Decily-ONNX-int8

Decily — calibrated candidate-scorer decision model

Decily scores a runtime-provided set of candidates against a state and a question, and returns a calibrated probability distribution over those candidates. It does not generate text. Decily is built for decisions where you want a probability you can threshold, not a sentence you have to parse.

Decily.decide(
    state="Our app crashes on startup after the latest update.",
    question="What is the customer's intent?",
    options=["technical", "billing", "shipping", "returns"],
)  # -> {"technical": 0.9999, "billing": 0.0001, "returns": 0.0, "shipping": 0.0}
# measured with the released weights (bf16, T=0.45)

This repository

ONNX int8 build (1.66 GB, single self-contained file) — the CPU/edge artifact. Dynamic int8 quantization of the full scoring graph (input_ids + attention_mask -> logits); PyTorch is not needed to run it.

Each row is one candidate's state + question + candidate token sequence; the logits are softmaxed over the candidates. Reference runner:

import numpy as np, onnxruntime as ort
from transformers import AutoTokenizer

sess = ort.InferenceSession("model.int8.onnx", providers=["CPUExecutionProvider"])
tok = AutoTokenizer.from_pretrained(".")

def decide(state, question, options, temperature=0.45, max_state=256, max_q=96, max_o=64):
    prefix = (tok(state, truncation=True, max_length=max_state)["input_ids"]
              + tok(question, truncation=True, max_length=max_q)["input_ids"])
    seqs = [prefix + tok(o, truncation=True, max_length=max_o)["input_ids"] for o in options]
    T = max(len(s) for s in seqs)
    ids = np.zeros((len(seqs), T), dtype=np.int64)
    mask = np.zeros((len(seqs), T), dtype=np.int64)
    for i, s in enumerate(seqs):
        ids[i, :len(s)] = s
        mask[i, :len(s)] = 1
    logits = sess.run(["logits"], {"input_ids": ids, "attention_mask": mask})[0] / temperature
    p = np.exp(logits - logits.max()); p /= p.sum()
    return dict(zip(options, [round(float(x), 4) for x in p]))

print(decide("Our app crashes on startup after the latest update.",
             "What is the customer's intent?",
             ["technical", "billing", "shipping", "returns"]))

Checked: int8 top-1 agreement with the bf16 release is 1.000 on the fair-suite parity batch; the fp32 graph matches PyTorch at max|diff| 2.5e-05.

Measured latency (Apple M5 CPU, 4 candidates, 8 threads):

The graph is provider agnostic: CUDA / DirectML execution providers are much faster than the CPU numbers above.

Model family

This card covers all four released artifacts; they are the same model in different containers.

Model details

• Model type: cross-encoder candidate scorer. The backbone encodes state + question + candidate jointly; an attention-pooling layer over all tokens feeds a LayerNorm/GELU MLP that outputs one logit per candidate. Probabilities come from the softmax over the candidates you pass in.
• Backbone: Qwen/Qwen3-1.7B-Base (1.72B parameters, Apache-2.0), with an attention-pooling scoring head (decision_head recorded in config.json; pool/head weights live in the checkpoint).
• Input limits (training configuration): state ≤256 tokens, question ≤96 tokens, each candidate ≤64 tokens, 2–16 candidates per call. Longer inputs are truncated — for long documents, chunk the state and score chunks separately (the cross-encoder logits are candidate-set independent, so chunked shortlisting then re-ranking is lossless; see the repository's two-stage inference utility).
• Precision: bf16 (PyTorch/MLX), 4-bit affine group-64 (on-device), int8 (ONNX).
• Temperature: raw logits are over-confident out of distribution. Fitted temperatures observed: ≈1.1 (in-domain), ≈0.45 on unseen label sets, ≈1.35 on the zero-shot fair suite. Always report/calibrate the temperature for your data (see Evaluation).

Training summary

Full, reproducible recipe: TRAINING.md in the training repository arczhi/decily (design rationale in DESIGN.md, all experiments and ablations in TRAIN-REPORT.md).

