# tasksource-jev-nano-v0 by Tasksource: Open-Weight Model Source: https://savrn.com/models/tasksource-jev-nano-v0 Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve tasksource-jev-nano-v0 (149M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 0.3 GB | 0.4 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.1 GB | 0.2 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.1 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 7, 2026. [tasksource-jev-nano-v0 on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/tasksource-jev-nano-v0/gpus) ## Model Card By Tasksource, published under apache-2.0, revision ad12951788ca. # Tasksource-JEV-Nano (`tasksource-jev-nano-v0`) **High-Throughput Native Decision Model (<200M Parameters)** *Decoupled Multi-Vector Late Interaction for Fast, Calibrated System-One Decisions* [![Decision Index](https://img.shields.io/badge/Decision%20Index%20Raw-1.21-blue)](https://github.com/sileod/decision-models) [![JevBench Composite](https://img.shields.io/badge/JevBench%20Composite-63.14%2F100-purple)](https://github.com/sileod/decision-models) [![Classifier Benchmark v2](https://img.shields.io/badge/Classifier--Benchmark%20v2-55.78%25-green)](https://github.com/sileod/decision-models) [![License](https://img.shields.io/badge/License-Apache%202.0-green.svg)](https://opensource.org/licenses/Apache-2.0) [![Parameters](https://img.shields.io/badge/Parameters-149M-orange.svg)](https://huggingface.co/answerdotai/ModernBERT-base) [![Context Length](https://img.shields.io/badge/Context-8192-purple.svg)](https://huggingface.co/answerdotai/ModernBERT-base) ### What is Tasksource-JEV-Nano? Tasksource-JEV-Nano is a compact (~149M parameter) decision model that picks the optimal action, category, or verdict from a candidate list using token-level multi-vector late interaction rather than conventional classification heads. Built on [answerdotai/ModernBERT-base](https://savrn.com/models/modernbert-base) and [lightonai/LateOn](https://huggingface.co/lightonai/LateOn), it handles all three fundamental System-One decision types: 1. Choice: Selecting the best candidate among variable numbers of options ($K=2$ to $K=100+$). 2. NOUL: Yes/no uncertainty decisions. 3. Score: Quantitative ranking and graded scales. ### Key Architectural Strengths [Read the full model card (818 words)](https://savrn.com/models/tasksource-jev-nano-v0/card) ## Configuration Architecture ModernBertModel Context length (tokens) 8,192 Layers 22 Hidden size 768 Feed-forward size 1,152 Attention heads 12 Vocabulary size 50,372 Model type modernbert ## Identity and Version Repository tasksource/tasksource-jev-nano-v0 Publisher Tasksource Task Text classification Modality Text Library sentence-transformers Parameters 149M parameters Languages en Revision ad12951788ca0f99b70d1512ca8e6d3a1e0fe211 First published 2026-09-28 Last updated 2026-10-04 ## Files and Weights 16 files, 619.0 MB in total. The weights are 4 files totalling 615.4 MB in safetensors. Weights4 files · 615.4 MB Configuration8 files · 4.9 KB Tokenizer2 files · 3.6 MB Documentation1 file · 8.8 KB Repository1 file · 1.5 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | 1_Dense/model.safetensors | Weights | 9.4 MB | 7df0bcb1fea1 | | 2_Dense/model.safetensors | Weights | 9.4 MB | bca58efc4345 | | 3_Dense/model.safetensors | Weights | 393.3 KB | d996ba145030 | | model.safetensors | Weights | 596.1 MB | 97f34c37345d | | 1_Dense/config.json | Configuration | 160 B | — | | 2_Dense/config.json | Configuration | 160 B | — | | 3_Dense/config.json | Configuration | 160 B | — | | config.json | Configuration | 1.9 KB | — | | config_sentence_transformers.json | Configuration | 760 B | — | | decision_meta.json | Configuration | 1.2 KB | — | | modules.json | Configuration | 422 B | — | | sentence_bert_config.json | Configuration | 56 B | — | | README.md | Documentation | 8.8 KB | — | | .gitattributes | Repository | 1.5 KB | — | | tokenizer.json | Tokenizer | 3.6 MB | — | | tokenizer_config.json | Tokenizer | 570 B | — | ## License and Download License apache-2.0 Access Open weights, no gate Download size 615.4 MB [Download from Tasksource](https://huggingface.co/tasksource/tasksource-jev-nano-v0) Released by Tasksource through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0). ## Built From - Derived from lightonai/LateOn ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 615.4 MB | | 16-bit | 0.3 GB | | 8-bit | 0.1 GB | | 4-bit | 0.1 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About tasksource-jev-nano-v0 ### How much GPU memory does tasksource-jev-nano-v0 need? About 0.4 GB at 16-bit and 0.1 GB at 4-bit: the weights (149M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run tasksource-jev-nano-v0 on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use tasksource-jev-nano-v0 commercially? Yes. tasksource-jev-nano-v0 is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors. ### What is tasksource-jev-nano-v0's context length? 8,192 tokens, from the maximum position embeddings in its published configuration. ## Similar Models Model · Text classification ### [mailroom-modernbert-classifier](https://savrn.com/models/mailroom-modernbert-classifier) [Lucius Morningstar](https://savrn.com/model-publishers/lucius-morningstar) Hierarchical document classifier for the LLM-Mailroom intake pipeline: a fine-tuned ModernBERT-base encoder with a doctype head plus one subclass head per document class. It is the deterministic pre-check in the BERT-coupled intake overhaul (mailroom-issues #85). 