Open-Weight Models Under Apache License 2.0

Model · Audio classification

distil-wav2vec2-adult-child-cls-37m

Bookbot

DistilWav2Vec2 Adult/Child Speech Classifier is an audio classification model based on the wav2vec 2.0 architecture. This model is a distilled version of wav2vec2-adult-child-cls on a private adult/child speech classification dataset. This model was trained using HuggingFace's PyTorch framework. All training was done on a Tesla P100, provided by Kaggle. Training metrics were logged via Tensorboard. The model achieves the following results on evaluation: The following hyperparameters were used during training: - learningrate: 3e-05 - trainbatchsize: 32 - evalbatchsize: 32 - seed: 42 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 128 - optimizer: Adam with betas=(0.9,0.999) and…

Model · Text generation

mvp

AI Box

The MVP model was proposed in MVP: Multi-task Supervised Pre-training for Natural Language Generation by Tianyi Tang, Junyi Li, Wayne Xin Zhao and Ji-Rong Wen. The detailed information and instructions can be found https://github.com/RUCAIBox/MVP. MVP is supervised pre-trained using a mixture of labeled datasets. It follows a standard Transformer encoder-decoder architecture. MVP is specially designed for natural language generation and can be adapted to a wide range of generation tasks, including but not limited to summarization, data-to-text generation, open-ended dialogue system, story generation, question answering, question generation, task-oriented dialogue system, commonsense…

Status: training in progress. No weights are published yet — this card describes the recipe and the pilot results that motivate it. A ~1B masked-diffusion language model decoded with confidence-targeted steps, then spend a few extra passes rewriting only the tokens the model is least sure about. The point is inference cost. An autoregressive model needs one sequential forward pass per token. This one needs ~20 passes for a whole sequence, regardless of its length. Cost is K + R forward passes. One refill pass fixes any number of positions at once, because the model processes the whole sequence in parallel — that is what makes targeted repair cheaper than more denoising. Draft and refill are…

Model · Translation

t5-small-openvino

Ella Charlaix

T5 is an encoder-decoder model pre-trained on a multi-task mixture of unsupervised and supervised tasks and for which each task is converted into a text-to-text format. For more information, please take a look at the original paper. Paper: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer Authors: Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu You can use this model with Transformers pipeline.

Model · Audio classification

ced-mini

Speech Team, Xiaomi MiLM Plus

CED are simple ViT-Transformer-based models for audio tagging, achieving sota performance on Audioset. Notable differences from other available models include: 1. Simplification for finetuning: Batchnormalization of Mel-Spectrograms. During finetuning one does not need to first compute mean/variance over the dataset, which is common for AST. 1. Support for variable length inputs. Most other models use a static time-frequency position embedding, which hinders the model's generalization to segments shorter than 10s. Many previous transformers simply pad their input to 10s in order to avoid the performance impact, which in turn slows down training/inference drastically. 1. Training/Inference…

The RT-DETRv2 model was proposed in RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer by Wenyu Lv, Yian Zhao, Qinyao Chang, Kui Huang, Guanzhong Wang, Yi Liu. RT-DETRv2 refines RT-DETR by introducing selective multi-scale feature extraction, a discrete sampling operator for broader deployment compatibility, and improved training strategies like dynamic data augmentation and scale-adaptive hyperparameters. These changes enhance flexibility and practicality while maintaining real-time performance. This model was contributed by @jadechoghari with the help of @cyrilvallez and @qubvel-hf This is RT-DETRv2 consistently outperforms its predecessor across all…

Model · Summarization

distilbart-cnn-6-6

Joshua

https://huggingface.co/sshleifer/distilbart-cnn-6-6 with ONNX weights to be compatible with Transformers.js. If you haven't already, you can install the Transformers.js JavaScript library from NPM using: Note: Having a separate repo for ONNX weights is intended to be a temporary solution until WebML gains more traction. If you would like to make your models web-ready, we recommend converting to ONNX using Optimum and structuring your repo like this one (with ONNX weights located in a subfolder named onnx).

