전체 검색
AI 모델 449
- tiny-lm-chat↗
sbintuitions
This repository provides a tiny 16M parameters language model for debugging and testing purposes. This is created by tuning sbintuitions/tiny-lm with oasset1 datasets in Japanese and English. The model was first pre-trained on English Wikipedia and Japanese Wikipedia to optimize a traditional language modelling objective for 25B tokens. And then it was fine-tuned on oasst1 datasets in Japanese and English for 15 epochs.
- Kimi-K2.7-Code-GGUF↗
unsloth
To run Kimi K2.7 Code in full precision lossless, run Q8 (UD-Q8_K_XL), which is 595GB and only 10GB bigger than Q4 (UD-Q4_K_XL). See our Kimi K2.7 Code guide for quantization analysis and instructions. Kimi K2.7 Code is a coding-focused agentic model built upon Kimi K2.6. With substantial improvements on real-world long-horizon coding tasks, it strengthens end-to-end task completion across complex software engineering workflows while improving token efficiency, reducing thinking-token usage by approximately 30% compared with Kimi K2.6. 1. **General Testing Details** - Unless stated otherwise, Kimi K2.7 Code and K2.6 were tested with thinking mode enabled via Kimi Code CLI at temperature = 1.0, top-p = 0.95, and a 262,144-token context length; GPT-5.5 ran in Codex with xhigh mode, and Opus 4.8 in Claude Code with xhigh mode. Aside from these differences, all benchmarks were evaluated under the same conditions. 2. **Coding Benchmarks** - Kimi Code Bench V2 is our in-house benchmark designed to evaluate coding agents on realistic tasks. It has diversed software engineering tasks across 10+ mainstream programming languages and a full production tech stack covering tasks from internal engineering use cases, production incidents, and real-world open-source projects, with emphasis on backend services, infrastructure, performance engineering, systems programming, security, frontend development, and ML/data engineering. - Program Bench evaluates code-generation agents by asking them to recreate a program’s behavior from only a compiled binary and its documentation. It spans 200 tasks, from small CLI tools to large systems like FFmpeg and SQLite. Submissions are judged against over 248,000 fuzz-generated behavioral tests. In each task, the agent is given an executable and its documentation, but no source code, decompilation, or internet access. It must choose its own implementation language, build the full program from scratch, and pass a behavioral test suite comparing its output against the original binary. - MLS-Bench evaluates whether AI systems can invent generalizable and scalable ML methods. MLS-Bench-Lite is the official 30-task subset of MLS-Bench, covering LLM pretraining and post-training, robotics, world models, computer vision, reinforcement learning, optimization, ML systems, AI for Science, and more. Agents are given 5 hours to explore before submitting
- FLUX.2-klein-4B-SDNQ-4bit-dynamic↗
Disty0
This model uses per layer fine grained quantization. What dtype to use for a layer is selected dynamically by trial and error until the std normalized mse loss is lower than the selected threshold. Minimum allowed dtype is set to uint4 and std normalized mse loss threshold is set to 1e-2. This created a mixed precision model with uint4 and int5 dtypes. SVD quantization is disabled.
