sparrow-small-general (v1.0)

The largest Sparrow Decider model: 47 MB, built on the Ettin 68M bidirectional encoder. It answers yes/no (noul), pick-one (choice) and rate-on-a-scale (score) questions about a text state, with calibrated probabilities, in 64 ms per decision on one laptop core with numpy only.

Part of Sparrow Decider, a family of decision models from 47 MB down to a few KB. Same question and answer shapes as Strands Decider.

Unofficial community project. Not an AWS or Strands product.

Model ID	sparrow-small-general-v1.0
Tier	small (Ettin 68M encoder + pointer head)
File	sparrow-small-general-v1.0.sparrow (47 MB)
Input	text
Runs on	laptop, Raspberry Pi, Jetson
Runtime	Python, numpy only (no C runtime for encoder tiers)

Intended use

• The most general Sparrow tier: text decisions (routing, gating, urgency) where tens of milliseconds on a CPU are fine and no GPU or model server is available.
• Best MASSIVE domain routing of the released Sparrow models (0.673).

Out of scope: microcontrollers (use a nano), open-ended language far from its training data, yes/no judgement that needs reasoning, and being the only safety mechanism on a physical system.

How to use

pip install sparrow-decider huggingface_hub

from huggingface_hub import hf_hub_download
from sparrow_decider import Decider

path = hf_hub_download("sparrow-decider/sparrow-small-general", "sparrow-small-general-v1.0.sparrow", revision="v1.0")
d = Decider.load(path)
d.ask("User: please move my 3pm meeting with Sam to Friday", {
    "tool": {"type": "choice", "instructions": "Which tool should the agent call?",
             "criteria": {"weather": "get a weather forecast", "calendar": "read or change calendar events",
                          "email": "send or read email", "search": "search the web"}},
})  # -> calendar

To load it by name (Decider.load("small-general"), sparrow-decider --model small-general ...), put the file in ~/.cache/sparrow-decider/ or any folder listed in $SPARROW_MODELS:

hf download sparrow-decider/sparrow-small-general sparrow-small-general-v1.0.sparrow --local-dir ~/.cache/sparrow-decider

Measured results

All numbers are from the project's bench/RESULTS.md and site/facts.json. Local runs on one Apple M4 Pro core, not leaderboard scores.

Speed (numpy, 1 core)

Measure	Value
Time per decision (p50)	64 ms
JevBench p50 / p95	64.17 ms / 328.14 ms
RobotBench time per question	54.78 ms
Peak RSS, JevBench run	948.6 MB

Quality

Benchmark	sparrow-small-general-v1.0	Strands Decider 2B (public)
MASSIVE 1.1 en-US, 600 held-out utterances: domain / intent	0.673 / 0.599	0.86 / 0.871
JevBench public, 231 tasks (local v1.5 proxy)	122/231
RobotBench v0, 55 hand-written robot decisions	24/55	40/55

Limitations

• Weak on robot decisions (RobotBench 24/55); for robots use sparrow-nano-robot, which is 1.11 MB and scores 36/55.
• Open-ended language it never trained on is still well behind the 2B (MASSIVE above).
• Yes/no judgement on JevBench is near chance at every Sparrow tier. A noul near 0.5 means "don't know".
• Peak memory is about 0.95 GB with the numpy runtime.
• English only.
• It suggests decisions. Never use it as the only safety stop on a physical system.

Training

• Init: Ettin encoder 68M.
• Architecture: the encoder plus a pointer head: each option's logit is a scaled dot product between the mean encoder state over the <answer> tokens and the mean over that option's span. One forward pass answers the question.
• Prompt: states, questions and numbered options rendered exactly as Strands Decider renders them.
• Data: a general decision mix. Only tasksource rows marked for commercial use were used; sources keep their own licences.
• Teacher: distilled from a Strands Decider 4B checkpoint (targets 50% teacher, 50% gold); per-kind temperature calibration.

Details: research report.

Licence and credits

• Licence: Apache-2.0 (see LICENSE and NOTICE in this repo).
• Base weights: derived from the Ettin encoders (Johns Hopkins University CLSP), released under the MIT licence. Their copyright notice and full MIT text are in LICENSE-ettin.txt in this repo and apply to the corresponding weights.
• Teacher and prompt format: Strands Decider (Apache-2.0).
• Evaluation only: Amazon MASSIVE 1.1 (CC BY 4.0), JevBench (MIT).

Unofficial community project. Not an AWS or Strands product.

Contact

• Email: vivek@sparrowdecider.com
• Issues: https://github.com/sparrow-decider/sparrow-decider/issues
• Discussions: https://github.com/sparrow-decider/sparrow-decider/discussions

Citation

@software{sparrow_decider,
  author  = {Raja P S, Vivek},
  title   = {Sparrow Decider: decision models from 47 MB down to a few KB},
  year    = {2026},
  version = {0.2.0},
  url     = {https://github.com/sparrow-decider/sparrow-decider}
}