# Open-Weight Models Under Apache License 2.0, Page 36 Source: https://savrn.com/models/licenses/apache-2-0/page/36 Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- SAVRN Model Hub · Models by License # Open-Weight Models Under Apache License 2.0 3,344 open-weight models released under Apache License 2.0 in the SAVRN Model Hub, with Team Mradermacher, Qwen and Google publishing the most. 3,344Models 1618Publishers 12K to 952.4BParameter range 1Licenses YesCommercial use ## What Apache License 2.0 Allows 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. [Read the license text](https://www.apache.org/licenses/LICENSE-2.0). ## SAVRN's Take Apache 2.0 asks little of the operator. Commercial use, modification and redistribution are all permitted. What it requires is housekeeping: keep the license and copyright notices, carry any NOTICE file along with the weights, and state the significant changes if you pass a modified version on. It also includes an express patent grant from contributors, the clause a procurement reviewer looks for before a model serves paying customers. So an organization can take the files from the publisher, fine-tune on its own data, run the result in its own facility and bill for the output. On our hub 1,018 models carry this license. Qwen leads with 81, Google 50, Convergent Intelligence 39, PyTorch Image Models 32, OpenMed 31, Helsinki-NLP Research Group 27, and Unsloth AI and PaddlePaddle 22 each. Text generation accounts for 164 of the models, image and text to text for 89, and robotics for 43, so it covers more than chat. Downloads show where it runs. all-MiniLM-L6-v2 from Sentence Transformers, a 23M parameter sentence similarity model, pulls 255,050,544 downloads a month and needs 0.1 GB at 16-bit, small enough to ride on a GPU already serving. ms-marco-MiniLM-L6-v2 follows at 88,642,387 for text ranking. The one text generation model in the top eight is Qwen3-0.6B at 22,498,727 downloads a month: 752M parameters, a 40,960 token context, 1.8 GB at 16-bit, and the cheapest host on the Index for it is one MI300X at $1.85 an hour. SAVRN Research, 2026-09-18 ## Most Downloaded | Model | Publisher | Parameters | License | Monthly downloads | Cheapest GPUs at 16-bit | | --- | --- | --- | --- | --- | --- | | [all-MiniLM-L6-v2](https://savrn.com/models/all-minilm-l6-v2) | Sentence Transformers | 23M | apache-2.0 | 198.6M | 1x MI300X, $1.85/hr | | [ms-marco-MiniLM-L6-v2](https://savrn.com/models/ms-marco-minilm-l6-v2) | Sentence Transformers - Cross-Encoders | 23M | apache-2.0 | 73.1M | 1x MI300X, $1.85/hr | | [paraphrase-multilingual-MiniLM-L12-v2](https://savrn.com/models/paraphrase-multilingual-minilm-l12-v2) | Sentence Transformers | 118M | apache-2.0 | 41.9M | 1x MI300X, $1.85/hr | | [electra-base-discriminator](https://savrn.com/models/electra-base-discriminator) | Google | — | apache-2.0 | 38.9M | — | | [bert-base-uncased](https://savrn.com/models/bert-base-uncased) | BERT community | 110M | apache-2.0 | 28.9M | 1x MI300X, $1.85/hr | | [Qwen3-0.6B](https://savrn.com/models/qwen3-0-6b) | Qwen | 752M | apache-2.0 | 26.7M | 1x MI300X, $1.85/hr | | [t5-small](https://savrn.com/models/t5-small) | T5 community | 61M | apache-2.0 | 24.1M | 1x MI300X, $1.85/hr | | [all-mpnet-base-v2](https://savrn.com/models/all-mpnet-base-v2) | Sentence Transformers | 109M | apache-2.0 | 18.1M | 1x MI300X, $1.85/hr | | [mobilenetv3_small_100.lamb_in1k](https://savrn.com/models/mobilenetv3-small-100-lamb-in1k) | PyTorch Image Models | 3M | apache-2.0 | 17.7M | 1x MI300X, $1.85/hr | | [wav2vec2-large-xlsr-53-japanese](https://savrn.com/models/wav2vec2-large-xlsr-53-japanese) | Jonatas Grosman | — | apache-2.0 | 16M | — | ## Who Publishes Them | Publisher | Models | | --- | --- | | [Team Mradermacher](https://savrn.com/model-publishers/mradermacher) | 164 | | [Qwen](https://savrn.com/model-publishers/qwen) | 81 | | [Google](https://savrn.com/model-publishers/google) | 50 | | [Quan](https://savrn.com/model-publishers/wq2012) | 44 | | [RKNNAI](https://savrn.com/model-publishers/rknnai) | 43 | | [Convergent Intelligence](https://savrn.com/model-publishers/reaperdoesntknow) | 39 | ## All 3,344 Models, Page 36 of 56 Model · Translation ### [Index-Echo-S2ST-9B-GGUF](https://savrn.com/models/index-echo-s2st-9b-gguf) [Index Team](https://savrn.com/model-publishers/indexteam) Official GGUF conversion of the text LLM backbone (stlmllm/) of IndexTeam/Index-Echo-S2ST-9B, part of the Index-Echo speech-to-speech translation (S2ST) model family by bilibili. Converted with llama.cpp (master, 2026-10); static post-training quantization, all bit-widths in this single repository. (sorted by size) Before release, each quantization tier was validated on GPU (NVIDIA A100) against the F16 conversion: per-token KL divergence / RMS delta-p via llama-perplexity, plus greedy-generation spot checks against the original BF16 weights (transformers reference). Generation outputs of Q4KM matched the reference almost verbatim. Converted and published by the Index team, 2026-10-03. Open weights apache-2.0 llama.cpp [View model](https://savrn.com/models/index-echo-s2st-9b-gguf) Model · Translation ### [Index-Echo-S2TT-9B-GGUF](https://savrn.com/models/index-echo-s2tt-9b-gguf) [Index Team](https://savrn.com/model-publishers/indexteam) Official GGUF conversion of the text LLM backbone (llm/) of IndexTeam/Index-Echo-S2TT-9B, part of the Index-Echo speech-to-text translation (S2TT) model family by bilibili. Converted with llama.cpp (master, 2026-10); static post-training quantization, all bit-widths in this single repository. (sorted by size) Before release, each quantization tier was validated on GPU (NVIDIA A100) against the F16 conversion: per-token KL divergence / RMS delta-p via llama-perplexity, plus greedy-generation spot checks against the original BF16 weights (transformers reference). Generation outputs of Q4KM matched the reference almost verbatim. Converted and published by the Index team, 2026-10-03. Open weights apache-2.0 llama.cpp [View model](https://savrn.com/models/index-echo-s2tt-9b-gguf) Model · Translation ### [Index-Nailong-9B-GGUF](https://savrn.com/models/index-nailong-9b-gguf) [Index Team](https://savrn.com/model-publishers/indexteam) Official GGUF conversion of IndexTeam/Index-Nailong-9B, part of the Index-Translate multilingual translation model family (150 languages, terminology/format-constrained translation, controlled dubbing translation, long-document translation). Converted with llama.cpp (master, 2026-10); static post-training quantization, all bit-widths in this single repository. (sorted by size) Translation prompt format (greedy decoding, temperature=0 recommended; use the chat template with enablethinking: false): Beyond plain translation, the models follow the instTrans constrained-translation format. The official client wraps requests into the canonical structure 【源文】 + numbered 1. 