Model · Audio classification

dasheng-base

Speech Team, Xiaomi MiLM Plus

Dasheng (Deep Audio-Signal Holistic Embeddings), or “大声” ("great sound"), is a general-purpose audio encoder trained on a large-scale self-supervised learning task. Dasheng is designed to capture rich audio information across various domains, including speech, music, and environmental sounds. The model is trained on 272,356 hours of diverse audio data with 1.2 billion parameters, and exhibits significant performance gains on the HEAR benchmark. Dasheng outperforms previous works on CREMA-D, LibriCount, Speech Commands, VoxLingua, and competes well in music and environmental sound classification tasks. examplefinetuneesc50.ipynb demonstrates how to train a linear head on the ESC-50 dataset…

Model · Object detection

rf-detr-large

Roboflow

RF-DETR is a real-time detection transformer family introduced in RF-DETR: Neural Architecture Search for Real-Time Detection Transformers by Robinson et al. and integrated in Transformers via PR #36895. RF-DETR is an end-to-end object detection model that combines ideas from LW-DETR and Deformable DETR: a DINOv2-with-registers style ViT backbone (with an RF-DETR windowing pattern for efficient attention), a multi-scale projector between encoder and decoder, and a multi-scale deformable DETR decoder for fast convergence and strong accuracy–latency tradeoffs. You can use the raw model for object detection. See the model hub to look for all available RF-DETR models. Here is how to use this…

Model · Object detection

dfine-small-coco

University of Science and Technology of China

The D-FINE model was proposed in D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu This model was contributed by VladOS95-cyber with the help of @qubvel-hf This is the HF transformers implementation for D-FINE coco -> model trained on COCO obj365 -> model trained on Object365 obj2coco -> model trained on Object365 and then finetuned on COCO D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global…

Model · Question answering

distilbert-onnx

Philipp Schmid

This model is a fine-tune checkpoint of DistilBERT-base-cased, fine-tuned using (a second step of) knowledge distillation on SQuAD v1.1. This model reaches a F1 score of 87.1 on the dev set (for comparison, BERT bert-base-cased version reaches a F1 score of 88.7).

This is a ported version of The base model is hubert-base-ls960, which is pretrained on 16kHz sampled speech audio. When using the model make sure that your speech input is also sampled at 16Khz. For more information refer to SUPERB: Speech processing Universal PERformance Benchmark Emotion Recognition (ER) predicts an emotion class for each utterance. The most widely used ER dataset IEMOCAP is adopted, and we follow the conventional evaluation protocol: we drop the unbalanced emotion classes to leave the final four classes with a similar amount of data points and cross-validate on five folds of the standard splits. For the original model's training and evaluation instructions refer to the…

Model · Translation

t5-small

Hugging Face Optimum

T5 is an encoder-decoder model pre-trained on a multi-task mixture of unsupervised and supervised tasks and for which each task is converted into a text-to-text format. For more information, please take a look at the original paper. Paper: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer Authors: Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu You can use this model with Transformers pipeline.

Model · Time series forecasting

chronos-t5-small

Autogluon

Update Feb 14, 2025: Chronos-Bolt & original Chronos models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code. Update Nov 27, 2024: We have released Chronos-Bolt models that are more accurate (5% lower error), up to 250 times faster and 20 times more memory-efficient than the original Chronos models of the same size. Check out the new models here. Chronos is a family of pretrained time series forecasting models based on language model architectures. A time series is transformed into a sequence of tokens via scaling and quantization, and a language model is trained on these…

Model · Object detection

dfine-xlarge-obj2coco

University of Science and Technology of China

The D-FINE model was proposed in D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu This model was contributed by VladOS95-cyber with the help of @qubvel-hf This is the HF transformers implementation for D-FINE coco -> model trained on COCO obj365 -> model trained on Object365 obj2coco -> model trained on Object365 and then finetuned on COCO D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global…

Model · Audio classification

spkrec-xvect-voxceleb

SpeechBrain

This repository provides all the necessary tools to extract speaker embeddings with a pretrained TDNN model using SpeechBrain. The system is trained on Voxceleb 1+ Voxceleb2 training data. For a better experience, we encourage you to learn more about SpeechBrain. The given model performance on Voxceleb1-test set (Cleaned) is: This system is composed of a TDNN model coupled with statistical pooling. The system is trained with Categorical Cross-Entropy Loss. First of all, please install SpeechBrain with the following command: Please notice that we encourage you to read our tutorials and learn more about The system is trained with recordings sampled at 16kHz (single channel). The code will…

