danielamitay/laya-en-fp32-swev

laya-en-fp32 · Swev

A Core ML export of convaiinnovations/laya for Swev, a Swift package for local typed decisions. Returns choices, ordinal scores, and probabilities for runtime-defined questions; this export does not generate free-form text.

• Package: laya-en-fp32-swev-l4096-k16.mlpackage (download the entire directory).
• Precision: FP32 weights and computation.
• Text context: up to 4,096 tokens including the question, options, and formatting. Swev selects the smallest fitting bucket: 128, 256, 512, 1,024, 2,048, or 4,096.
• Questions: choice, score, and noul; up to 16 answer options per question and 64 questions per request, evaluated independently.
• Runtime: Swift 6, macOS 15+ or iOS 18+, with a Swev version supporting schema 2.0. Export version 0.3.0. Validated locally on macOS with CPU-only Core ML; memory and compute-unit support depend on the device.

Text-only export. Per-field prompt limits also apply; oversized inputs are rejected rather than silently truncated.

Swift usage

Add Swev as a Swift package dependency, then:

import Swev

let model = try await SwevModel.load(from: HuggingFaceModel(
    repository: "danielamitay/laya-en-fp32-swev",
    package: "laya-en-fp32-swev-l4096-k16.mlpackage"
))
let response = try await model.predict(
    state: "apple",
    questions: [.noul(id: "edible", instructions: "Is this an edible food?")]
)
print(try response.noul("edible").noul)

The first call downloads the package; later calls reuse Swev’s local download cache. Keep the loaded model resident to avoid recompilation and initialization on each request. See loading and caching and model support. No Python runtime or separate tokenizer/adapter files are required.

Source and conversion

Source checkpoint: convaiinnovations/laya at 1c5edc17a7ac. Tokenizer, formatting, and inference settings are bundled in the package. This is a converted export, not a new fine-tune.

Uses the English checkpoint at the repository root, built on ModernBERT-large. Source implementation: Laya. The export extends its configured context to 4,096 tokens with a 256-token question-prefix budget and retains its checkpoint temperature settings. This is not the multilingual or typed-decisions checkpoint.

Conversion tooling and the package contract are documented in Swev’s conversion guide and schema reference. Exported capacity is not a guarantee of task accuracy; evaluate the checkpoint on your own inputs.

License

Apache License 2.0; see LICENSE. Source model attribution and pinned revisions are listed above. This is an independent Swev-compatible export, not an official release by the original model authors.