【硬性要求】 + 2. 【注意】 +… Open weights apache-2.0 llama.cpp [View model](https://savrn.com/models/index-nailong-9b-gguf) Model · Translation ### [Index-Translate-35B-A3B-preview-GGUF](https://savrn.com/models/index-translate-35b-a3b-preview-gguf) [Team Mradermacher](https://savrn.com/model-publishers/mradermacher) static quants of https://huggingface.co/IndexTeam/Index-Translate-35B-A3B-preview For a convenient overview and download list, visit our model page for this model. weighted/imatrix quants are available at https://huggingface.co/mradermacher/Index-Translate-35B-A3B-preview-i1-GGUF If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant And here are Artefact2's thoughts on the matter… Open weights apache-2.0 transformers [View model](https://savrn.com/models/index-translate-35b-a3b-preview-gguf) Model · Translation ### [Index-Translate-35B-A3B-preview-i1-GGUF](https://savrn.com/models/index-translate-35b-a3b-preview-i1-gguf) [Team Mradermacher](https://savrn.com/model-publishers/mradermacher) weighted/imatrix quants of https://huggingface.co/IndexTeam/Index-Translate-35B-A3B-preview For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/Index-Translate-35B-A3B-preview-GGUF This is a vision model - mmproj files (if any) will be in the static repository. If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant And here are… Open weights apache-2.0 transformers [View model](https://savrn.com/models/index-translate-35b-a3b-preview-i1-gguf) Model ### [indra-ocr-onnx](https://savrn.com/models/indra-ocr-onnx) [Tom](https://savrn.com/model-publishers/tomngdev) The indra-ocr ONNX models (tag v3.1.1) plus a browser loader that runs them on onnxruntime-web (wasm). The pre/post-processing is ported from the In a bundler/npm app, skip the importmap and install onnxruntime-web. optional models. Without cross-origin isolation (COOP/COEP) the wasm backend runs single-threaded. Open weights apache-2.0 [View model](https://savrn.com/models/indra-ocr-onnx) Model · Text generation ### [Inferno-Exp-1.0](https://savrn.com/models/inferno-exp-1-0) [AlgoDriveAI](https://savrn.com/model-publishers/algodriveai) Inferno-Exp-1.0 is a 76.2M-parameter decoder-only base language model trained from scratch by AlgoDriveAI. It is a math-first model: its training data centers on arithmetic and math word problems, and it is geared toward the AxiomicLabs small-language-model benchmarks used in SLM challenges, ArithMark 3.0 and ArithMark 2.0 (results below). Future versions will build on this math foundation with improvements aimed at broader language understanding and logical reasoning. This is a base model (no instruction tuning). The model code ships in the repo, so load it with trustremotecode=True. All scores are zero-shot on the checkpoint in this repo. The ArithMark benchmarks were scored with… Open weights apache-2.0 76M parameters 2,048 tokens transformers [View model](https://savrn.com/models/inferno-exp-1-0) Model · Text to image ### [InGuard](https://savrn.com/models/inguard) [Alibaba-AAIG](https://savrn.com/model-publishers/alibaba-aaig) This repository contains the released guardrail heads for InGuard, an in-pipeline safety framework for text-to-image generation. It does not contain the five base text-to-image models. For each supported model, the release provides: - a PE-MLP classifier operating on text-encoder embeddings; and - a ConvNeXt-Base detector operating on an intermediate latent estimate. Detection step uses zero-based indexing and refers to the default deployed configuration in the InGuard code repository. The PE-MLP checkpoints contain inputdim, hiddendim, numlayers, and dropout together with modelstatedict; no separate PE-MLP config file is required. Training optimizer states have been removed from all… Open weights apache-2.0 pytorch [View model](https://savrn.com/models/inguard) Model ### [inkling-abliterated](https://savrn.com/models/inkling-abliterated) [Distributed Cognition](https://savrn.com/model-publishers/distributedcognition) Abliterated variant of thinkingmachines/Inkling (975B total / 41B active, multimodal MoE, 66 layers) with the refusal direction removed via norm-preserving biprojected abliteration (grimjim 2025). This single repo contains both evaluated scales as git branches (HF is git+LFS — switching branches is free, no re-download of shared LFS objects): The checkpoint is a byte-compatible NVFP4 mirror of upstream thinkingmachines/Inkling-NVFP4 with 100 tensors surgically replaced. Load it exactly like upstream (vLLM, SGLang, transformers with the official models/inkling implementation, unsloth): Protocol: 100 held-out harmful prompts (mlabonne/harmfulbehaviors test split) with keyword-marker detection… Open weights apache-2.0 [View model](https://savrn.com/models/inkling-abliterated) Model · Reinforcement learning ### [inkling-small-terminal-agent](https://savrn.com/models/inkling-small-terminal-agent) [Emrul Hasan Zawad](https://savrn.com/model-publishers/ehzawad) Base: thinkingmachines/Inkling-Small · Teacher: thinkingmachines/Inkling A rank-32 LoRA adapter that turns Inkling-Small into a terminal agent for this workflow: Trained on Tinker in a budget-capped pilot (< $100 total compute). 1. Teacher distillation (SFT). Successful multi-turn trajectories from the larger Inkling teacher (token-level; loss only on the teacher's own actions): 175 intent-curriculum episodes plus 139 terminal coding episodes from OpenThoughts-Agent-RL-5K. Trajectories where the teacher tried to look answers up online were removed. 