Open-Weight Models

Model · Text generation

DualMinded-Qwen3-1.7B

Convergent Intelligence

A 1.7B parameter dual-cognition model trained on Opus 4.6 reasoning traces. The model implements a three-phase cognitive loop — explore, examine, respond — where it reasons freely, critiques its own reasoning, then synthesizes a clean answer. This is the multi-model collision array collapsed into a single architecture. The dialectical structure that produces novel insights from architectural diversity is recreated through role-conditioned generation on shared weights. No extra parameters, no routing — same weights, different cognitive modes. DualMinded-Qwen3-1.7B is the product of a four-stage pipeline: Stage 1 — Multi-Teacher Distillation: Qwen3-30B-A3B in three variants (Instruct…

Model · Text generation

LFM2.5-1.2B-Distilled-SFT

Convergent Intelligence

A 1.2B hybrid model (SSM + attention) built in two stages: knowledge distillation from a 24B MoE hybrid teacher on STEM chain-of-thought data, then supervised fine-tuning on logical inference. The first proof-weighted distillation + SFT pipeline on a non-transformer architecture. Liquid Foundation Models run at 239 tok/s on AMD CPU and fit under 1GB of RAM. This model adds structured STEM reasoning and formal logical inference to that efficiency substrate. LFM2.5-1.2B distilled from LFM2-24B-A2B — a 24B MoE hybrid (SSM + attention) with only 2B active parameters per token. Teacher and student share the LFM hybrid architecture, so the KL divergence transfers reasoning patterns between…

Model · Text generation

GPT-X2-125M-CIx-Long-Context

Convergent Intelligence

This model is a custom-code derivative of AxiomicLabs/GPT-X2-125M, adapted for experimental long-context causal language modeling and architecture research. The repository includes a Hugging Face Transformers-compatible GPT-X2 implementation with optional Symplectic Metric-RoPE Governor support and training utilities built around CIxOpt, a heterogeneous optimizer developed for efficient parameter routing across large projection matrices, sensitive normalization parameters, and optional governor modules. The model is intended as a research checkpoint for compact long-context generation, positional encoding experiments, optimizer testing, and continued fine-tuning. This implementation uses a…

Model · Text generation

DualMind

Convergent Intelligence

Single Architecture, Dual Cognition — The Multi-Model Collision Array on Shared Weights DualMind is a 1.7B parameter model that implements dual-mental-modality reasoning — a single model with two internal voices sharing the same weights, differentiated only by role tokens: - — Unconstrained reasoning. Derivation, speculation, working through the problem freely. - — Adversarial self-response. The model reads its own explore output and critiques it. Error detection, verification, refinement. - — Clean synthesis. The final answer distilled from the internal dialogue. This is the multi-model collision array collapsed into a single architecture. The dialectical structure that produces novel…

Model · Summarization

bigbird-pegasus-large-pubmed

Google

BigBird, is a sparse-attention based transformer which extends Transformer based models, such as BERT to much longer sequences. Moreover, BigBird comes along with a theoretical understanding of the capabilities of a complete transformer that the sparse model can handle. BigBird was introduced in this paper and first released in this repository. Disclaimer: The team releasing BigBird did not write a model card for this model so this model card has been written by the Hugging Face team. BigBird relies on block sparse attention instead of normal attention (i.e. BERT's attention) and can handle sequences up to a length of 4096 at a much lower compute cost compared to BERT. It has achieved SOTA…

Model · Question answering

xlm-roberta-base-squad2-distilled

Deepset

Haystack's distillation feature was used for training. deepset/xlm-roberta-large-squad2 was used as the teacher model. Haystack is an AI orchestration framework to build customizable, production-ready LLM applications. You can use this model in Haystack to do extractive question answering on documents. To load and run the model with Haystack: For a complete example with an extractive question answering pipeline that scales over many documents, check out the corresponding Haystack tutorial. Evaluated on the SQuAD 2.0 dev set Timo Möller: [email protected] deepset is the company behind the production-ready open-source AI framework Haystack. We also have a Discord community open to…

Model · Translation

plt5-base

Allegro ML Research

plT5 models are T5-based language models trained on Polish corpora. The models were optimized for the original T5 denoising target. plT5 was trained on six different corpora available for Polish language: The training dataset was tokenized into subwords using a sentencepiece unigram model with vocabulary size of 50k tokens. If you use this model, please cite the following paper: The model was trained by Machine Learning Research Team at Allegro and Linguistic Engineering Group at Institute of Computer Science, Polish Academy of Sciences. You can contact us at: [email protected]

Model · Robotics

openvla-oft-libero

Jiaming Tang

This repository contains the OpenVLA-OFT checkpoint for LIBERO-Spatial, as described in Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success. OpenVLA-OFT significantly improves upon the base OpenVLA model by incorporating optimized fine-tuning techniques. See here for other OpenVLA-OFT checkpoints: https://huggingface.co/moojink?searchmodels=oft This example demonstrates generating an action chunk using a pretrained OpenVLA-OFT checkpoint. Ensure you have set up the conda environment as described in the GitHub README.

