Text classification with Laya
A Laya `choice` question is a zero-shot classifier. You name the labels, optionally describe each one, and get the top label with a probability for every option in one forward pass. Labels are defined at request time, so there is no training step to get started.
252 projects
laya-onnx
tozp
Hosts ONNX-converted English Laya model weights in FP32, FP16, and INT8 formats
laya-multilingual-coreai
smdesai
Provides a Core AI conversion of multilingual Laya for typed decisions on Apple devices
laya-multilingual-onnx
soyelmismo
Provides CPU-optimized, quantized ONNX checkpoints for multilingual Laya
laya-banking77-v1
Cahol
Fine-tunes Laya for classifying customer messages into the 77 BANKING77 banking intents
laya-issue-triage
harikarthikmanyam
Fine-tunes Laya to classify GitHub issue types and assess whether more information is needed
cut-laya-onnx
harshpreet931
An 8-bit ONNX conversion runs Laya text classification in browser-based editors
laya-prompt-guard
16sulphur
A Laya fine-tune detects prompt injection and jailbreak attempts in text
laya-idjvsuen-v1
faall7479
A multilingual Laya fine-tune returns calibrated typed decisions on Indonesian, Javanese, Sundanese and English text
nlp-serving
saugataroyarghya
A BentoML playground serving and comparing focused NLP models, including Laya
decision-model-playground
tedliou
A local web playground comparing Laya and Jev for recommending articles
jev-vs-open-decision-models
elcronos
Benchmarks Jev against Laya and PrismNLI on zero-shot text classification tasks
sokudan
hiroki-abe-58
Builds a Japanese typed-decision model and benchmarks it against Laya
chatassistant-laya
HarrisXiu
Analyzes pasted chat messages locally with Laya and can generate optional reply suggestions
jev-vs-laya
DDnim
Benchmarks Jev and Laya on typed decisions for reviewing SQL statements
laya-mac-serve
chrisns
Serves Laya from a macOS menu bar app through an OpenAI-compatible HTTP endpoint
chinese-workflow-decision-bench
Adkid-Zephyr
Benchmarks Jev and Laya on synthetic Feishu message classification scenarios
SmartMom
clydechen0228
Uses Laya for edge classification and training within a smart-factory operations platform
jev-laya-benchmark
EnesDemir143
Benchmarks TypeSafe Jev and Laya-MLX on structured issue classification
laya-mlx-per-turn-classifier
M37Labs
Demonstrates per-turn customer message classification with Laya on Apple Silicon
layaAsRagJudge
vishalmysore
Verifies RAG claims in the browser using Laya decisions and evaluates accuracy against labeled claims
laya-main
bill9924
Maps patent texts to OpenAlex concepts using TF-IDF retrieval and a Laya decision model
laya-adk-toolkit
Ashfaqbs
Google ADK tools expose Laya's typed-decision engine for classification, scoring, and detection
decision-making
tzt-company
A local demo runs Laya for multilingual text decisions and Git commit risk checks
laya-ko-decision-onnx
2nugu
Fine-tunes Laya for Korean decisions and provides PyTorch training and ONNX export scripts
Example
From the core README:
from laya import Router
router = Router()
questions = {
"department": {"type": "choice", "instructions": "Which department should handle this?",
"criteria": {"billing": "invoices, payments, refunds",
"technical": "bugs, outages, system errors",
"other": "everything else"}},
}
result = router.predict("La aplicación se cierra cada vez que abro la configuración.", questions)
print(result["answers"]["department"]["choice"]) # technical
Each answer includes probabilities for every label and a confidence value you can gate on.
Label count matters
All options share a token budget (head_max_len: 192 on laya, 256 on the other two checkpoints). The README reports Banking77 (77 labels) at 0.425 for Laya against 0.870 for Jev on 72 labels, because each label gets only 3 to 4 tokens. Keep a question under about 20 options, or raise head_max_len, split into coarse and fine questions, or use predict_shortlist with embeddings.
Accuracy you can expect
- The core README lists AG News 0.947 and DAIR Emotion 0.573 for the English checkpoint.
- sysone-bench ran Laya and Jev on byte-identical inputs: AG News 0.940 vs 0.910, emotion 0.540 vs 0.550, and a 12-intent Banking77 subset 0.802 vs 0.906.
- jevbench (dhruvmehra) compares Jev, Laya, LLMs, fine-tuned DistilBERT and zero-shot NLI on SST-2, AG News and Banking77.
Fine-tune when zero-shot is not enough
The base checkpoints score 0.362 and 0.352 on the README's typed-decisions benchmark. The fine-tuned laya-typed-decisions checkpoint reaches 0.766. The fine-tuning notebook runs on Kaggle's free 2x T4 GPUs.
Projects
ollaya serves Laya alongside zero-shot NLI and GLiClass classifiers behind one API. edgejev reports AG News accuracy for fp32 and INT8 CPU builds.
Caveats
The README warns against boolean-word labels such as yes/no or true/false in choice questions, and it reports failures on negated requests. Use descriptive labels and test on your own data.
More ways to use Laya