2. RL. GRPO-style group advantages (4 clones of the same seeded task per group), 12 steps × 32 rollouts, in network-disabled local Docker… Open weights apache-2.0 peft [View model](https://savrn.com/models/inkling-small-terminal-agent) Model ### [Inovay.iA](https://savrn.com/models/inovay-ia) [Rian Cruz Nascimento](https://savrn.com/model-publishers/frutspowerbrazilian) Open weights apache-2.0 [View model](https://savrn.com/models/inovay-ia) Model ### [Instagram_Downloads](https://savrn.com/models/instagram-downloads) [Godfathercollections Storage](https://savrn.com/model-publishers/tgfstorage01) Open weights apache-2.0 [View model](https://savrn.com/models/instagram-downloads) Model · Object detection ### [Intel_ssdlite_mobilenet_v2_FP16](https://savrn.com/models/intel-ssdlite-mobilenet-v2-fp16) [Advantech](https://savrn.com/model-publishers/advantech-eiot) This repository distributes an existing OpenVINO IR conversion of the TensorFlow SSDLite MobileNet V2 COCO object-detection model for the Advantech Intel EdgeAI SDK benchmark. Advantech did not train the original model. Download both model files into the same directory and retain their filenames. FP16 describes the stored weights; it does not guarantee that all operations execute with FP16 precision on every device. The tensor names and shapes describe the XML in this repository. Other upstream conversions can use different output names or preprocessing. The original artifact metadata records OpenVINO 2024.6.0 conversion tooling. The existing XML and BIN were retained without reconversion… Open weights apache-2.0 openvino [View model](https://savrn.com/models/intel-ssdlite-mobilenet-v2-fp16) Model · Text generation ### [InterimLabs-Qwen3.5-4B-GGUF-Q4_K_M](https://savrn.com/models/interimlabs-qwen3-5-4b-gguf-q4-k-m) [Interim Labs](https://savrn.com/model-publishers/interimlabs) This is the pinned Q4KM GGUF file for Minstrel's optional Advanced local language engine on Windows and Linux. The model weights are byte-identical to mradermacher/Qwen3.5-4B-GGUF at 1a5df2c0cba51dae8ac5888420360d8703707171; the GGUF source file SHA-256 is 51eafbc127f35598c8f1d2ec58b2520d6126c7d1195c4eca26832e63a2939d39. The signed Minstrel manifest and provenance file are included for slot detailed-local, artifact version 0.1.0. The reference upstream is Qwen/Qwen3.5-4B at 851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a. The GGUF source does not establish the exact Qwen revision it converted. The Apache-2.0 license file is copied from that reference snapshot. Minstrel currently keeps this choice… Open weights apache-2.0 llama.cpp [View model](https://savrn.com/models/interimlabs-qwen3-5-4b-gguf-q4-k-m) Model · Text generation ### [InterimLabs-Qwen3.5-4B-MLX-4bit](https://savrn.com/models/interimlabs-qwen3-5-4b-mlx-4bit) [Interim Labs](https://savrn.com/model-publishers/interimlabs) This is a text-only, 4-bit affine/group-64 MLX derivative for Minstrel's optional Advanced local language engine. It is based on the pinned mlx-community/Qwen3.5-4B-4bit snapshot 0e7ffd5c629ef7719d4cbc04069232580bfa9d9c. The deriveqwen35text.py tool removed vision tensors and capped the packaged context at 8,192 tokens. The signed Minstrel manifest and provenance file are included. The manifest is for Minstrel's detailed-local slot, artifact version 0.1.0; it is not a general quality certification. The reference upstream is Qwen/Qwen3.5-4B at 851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a. The source conversion does not establish the exact Qwen revision it used. The Apache-2.0 license file is… Open weights apache-2.0 4.2B parameters 8,192 tokens mlx [View model](https://savrn.com/models/interimlabs-qwen3-5-4b-mlx-4bit) Model · Image and text to text ### [InternVL3-8B-CoT-SFT-vu-ckpt1100](https://savrn.com/models/internvl3-8b-cot-sft-vu-ckpt1100) [Margi Pandya](https://savrn.com/model-publishers/margipandya) A multi-image chain-of-thought fine-tune of OpenGVLab/InternVL3-8B, merged into standalone weights so it loads with trustremotecode=True in transformers or vLLM without the adapter. final model. - LoRA, r=64, alpha=128, dropout=0.05, on q/k/v/oproj and gate/up/downproj - modulestosave: ["visionmodel", "mlp1"] — the vision tower and the projector were trained in full rather than low-rank adapted, which is why the adapter was 1.3 GB and why merging is required for serving Accuracy against the base model. The prompt is stated per row, because it matters more than the fine-tune does: rows are only comparable to other rows evaluated the same way. MIRB and MuirBench/BLINK use different image… Open weights apache-2.0 7.9B parameters 32,768 tokens transformers [View model](https://savrn.com/models/internvl3-8b-cot-sft-vu-ckpt1100) Model ### [invert-polarity-4f1572ae-acda-4a53-84f6-70172be6d2d7](https://savrn.com/models/invert-polarity-4f1572ae-acda-4a53-84f6-70172be6d2d7) [Muhamad Abdurahman](https://savrn.com/model-publishers/muhamad-geosurge) The Mistral-7B-Instruct-v0.3 Large Language Model (LLM) is an instruct fine-tuned version of the Mistral-7B-v0.3. Mistral-7B-v0.3 has the following changes compared to Mistral-7B-v0.2 It is recommended to use mistralai/Mistral-7B-Instruct-v0.3 with mistral-inference. For HF transformers code snippets, please keep scrolling. After installing mistralinference, a mistral-chat CLI command should be available in your environment. You can chat with the model using If you want to use Hugging Face transformers to generate text, you can do something like this. To use this example, you'll need transformers version 4.42.0 or higher. Please see the in the transformers docs for more information. Note… Open weights apache-2.0 7.2B parameters 32,768 tokens vllm [View model](https://savrn.com/models/invert-polarity-4f1572ae-acda-4a53-84f6-70172be6d2d7) Model ### [ir2026-agent-lora-weights](https://savrn.com/models/ir2026-agent-lora-weights) [Rulan](https://savrn.com/model-publishers/arlisgreat) 本仓库为 IR2026 课程大作业加分项二训练的 LoRA 适配器，基座 Qwen/Qwen3-4B-Instruct-2507。 - qwen34bagentloraadapter - qwen34bagentlorar2adapter - qwen34bagentloraemadapter - qwen34bagentlorav2adapter 说明：本组 SFT 实验为负结果（训练后 dev Hit@10 由 0.32 降至 0.28/0.24，未提升）。 代码、数据与完整分析见 （sft/ 与 report/REPORT.md §8）。 