wowsuch ai
Live

ConvAI InnovationsModel

Laya Typed Decisions

Laya Typed Decisions is a checkpoint fine-tuned for four typed-decision workflows. The source reports 0.766 accuracy on its typed-decisions benchmark, compared with 0.362 for the base English checkpoint on the same decisions.

What it does

Developers can use it to run local choice, ordinal-score, and boolean decisions, or route requests among Core ML model bundles.

Availability
Install from PyPI; model bundles available on Hugging Face
Pricing model
Apache-2.0
Deployment
Local inference on Apple Silicon; download once, then run offline
Privacy
Subsequent predictions stay local
Licence
Apache-2.0
Capability
analysisextractionevaluationautomation
Modality
text
Access
APISDKCLIopen weights
Deployment
on-device
Pricing
open-source
Target user
developerresearcherenterprise

Evidence

10 claims
  • Laya Typed Decisions is a fine-tuned checkpoint for four typed-decision workflows. [1]
    On the typed-decisions benchmark (2,000 decisions across four workflows), the fine-tuned [`laya-typed-decisions`](https://huggingface.co/convaiinnovations/laya-typed-decisions) checkpoint scores **0.766** accuracy, against **0.362** for the base English checkpoint on the same decisions.
    Hugging Face. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • It scores 0.766 accuracy on the typed-decisions benchmark. [1]
    On the typed-decisions benchmark (2,000 decisions across four workflows), the fine-tuned [`laya-typed-decisions`](https://huggingface.co/convaiinnovations/laya-typed-decisions) checkpoint scores **0.766** accuracy, against **0.362** for the base English checkpoint on the same decisions.
    Hugging Face. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • The base English checkpoint scores 0.362 on the same benchmark decisions. [1]
    On the typed-decisions benchmark (2,000 decisions across four workflows), the fine-tuned [`laya-typed-decisions`](https://huggingface.co/convaiinnovations/laya-typed-decisions) checkpoint scores **0.766** accuracy, against **0.362** for the base English checkpoint on the same decisions.
    Hugging Face. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • The checkpoint is available in the repository's typed-decisions subfolder. [1]
    On the typed-decisions benchmark (2,000 decisions across four workflows), the fine-tuned [`laya-typed-decisions`](https://huggingface.co/convaiinnovations/laya-typed-decisions) checkpoint scores **0.766** accuracy, against **0.362** for the base English checkpoint on the same decisions.
    Hugging Face. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • Version 0.4.0 is available as the laya-coreml Python package on PyPI.
  • The runtime supports batch and long-text prediction, hooks, structured decisions, calibration, HTTP/MCP, evaluation, and Router behavior.
  • After downloading a model, users can run inference offline on Apple Silicon.
  • The runtime returns probabilities for choice, ordinal-score, and boolean questions without autoregressive decoding. [2]
    Laya returns probabilities for **choice**, ordinal **score**, and boolean **noul**
    GitHub new repositories. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • Subsequent predictions stay local.
  • The project is an independent port and is not an official Convai Innovations or Apple release. [2]
    Not an official Convai Innovations or Apple release.
    GitHub new repositories. Quoted text is outside CC BY 4.0 and keeps its original rights.

Sources

2
1
Hugging FaceOwner-controlled
https://huggingface.co/convaiinnovations/laya Retrieved 2026-10-29 00:21 UTC
Verified
2
GitHub new repositoriesOwner-controlled
https://github.com/mizorewww/laya-coreml Retrieved 2026-10-29 00:21 UTC
Verified

Versions

9
VersionEffectiveChangeChanged fieldsState
92026-10-20 13:21 UTCevidence updatescoreCurrent
82026-10-18 15:10 UTCevidence updatescoreCurrent
72026-10-11 21:20 UTCevidence updatescoreCurrent
62026-10-10 22:14 UTCmaterial changecapability, access, pricing, availability, deployment, privacyCurrent
52026-10-03 05:35 UTCevidence updatescoreCurrent
42026-10-02 06:29 UTCmaterial changeaccess, availability, deployment, privacyCurrent
32026-09-26 11:47 UTCevidence updatescoreCurrent
22026-09-25 12:41 UTCmaterial changecapability, access, deployment, pricing, availability, pricing model, deployment, privacyCurrent
12026-09-24 13:37 UTCpublicationNoneCurrent

Open in the catalog

Original catalog content is licensed under CC BY 4.0, attributed to wowsuchai. Source quotations and linked content keep their original rights.