nli for Ollaya

Ollaya package of MoritzLaurer/deberta-v3-large-zeroshot-v2.0 and MoritzLaurer/ModernBERT-large-zeroshot-v2.0 by Moritz Laurer. Ollaya runs open decision models locally, the way Ollama runs LLMs: typed questions in, calibrated answers out, behind a TypeSafe-compatible API.

ollaya run nli

What is in this repository

This repository holds only the files Ollaya derives, with no weights. Each graph is an ONNX export of the original model whose weights reference the author's own model.safetensors by byte offset, so ollaya pull downloads the weights from the upstream repository, unmodified and pinned to a commit, and verifies their sha256.

Tag	Upstream	Files
nli:deberta-v3-large	MoritzLaurer/deberta-v3-large-zeroshot-v2.0@cf44676	deberta-v3-large/model-fp32.onnx, deberta-v3-large/decision.json, deberta-v3-large/calibration.json
nli:modernbert-large	MoritzLaurer/ModernBERT-large-zeroshot-v2.0@a51e07b	modernbert-large/model-fp32.onnx, modernbert-large/decision.json, modernbert-large/calibration.json

Each tag has an fp32 graph, used on CPU and GPU. Each tag also has decision.json (sequence layout, special tokens) and calibration.json (temperatures).

Parity

Ollaya's Rust runtime matches the Python reference exactly on 483 questions (1,513 hypothesis rows) per model. The token ids are identical, and so is the decision on every question. Probabilities are within 2.5e-5, on CPU and CUDA.

License

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