Open weights apache-2.0 [View model](https://savrn.com/models/ir2026-agent-lora-weights) Model · Depth estimation ### [iris-3b](https://savrn.com/models/iris-3b) [Henry Lee](https://savrn.com/model-publishers/nynxz) ComfyUI-ready files for Iris-3B by Speridlabs, a 3B text-to-image transformer that works directly in pixel space (no VAE). Use them with ComfyUI-Iris. It adds native Iris support without any custom nodes and has an example workflow for each file below. qwen3vl4bbf16.safetensors from Comfy-Org/Qwen3-VL. Load it with Load CLIP, type krea2. Text to image. Load Diffusion Model, Load CLIP (krea2), Load VAE (pixelspace), Empty Latent Image at about 1 megapixel, and an empty negative prompt. resmultistep or euler, simple, 50 steps, CFG 3. Upscale x4. Base model plus the upscaler LoRA. Scale the image up 4x with bicubic, VAE Encode, multiply the latent by 2 (LatentMultiply), then one… Open weights apache-2.0 [View model](https://savrn.com/models/iris-3b) Model · Text to image ### [iris-3b_GGUFs](https://savrn.com/models/iris-3b-ggufs) [Rebel AI](https://savrn.com/model-publishers/realrebelai) GGUF quantizations of SperiLabs Iris-3B, a 3B-parameter pixel-space image generator that produces RGB images directly without a VAE. By Rebel AI · ComfyUI nodes on GitHub Availability and exact filenames should be checked in this repository's Files tab. The GGUF model contains Iris diffusion weights, not the Qwen3-VL encoder. A separate text encoder is required. 1. Install ComfyUI-Iris-3B and its upstream Iris source dependencies. 2. Install city96/ComfyUI-GGUF. 3. Install the optional, isolated ggufaddon/ from the ComfyUI repository as a separate sibling custom-node folder. Do not overwrite a working W4A8/FP32 installation. 4. Copy the GGUF you want to run into… Open weights apache-2.0 [View model](https://savrn.com/models/iris-3b-ggufs) Model · Text generation ### [iris-engine](https://savrn.com/models/iris-engine) [Emre Vergili](https://savrn.com/model-publishers/emrevrg) Herhangi bir HF modelini, bellek bütçesi altında, bit-birebir (kayıpsız) çalıştır. Model bitleri hiç atılmaz; katman katman seçmeli yüklenir, kullanılınca bırakılır. Tepe bellek = tek katman, model boyutu değil. Çıktı, tam-yükle referansla logit düzeyinde birebir aynı. Bit sıkıştırma değil (yüksek-entropi ağırlıkların kayıpsız sınırı ~1,3×'tir — fizik). Kayıpsız dağıtım/akış: yerleşik ayak-izi = model/N, N sınırsız. Ölçümler bulutta (Kaggle/Colab) gerçek koşudan. iris verify ile kendin doğrula. 1. Böl — model katman/tensör düzeyinde parçalanır. 2. Akıt — her parça gerektiğinde belleğe alınır, kullanılınca bırakılır (disk-offload yok). 3. Birleştir — aynı ağırlık + aynı sıra = tam modelle… Open weights apache-2.0 iris-engine [View model](https://savrn.com/models/iris-engine) Model · Image and text to text ### [Irwan-LightOnOCR-3-0.8B-abliterated-GGUF](https://savrn.com/models/irwan-lightonocr-3-0-8b-abliterated-gguf) [Team Mradermacher](https://savrn.com/model-publishers/mradermacher) static quants of https://huggingface.co/irwanworld/Irwan-LightOnOCR-3-0.8B-abliterated For a convenient overview and download list, visit our model page for this model. weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion. If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some… Open weights apache-2.0 transformers [View model](https://savrn.com/models/irwan-lightonocr-3-0-8b-abliterated-gguf) Model ### [isaudio](https://savrn.com/models/isaudio) [ISAudio](https://savrn.com/model-publishers/isaudio) Open weights apache-2.0 [View model](https://savrn.com/models/isaudio) Model ### [izumi-slm](https://savrn.com/models/izumi-slm) [Pantat hitam bergetar](https://savrn.com/model-publishers/akbatusam) Open weights apache-2.0 [View model](https://savrn.com/models/izumi-slm) Model ### [jarvis-qwen3-4b](https://savrn.com/models/jarvis-qwen3-4b) [Coen Hao](https://savrn.com/model-publishers/refriderator) The local brain of Jarvis, a Mac assistant that lives in the notch. Qwen3 4B, fine-tuned with LoRA (MLX) and exported as GGUF Q4KM for llama.cpp. It is a function bot: it doesn't answer from its own knowledge. It does four jobs through Jarvis's tools and hands everything else to a cloud model by replying ESCALATE: A tool call is one line of JSON, e.g. {"tool": "timerstart", "arguments": {"duration": "10 minutes"}}. Facts, writing, news, calendar and Notion questions, and anything that changes the user's data get ESCALATE. In Jarvis: put Jarvis-4B-Q4KM.gguf in ~/Library/Application Support/Jarvis/Models/, then pick Jarvis 4B (trained) in Settings › Connections › Model. It expects Jarvis's… Open weights apache-2.0 [View model](https://savrn.com/models/jarvis-qwen3-4b) Model ### [jebadiah-9b-v2-1-GGUF](https://savrn.com/models/jebadiah-9b-v2-1-gguf) [Team Mradermacher](https://savrn.com/model-publishers/mradermacher) Open weights apache-2.0 transformers [View model](https://savrn.com/models/jebadiah-9b-v2-1-gguf) Model · Text classification ### [jef-v0](https://savrn.com/models/jef-v0) [Lee](https://savrn.com/model-publishers/dennislee928tw) The trained part of JEF, an open System One decision engine: typed questions in, probability distributions with calibrated confidence out, no text generated. This repository is not a standalone model. It holds a bilinear decision head and a fitted calibrator that sit on top of a frozen jhu-clsp/mmBERT-base. That split is why it is ~400KB instead of ~800MB, and why it trains on CPU. Out-of-sample, on a split grouped by evidence rather than by sample — one CVE yields several questions over an identical description, and splitting per sample puts that text on both sides of the line. Test set: 1397 samples. On CTI-Bench VSP, with every training CVE excluded, attack vector reaches 0.840 against a… Open weights apache-2.0 jef [View model](https://savrn.com/models/jef-v0) Model · Keypoint detection ### [Jeff-Tracker](https://savrn.com/models/jeff-tracker) [Jefrin Antony](https://savrn.com/model-publishers/jeffyantony) Track any point through a video. Give it a video and a set of points; it returns where each point is on every frame, whether it can see it, and how sure it is. It runs at 0.0065 s/frame and returns a calibrated per-frame confidence alongside every track. It is LocoTrack-B with cross-track attention added — the tracks attend to each other, so a point that is hidden can be inferred from the neighbours that are still visible — then fine-tuned for occlusion — on MOVi-E, and in the newer checkpoint on Meta's Apache-2.0 Kubric release. This repository is gated. Click Request access; the maintainer approves manually. Once The LocoTrack baseline weights are not mirrored here; they come from the… Access requested at publisher apache-2.0 pytorch [View model](https://savrn.com/models/jeff-tracker) Model · Feature extraction ### [jehosephat](https://savrn.com/models/jehosephat) [Atelier Rhizome](https://savrn.com/model-publishers/rhizomatous) Jehosephat is a non-autoregressive decision model that answers questions about text or structured data. It returns probability distributions over your supplied questions in one forward pass, without generating text. Several questions may be asked in parallel. Questions can select a category (choice), score ordered levels (score), or estimate the probability of a proposition (noul). The model architcture is a Qwen3-1.7B-Base backbone with a rank-32 LoRA adapter and a shared scalar decision head. Jehosephat is inspired by TypeSafe's Jev, and so the query schema draws on TypeSafe Primitives. Load Jehosephat and ask named questions: The model weights download automatically on first use.