Model · Text generation

Fastino-Nemotron-3.5-Lightning-Finance

Fastino

Fastino-Nemotron-3.5-Lightning-Finance is a 30B-parameter, 3B-active mixture-of-experts model specialized for financial reasoning, extraction, and research fine-tuned on LoRA with the Fastino Fine-Tuning Agent. The model targets financial document reasoning, numerical question answering over filings and tables, numeric span extraction, financial entity recognition, conversational analysis, and source-grounded financial research. The evaluation suite includes FinQA, TAT-QA, SEC-Num, FinEntity, BizFinBench, BigFinanceBench, ConvFinQA, and FiQA. The published weights are BF16 and require about 66 GB before runtime overhead. An 80 GB or larger GPU, or tensor parallelism across multiple GPUs, is…

Model · Zero-shot classification

nli-roberta-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 · Object detection

dfine-medium-coco

University of Science and Technology of China

The D-FINE model was proposed in D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu This model was contributed by VladOS95-cyber with the help of @qubvel-hf This is the HF transformers implementation for D-FINE coco -> model trained on COCO obj365 -> model trained on Object365 obj2coco -> model trained on Object365 and then finetuned on COCO D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global…

Model · Text generation

Qwen3.5-0.8-Cyber-GGUF

Convergent Intelligence

A compact Qwen3.5 0.8B repository with a practical GGUF quantization ladder for local inference. This card describes what is present in the repository. The public files do not document the fine-tuning dataset or provide evaluation results, so the Cyber label should be read as the repository variant name—not as a verified capability claim. With a recent llama.cpp build: The repository includes a BF16 mmproj file and its configuration includes vision components. That establishes that a projector artifact is present; it does not establish that the end-to-end multimodal path was validated for this release. Verify image input locally before depending on it. - Local experimentation with a small…

Model · Robotics

smolvla_robotwin

LeRobot

SmolVLA is a compact, efficient vision-language-action model that achieves competitive performance at reduced computational costs and can be deployed on consumer-grade hardware. 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.

This is a ported version of The base model is hubert-large-ll60k, which is pretrained on 16kHz sampled speech audio. When using the model make sure that your speech input is also sampled at 16Khz. For more information refer to SUPERB: Speech processing Universal PERformance Benchmark Emotion Recognition (ER) predicts an emotion class for each utterance. The most widely used ER dataset IEMOCAP is adopted, and we follow the conventional evaluation protocol: we drop the unbalanced emotion classes to leave the final four classes with a similar amount of data points and cross-validate on five folds of the standard splits. For the original model's training and evaluation instructions refer to the…

Model · Audio classification

Common-Voice-Gender-Detection-ONNX

Prithiv Sakthi

This is an ONNX version of prithivMLmods/Common-Voice-Gender-Detection. It was automatically converted and uploaded using this space. Wav2Vec2: Self-Supervised Learning for Speech Recognition: https://arxiv.org/pdf/2006.11477 Common-Voice-Gender-Detection is designed for: Speech Analytics – Assist in analyzing speaker demographics in call centers or customer service recordings. Conversational AI Personalization – Adjust tone or dialogue based on gender detection for more personalized voice assistants. Voice Dataset Curation – Automatically tag or filter voice datasets by speaker gender for better dataset management. Research Applications – Enable linguistic and acoustic research involving…

Model · Zero-shot classification

bart-large-mnli-yahoo-answers

Joe Davison

This model takes facebook/bart-large-mnli and fine-tunes it on Yahoo Answers topic classification. It can be used to predict whether a topic label can be assigned to a given sequence, whether or not the label has been seen before. You can play with an interactive demo of this zero-shot technique with this model, as well as the non-finetuned facebook/bart-large-mnli, here. This model was fine-tuned on topic classification and will perform best at zero-shot topic classification. Use hypothesistemplate="This text is about {}." as this is the template used during fine-tuning. For settings other than topic classification, you can use any model pre-trained on MNLI such as facebook/bart-large-mnli…

Model · Robotics

GigaBrain-0.7-3.5B-Base

GigaAI

Vision-language-action (VLA) models have become a dominant paradigm for generalist embodied agents, demonstrating strong complex and long-horizon task completion in structured settings. Yet it remains an open question whether current VLA systems can benefit from more effective architectural design, scale to substantially larger and more heterogeneous data regimes, and achieve broader generalization across tasks and embodiments. To this end, we present GigaBrain-0.7, an embodied foundation model with substantially improved generalization across diverse robot embodiments. Specifically, GigaBrain-0.7 unifies understanding, prediction, and action through a three-system architecture, scales…