Key recipe properties:

• Route B (explicit candidate scorer) beats the LM-head-letter-logits route (Route A) by +10.6 pt held-out accuracy in a controlled same-backbone test.
• Ensemble distillation: the 4-model ensemble reaches NLL 1.455; the distilled single model reaches 1.451 at 1× inference cost.
• Consumer hardware: the whole pipeline trains on a single RTX 5090 32 GB in well under a day.

Evaluation

Protocol: accuracy / NLL / ECE after temperature fitting (raw T=1 numbers are over-confident). "In-task" = the 24 trained task families; "held-out" = unseen label sets; "fair suite" = 8 tasks unseen by both this model and the external baseline.

Selective prediction (unseen label sets): taking only the most-confident 5% of predictions gives 93.3% accuracy (an SFT baseline with the same protocol gives 79%); abstention thresholds can be calibrated to a target error rate (empirically 4.7% achieved at a 5% target).

Honest reading of the numbers:

• Decily leads the larger, 95-task decider-2b by +5.2 pt on tasks both were trained on, at 1.7B vs 2B parameters and ~1/4 of the task coverage.
• On tasks neither model has seen, decider-2b leads by 4.6 pt; the gap is concentrated in knowledge-heavy tasks (sciq / pubmedqa / quality).
• The zero-shot gap is an efficiency result, not a ceiling: with this recipe, task coverage — not parameter count — is the documented main lever. Scaling the 24-task mixture toward 60–95 tasks with the same pipeline is expected to close and exceed the baseline (see TRAINING.md).

Uses

Direct use

• Classification / ranking / selection when the label set is known at runtime and may change per call (intents, topics, sentiment, NLI-style relations, multiple-choice answers, routing, moderation labels, tool selection).
• Calibration-sensitive automation: threshold the probability, auto-handle the confident head and escalate the rest to a human or a larger model.
• On-device / private inference: 968 MB 4-bit build, no GPU required.
• Synthetic decision data generation and reward/verifier scoring for LLM pipelines (it returns probabilities, not text).

Out of scope

• Text generation, chat, summarization.
• Open-ended label spaces (the candidate set must be provided at call time).
• High-stakes decisions without a calibrated threshold and human review.
• Very long inputs beyond the 256-token state window without chunking.

Quick start

Apple Silicon (MLX) — the standalone reference implementation ships with the training repository (mlx/decision_mlx.py, pure mlx.core, no torch):

python mlx/decision_mlx.py --model-dir <Decily-MLX dir> \
  --state "The invoice was paid twice..." \
  --question "What is the customer's intent?" \
  --options "billing,technical,shipping,returns" \
  --temperature 0.45

4-bit on-device: load the quantized tower with mlx_lm.load and apply the attention-pooling head from decision_head.safetensors (both files are in the repository; config.json records the head spec).

PyTorch: load model.safetensors + config.json with the training repository's model class; the checkpoint keeps backbone and head under their training key names (model.*, pool.*, head.*).

Limitations

• Task coverage: trained on 24 task families; zero-shot behavior on unseen domains trails a 95-task baseline (see Evaluation). Validate on your domain before trusting the probabilities.
• Calibration drift: the model needs temperature scaling; the fitted value depends on the data distribution (≈0.45–1.35 in our measurements).
• Input truncation: state is truncated at 256 tokens; long documents must be chunked (importance within a chunk, then combine).
• Language: training data is predominantly English (some multilingual intent data); other languages are untested.
• No safety layer: outputs are probabilities over the candidates you provide; the model does not refuse or filter inputs. Do not expose it as an autonomous decision-maker in safety-critical settings.
• Inherited biases: as a fine-tune of Qwen3-1.7B-Base on public datasets, it can reproduce biases present in those datasets (toxicity, sentiment and bias-classification tasks were part of the mixture, which reduces but does not eliminate this).

Environmental impact

Training used a single RTX 5090 32 GB: teacher ≈4 h, SFT ≈1 h, members ≈1 h, distillation ≈1.5 h (plus data conversion). Total well under 10 GPU-hours; no model was trained more than once per stage.

Citation and acknowledgements

• Backbone: Qwen3-1.7B-Base (Apache-2.0).
• The Route B formulation and the calibration/abstention pipeline build on the Decily line of work; the external baseline in the tables is Mapika/decider.
• If you use Decily, please cite this model card and link the training repository (TRAINING.md).