Model · Image segmentation

sae-2026-hang-all-yolo26n-seg-v2-960

Black Bee Drones

Instance-segmentation model for the "hang the wire" hook mission of the SAE Brasil Eletroquad 2026 competition, trained by Black Bee Drones. It segments two classes - rose (each visible segment of the suspended rope) and sphere (the orange sphere on the rope) - and runs on every control-loop tick of the mission. The team finished 2nd overall The drone takes off, finds the orange sphere mounted on one of two suspended ropes, parks a fixed distance from it, picks which side of the rope to fly along, turns perpendicular to the rope, descends on LIDAR, releases a hook with a servo, and lands. Because each visible rope segment is its own rose instance, the controller can measure both rope arms…

Model · Summarization

bert-mini2bert-mini-finetuned-cnn_daily_mail-summarization

Manuel Romero

This model is a warm-started BERT2BERT (mini) model fine-tuned on the CNN/Dailymail summarization dataset. The model achieves a 16.51 ROUGE-2 score on CNN/Dailymail's test dataset. For more details on how the model was fine-tuned, please refer to this notebook.

Model · Text generation

Disctil-Qwen3-1.7B

Convergent Intelligence

This model is a fine-tuned version of reaperdoesntknow/DiStil-Qwen3-1.7B-uncensored. It has been trained using TRL. This model was trained with SFT. This model is the DISC-refined node in the DistilQwen distillation chain. Discrepancy Calculus is a measure-theoretic framework that quantifies mismatch between integration and differentiation via the discrepancy operator: $$Df(x) = \lim{\varepsilon \downarrow 0} \frac{1}{\varepsilon} \intx^{x+\varepsilon} \frac{|f(t) - f(x)|}{|t - x|}\, dt$$ DISC refinement applies the Mesh Fundamental Identity decomposition ($f = \text{AC} + \text{jumps} + \text{Cantor}$) to the model's weight space, identifying and preserving structural boundaries that…

Model · Text generation

DeepReasoning_1R

Convergent Intelligence

Part of the Standalone Models by Convergent Intelligence LLC: Research Division DistilQwen Collection — Our only BF16 series. Proof-weighted distillation from Qwen3-30B-A3B → 1.7B and 0.6B on H100. Three teacher variants (Instruct, Thinking, Coder), nine models, 2,788 combined downloads. The rest of the portfolio proves structure beats scale on CPU. This collection shows what happens when you give the methodology real hardware. This model is part of the Convergent Intelligence LLC: Research Division portfolio. All models in this portfolio are developed under the Discrepancy Calculus (DISC) framework — a measure-theoretic approach to understanding and controlling the gap between what a model…

Model · Tabular classification

tabpfn_3_5

Prior Labs

TabPFN-3.5 is a transformer-based foundation model that uses in-context learning to solve tabular prediction problems in a forward pass. One checkpoint serves both classification and regression. Inference code can be found at https://github.com/PriorLabs/TabPFN. Fitting a classifier and predicting looks like this: TabPFNRegressor works the same way, off the same checkpoint. For more examples, see the github repo: https://github.com/PriorLabs/TabPFN. To use a specific checkpoint, pass its filename via modelpath: Every checkpoint ships with basic text and datetime column handling enabled (TRANSFORMTEXT, TRANSFORMDATES) and a feature limit of 20,000. Developed by Prior Labs. Regression and…

Model · Zero-shot classification

deberta-v3-large-zeroshot-v2.0-c

Moritz Borrett-Laurer (formerly Laurer)

Models in this series are designed for efficient zeroshot classification with the Hugging Face pipeline. These models can do classification without training data and run on both GPUs and CPUs. An overview of the latest zeroshot classifiers is available in my Zeroshot Classifier Collection. The main update of this zeroshot-v2.0 series of models is that several models are trained on fully commercially-friendly data for users with strict license requirements. These models can do one universal classification task: determine whether a hypothesis is "true" or "not true" given a text (entailment vs. notentailment). This task format is based on the Natural Language Inference task (NLI). The task is…

Model · Zero-shot classification

nli-deberta-base

Sentence Transformers - Cross-Encoders

This model was trained using SentenceTransformers Cross-Encoder class. The model was trained on the SNLI and MultiNLI datasets. For a given sentence pair, it will output three scores corresponding to the labels: contradiction, entailment, neutral. For evaluation results, see SBERT.net - Pretrained Cross-Encoder. Pre-trained models can be used like this: You can use the model also directly with Transformers library (without SentenceTransformers library): This model can also be used for zero-shot-classification

Model · Robotics

openvla-7b-oft-finetuned-libero-spatial-object-goal-10

Moo Jin Kim

This repository contains the OpenVLA-OFT checkpoint trained on 4 LIBERO task suites combined (-Spatial, -Object, -Goal, -Long), as described in Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success. OpenVLA-OFT significantly improves upon the base OpenVLA model by incorporating optimized fine-tuning techniques. See here for other OpenVLA-OFT checkpoints: https://huggingface.co/moojink?searchmodels=oft This example demonstrates generating an action chunk using a pretrained OpenVLA-OFT checkpoint. Ensure you have set up the conda environment as described in the GitHub README.

