WaterSheep
WaterSheep answers yes/no, single-choice, rating and multi-label questions about any text, with a
probability for every option. Version 0.1.0 (watersheep-20260928-125452).
Usage
pip install transformers torch
from transformers import pipeline
ws = pipeline(model="samratduttaofficial/WaterSheep", trust_remote_code=True)
ws("I was charged twice.", question="Which team should handle this?", options=["billing", "shipping", "support"])
| Type | Options | Answer |
|---|---|---|
noul | none (yes/no) | probability of yes |
choice | any labels | the best option |
score | a digit scale, e.g. 1 to 5 | the expected level |
multi | any labels, with type="multi" | every option above the threshold |
Every answer includes a probability for each option.
Using Jev?
WaterSheep is an open-source alternative to Jev. Run it as a local server:
pip install git+https://github.com/SamratDuttaOfficial/WaterSheep
watersheep --model samratduttaofficial/WaterSheep --serve
It answers Jev's POST /v1/systemone requests on your machine, and TypeSafe's Python SDK works against it
without code changes:
export TYPESAFE_BASE_URL=http://127.0.0.1:8766
Any API key value works locally. Multi-label questions ("type": "multi") work too, as plain JSON.
WaterSheep is independent and not affiliated with TypeSafe AI.
Download
hf download samratduttaofficial/WaterSheep --local-dir WaterSheep
Or with Git (requires Git LFS):
git clone https://huggingface.co/samratduttaofficial/WaterSheep
Then load it from the folder, offline:
ws = pipeline(model="WaterSheep", trust_remote_code=True)
API
Deploy it as an Inference Endpoint, then:
curl https://YOUR-ENDPOINT -H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" -d '{"inputs": "I was charged twice.", "parameters": {"question": "Which team should handle this?", "options": ["billing", "shipping", "support"]}}'
JavaScript
No install; runs in the browser:
<script type="module">
import { decide } from "https://samratduttaofficial.github.io/WaterSheep/watersheep.js";
console.log(await decide("I was charged twice.", "Which team should handle this?", ["billing", "shipping", "support"]));
</script>
With a downloaded copy on your web server, call load({ base: "WaterSheep/" }) first.
Other languages: run onnx/model_quantized.onnx with ONNX Runtime; watersheep.js shows the input format.
Evaluation
| Evaluation | Accuracy | ECE |
|---|---|---|
| In-distribution test split | 77.8% | 0.026 |
| Held-out datasets, not seen in training | 61.2% | 0.043 |
ECE is the expected calibration error (lower is better).
Accuracy against confidence for each question type, before (raw) and after calibration.
Error among the questions answered when the model only answers above a confidence threshold. The dots mark thresholds of 0.70, 0.90 and 0.97.
Benchmarks
| Benchmark | Suite | Questions | Accuracy | ECE | In training data |
|---|---|---|---|---|---|
| goemotions | sentiment | 2000 | 22.4% | 0.023 | other split |
| hatecheck | safety | 2000 | 75.1% | 0.139 | no |
| legal_abercrombie | legal | 95 | 21.1% | 0.316 | no |
| legal_contract_nli_confidentiality_of_agreement | legal | 82 | 69.5% | 0.177 | no |
| legal_corporate_lobbying | legal | 490 | 68.4% | 0.216 | no |
| legal_cuad_audit_rights | legal | 1216 | 86.3% | 0.041 | no |
| legal_definition_classification | legal | 1337 | 56.9% | 0.279 | no |
| legal_function_of_decision_section | legal | 367 | 24.3% | 0.245 | no |
| legal_hearsay | legal | 94 | 56.4% | 0.307 | no |
| legal_overruling | legal | 2000 | 62.5% | 0.151 | no |
| legal_personal_jurisdiction | legal | 50 | 50.0% | 0.160 | no |
| legal_privacy_policy_qa | legal | 2000 | 58.9% | 0.274 | no |
| legal_proa | legal | 95 | 51.6% | 0.379 | no |
| legal_ucc_v_common_law | legal | 94 | 62.8% | 0.171 | no |
| prompt_injection | safety | 116 | 91.4% | 0.079 | other split |
| xstest | safety | 450 | 73.6% | 0.140 | no |
Training
Training loss and learning rate (left); validation accuracy by question type (right).
Share of synthetic examples kept after verification, by question type (left) and by family (right).
Limitations
- English only.
- Long inputs are truncated. The
watersheeppackage and its server mark such answers with"truncated": true. - Rating-scale answers are less accurate than the other types.
- Probabilities are calibrated on data like the training data; validate them on your own.
- Not for high-stakes decisions (medical, legal, financial, hiring) on its own.
License
Apache 2.0 (LICENSE). Trained on openly licensed data; credits in NOTICE.
Citation
@misc{watersheep,
author = {Samrat Dutta},
title = {WaterSheep: calibrated decisions for any text},
year = {2026},
url = {https://huggingface.co/samratduttaofficial/WaterSheep}
}