Model · Text generation

Qwen3.5-2B-CyberSec-GGUF

Convergent Intelligence

GGUF exports of reaperdoesntknow/Qwen3.5-2B-CyberSec for local inference with llama.cpp-compatible runtimes. The source model is associated with the Trendyol Cybersecurity Instruction Tuning Dataset. No benchmark or safety-evaluation results are published with this GGUF release. A BF16 projector file is present, and the source-model configuration includes vision components. This release does not include a documented multimodal smoke-test receipt. Verify the projector, prompt format, runtime version, and image path before claiming multimodal support. - Local qualitative evaluation of the source checkpoint. - CPU or consumer-GPU experimentation. - Comparison of BF16, Q80, and Q4KM output…

Model · Zero-shot classification

gliner-guard-omni

HiveTraceLab

One encoder model that replaces your entire guardrail stack: safety classification, PII detection, adversarial attack detection, intent and tone analysis — all in a single forward classification, NER and more · no LLM required Install dependencies Classify Harmful messages and Detect PII via single forward pass GLiNER Guard Omni fine-tunes fastino/gliner2-multi-v1 on our guardrail taxonomy while preserving its multilingual zero-shot generalization. You get GLiNER Guard's safety understanding on top of the base model's ability to handle labels and domains beyond the training set — so you can define custom policies with nothing but natural language descriptions. For specific usecases you can…

Model · Audio classification

Qwen3-ForcedAligner-0.6B-4bit

Ivan

4-bit quantized version of Qwen/Qwen3-ForcedAligner-0.6B for Apple Silicon inference via MLX. Predicts word-level timestamps for audio+text pairs in a single non-autoregressive forward pass. Unlike ASR (autoregressive, token-by-token), the forced aligner runs the entire sequence in one forward pass through the decoder. The classify head predicts a timestamp class (0–4999) at each token position, which maps to time via classindex × 80ms. This model is designed for use with speech-swift: Text decoder (attention projections, MLP, embeddings) quantized to 4-bit using group quantization (groupsize=64). Audio encoder and classify head kept as float16 for accuracy.

Model · Summarization

rut5_base_headline_gen_telegram

Ilya Gusev

Based on rut5-base model

Model · Image segmentation

manga109-segmentation-bubble

Vũ tiến huy

A high-performance YOLO11n instance segmentation model fine-tuned for detecting and segmenting speech bubbles in manga/comic images. Left: Segmentation Loss (Train vs Val) | Right: Mask mAP Metrics over epochs This model was trained on a combined dataset of: 1. MS92/MangaSegmentation - Manga panel and bubble segmentation dataset 2. Manga109 - Large-scale manga dataset with speech bubble annotations If you use this model in your research, please cite: This model is released under the Apache 2.0 License. - Ultralytics for the YOLO framework - MS92/MangaSegmentation dataset - Manga109 dataset

Model

octomind-embed

Muvon Un Limited

Embedding model for octomind capability / skill auto-activation: ibm-granite/granite-embedding-30m-english (30M params, 6 layers, 384-dim, CLS-pooled, prefix-free, English) fine-tuned on trigger phrases from the octomind-tap capabilities + skills catalog and blended back into the base as a WiSE-FT model soup, which beats both the base and the raw fine-tune on the runtime gate (mean-of-top-3 cosine + threshold + margin). CachedMultipleNegativesRankingLoss (scale 10) on in-class pairs and positive-aware hard-negative triplets, MatryoshkaLoss over [384, 256, 192, 128, 96], then weight interpolation with the base. - model.safetensors + 1Pooling/ — sentence-transformers layout (fp32).…

Model · Text generation

TopologicalQwen

Convergent Intelligence

Topology-Aware Knowledge Distillation from Qwen3-30B-A3B → 1.7B TopologicalQwen is a 1.7B parameter model distilled from Qwen3-30B-A3B using Topological Knowledge Distillation (TKD) — a methodology that treats the teacher's output distribution over a concatenated token stream as a bounded variation (BV) function and decomposes knowledge transfer into three channels via the Mesh Fundamental Identity: 1. Smooth distillation (AC component) — Standard KL divergence over regions where the teacher's distribution varies continuously. This is what every other KD method does and stops at. 2. Jump corrections (D^j f) — Explicit correction terms at conceptual boundaries where the teacher's…