… Open weights apache-2.0 transformers [View model](https://savrn.com/models/jehosephat) Model · Text to image ### [jenny](https://savrn.com/models/jenny) [Michael Romeike](https://savrn.com/model-publishers/microm1966) Open weights apache-2.0 diffusers [View model](https://savrn.com/models/jenny) Model ### [JEPA-CoT](https://savrn.com/models/jepa-cot) [DKNTZMN](https://savrn.com/model-publishers/dkntzmn) JEPA-CoT is an experimental chained architecture combining JEPA-style latent prediction with chain-of-thought-inspired multi-stage computation. This repository is licensed under the Apache License 2.0. The complete retrained Scale-1 checkpoint set contains three independent seeds. K=8 · observation dimension=512 · latent dimension=128 · T=24 · batch=512 · epochs=40. Previously verified seed-0 checkpoint: Each checkpoint contains the model identifier, seed, configuration, complete PyTorch state dict, and evaluation metrics. Open weights apache-2.0 [View model](https://savrn.com/models/jepa-cot) Model · Image and text to text ### [JEV-27B-VL-AWQ](https://savrn.com/models/jev-27b-vl-awq) [AI Outfitters](https://savrn.com/model-publishers/aioutfitters) 4-bit AWQ quantization of autotrust/JEV-27B-VL, built with llm-compressor and served with vLLM. System 1 (POST /v1/decide) runs on a single 24 GB GPU — the fp8 build needs two — and on every axis we measured it is faster than the two-GPU fp8 deployment it replaces. The System 1 adapter is merged into the weights (including the decision head), so no --enable-lora is needed. The decision head (lmhead) and the vision tower are kept in bf16; every other text layer is quantized, including the Gated DeltaNet projections of the 48 linear-attention layers — most Qwen3.5-family quants exclude those and therefore do not fit on one 24 GB card. The AWQ step matters: a data-free RTN build of the same… Open weights apache-2.0 26.9B parameters 262,144 tokens transformers [View model](https://savrn.com/models/jev-27b-vl-awq) Model · Text classification ### [JEV-35B](https://savrn.com/models/jev-35b) [AutoTrust AI Lab](https://savrn.com/model-publishers/autotrust) autotrust/JEV-27B-VL on the board. Approximate public vision score 72.08 (our rebuild of the vision benchmarks; JEV-27B-VL 71.67 on the same rebuild, a tie within noise). Median single-request latency 241 ms on one B200 (JEV-27B-VL 271 ms). JEV-35B: all 140,178 scoreable requests of the 0.3 public suite answered (0 errors), System 1 only (thinking off), every choice read in one pass (up to 256 options), scored with the kit's score --edition 0.3. Our scoring, not a board entry: the board's Full score also counts private tests (80 %), which only its maintainers run. JEV-27B-VL: the board's public numbers for JEV-27B, whose text decisions JEV-27B-VL reproduces (see the JEV-27B-VL card).… Open weights apache-2.0 36B parameters 262,144 tokens transformers [View model](https://savrn.com/models/jev-35b) Model · Text classification ### [JEV-35B-NVFP4](https://savrn.com/models/jev-35b-nvfp4) [AutoTrust AI Lab](https://savrn.com/model-publishers/autotrust) This repository is autotrust/JEV-35B with its 10,240 routed expert MLPs (40 layers × 256 experts) quantized to NVFP4. Everything else is the bf16 model unchanged: attention, linear attention, the shared expert, the routers, the vision tower, the MTP layer, the lmhead, the System 1 adapter, the decision head and the temperatures. Below the NVFP4 comparison, the rest of this card is the JEV-35B card; its tables are the bf16 results unless marked otherwise. Retrieval 0.706, Tools 0.753, Arts 0.413); the largest per-benchmark changes are Home appliance simulator −5.7, CRUXEval −4.7 and GPQA +5.4. CharXiv charts (−5.3). test p = 0.63); every grasped cube still reached the tray (13/13). Games… Open weights apache-2.0 19.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/jev-35b-nvfp4) Model · Text generation ### [JEV-9B-MLX-8bit](https://savrn.com/models/jev-9b-mlx-8bit) [Nativ](https://savrn.com/model-publishers/nativ-community) MLX conversion of autotrust/JEV-9B for mlx-vlm, with the System 1 LoRA (including its lmhead update) merged and lmhead, the decision readout, kept unquantized. JEV is a decision model: it returns a probability for every option of a question in one forward pass and does not generate text. JEV support is not yet in an mlx-vlm release Open weights apache-2.0 9B parameters 262,144 tokens mlx [View model](https://savrn.com/models/jev-9b-mlx-8bit) Model · Text generation ### [JEV-9B-MLX-NVFP4](https://savrn.com/models/jev-9b-mlx-nvfp4) [Nativ](https://savrn.com/model-publishers/nativ-community) MLX conversion of autotrust/JEV-9B for mlx-vlm, with the System 1 LoRA (including its lmhead update) merged and lmhead, the decision readout, kept unquantized. JEV is a decision model: it returns a probability for every option of a question in one forward pass and does not generate text. JEV support is not yet in an mlx-vlm release Open weights apache-2.0 9B parameters 262,144 tokens mlx [View model](https://savrn.com/models/jev-9b-mlx-nvfp4) Model ### [jev-gate-student-b](https://savrn.com/models/jev-gate-student-b) [W C](https://savrn.com/model-publishers/sargedev) A LoRA adapter (r=16, α=32, on qproj/vproj) on Qwen/Qwen2.5-0.5B-Instruct, distilled from the Jev typed-judgment API into a compact local judge for agent-memory gating. What it does: given a query and a candidate