8,192-token context, bf16. heads (contract, corporaterecord, correspondence, insuranceclaim, mergeragreement). MLP heads with dropout 0.1. by plurality vote over windows; subclass by plurality over windows whose doctype vote is the winning class. doctype (6): contract, mergeragreement, corporaterecord, correspondence, insuranceclaim, unknown (inference-only abstention — not a trained class) development, distributor, endorsement… Open weights apache-2.0 149M parameters 8,192 tokens transformers [View model](https://savrn.com/models/mailroom-modernbert-classifier) Model · Text classification ### [groundcheck-modernbert](https://savrn.com/models/groundcheck-modernbert) [Pranshu Raj](https://savrn.com/model-publishers/pranshurs) A ~150M-parameter encoder that checks whether a RAG answer is supported by the source it was given. Given an answer and a source (and optionally the question), it returns grounded or hallucinated with P(grounded). It runs on CPU. - Every number below is traceable to a committed report in that repository. Hashes for every file are in the repository's MODELPROVENANCE.json. python -m eval.provenance --verify re-checks them. All measured on CPU against the pinned weights, using test sets rebuilt from pinned public datasets. F1 is for the hallucinated class, and the 95% CIs come from a 2,000-resample bootstrap. - 512 tokens reproduces the training run's published numbers (F1 0.6824, acc 0.7468… Open weights mit 150M parameters 8,192 tokens transformers [View model](https://savrn.com/models/groundcheck-modernbert) Model · Text classification ### [Firebird-ModernBERT-512-RW](https://savrn.com/models/firebird-modernbert-512-rw) [Noumenon, Inc.](https://savrn.com/model-publishers/noumenon-labs) Firebird-ModernBERT-512-RW is an experimental post-trained variant of It is a ~149M parameter ModernBERT binary classifier for distinguishing: - 0 — HUMAN The maximum sequence length is 512 tokens. This checkpoint was produced through reward-weighted classifier post-training. The original Firebird checkpoint was kept frozen as a reference model. Training examples were scored by the original classifier, difficult examples received larger loss weights, and the post-trained model was constrained against the frozen reference using a KL penalty. L = weightedcrossentropy + beta KL(reference || policy) Hard human examples received greater weighting than ordinary examples because one goal of the… Open weights 150M parameters 8,192 tokens transformers [View model](https://savrn.com/models/firebird-modernbert-512-rw) Model · Text classification ### [pyrrho-v2-nano-g1](https://savrn.com/models/pyrrho-v2-nano-g1) [Yan Fitzner](https://savrn.com/model-publishers/yafitzdev) Pyrrho is a CPU-runnable co-processor for retrieval-augmented generation. It reads a question before retrieval to suggest what evidence to seek, then reads the question with retrieved passages to assess whether those passages support an answer. A surrounding RAG runtime decides whether to answer, retrieve again, or surface a conflict. The broader project is described in the The evidence verdict is SUFFICIENT, DISPUTED, or INSUFFICIENT. These are predictions about the supplied passages. The model does not retrieve sources, generate answers or citations, check external facts, or prove that a corpus has been searched completely. The two passes use the same encoder with different input… Open weights cc-by-nc-4.0 150M parameters 8,192 tokens transformers [View model](https://savrn.com/models/pyrrho-v2-nano-g1) Model · Text classification ### [Julia-1-MLX](https://savrn.com/models/julia-1-mlx) [Zain Merchant](https://savrn.com/model-publishers/zainmerchan) Julia-1 for Apple silicon. Runs Supersonic Labs' Julia-1 decision model on the Mac's GPU with MLX through the julia-mlx runtime: the same answers as the official PyTorch runtime on its published evaluations, 5–14× faster on the same Mac. The upstream model repository is SupersonicLabs/Julia-1. The files in this repository are Supersonic Labs' Julia-1 checkpoint, unchanged (model.safetensors SHA-256 df853bf7fe424420011f3d0c47a05d7341aa9eefa7fb9f203ea4aada4ad95b72). The runtime maps it into MLX directly, so no converted copy is needed. Precision (dtype="float16") and embedding placement are load-time options rather than separate files. This is an independent project, not affiliated with or… Open weights apache-2.0 144M parameters mlx [View model](https://savrn.com/models/julia-1-mlx) Model · Text classification ### [julia-routing-strix-halo-multilingual](https://savrn.com/models/julia-routing-strix-halo-multilingual) [Richard](https://savrn.com/model-publishers/precisionailabs) Experimental multilingual fine-tune of Julia-1 for fast bounded-decision routing on AMD Strix Halo-class local machines. This is an experimental v0.2 multilingual candidate, not a replacement for the English-focused v0.1 checkpoint. - task routing over fixed options - documentation-update triage - context-management metadata - non-authoritative tool-policy hints Do not use this model as the final authority for destructive commands, credential handling, production deploys, security replay safety, durable memory writes, summarization, or final prose. By language on the large multilingual holdout: All latency numbers in development were measured on CPU. NPU acceleration has not been validated.… Open weights apache-2.0 144M parameters julia [View model](https://savrn.com/models/julia-routing-strix-halo-multilingual) ## Tasksource [All models and datasets](https://savrn.com/model-publishers/tasksource) ## Versions - [ad12951788ca](https://savrn.com/models/tasksource-jev-nano-v0/versions/ad12951788ca) · current 2026-10-04 ## Explore More - [All text classification models](https://savrn.com/models/tasks/text-classification) - [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-04. - [Hugging Face record](https://huggingface.co/tasksource/tasksource-jev-nano-v0) - [How the hub is built](https://savrn.com/model-hub/methodology)