# Amada decision model by LiamDB

> Amada runs typed text decisions on a CPU through ONNX. Read its noncommercial licence, context limits and maker-reported evaluation scope.

Updated 2026-11-12 (Australia/Adelaide)

## Quick answers

### What is Amada by LiamDB?

[Amada](<https://decisionmodels.ai/models/amada>): A CPU decision model family.

[Amada facts](<https://decisionmodels.ai/models/amada>)

[Source](<https://github.com/DevLiamDb/Amada>)

### Who makes Amada?

[Amada](<https://decisionmodels.ai/models/amada>): LiamDB

[Amada facts](<https://decisionmodels.ai/models/amada>)

[Source](<https://github.com/DevLiamDb/Amada>)

By LiamDB · GitHub weight release: 2026-11-08 · NON-COMMERCIAL

A CPU decision model family. [Amada](<https://decisionmodels.ai/models/amada>) 3.1 uses mmBERT-base to answer yes/no, choice and ordered-score questions, with a separate probability that the text does not provide the answer for yes/no questions. Downloads pair an ONNX model with its tokenizer.

## Facts

- Base: mmBERT-base for 2.0 and later; ModernBERT-large for 1.0
- Context: 8,192 tokens for 3.x; 512 for 1.0 and 2.0
- Licence: CC BY-NC-ND 4.0 for weights and repository code; non-commercial, no redistribution of modified versions
- Access: Public release ZIP files and Python ONNX runner

## Claims and sources

- **VENDOR-REPORTED**: The maker reports 0.961 yes/no, 0.915 choice and 0.760 score accuracy on short English tests, using GPT-6 Luna labels. Human-labelled public-data results include SST-2 0.899 and Banking77 0.611. Half of the bake-off sets informed calibration, so the figures are not one untouched general benchmark. [source](<https://github.com/DevLiamDb/Amada/releases/tag/v3.1>)

## Links

- [Source: Code, model description and limits](<https://github.com/DevLiamDb/Amada>)
- [Source: 3.1 release and weights](<https://github.com/DevLiamDb/Amada/releases/tag/v3.1>)
- [Source: Licence](<https://github.com/DevLiamDb/Amada/blob/main/LICENSE>)

## Related projects

No related projects are listed in the registry.