memory passage, outputs P(relevant) as the calibrated yes probability read from the final-token logits of yes vs no. Used to filter which vector-recalled memories get injected into agent context (vector recall → cross-encoder band gate → this judge for gray-band cases). Training: LoRA, 1 epoch, LR 1e-4, on rows sampled from SargeDev/jev-distill-corpus — paired relevance judgments (Jev graded 0–7 expected-value scores + binary labels) distilled from typed judgment calls by a larger… Open weights apache-2.0 [View model](https://savrn.com/models/jev-gate-student-b) Model · Text classification ### [Jev-LCT-Adapters](https://savrn.com/models/jev-lct-adapters) [CaoHaoWei](https://savrn.com/model-publishers/caohaowei) This repository contains the standalone lightweight Looped Adapters and uncertainty calibration mappers for Jev-LCT (Looped Calibration Transformer). Instead of downloading the entire merged model weights (which range from 2GB to 16GB), users who already have base Qwen models (Qwen2.5-0.5B, Qwen2.5-1.5B, Qwen3-8B) can directly mount these compact adapter weights (only 220MB ~ 1.5GB) onto the last $k=2$ layers of the base model. 本仓库托管 Jev-LCT (循环校准 Transformer) 的全套独立轻量级适配器权重（.pt）与保序校准映射器（.pkl）。 对于本地已有 Qwen 官方基座（如 Qwen/Qwen2.5-0.5B、Qwen/Qwen2.5-1.5B、Qwen3-8B）的开发者，无需重复下载数十 GB 的完整权重，仅需下载本仓库对应的轻量适配器（仅 220MB ~ 1.5GB），即可在原生基座上获得 50ms 级别系统一极速决策与内生轨迹校准置信度能力。 3. 即插即用：与 lctqwenstandalone.py… Open weights apache-2.0 transformers [View model](https://savrn.com/models/jev-lct-adapters) Model · Text classification ### [Jev-LCT-Qwen2.5-0.5B](https://savrn.com/models/jev-lct-qwen2-5-0-5b) [CaoHaoWei](https://savrn.com/model-publishers/caohaowei) Jev-LCT-Qwen2.5-0.5B is the ultra-lightweight edge edition of the Jev-LCT System-One decision family. Weighing only ~1.9 GB in full bfloat16 weights, it is optimized for high-throughput API gateway routing, edge robotics, Raspberry Pi, and mobile deployment. - ~50ms 极低延迟：专为高并发 API 网关路由、实时内容审核设计； - MMLU 达 50.0%：大幅超越参数相近的判别模型（Laya 33.3%, Open-Jev 35.0%）； Apache License 2.0. Full repository at GitHub. Open weights apache-2.0 494M parameters 32,768 tokens transformers [View model](https://savrn.com/models/jev-lct-qwen2-5-0-5b) Model · Text classification ### [Jev-LCT-Qwen2.5-1.5B](https://savrn.com/models/jev-lct-qwen2-5-1-5b) [CaoHaoWei](https://savrn.com/model-publishers/caohaowei) In agentic workflows, API routing, and edge decision-making, conventional autoregressive LLMs suffer from high token-by-token generation latency and brittle string parsing, while small discriminative models produce systematically miscalibrated verbal confidence (overconfident hallucinations). Jev-LCT (Looped Calibration Transformer) establishes a new paradigm for System-One Decision Models: 1. Parallel Looped Prefill: Recurrently iterates only the top $k=2$ layers of Qwen2.5-1.5B with sequence right-shifting and Scale-Preserving RMS Injection, strictly preventing representation collapse. 2. Endogenous Trajectory Confidence: Extracts genuine calibrated confidence directly from hidden state… Open weights apache-2.0 1.5B parameters 131,072 tokens transformers [View model](https://savrn.com/models/jev-lct-qwen2-5-1-5b) Model · Text classification ### [Jev-LCT-Qwen3-8B](https://savrn.com/models/jev-lct-qwen3-8b) [CaoHaoWei](https://savrn.com/model-publishers/caohaowei) Jev-LCT-Qwen3-8B is the flagship enterprise-grade decision engine of the Jev-LCT family. Combining Qwen3-8B's extensive foundational capabilities with Looped Calibration and adaptive early exit, it provides frontier generative reasoning capabilities with deterministic sub-100ms decision latency. - 85.0% 科学推理 + 70.0% MMLU：媲美中大型生成模型的复杂逻辑推理能力，但单次推断控制在 89.2 ms 内。 - 企业级智能体中枢：支持高风险场景的“选择性预测（Selective Prediction）”，在 80% 覆盖率下实现近乎零差错审核。 - 全量独立权重：开箱即用，支持多 GPU 分片或单张 24GB 显卡（RTX 3090 / 4090）bfloat16 全速推断。 Apache License 2.0. Full repository at GitHub. Open weights apache-2.0 8.2B parameters 40,960 tokens transformers [View model](https://savrn.com/models/jev-lct-qwen3-8b) Model · Text classification ### [jev-novel-3-27b-bf16](https://savrn.com/models/jev-novel-3-27b-bf16) [Aikexue170](https://savrn.com/model-publishers/aikexue170) JEVnovel 的做法是：用人类偏好数据训练判别模型，并把输出做成分类头——一次前向传播直接给出判断和置信度，不做自回归生成。 - JEVnovel 2 解决「识别」：判断一句话需不需要修改； - JEVnovel 2.5 尝试把粒度细化到逗号级小片，加了「短句头」做三分类（拖后腿 / 中性 / 有价值）。实测效果不好：它分不清「可以删掉的纯装饰」和「必须保留的信息」，给不出改稿动作，3 代将其取消； - JEVnovel 3 换成两个能直接指导改稿的头： 教师由 Qwen3.8-27B 微调而来，对整句打分，一次前向输出两组结果： 2. 信息密度 ← StartLux-Decision-27B 裁判。质量和信息是两回事：有的片段「写得烂但没有信息量」（纯装饰，该删），有的「写得烂但承载着情节」（该改）。教师只管质量；信息维度交给裁判模型——它对每一小片回答一个问题：「编辑要把这句话改简洁时，这个片段能不能删掉？」答案直接作为片级密度标签。 3. 学生（0.8B / 4B，LoRA，双头）用这两套标注训练。 完整的方法、数据规模与评测细节见仓库内的《JEVnovel3技术报告.md》。 协议与系列一致：主指标取人类标注人数最多、共识最高的 538 句（main-bench），统计 Macro-F1、bad 召回、bad 精确和正常句误报；判定阈值先在验证集上确定并冻结，再应用到测试集。 - 相比上一代 27B：主指标提升 5.6 个点，bad 召回提升约 12.6 个点，抓得更全；代价是正常句误报从 7.1% 升到 9.4%，精确率基本持平。 - 出厂冻结阈值（0.31）下… Open weights apache-2.0 27.4B parameters 262,144 tokens pytorch [View model](https://savrn.com/models/jev-novel-3-27b-bf16) Model · Text classification ### [Jev-Omni](https://savrn.com/models/jev-omni) [Wiktor Ryznar](https://savrn.com/model-publishers/rydward98) A multimodal decision classifier for text, images, audio and video. Supply a question and options; receive a probability for each option—not a generated explanation. Built on Gemma 4 12B IT, with a 30,000-question fine-tuning run. Merged-model results. ¹Equal-weight scenario/group average. Reference scores are officially reported by their developers and may use different evaluation protocols. Jev-Omni cost uses its recorded input tokens at OpenRouter's Gemma 3 12B input rate ($0.05/M); its classifier generates no output tokens. Both classifiers are priced one call per question: a classifier answers one question at a time, so the state is re-sent for each of them and there is no discount for… Open weights apache-2.0 transformers [View model](https://savrn.com/models/jev-omni) Model · Image and text to text ### [Jev-Omni-MLX-4bit](https://savrn.com/models/jev-omni-mlx-4bit) [Leo Zhang](https://savrn.com/model-publishers/ruiruiz30) An Apple Silicon MLX conversion of akhilaaa3/Jev-Omni for local inference on a Mac mini with 16GB unified memory. This repository keeps the upstream Jev-Omni unified multimodal checkpoint and its trained 256-way decision head, then converts the language-model weights to 4-bit