- unispeech-sat-base-100h-libri-ft↗
Microsoft
A unispeech-sat-base model that was fine-tuned on 100h hours of Librispeech on 16kHz sampled speech audio. When using the model make sure that your speech input is also sampled at 16Khz. Authors: Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu **Abstract** *Self-supervised learning (SSL) is a long-standing goal for speech processing, since it utilizes large-scale unlabeled data and avoids extensive human labeling. Recent years witness great successes in applying self-supervised learning in speech recognition, while limited exploration was attempted in applying SSL for modeling speaker characteristics. In this paper, we aim to improve the existing SSL framework for speaker representation learning. Two methods are introduced for enhancing the unsupervised speaker information extraction. First, we apply the multi-task learning to the current SSL framework, where we integrate the utterance-wise contrastive loss with the SSL objective function. Second, for better speaker discrimination, we propose an utterance mixing strategy for data augmentation, where additional overlapped utterances are created unsupervisely and incorporate during training. We integrate the proposed methods into the HuBERT framework. Experiment results on SUPERB benchmark show that the proposed system achieves state-of-the-art performance in universal representation learning, especially for speaker identification oriented tasks. An ablation study is performed verifying the efficacy of each proposed method. Finally, we scale up training dataset to 94 thousand hours public audio data and achieve further performance improvement in all SUPERB tasks..*
- LFM2-350M↗
LiquidAI
LFM2 is a new generation of hybrid models developed by Liquid AI, specifically designed for edge AI and on-device deployment. It sets a new standard in terms of quality, speed, and memory efficiency. We're releasing the weights of four post-trained checkpoints with 350M, 700M, 1.2B, and 2.6B parameters. They provide the following key features to create AI-powered edge applications: * **Fast training & inference** – LFM2 achieves 3x faster training compared to its previous generation. It also benefits from 2x faster decode and prefill speed on CPU compared to Qwen3. * **Best performance** – LFM2 outperforms similarly-sized models across multiple benchmark categories, including knowledge, mathematics, instruction following, and multilingual capabilities. * **New architecture** – LFM2 is a new hybrid Liquid model with multiplicative gates and short convolutions. * **Flexible deployment** – LFM2 runs efficiently on CPU, GPU, and NPU hardware for flexible deployment on smartphones, laptops, or vehicles.
- Wan 2.7 R2V↗
待核实厂商
Wan 2.7 R2V API（APIMart）：wan2.7-r2v 用最多 5 个参考图片或视频生成 2–15 秒 720P/1080P 视频，保持角色和音色一致。开通会员，API 调用立省20%起。
- adetailer↗
Bingsu
Since `getattr` is classified as a dangerous pickle function, any segmentation model that uses it is classified as unsafe. All models were created and saved using the official ultralytics library, so it's okay to use files downloaded from a trusted source.
- Qwen3.5-abliterated-MAX-AIO-GGUF↗
prithivMLmods
> prithivMLmods/Qwen3.5 Unredacted-MAX models (0.8B, 2B, 4B, 9B, 27B variants) are experimental "abliterated" evolution of Alibaba's official Qwen3.5 series, applying advanced refusal-minimization techniques to remove safety alignments and internal censorship patterns while preserving or enhancing core reasoning, coding, math, and multimodal capabilities across scales. Created by prithivMLmods as part of their "Unredacted MAX" collection—similar to their Qwen3-VL abliterated releases like Qwen3-VL-8B-Instruct-Unredacted-MAX—these models undergo continual "abliteration" training for unrestricted instruction adherence, dense output generation, and reduced hallucinations on diverse prompts without standard safety refusals, making them suitable for edge deployment (FP8/GGUF quants available) in research, dataset creation, or applications needing uncensored reasoning. Scaling from ultra-light 0.8B (mobile/Pi) to powerful 27B, they retain Qwen3.5's hybrid Gated DeltaNet architecture, 262K context, 201-language vocab, and toggleable thinking modes while prioritizing raw capability over guardrails, outperforming vanilla counterparts in open-ended tasks per community benchmarks though lacking official evaluations. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
- bleurt-tiny-512↗
Elron
- **Developed by:** Elron Bandel, Thibault Sellam, Dipanjan Das and Ankur P. Parikh of Google Research - **Shared by [Optional]:** Elron Bandel - **Model type:** Text Classification - **Language(s) (NLP):** More information needed - **License:** More information needed - **Parent Model:** BERT - **Resources for more information:** - GitHub Repo - Associated Paper - Blog Post The model should not be used to intentionally create hostile or alienating environments for people. Significant research has explored bias and fairness issues with language models (see, e.g., Sheng et al. (2021) and Bender et al. (2021)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
- Wan2.2-Distill-Models↗
lightx2v
*Distillation-accelerated version of Wan2.2 - Dramatically faster speed with excellent quality* > 🌐 **Create online with LightX2V Studio** — > an AI avatar and multimodal content production platform built on LightX2V. > Use Wan2.2 video generation in your browser, with no local setup required. > See the online usage guide below. --- - 2026.04.12: We are excited to release the Wan2.2-I2V-A14B-4step-720p-high and Wan2.2-I2V-A14B-4step-720p-low models. Compared to previous iterations, this version was trained on a high-quality 720p dataset and features an optimized low-noise training algorithm. These enhancements significantly boost the model's performance in fine-grained detail rendering and visual texture.