
18.4
Knowledge
BoolQ
31.3
SQuAD 2
-1.7
PubMedQA
2.8
Judging
HelpSteer 2
-11.4
SummEval rel.
3.6
SummEval cons.
-16.3
Accuracy and calibration
| Benchmark | Accuracy | Choice | Noul | Score | ECE | Brier | Rejected |
|---|---|---|---|---|---|---|---|
| Public benchmark | 49.2(47.8–50.5) | 60.4 | 59.4 | 13.6 | 0.067 | 0.522 | 301 / 3,880 |
| typed-decisions | 47.4(45.0–50.0) | 37.5 | 59.0 | 46.3 | 0.061 | 0.619 | 7 / 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 | 26.4 ms | 33.1 ms | 3.36 s | 2.9 GB | 1.8 GB | F32 |
| Threadripper 3960X CPU | 1.20 s | 1.27 s | 4.44 s | – | 1.1 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
| Device | Same decision | Max logit diff. | Result | Record |
|---|---|---|---|---|
| RTX 5090 CPU | 599/599 | 3.2e-5 | pass | results/runs/2026-10-24-parity-rtx5090-cpu-cob560.json |
| RTX 5090 CUDA | 599/599 | 1.8e-5 | pass | results/runs/2026-10-24-parity-rtx5090-cuda-cob560.json |
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 decima:base-clean
# 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 decima:base-clean \
--url http://127.0.0.1:11435 --data typed.jsonl --out typed/decima-base-clean
# Bespoke Labs' public benchmark; build its 13 subsets first
NIMBLE_SRC=nimble python -m ollaya_convert.bench_public run --model decima:base-clean \
--url http://127.0.0.1:11435 --data public --out public/decima-base-clean
# speed and memory
python -m ollaya_convert.bench_latency --models decima:base-clean --memory \
--machine my-machine --device cuda --out latency.json