Model · Text generation

TameForCasualLM

Convergent Intelligence

With Blackhole Rope Dynamics This model builds on the original 421M TAMELM-AFMoER by introducing the Blackhole Rope (BHR) mechanism—a dynamic field-based routing system designed to stabilize, amplify, and concentrate information flow across multiple temporal scales. While the original AFMoER established efficiency in routing-based intelligence, the BHR variant explores how structured gravitational-like attractors can further enhance reasoning depth without exponential increases in computation or parameters. The Blackhole Rope is a symplectic, multiscale vortex mechanism inside AFMoER that: If AFMoER routes are like neuronal pathways, the Blackhole Rope is the myelinated tether that keeps…

Model · Summarization

rut5_base_sum_gazeta

Ilya Gusev

This is the model for abstractive summarization for Russian based on rut5-base. Source maxlength: 600 Target maxlength: 200 norepeatngramsize: 4 numbeams: 5 Source maxlength: 600 Target maxlength: 200 norepeatngramsize: 4 numbeams: 5

Model · Text generation

Shepherd-Alpha

Convergent Intelligence

The first defense AI reasoning model on Hugging Face. Shepherd-Alpha is a tactical reasoning model fine-tuned on dual-perspective military scenario analysis using BiCell Depth Dispersal — a novel training methodology that partitions transformer layers by abstraction depth and trains them asymmetrically to separate representation encoding from task-specific reasoning. Developed by Convergent Intelligence LLC: Research Division Given a tactical scenario, Shepherd-Alpha produces structured dual-perspective analysis: - Attack reasoning — how an adversary would exploit the situation - Defense reasoning — how to counter, mitigate, and survive The model is trained to think like both attacker and…

Model · Text generation

SMOLM2Prover

Convergent Intelligence

SmolLM2Prover is a specialized, fine-tuned version of prithivMLmods/SmolLM2-CoT-360M. While retaining the strong conversational abilities of its base model, this version has been specifically enhanced to excel at deep thinking, logical reasoning, and higher-level mathematics, with a focus on generating step-by-step proofs and explanations (Chain-of-Thought). The model was fine-tuned using multiple rounds of Supervised Fine-Tuning (SFT) with the TRL library on a curated dataset, enhancing its ability to follow complex instructions and reason through problems. This model is intended to be used for text generation tasks that require logical reasoning or advanced conversation. The easiest way…

This model is a fine-tuned version of t5-small on the cnndailymail dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 5.6e-05 - trainbatchsize: 8 - evalbatchsize: 8 - lrschedulertype: linear - numepochs: 8 - Transformers 4.14.0 - Pytorch 1.5.0 - Datasets 2.3.2 - Tokenizers 0.10.3

Model · Text generation

SharperSwarm

Convergent Intelligence

SAGI (Swarm AGI) is a novel causal language model that integrates swarm intelligence dynamics with transformer architecture. The model treats cognition as a dynamic, adaptive system where multiple internal "agents" collaborate through differentiable routing, trust mechanisms, and shared memory. V3.2 introduces a revolutionary Self-Assessment Layer, allowing the system to predict its own performance, identify skill gaps, and autonomously design its own learning curriculum. 1. Pre-Assessment: Predict success, identify risks, recommend strategy. 2. Execution: Generate with selected strategy. 3. Real-Time Monitoring: Catch and correct errors during generation. 4. Post-Assessment: Update skill…

Model · Robotics

INTACT-pi0-finetune-bridge

Juexiao Zhang

This repository contains a checkpoint of the Pi0 model (HF implementation | Paper) finetuned on the BridgeV2 dataset for robotic manipulation tasks. The model is later used for testing on the Simpler Environment and our INTACT Probing Suite for the generalization boundaries of VLA models. Paper: From Intention to Execution: Probing the Generalization Boundaries of Vision-Language-Action Models Or directly in python with Lerobot, see blow: First, install lerobot Then For more details please refer to our paper and code Checkpoint choice After training 15 epochs, we sweep the checkpoint at epoch 1, 2, 3, 4, 5, 10, 15 for performance on the original 4 Bridge tasks in the SimplerEnv, and choose…

