Open-Weight Models Under Apache License 2.0

weighted/imatrix quants of https://huggingface.co/shuff57/lfm2-24b-phase1-reasoning For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/lfm2-24b-phase1-reasoning-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…

Model · Text generation

LFM2.5-24B-A2B-JANGH2

Osaurus-AI

A JANG bundle of LiquidAI/LFM2.5-24B-A2B: Liquid AI's 24B mixture-of-experts reasoning model with 2B active parameters (40 layers: 30 short-convolution and 10 full-attention, 64 routed experts with 4 active, two dense layers), built for on-device agents. It always thinks before answering and calls tools in Liquid's Pythonic format. 44.7 GiB of bf16 weights in 7.3 GiB. Every quantized matrix with an input Hessian was rounded with GPTQ. The Hessians come from the model's own outputs: the bf16 model wrote its thinking and answers for math, finance, code in eleven languages, terminal and coding-agent sessions (with real tool results), multi-step tool calls, networking, cybersecurity, science…

Model · Text generation

LFM2.5-8B-A1B-TAP-DPQ-v7-GGUF

DuoNeural

DuoNeural Systems & Production Quantization Architecture | Daedalus Protocol Produced by Jesse Caldwell, Archon, Aura (DuoNeural Research Lab) This repository presents the official TAP-DPQ v7 (Thouless-Anderson-Palmer Driven-Dissipative Post-Training Quantization v7) release of Liquid AI's LFM-2.5-8B-A1B. LFM-2.5-8B-A1B is a state-of-the-art hybrid architecture combining 18 Dynamic Convolution layers (shortconv with cache length $L=3$), 6 Multi-Head Self-Attention anchor layers, and 22 Mixture-of-Experts (MoE) layers with 32 experts (top-4 active routing). 1. Hurwitz Recurrent Core Stability: - The dynamic convolution temporal filtering kernels ($3 \times 2048$ per layer) govern the…

Model · Text generation

LibraMindMini

Alby13

LibraMind is a 565M-parameter chat model pretrained from scratch by alby13 on a single consumer GPU (NVIDIA RTX 4090). It is a hybrid architecture that interleaves 18 Gated DeltaNet linear-recurrent layers with 6 Gated Attention layers (a repeating DDDA × 6 pattern), with Block Attention Residuals and logit soft-capping, trained with the Muon optimizer. It went through a full five-stage pipeline: pretraining, midtraining, supervised fine-tuning, preference tuning, and RL with verifiable rewards. - Research on small hybrid recurrent/attention language models - Studying low-resource, single-GPU training pipelines (pretraining through RL) and the Muon optimizer at small scale - Experiments…

Model · Image to video

LIFT

Shengxiang Ji

We introduce LIFT, a unified image-to-video generation framework that complements camera control with Layout-In-FuTure control, enabling users to specify what should appear in a future view and where it should appear. Our models are built on Wan2.1-Fun-V1.1-1.3B-Control-Camera.

Model

LightJev-0.6B-v0.1

Rongxinzy

A trained finite-candidate decision prototype built on Qwen3-0.6B by RongxinAI. Full weights, tokenizer, training logs, frozen gold-only data and held-out predictions are included. This is a synthetic-domain research release, not a general business decision model or an official Jev reproduction. Selected arm: CE, checkpoint step 250, chosen only by development cross-entropy. This selected checkpoint contains 20 head-only plus 230 full-model updates (4,000 sampled questions); each completed training arm ran the longer 20 + 300 update budget (5,120 sampled questions). It has 596,052,993 parameters, including a shared 3,073-parameter LayerNorm/scalar scoring head. Choice, Boolean and ordinal…

Model · Image and text to text

LightOnOCR-3-0.8B-ONNX-web

superPDF

lightonai/LightOnOCR-3-0.8B by LightOn, converted to ONNX for superPDF, where it powers high-accuracy reading in the browser (WebGPU, superpdf.ai) and the command-line tool (default model). It runs with Transformers.js on ONNX Runtime (WebGPU), in the browser and in Node. from the unused multi-token-prediction layer). Every weight in those graphs is recomputed from LightOn's weights and written at the same place, so the graphs themselves are unchanged. asymmetric with zero points, computed with ONNX Runtime's own quantizematmul4bits); vision encoder float16. Norm weights are stored as 1 + w and the gated delta net's A as -exp(Alog), as the builder expects. - These are the same bytes as…

