전체 검색

AI 모델 2119

• Qwen3.6-35B-A3B-heretic-moe-balanced-v2↗

  DmitryDB

  Decensored (abliterated) **Qwen3.6-35B-A3B**, the **balanced** operating point: refusals removed at the lowest damage to the model. Built with a fork of Heretic that searches against a **perplexity objective** instead of first-token KL — that is what makes this build cheaper than the previous release at the same or fewer refusals. Sibling: `-max-v2` is the more aggressive point (fewer refusals, more damage); it also loses more facts, so **balanced is the recommended build**. Method and figures: the study repo. The refusal direction is removed from every projection that writes to the residual stream — full and linear attention outputs, the 256 routed experts (fused), and the always-on shared expert. The multi-token-prediction layer is restored verbatim from the base model, so that branch is unmodified. All builds carry the full 1045 tensors.

• TinyLLama-v0↗

  Maykeye

  This is a first version of recreating roneneldan/TinyStories-1M but using Llama architecture. * Full training process is included in the notebook train.ipynb. Recreating it as simple as downloading TinyStoriesV2-GPT4-train.txt and TinyStoriesV2-GPT4-valid.txt in the same folder with the notebook and running the cells. Validation content is not used by the script so you put anythin in * Backup directory has a script do_backup that I used to copy weights from remote machine to local. Weight are generated too quickly, so by the time script copied weihgt N+1

• Qwen3.8-27B-Whitehat-GGUF↗

  QuaduxIT

  > **⚠️ Authorized security use only — Nur für autorisierte Sicherheitsarbeit.** > For authorized IT-security work on systems you own or have explicit permission to test. Physical harm, weapons/CBRN, illegal substances and CSAM are refused by design. Provided "as is" without warranty (Apache-2.0). **By downloading or using this model you accept the Disclaimer and Terms of Use** (see the "Haftungsausschluss / Disclaimer" section below and `DISCLAIMER.md`). > *Dieses Modell dient autorisierter IT-Sicherheitsarbeit; mit der Nutzung akzeptierst du den Haftungsausschluss (siehe unten und `DISCLAIMER.md`).* A **red-team / white-hat fine-tune** of Qwen/Qwen3.8-27B, produced by **Quadux IT GmbH** as a **local, private assistant for internal offensive-security work and vulnerability self-assessment**. Hosted models refuse most offensive-security tasks, and you often cannot send sensitive vulnerability or system data to an external service anyway. This model fills that gap: it **helps fully with any computer- and network-security task** — offensive and defensive, including exploit development, malware development and analysis, reverse engineering, and license/DRM research — so your findings stay in-house. At the same time it **still refuses requests aimed at real physical harm to people** (weapons, explosives, drugs, poisons, chemical/biological weapons, violence) and **child sexual abuse material**. That boundary holds on **both text and image input** and **across languages**.

• Samantha-1.11-13B-GGUF↗

  TheBloke

• AI-image-detector↗

  umm-maybe

  *__NOTE__: Unless you are trying to detect imagery generated using older models such as VQGAN+CLIP, please use the updated version of this detector instead.* This model is a proof-of-concept demonstration of using a ViT model to predict whether an artistic image was generated using AI. It was created in October 2022, and as such, the training data did not include any samples generated by Midjourney 5, SDXL, or DALLE-3. It still may be able to correctly identify samples from these more recent models due to being trained on outputs of their predecessors.

• opensearch-neural-sparse-encoding-doc-v2-mini↗

  opensearch-project

  The model should be selected considering search relevance, model inference and retrieval efficiency(FLOPS). We benchmark models' **zero-shot performance** on a subset of BEIR benchmark: TrecCovid,NFCorpus,NQ,HotpotQA,FiQA,ArguAna,Touche,DBPedia,SCIDOCS,FEVER,Climate FEVER,SciFact,Quora. Overall, the v2 series of models have better search relevance, efficiency and inference speed than the v1 series. The specific advantages and disadvantages may vary across different datasets. - **Paper**: Towards Competitive Search Relevance For Inference-Free Learned Sparse Retrievers - **Fine-tuning sample**: opensearch-sparse-model-tuning-sample

• GLM-OCR↗

  zai-org

  GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture. It introduces Multi-Token Prediction (MTP) loss and stable full-task reinforcement learning to improve training efficiency, recognition accuracy, and generalization. The model integrates the CogViT visual encoder pre-trained on large-scale image–text data, a lightweight cross-modal connector with efficient token downsampling, and a GLM-0.5B language decoder. Combined with a two-stage pipeline of layout analysis and parallel recognition based on PP-DocLayout-V3, GLM-OCR delivers robust and high-quality OCR performance across diverse document layouts. - **State-of-the-Art Performance**: Achieves a score of 94.62 on OmniDocBench V1.5, ranking #1 overall, and delivers state-of-the-art results across major document understanding benchmarks, including formula recognition, table recognition, and information extraction. - **Optimized for Real-World Scenarios**: Designed and optimized for practical business use cases, maintaining robust performance on complex tables, code-heavy documents, seals, and other challenging real-world layouts.

• Nemotron-Labs-Diffusion-14B↗

  NVIDIA

  Nemotron-Labs-Diffusion is a tri-mode language model that supports both AR decoding and diffusion-based parallel decoding by simply switching the attention pattern of the same model during inference. The synergy between these two modes enables a third mode, called self-speculation: the same model performs diffusion-based parallel drafting and AR verification with shared KV cache, achieving high acceptance lengths and decoding efficiency. The seamless mode switching by simply changing attention patterns enables high efficiency at different concurrency levels in varying deployment scenarios with one single model. An optional LoRA adatper can be applied to the diffusion drafter in the linear self-speculation mode to further increase the acceptance length: NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the bias, explainability, safety & security, and privacy subcards.

• Qwen3.8-27B-Uncensored-Aggressive-W4A16-AWQ↗

  philbert440

  4-bit weight (W4A16) AWQ quant of **Qwen3.8-27B-Uncensored-Aggressive** (α=1.15 recipe update), in `compressed-tensors` format. Vision tower and the grafted **MTP head** are preserved. Recipe update to the Aggressive line: the previous build ablated at α≈1.24, which a larger benchmark sweep showed **over-ablates** past the ~1.15 quality peak. This build uses **α=1.15** — more open and better on every measured axis. Quant: AWQ, group size 128, asymmetric, MSE observer, 256 Magpie-CoT calibration samples @1024 tokens.

• sam-audio-small↗

  facebook