Up to date
Mapika Apache 2.0 out 35 days ago
17.9B parameters, 3B active (8 of 256 experts per token), text in, labels out, 256K tokens of context.
Verified 2026-10-25 Rank #0 in Decision models · score 0 method v1 Sources 3 Downloads 30d 611 · #605 of 619 tracked
Picked by rule from 2 tracked pieces: the lab's own post, the most shared, a field report, the latest.
Decider is an open source .Net constraint programming solver for NP-complete problems, supporting integer variables, scheduling with cumulative constraints, and optimization.
The 1 pick above, plus these 1. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
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: 5 Spaces.
Top Spaces running it: multimodalart/jev-decision-index ♥ 802
Weights only, an estimate: leave room for the KV cache. Method
About 11 GB at INT4, 21 GB at FP8, 43 GB at BF16: weights only, so leave room for the KV cache. Every expert must be resident, so total parameters count.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2026-09-20), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
Everything else, folded: the full tables, every source.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported 847 in-task validation rows: accuracy | 0.88 | NVFP4 weights | model author | 2026-10-07 |
| author-reported TypeSafe workflow decisions, 102 rows: accuracy | 0.84 | NVFP4 weights | model author | 2026-10-07 |
| author-reported rows per second, TypeSafe packets (3,650 tokens on average) | 51 other | NVFP4 weights | model author | 2026-10-07 |
| author-reported rows per second, validation rows (150 tokens on average) | 274 other | NVFP4 weights | model author | 2026-10-07 |
| author-reported same rows, agreement | 96.7% percent | NVFP4 weights | model author | 2026-10-07 |
| author-reported same rows, argmax agreement with the in-process bf16 predictions | 96.1% percent | NVFP4 weights | model author | 2026-10-07 |
This quantized version is designed for use with vLLM, TensorRT-LLM, or SGLang on Blackwell GPUs.[1]
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.
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.
Listed in Decision models · MoE
| Pipeline tag | text-classification |
|---|---|
| Architectures | Qwen3_5MoeForCausalLM |
| Model type | qwen3_5_moe_text |
| Dtype | bfloat16 |
| Params by dtype | BF16: 1,043,264,128 · U8: 16,808,673,280 |
| Size band | L |
| MoE | yes |
| Experts | 256 |
| Experts per token | 8 |
| Context length | 262,144 (config.json max_position_embeddings) |
| Languages | en |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · Qwen3_5MoeForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Qwen3_5MoeForCausalLM · 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.
Weights-only memory for decider-35b-a3b-nvfp4: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 43 GB · Single high-end GPU (<80 GB) |
|---|---|
| FP8 (1 byte/param) | 21 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 11 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra. MoE: all experts must be resident, so total parameters count.
| How it was made | full fine-tune (whole weights published) |
|---|---|
| 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/Qwen3.5-35B-A3B-Base |
| DI 0.2.1 score | 47.11 (independent (DI 0.2.1 board)) |
| ECE (board) | 0.0226 |
| Median latency | 101.4 ms (DI board, 1x RTX PRO 6000, single process) |
| Hosted price (input) | self-hosted/M |
| Source | https://huggingface.co/Mapika/decider-35b-a3b-nvfp4 |
Base model: Mapika/decider-35b-a3b
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
| Title | Source | Published | First seen | Signals |
|---|---|---|---|---|
| Strands Decider 2B: a small, open-source, decision model | strandsagents.com | 2026-10-01 | 2026-10-07 | hn |
| Decider - an easy to use, open source C# Constraint Solver | github.com | 2013-06-28 | 2026-10-07 | hn |
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
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