Model · Image and text to text

LightOnOCR-3-4B-ONNX

superPDF

lightonai/LightOnOCR-3-4B by LightOn, converted to ONNX for superPDF, where it powers high-accuracy reading in the command-line tool and MCP server (--model 4b, model: "4b"). It runs with Transformers.js on ONNX Runtime (WebGPU), in the browser and in Node. from the unused multi-token-prediction layer). Every weight in those graphs is recomputed from LightOn's weights and written at the same place, so the graphs themselves are unchanged. asymmetric with zero points, computed with ONNX Runtime's own quantizematmul4bits); vision encoder float16. Norm weights are stored as 1 + w and the gated delta net's A as -exp(Alog), as the builder expects. - Converted by superPDF with the converter below.…

static quants of https://huggingface.co/TCMLLM/Lingdan-4B-Base 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 lower-quality quant And…

static quants of https://huggingface.co/TCMLLM/Lingdan-4B-R1 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 lower-quality quant And…

static quants of https://huggingface.co/TCMLLM/Lingdan-8B-R1 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 lower-quality quant And…

Model · Image and text to text

linne-oak-1-0

Hert4

linne-oak-1.0 is a 31B decision model. Give it a state (text, a document, JSON, a conversation or an agent trajectory) and a question with a set of options, and it returns one option together with a probability for every option, in a single forward pass. It is the larger sibling of linne-fern-1.0: linne-fern is a 27B causal model that also chats, linne-oak is a dedicated decision model. - Decision Index public score 63.94, measured with the official kit (edition 0.3, all 37 public benchmarks, 140,620 requests, complete). - Reads the whole prompt at once: the full-attention layers attend in both directions over the prompt (sliding-window layers keep their native behavior), so the decision…

WebAssembly + WebGPU runtime binaries and the matching @litert-lm/core TypeScript SDK for running Gemma 4 (E2B, E4B, 12B, 26B-A4B, 31B) in the browser with: - Text decode with shared-KV neural Multi-Token Prediction (MTP) speculative decoding (GPUARTISAN backend, gamma 1-3, greedy token-for-token equivalence with standard decode). - Multimodal vision (SigLIP encoder on WebGPU, 70 / 140 / 280 / 560 / 1120 token budgets, zero-copy RGBA8 ImageData preprocessing, per-budget executor memory release). - Multimodal audio (Conformer encoder, streaming markers, zero-copy audio ingestion). - Streamed model loading for 2-20 GB.litertlm bundles under wasm32 (MAXIMUMMEMORY=4GB), with section filtering…

static quants of https://huggingface.co/WasamiKirua/Llama-3.2-3B-Alucard-IT 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…

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.

LoRA adapters fine-tuning Qwen2.5-7B-Instruct to emit a full CIF crystal structure from a prompt of reduced composition + target space-group number. Part of a controlled composition-sweep study (MP-20: MPTS-52 training ratio at fixed volume/steps). Trained with Unsloth, LoRA r=32 / α=64, lr 1e-4, 24,000 unique training crystals, pinned 4,500 steps (no early stopping), 16-bit. Adapter base: unsloth/Qwen2.5-7B-Instruct (loads fine on stock Qwen/Qwen2.5-7B-Instruct). Finding: at matched volume + steps, match declines monotonically as the MP-20 fraction rises — the gain reported for "combined" data was volume, not symmetry composition. Each subfolder contains: model/ (LoRA adapter +…

A 97.6M-parameter language model trained from scratch, then fine-tuned for short, polite, everyday English conversation. It is a small, open research and learning model: you can read every line of its training code, run it on a laptop CPU, and see exactly where a model of this size is good and where it fails. - HellaSwag 33.6% (accnorm). That is above GPT-2 small (~30%) and below SmolLM2-135M (43.1%), which saw about 250x more training text. - The custom PyTorch code is included. The model does not use transformers; see How to use. This is QuickTalk run 8. "LMLM97M1" is its published name. 1. Learning and teaching how LLMs work. It is a complete, small, from-scratch GPT, with its training…

Model · Image and text to text

Lodestar-4B

StartLux

Lodestar-4B is a 4-billion-parameter decision model. You give it a state (plain text, JSON or a long document) and one or more typed questions; it returns a probability for every listed option. Each answer is read from a single forward pass. The model never generates free text, so there is nothing to parse and no output length to budget for. (Probabilities rounded to three decimals.) The same call from Python, run inside the downloaded folder: A score question takes its levels as a list, lowest first, and returns {"type": "score", "score":, "probabilities": {"0": p0, "1": p1,...}}. Every question is rendered into one chat prompt (thinking disabled): The probability of each option is the…

a page (a book, a paper, a report) and it returns structured HTML: headings, paragraphs, tables and formulas marked separately, flowcharts as Mermaid, sheet music as ABC notation. It runs entirely on your own machine. The original GGUF lives under a single personal account, on a single host. If it is renamed, deleted or re-uploaded, people who depend on it lose their download, and access to Hugging Face from mainland China is unreliable. So we keep the original files, unmodified, on ModelScope and on our own Hugging Face account. File names, sizes and sha256 digests all match the original. Reproducible: we also converted and quantized the official Alibaba weights ourselves. The quantized…