Model · Fill mask

barthez

Moussa Kamal Eddine

A french sequence to sequence pretrained model based on BART. BARThez is pretrained by learning to reconstruct a corrupted input sentence. A corpus of 66GB of french raw text is used to carry out the pretraining. Unlike already existing BERT-based French language models such as CamemBERT and FlauBERT, BARThez is particularly well-suited for generative tasks (such as abstractive summarization), since not only its encoder but also its decoder is pretrained. In addition to BARThez that is pretrained from scratch, we continue the pretraining of a multilingual BART mBART which boosted its performance in both discriminative and generative tasks. We call the french adapted version mBARThez.

Model · Video classification

vi-sign-language-videomae-base

Star Duong

This repository houses a fine-tuned VideoMAE (Base) model optimized for multi-class Vietnamese Sign Language Recognition (VSLR). The model architecture adapts self-supervised video representations to accurately classify short video clips of sign gestures into distinct Vietnamese text labels. The model processes short video sequences by partitioning them into spatiotemporal patches, mapping sequential gestures (such as "Ăn", "Bệnh viện", "Xin lỗi") to their corresponding semantic classes. The training routine was monitored closely across key evaluation metrics to prevent overfitting while maximizing classification accuracy on the validation split. The plot below illustrates the progression…

Model · Robotics

act_aloha_sim_transfer_cube_human

LeRobot

Action Chunking Transformer Policy (as per Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware) trained for the AlohaTransferCube environment from gym-aloha. See the LeRobot library (particularly the evaluation script) for instructions on how to load and evaluate this model. Trained with LeRobot@3c0a209. The model was trained using LeRobot's training script and with the alohasimtransfercubehuman dataset, using this command: The training curves may be found at https://wandb.ai/aliberts/lerobot/runs/720l37xb. The current model corresponds to the checkpoint at 80k steps. This took about 1h45 to train on an Nvida A100. The model was evaluated on the AlohaTransferCube task from…

Model · Robotics

diffusion_pusht

LeRobot

Diffusion Policy (as per Diffusion Policy: Visuomotor Policy Learning via Action Diffusion) trained for the PushT environment from gym-pusht. See the LeRobot library (particularly the evaluation script) for instructions on how to load and evaluate this model. Trained with LeRobot@3c0a209. The model was trained using LeRobot's training script and with the pusht dataset, using this command: The training curves may be found at https://wandb.ai/aliberts/lerobot/runs/s7elvf4r. The current model corresponds to the checkpoint at 175k steps. The model was evaluated on the PushT environment from gym-pusht and compared to a similar model trained with the original Diffusion Policy code. There are two…

Model · Zero-shot classification

deberta-base-long-nli

Tasksource

deberta-v3-base with context length of 1280 fine-tuned on tasksource for 250k steps. I oversampled long NLI tasks (ConTRoL, doc-nli). Training data include helpsteer v1/v2, logical reasoning tasks (FOLIO, FOL-nli, LogicNLI...), OASST, hh/rlhf, linguistics oriented NLI tasks, tasksource-dpo, fact verification tasks. This checkpoint has strong zero-shot validation performance on many tasks (e.g. 70% on WNLI), and can be used for: - Zero-shot entailment-based classification for arbitrary labels [ZS]. - Natural language inference [NLI] - Further fine-tuning on a new task or tasksource task (classification, token classification, reward modeling or multiple-choice) [FT]. Zero-shot GPT-4 scores…

Model · Text generation

next-1b

Lamapi

Next-1B is a 1-billion parameter causal language model based on Gemma 3, designed for efficiency, low-resource deployment, and reasoning-focused natural language understanding. Extremely lightweight — can run on consumer GPUs with low VRAM. Optimized for text reasoning, summarization, and creative generation. Supports Turkish natively while remaining multilingual. Open-source and transparent for research and applications. Ideal for developers, students, and organizations needing fast, reliable, and low-resource text-generation. 1. Lightweight Efficiency: Run smoothly on low-resource devices. 2. Reasoning-Focused: Provide logical and coherent text outputs. 3. Accessibility: Fully open-source…

This policy has been trained and pushed to the Hub using LeRobot. See the full documentation at LeRobot Docs. For a complete walkthrough, see the training guide. Below is the short version on how to train and run inference/eval: Writes checkpoints to outputs/train/ /checkpoints/. Prefix the dataset repo with eval\ and supply --policy.path pointing to a local or hub checkpoint.

Model · Text generation

Discovery

Convergent Intelligence

A 70M parameter causal language model built on the Mixture-of-Attentions (MoA) architecture — distance-based metric attention that respects the triangle inequality by construction, not approximation. Every attention head operates in a proper metric space. The geometry is enforced, not hoped for. Standard transformers compute attention as a dot product: Q·Kᵀ. This has no geometric meaning — it's a bilinear form, not a distance. Two tokens can be "close" by dot product while violating basic metric properties. MoA replaces this with negative squared distance under a learned diagonal Mahalanobis metric, then enforces the triangle inequality through a regularizer over random triples sampled…