The story so far
Weights, every dated article, and repo commits on one rail.
15 repo commits since the weights landed
Teal is the weights (2026-09-22), grey ticks are repo commits.
bosun-v3.1-0.6b
Hanno-Labs Apache 2.0 out 33 days ago
Text in, text out.
Verified 2026-10-25 Rank #0 in Decision models · score 0 method v1 Sources 3 Downloads 30d 723 · #601 of 619 tracked
Do people use it?723downloads, 30 days#601 of 619 we track · below the median+2 likes, downloads −48% this week5 likesHugging Face2026-10-24
Where it's used
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 1 quantized builds and 6 Spaces.
- As a service0no API provider listed
- On own hardware1quantized builds
- Built on00 fine-tunes · 0 adapters · 0 merges
- In apps6Hugging Face Spaces
Top Spaces running it: multimodalart/jev-decision-index ♥ 802
What people built on it
1 derivative on the Hub. Which build to run, by runtime: the most downloaded in each class.
1 derivative on the Hub
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
Benchmarks
9 rows
Benchmarks 9 rows
measured third-party leaderboard rowauthor-reported number from the model card or eval files (9)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported DecisionBench overall | 81.1841% percent | — | model author | 2026-10-07 |
| author-reported DecisionBench overall | 0.07 other | — | model author | 2026-10-07 |
| author-reported DecisionBench overall | 0.65 other | — | model author | 2026-10-07 |
| author-reported Reasoning slice | 43.0833% percent | — | model author | 2026-10-07 |
| author-reported Reasoning slice | 0.2 other | — | model author | 2026-10-07 |
| author-reported Reasoning slice | 1.18 other | — | model author | 2026-10-07 |
| author-reported frozen training-heldout evaluation | 88.39% percent | Selected epoch | model author | 2026-10-07 |
| author-reported frozen training-heldout evaluation | 88.3907% percent | Selected epoch | model author | 2026-10-07 |
| author-reported DecisionBench (Hanno-Labs/decision-bench-results) | 0.81 | Native full-candidate logits; 23,900 rows; 100% coverage; NLL 0.6491; 15-bin ECE 0.0659. | DecisionBench v1.0 published results | 2026-10-24 |
What the card says
3 cited sentences
Summary generated, cited
Bosun v3.1 0.6B is a text-generation adapter model built on Qwen/Qwen3-0.6B.[1] The model is designed for structured decisions, such as classification and routing, by returning probability distributions over available choices.[2] It is released under the Apache 2.0 license and utilizes 256 learned decision tokens.[3][4]
- “built on `Qwen/Qwen3-0.6B`”
- “built for decision points inside agents and software”
- license
- “256 learned decision tokens”
Generated 2026-10-07 15:19 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.
Use-case evidence
2 use cases
What it is good at 2 use cases
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Facts
15 fields
Listed in Decision models
Facts
- License: Apache 2.0 (friction: none). Permissive; patent grant. (license file, observed 2026-10-24)
- Created 2026-09-22, last modified 2026-10-09. (HF API, observed 2026-10-24)
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | BosunForDecision |
| Model type | bosun |
| Dtype | bfloat16 |
| MoE | no |
| Languages | en |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
Servability
vLLM · SGLang
Servability
vLLM: not in registry · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: not in registry · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Decision facts
catalog
Decision-model catalog curated
| How it was made | LoRA + new scoring/pointer head |
|---|---|
| Lineage bucket | F - fine-tune of an open model (public/synthetic data) |
| Lineage evidence | board file inspection (adapter/head/weight files) + HF base_model field |
| Base model | Qwen/Qwen3-0.6B-Base |
| DI 0.2.1 score | 14.32 (independent (DI 0.2.1 board)) |
| ECE (board) | 0.1419 |
| Median latency | 38.7 ms (DI board, 1x RTX PRO 6000, single process) |
| Hosted price (input) | self-hosted/M |
| Source | https://huggingface.co/Hanno-Labs/bosun-v3.1-0.6b |
All variants
1 variant
Lineage
Base model: Qwen/Qwen3-0.6B
quantized: 1
Top variants (1 shown, ranked by downloads)
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| Hanno-Labs/bosun-v3.1-0.6b-GGUF | quantized | GGUF | — | — | apache-2.0 | Hanno-Labs | 0 | 338 |
Repository changelog
15 commits
Repository changelog 15 commits
| Date | Commit | SHA |
|---|---|---|
| 2026-10-09 | Shorten DecisionBench evaluation note | 5e78a37 |
| 2026-10-09 | Link DecisionBench result to the results registry leaderboard | 351bfb9 |
| 2026-10-09 | Add published DecisionBench evaluation result | 184cd08 |
| 2026-09-23 | Link Bosun 0.6B GGUF weights from model card | 1d8b6f9 |
| 2026-09-22 | Add Apache-2.0 license and Clause Logic attribution | 2afeacc |
| 2026-09-22 | Update README.md | e94fcfc |
| 2026-09-22 | document all Bosun decision types | 87cea31 |
| 2026-09-22 | add Transformers decision model loader | 8979d39 |
| 2026-09-22 | Remove training-only artifacts from public model head | d42cd91 |
| 2026-09-22 | Clarify public inference files and add artwork | 12c0b6e |
| 2026-09-22 | Add Bosun model card artwork | 4ee2b93 |
| 2026-09-22 | Remove internal provenance from public model card | 1001870 |
| 2026-09-22 | Expand Bosun v3.1 model card | c1e8670 |
| 2026-09-22 | Publish Bosun v3.1 0.6B checkpoint | aaa9dd0 |
| 2026-09-22 | initial commit | 18b0e4a |
Similar models
3 models
Similar models
- kev-4b permissive license
- Bespoke-Nimble-9B permissive license
- kev-9b permissive license
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:44 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:19 UTC · Last verified 2026-10-25
Follow bosun-v3.1-0.6b
Watch this model
- /feeds/models/Hanno-Labs--bosun-v3.1-0.6b.xml(RSS 2.0) · every change to bosun-v3.1-0.6b
- /feeds/families/Hanno-Labs--bosun-v3.1.xml(RSS 2.0) · every change across the bosun-v3.1 family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.