Model · Reinforcement learning

Logics-SWE-Qwen3.6-27B

Logics-MLLM

[2026.09.18] Released Logics-SWE-Qwen3.6-27B under the Apache-2.0 license. - The technical report is in preparation. A link will be added when available. Logics-SWE-Qwen3.6-27B is a 27B-parameter model developed for repository-level software engineering agents. Starting from the Qwen3.6-27B model used in our study, it combines category-aware expert development with multi-teacher on-policy distillation into a single deployment policy. Repository-level tasks require agents to navigate code, edit files, execute commands, inspect feedback, and iteratively repair their solutions. Our work starts from the category see-saw: aggregate progress during joint RL can conceal opposing changes across…

Model · Text generation

Longleaf-35B-A3B

Red Hills Media

Longleaf is our local model for coding and agent work: repository edits, debugging, structured tools, and clear answers. Built on Qwen3.6-35B-A3B. - 35B total / 3B active parameters. - MLX 4-bit weights. - 262,144-token configured context. EvalPlus 0.3.1, 164 tasks, pass@1. Thinking off, greedy sampling, 2,048-token output cap, 262,144-token configured context. Generation used our M1 Ultra 231 / 198 / 195 tokens, respectively. CLI sample: five Terminal-Bench 2 tasks on ARM64, thinking off. Per task: 300 seconds, 24 model calls, 32 tool calls, 2,048 output tokens per call. Our M1 Ultra Splash runtime, medium thinking, model already loaded: Read, edit, and test completed in about 6 seconds…

Model · Text generation

Longleaf-35B-A3B-Abliterated

Red Hills Media

The abliterated version of Longleaf, our local model for coding, repository edits, debugging, and structured tools. - 35B total / 3B active parameters. - MLX 4-bit weights. - 262,144-token configured context. EvalPlus 0.3.1, 164 tasks, pass@1. Thinking off, greedy sampling, 2,048-token output cap, 262,144-token configured context. Generation used our M1 Ultra 231 / 198 / 195 tokens, respectively. CLI sample: five Terminal-Bench 2 tasks on ARM64, thinking off. Per task: 300 seconds, 24 model calls, 32 tool calls, 2,048 output tokens per call. Apache 2.0. This release contains modified model weights, based on Qwen3.6-35B-A3B, and the Abliteration changes refusal behavior, not output…

Model

looped-moe-329m-ablation-tensorboard

Pengxiang Li

TensorBoard scalar logs and per-loop activation diagnostics for seven ablations of a 329M looped MoE transformer. Each run changes exactly one switch of the base recipe. - 10 layers, hidden 512, 8 heads / 4 KV heads - Global batch 1024 sequences x 1024 tokens, micro-batch 16, AdamW (0.9, 0.95), weight decay 0.1, gradient clip 1.0, bf16 - Loop residual scale gamma = lambda / (H sqrt(L)) and embedding re-injection g = lambda / (t sqrt(L)), with lambda = 0.5 frozen bpsegmentlen=0 runs all H loops as one graph and backpropagates once per step. It also removes the intermediate-loop losses and the extra optimizer steps of the segmented scheme. With H=3 the segmented scheme has a single segment…

static quants of https://huggingface.co/cagataydev/loopl-0.8b For a convenient overview and download list, visit our model page for this model. weighted/imatrix quants are available at https://huggingface.co/mradermacher/loopl-0.8b-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: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9 See…

weighted/imatrix quants of https://huggingface.co/cagataydev/loopl-0.8b For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/loopl-0.8b-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 Artefact2's thoughts on the matter…

static quants of https://huggingface.co/cagataydev/loopl-2b 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 lower-quality quant And…

static quants of https://huggingface.co/cagataydev/loopl-vl-2b 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 lower-quality quant And…

A LoRA adapter for Qwen/Qwen3.5-4B that makes typed decisions about a document in one forward pass and returns calibrated probabilities over the options: a System-1 decision engine, fast judgment rather than generated reasoning. Served with the code at github.com/tuantong/loupe-serving, which reads the answer from the next-token probabilities of the option labels; nothing is generated. An exact average (model soup) of LoRA adapters on the attention, linear-attention and MLP projections of all 32 layers (three seeds of one recipe, a continuation of that recipe on written rationales, and a run whose targets include a second teacher), stored as one rank-96 adapter. Merge it into the base…

LT-OPD trains vision-language models with fewer visual tokens through on-policy self-distillation. This Qwen3.5-4B model combines CDPruner with an MLP that aggregates discarded tokens into their nearest retained tokens, preserving their original positions without adding tokens. Use Python 3.12 and the CUDA environment in the installation guide. Load the model with LT-OPD's visual-token compression runtime

A few LoRAs I've collected over the months. Best used with rgthree's Power LoRA Loader node so you can see each LoRA's info, strength, advice, etc.

weighted/imatrix quants of https://huggingface.co/Cyclone-Labs/Lucent-Witch-31B For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/Lucent-Witch-31B-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 Artefact2's thoughts on the…

Model · Text generation

LUHANADJolTctKlt

Priyansh Singhal

This seedoss model was trained 2x faster with Unsloth and Huggingface's TRL library.

