snap for Ollaya

Ollaya package of logitlab/snap1-2b-GGUF by logitlab (snap1-2b, with emnlmn's snap) and OpenBMB (MiniCPM5). Ollaya runs open decision models locally, the way Ollama runs LLMs: typed questions in, calibrated answers out, behind a TypeSafe-compatible API.

ollaya run snap

What is in this repository

This repository holds only the files Ollaya derives, with no weights. The model is the authors' own GGUF file: ollaya pull downloads it from their repository, unmodified and pinned to a commit, verifies its sha256, and Ollaya runs it on llama.cpp.

Tag	The authors' GGUF	Files
snap:2b	logitlab/snap1-2b-GGUF@3932191 snap1-2b-q8_0.gguf	2b/decision.json, 2b/calibration.json
snap:2b-q4_k_m	logitlab/snap1-2b-GGUF@3932191 snap1-2b-q4_k_m.gguf	2b-q4_k_m/decision.json, 2b-q4_k_m/calibration.json

Each tag has decision.json (the prompt, the option labels Ollaya reads and llama.cpp's settings) and calibration.json (temperatures).

Parity

Ollaya's runner matches stock llama-server of the pinned build (b11146) on the same GGUF, CPU and CUDA (RTX 5090 for Q8_0, RTX 4090 for Q4_K_M): 573 questions per file, every decision the same, option logits within 7.4e-6 and probabilities within 1.9e-6. The prompt token ids are snap's own, from its export-prompts.

License

Same as the upstream model (Apache-2.0). Ollaya itself is Apache-2.0.