ollaya-dev/sura-2b

Sura 2b

Planned. The weights are not published yet. They will appear in this repository once training finishes and the model passes Ollaya's parity checks and benchmark.

Sura 2b is a decision model by Ollaya. The larger sura, for accuracy on a GPU. Give it a state (a message, a support ticket, a JSON object, an agent's next tool call) and questions, and it answers every question in one forward pass with calibrated probabilities:

• noul: yes or no, as the probability of yes;
• choice: one of up to 255 options, each described in plain words;
• score: a level on a scale of described levels.

Ollaya's model families

Family	Network	Built for
coba	encoder (Ettin), 68M to 1B	speed: CPUs, and the browser at 35 to 147 MB
sura	decoder (Qwen3.5)	accuracy

Model	Base	Status
sura-0.8b	Qwen/Qwen3.5-0.8B-Base	in training
sura-2b	Qwen/Qwen3.5-2B-Base	planned

How it is built

• Network: the joint inline-readout protocol of Perplexity's pplx-decider (version 3). Every question of a request shares one sequence with the state; each question is followed by its options and a marker, and a 255-way linear readout scores the options at that marker. The full-attention layers see the whole sequence; the Gated DeltaNet layers stay causal.
• Teacher: perplexity-ai/pplx-decider-flash-v1.2 (Qwen3.5-4B, Apache-2.0). The student shares its tokenizer and prompt, so it learns the teacher's full distribution over the options, not only its answers.
• Data: the mixture of coba's zero-shot models, about 340,000 states with gold and teacher labels. It holds none of the datasets of the benchmarks Ollaya reports, so the results will be zero-shot.
• License: Apache-2.0, like the base model and the teacher.

Run it

Once the weights are published, with Ollaya (Linux, macOS, Windows; CPU, CUDA, Metal):

curl -fsSL https://ollaya.dev/install.sh | sh
ollaya run sura:2b --ask "Is the customer asking for a refund?" < ticket.txt