Model

LUHANADJolTctKlt-lora

Priyansh Singhal

This seedoss model was trained 2x faster with Unsloth

LoRA adapter 已融合进基座，权重是 bf16 全精度，没有经过任何量化。 MLX 和 transformers 都能直接加载——它是标准 safetensors 格式，所以一套权重两种用法。 一句话：要在它身上"再加工"，就必须用这个版本；只是想用它，4bit 更省。 命令行也行（mlxlm.generate 自己会套模板）： 和 4bit 版完全一致——模型输出裸 JSON： 完整接入代码（含多步循环、JSON 兜底解析）见主仓库的 1. 不套 chat template 会复读。 直接 generate(model, tok, prompt="递归是什么？") 会得到类似这样的东西： 这不是模型坏了，是你喂的是裸补全而不是对话。用 tok.applychattemplate(...)， 或者干脆用 mlxlm.generate 命令行。 2. enablethinking 记得显式传 False。 模板里有 开关，不传的话没有默认值，输出可能带游离的 。 想要思考内容就传 True，它会单独出现（不混在回答里）。 3. 加载比 4bit 慢、占内存多——这是正常的。 实测对比（MacBook Pro M4 / 16GB）： ① 无工具时算术会算错 → 在 system 里要求"一律调 calculate"； ② 通用 system 下身份可能答错 → system 里写清身份； - 基座：MiniCPM5-2B（面壁智能 OpenBMB） - 许可：Apache 2.0 个人项目，欢迎提 issue。有用的话 Bilibili 关注一下

Our model's name is YuE (乐). In Chinese, the word means "music" and "happiness." Some of you may find words that start with Yu hard to pronounce. If so, you can just call it "yeah." We wrote a song with our model's name. YuE is a groundbreaking series of open-source foundation models designed for music generation, specifically for transforming lyrics into full songs (lyrics2song). It can generate a complete song, lasting several minutes, that includes both a catchy vocal track and accompaniment track. YuE is capable of modeling diverse genres/languages/vocal techniques. Please visit the Demo Page for amazing vocal performance. Join Us on Discord! 2025.03.12 Paper Released: We now release…

Model · Image to image

MadStickArt-v0.1.0

Danger Room

Try MadStickArt online → — generate a storyboard in your browser before downloading. Adjust composition and camera controls, then export clean or annotated PNGs and scene JSON. The playground runs the released neural decoder locally and uses optional rule-based scene suggestions; Strands Decider is available in the downloadable Python pipeline. MadStickArt is a small, trained 12,192-parameter image model for clean stick-figure storyboard scenes at 956 × 400 pixels, exactly 2.39:1. This first public release is danger-room/MadStickArt-v0.1.0, licensed under Apache 2.0. Pin revision v0.1.0 for reproducible downloads. The model is a custom PyTorch layout-conditioned neural raster decoder. A…

Model · Image to image

MadStickArt-v0.8.0

Danger Room

MadStickArt is a trained, 61,296-parameter neural raster decoder for stick-figure storyboard scenes. It reconstructs 956 × 400 grayscale frames from four explicit layout masks: actors, setting, props, and motion. Its output contract is 2.39:1. Try MadStickArt online → — the browser trial is deployed separately. Its deployed model/version and renderer are recorded in the Space's model-provenance.json. The renderer included with this release supports 186 settings and 320 props; the deployed trial has its own model and renderer pins and is updated through a separate reviewed stage. Browser suggestions use transparent rules, while Strands decisions run in the downloadable Python pipeline. The…

Model

mae-baseline-2024

Tobias Schwarz

A small Mae implementation for Multitask, packaged with an explicit configuration and an initialization checkpoint. The huge variant is a reproducible starting point, not a trained model release. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses adam with a constant warmup schedule. These are starting values in the script, not…

Model

mae-finetuned

Lauren Wood

This is an experimental Mae codebase for Retrieval. It keeps the base setup intentionally manageable so architecture changes can be inspected before a full training run. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses rmsprop with a constant warmup schedule. These are starting values in the script, not evidence of a completed run.…

Model

mae-generation-large

Chunhung Lee

A research-oriented Mae prototype targeting Generation. The included large setup documents defaults and file formats without presenting unverified performance numbers. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses sgd with a exponential schedule. These are starting values in the script, not evidence of a completed run. For a…