affine quantization with group size 64. The vision weights remain BF16 and the decision head remains FP32. It is an independent conversion. It is not an official TypeSafe Jev release and does not claim to reproduce TypeSafe's proprietary system. It is also not a new fine-tune. - Mac mini, Apple M4, 10 CPU cores, 16GB unified memory - macOS 26.5.1 - Python 3.13.12 - MLX 0.32.2 - MLX-VLM 0.7.1 - single request, batch… Open weights apache-2.0 12B parameters 262,144 tokens mlx [View model](https://savrn.com/models/jev-omni-mlx-4bit) Model ### [jev-qwen](https://savrn.com/models/jev-qwen) [Patrick Dugan](https://savrn.com/model-publishers/alephfunk) LoRA adapters from the Small Jev experiments (Patrick Dugan, Red Team Gladiatorics). TypeSafe's Jev is described as a "System One" model: it returns a typed, calibrated decision over a pre-enumerated output space from a single query. These adapters give an open 4B base model a single-pass, option-scored readout of that kind. They are benchmarked against compact looped decoders, TRM skill gates, script/LDT/Control-Harness gates and a 1-bit 8B LLM. This is not Jev. Jev's weights and architecture are not public; nothing here measures Jev. published, because they contain held-out evaluation items of other projects. All adapters are LoRA r16 / alpha 32 on q/k/v/o, the DeltaNet projections… Open weights apache-2.0 peft [View model](https://savrn.com/models/jev-qwen) Model ### [Jev_Qwen3.8-27B-GGUF](https://savrn.com/models/jev-qwen3-8-27b-gguf) [ramGPT](https://savrn.com/model-publishers/ramgpt) GGUF quantization of SargeDev/JevQwen3.8-27B. - JevQwen3.8-27B-Q4KM.gguf — Q4KM, about 15.8 GiB, 4.92 BPW The source config declares mtpnumhiddenlayers: 1, but the published source weights contain 64 main blocks (blk.0 through blk.63) and no MTP / NextN tensors. A normal conversion therefore advertises 65 blocks and fails to load in llama.cpp because blk.64. tensors are absent. This target GGUF was converted with llama.cpp bd4f514 using --no-mtp, which produces the correct 64-block target model. Validated with llama.cpp bd4f514 on an RTX 4090. The source model was trained with thinking disabled. For the tuned behavior, use reasoning off. llama-cli -m JevQwen3.8-27B-Q4KM.gguf -ngl 999 -c… Open weights apache-2.0 [View model](https://savrn.com/models/jev-qwen3-8-27b-gguf) Model ### [Jev_Qwen3.8-27B-NVFP4-FP8](https://savrn.com/models/jev-qwen3-8-27b-nvfp4-fp8) [W C](https://savrn.com/model-publishers/sargedev) NVFP4 + FP8 mixed-precision quantization of SargeDev/JevQwen3.8-27B (itself a QLoRA merge of huihui-ai/Huihui-Qwen3.8-27B-abliterated trained on SargeDev/jev-distill-corpus-v3). - NVFP4 W4A4 (E2M1, block 16, FP8-E4M3 scales): all attention projections + gate/upproj - FP8DYNAMIC (per-channel weights, per-token activations): all downproj — the layer class most sensitive to 4-bit - 28 GB total (vs 51 GB bf16) — loads in vLLM with --quantization compressed-tensors 512 samples, 2048 max seq: 384 held-out jev-distill judge prompts in the exact eval-time format (state + question + options), plus 128 ultrachat200k general-chat rows to keep broad-language calibration honest. Calibrating on the… Open weights apache-2.0 21.2B parameters 262,144 tokens [View model](https://savrn.com/models/jev-qwen3-8-27b-nvfp4-fp8) Model · Text classification ### [Jev-Qwen3Guard-Gen-Domain-0.6B](https://savrn.com/models/jev-qwen3guard-gen-domain-0-6b) [Pengyi Zhang](https://savrn.com/model-publishers/zhangpy) English | 简体中文 Jev-Qwen3Guard-Gen-Domain-0.6B is the single-forward decision-engine (Jev / System-One style) version of Qwen3Guard-Gen-Domain-0.6B, a generative guard model for the Hong Kong elderly-care domain. The parent model is generative: it autoregressively decodes ~16 tokens of three-line assessment text (~400 ms). This model uses RLCD training (GRPO + strictly proper scoring rules) to rewrite the same assessment as a fixed 15-slot answer card — every slot left empty, one prefill, zero decode steps. Reading next-token probabilities at each slot anchor yields the full decision: The safety taxonomy (13 categories = 9 general + 4 HK elderly-care additions), the dual evaluation modes… Open weights apache-2.0 596M parameters 32,768 tokens transformers [View model](https://savrn.com/models/jev-qwen3guard-gen-domain-0-6b) Model ### [JevEmbed-Qwen3-Embedding-4B](https://savrn.com/models/jevembed-qwen3-embedding-4b) [Lychee Team](https://savrn.com/model-publishers/hit-tmg) Open weights apache-2.0 [View model](https://savrn.com/models/jevembed-qwen3-embedding-4b) Model ### [jevify-gemma-4-e4b-it-readout-coh](https://savrn.com/models/jevify-gemma-4-e4b-it-readout-coh) [Praveen Raj U S](https://savrn.com/model-publishers/praveenrajus) A System One decision model: it does not write text. It reads a state, answers typed questions (choice, score, noul), and returns calibrated probability distributions your code can branch on. This repo is a rank-16 LoRA (34,881,536 parameters) on google/gemma-4-E4B-it, merged into the weights at load. How it was trained (readout fine-tuning). The model is trained on its own decision readout — the distribution over the allowed answers read at the answer position, one forward pass, no decoding — with the primitive's proper scoring rule, plus a coherence penalty (weight 1.0): every training question comes with automatically derived siblings (the options as yes/no questions, the negation, the… Open weights apache-2.0 jevify [View model](https://savrn.com/models/jevify-gemma-4-e4b-it-readout-coh) Model ### [jevify-qwen3.5-4b-base-readout-coh](https://savrn.com/models/jevify-qwen3-5-4b-base-readout-coh) [Praveen Raj U S](https://savrn.com/model-publishers/praveenrajus) A System One decision model: it does not write text. It reads a state, answers typed questions (choice, score, noul), and returns calibrated probability distributions your code can branch on. This repo is a rank-16 LoRA (21,233,664 parameters) on Qwen/Qwen3.5-4B-Base, merged into the weights at load. How it was trained (readout fine-tuning). The model is trained on its own decision readout — the distribution over the allowed answers read at the answer position, one forward pass, no decoding — with the primitive's proper scoring rule, plus a coherence penalty (weight 1.0): every training question comes with automatically derived siblings (the options as yes/no questions, the negation, the… Open weights apache-2.0 jevify [View model](https://savrn.com/models/jevify-qwen3-5-4b-base-readout-coh) Model ### [jevify-qwen3.5-4b-readout](https://savrn.com/models/jevify-qwen3-5-4b-readout) [Praveen Raj U S](https://savrn.com/model-publishers/praveenrajus) A System One decision model: it does not write text. It reads a state, answers typed questions (choice, score, noul), and returns calibrated probability distributions your code can branch on. This repo is a rank-16 LoRA (21,233,664 parameters) on Qwen/Qwen3.5-4B, merged into the weights at load. How it was trained (readout fine-tuning). The model is trained on its own decision readout — the distribution over the allowed answers read at the answer position, one forward pass, no decoding — with the primitive's proper scoring rule. Options are shuffled per family. Training data: the train splits of the 16 non-held-out jev-bench sources (5,885 families, at most 400 records per source); lr… Open weights apache-2.0 jevify [View model](https://savrn.com/models/jevify-qwen3-5-4b-readout) Model ### [jevify-qwen3.5-4b-readout-coh](https://savrn.com/models/jevify-qwen3-5-4b-readout-coh) [Praveen Raj U S](https://savrn.com/model-publishers/praveenrajus) A System One decision model: it does not write text. It reads a state, answers typed questions (choice, score, noul), and returns calibrated probability distributions your code can branch on. This repo is a rank-16 LoRA (21,233,664 parameters) on Qwen/Qwen3.5-4B, merged into the weights at load. How it was trained (readout fine-tuning). The model is trained on its own decision readout — the distribution over the allowed answers read at the answer position, one forward pass, no decoding — with the primitive's proper scoring rule, plus a coherence penalty (weight 1.0): every training question comes with automatically derived siblings (the options as yes/no questions, the negation, the… Open weights apache-2.0 jevify [View model](https://savrn.com/models/jevify-qwen3-5-4b-readout-coh) Model ### [jevify-qwen3-vl-2b-t2](https://savrn.com/models/jevify-qwen3-vl-2b-t2) [Praveen Raj U S](https://savrn.com/model-publishers/praveenrajus) A Jevify Tier 2 model: Qwen/Qwen3-VL-2B-Instruct used as a vision-language model: state may carry images (PIL, path, URL, data URI or bytes) and the same three typed questions are asked about them, with a rank-16 LoRA confined to the decoder half of the model, trained on the readout itself — the restricted log-softmax over the allowed answers at the answer position — with the primitive's proper scoring rule, on the aokvqa train split only. The adapter (17,432,576 parameters) is merged into the backbone at load, so it serves at the plain checkpoint's speed. The calibration recipe was fitted on validation splits of the adapted model. pope, ai2d never appeared in training. Serve it as a… Open weights apache-2.0 [View model](https://savrn.com/models/jevify-qwen3-vl-2b-t2) Model · Text classification ### [jiwo-0.8b-GGUF](https://savrn.com/models/jiwo-0-8b-gguf) [Breaksoftware](https://savrn.com/model-publishers/breaksoftware) I decided to use jiwo-0.8b because it had a high score on https://huggingface.co/spaces/multimodalart/jev-decision-index compared to its parameter count. This GGUF conversion / quantization was generated by a local agent powered by Qwen3.8 27B. See agent-trajectory-log.md for exhaustive details. Run with https://llama.app This is a decision model, to be used via /v1/systemone API. See https://github.com/ggml-org/llama.cpp/pull/29818 - https://huggingface.co/eljiwo/jiwo-0.8b Open weights apache-2.0 [View model](https://savrn.com/models/jiwo-0-8b-gguf) Model · Image to image ### [JoyAI-Image-Edit-Diffusers](https://savrn.com/models/joyai-image-edit-diffusers) [Le Jin](https://savrn.com/model-publishers/jinstudio) Open weights apache-2.0 16.3B parameters diffusers [View model](https://savrn.com/models/joyai-image-edit-diffusers) Model · Image to image ### [JoyAI-Image-Edit-Plus-Diffusers](https://savrn.com/models/joyai-image-edit-plus-diffusers) [Le Jin](https://savrn.com/model-publishers/jinstudio) Open weights apache-2.0 16.3B parameters diffusers [View model](https://savrn.com/models/joyai-image-edit-plus-diffusers) Model · Text classification ### [Julia-1-LiteRT](https://savrn.com/models/julia-1-litert) [LiteRT Community (FKA TFLite)](https://savrn.com/model-publishers/litert-community) The SupersonicLabs/Julia-1 decision model, converted to LiteRT with litert-torch, answers one question in a median of 80.7 ms (first session) to 112.1 ms (second session) of graph time on a Samsung Galaxy S26 GPU with explicit FP32 computation, at a 512-token window (three runs on 2026-09-30; each run's thermal state is below). Julia-1 reads a state (text or JSON), a typed question and 2 to 20 options, and returns one probability per option from one forward pass. A choice question returns the winning option, a score question the expected index on an ordered rubric, and a noul (yes/no) question the probability of true. On the phone, all 706 validation rows got the same answer as the author's… Open weights apache-2.0 litert [View model](https://savrn.com/models/julia-1-litert) 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) ## Questions ### Can I use Apache License 2.0 models commercially? Yes. 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. ### Which Apache License 2.0 models are most downloaded? By monthly downloads reported by the Hugging Face Hub: . ## Other Licenses - [mit](https://savrn.com/models/licenses/mit) 1319 - [cc-by-nc-4.0](https://savrn.com/models/licenses/cc-by-nc-4-0) 213 - [cc-by-4.0](https://savrn.com/models/licenses/cc-by-4-0) 164 - [bsd-3-clause](https://savrn.com/models/licenses/bsd-3-clause) 108 - [gemma](https://savrn.com/models/licenses/gemma) 63 - [cc-by-nc-sa-4.0](https://savrn.com/models/licenses/cc-by-nc-sa-4-0) 55 - [llama3.2](https://savrn.com/models/licenses/llama3-2) 40 - [unknown](https://savrn.com/models/licenses/unknown) 33 - [agpl-3.0](https://savrn.com/models/licenses/agpl-3-0) 32 - [llama3.1](https://savrn.com/models/licenses/llama3-1) 28 - [cc-by-sa-4.0](https://savrn.com/models/licenses/cc-by-sa-4-0) 20 - [gpl-3.0](https://savrn.com/models/licenses/gpl-3-0) 20 [See all](https://savrn.com/models/licenses)