Model · Text generation

SAGI

Convergent Intelligence

SAGI is a novel causal language model that integrates swarm intelligence dynamics with transformer architecture. The model treats cognition as a dynamic, adaptive system where multiple internal "agents" collaborate through differentiable routing, trust mechanisms, and shared memory. - Episodic + Semantic Memory: Dual memory system with trainable retrieval utility The swarm processes observations derived from token embeddings, updating its internal state S. This state conditions the transformer's attention patterns and feed-forward activations via learned projections, creating bidirectional information flow between symbolic (tokens) and subsymbolic (swarm dynamics) processing. - Educational…

Model · Text generation

Qwen3.5-2B-Opus-Distil

Convergent Intelligence

A 2B-parameter Qwen3.5 fine-tune, part of the Opus-Distil line in the reaperdoesntknow open-weight portfolio. Text-generation / reasoning model, trained with Unsloth + Hugging Face TRL.

Model · Video classification

vjepa2-vitg-fpc64-384

AI at Meta

A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository. To run V-JEPA 2 model, ensure you have installed the latest transformers: V-JEPA 2 is intended to represent any video (and image) to perform video classification, retrieval, or as a video encoder for VLMs. To load a video, sample the number of frames according to the model. For this model, we use 64. To load an image, simply copy the image to the desired number of frames. For more code examples, please refer to the V-JEPA 2…

Model · Text generation

Qwen3-1.7B-Thinking-Distil

Convergent Intelligence

Extended Reasoning Distillation from Qwen3-30B-A3B-Thinking → 1.7B The most downloaded model in the Convergent Intelligence portfolio. Qwen3-1.7B-Thinking-Distil captures extended deliberation patterns from the Qwen3-30B-A3B Thinking teacher — the variant that generates long-form reasoning chains before committing to an answer — and compresses them into a 1.7B student via supervised fine-tuning on the longwriter-6k dataset. The Thinking teacher produces the richest signal of the three teacher variants in the DistilQwen family (Instruct, Thinking, Coder). Where Instruct distillation captures clean instruction-following and Coder captures hierarchical decomposition, Thinking distillation…

Model · Text generation

Qwen3.5-2B-CyberSec

Convergent Intelligence

An English Qwen3.5 2B checkpoint associated with the Trendyol Cybersecurity Instruction Tuning Dataset and exported in Transformers / Safetensors format. This release is intended for research and local experimentation. The repository does not currently publish benchmark or safety-evaluation results, so the model should not be treated as a validated cybersecurity authority. The configuration identifies a Qwen3.5 conditional-generation architecture with text and vision components. Use a recent Transformers release that supports this architecture. Dependency and device behavior can vary across Transformers versions. Pin a tested environment for reproducible use. - Research on small-model…

Model · Text generation

Qwen3-1.7B-Distilled-30B-A3B

Convergent Intelligence

A 1.7B-parameter causal language model distilled from Qwen3-30B-A3B on 6,122 STEM chain-of-thought samples using discrepancy-informed knowledge distillation. The training objective emphasizes proof structure, detects reasoning pivot tokens through token-level divergence dynamics, smooths high-entropy student singularities before distillation, and monitors structural drift through discrepancy energy. Standard knowledge distillation treats all tokens uniformly. Even proof-weighted approaches typically apply a static multiplier over the entire derivation span. That helps, but it still misses the internal structure of reasoning: some regions are smooth procedural continuation, while others are…

Model · Text generation

Qwen3-1.7B-Coder-Distilled-SFT

Convergent Intelligence

A 1.7B model built in two stages: knowledge distillation from a 30B Coder teacher to establish a structured reasoning backbone, then supervised fine-tuning on ~54,600 logical inference problems. The Coder teacher's decomposition patterns meet formal propositional logic. The hypothesis: a model that learned STEM derivation from a Coder teacher (Stage 1) already has latent structure for sequential logic, state tracking, and compositional reasoning. Logical inference SFT (Stage 2) activates that structure explicitly — the model doesn't learn logic from scratch, it surfaces what the Coder teacher already gave it. Qwen3-1.7B distilled from Qwen3-Coder-30B-A3B-Instruct — the coding-specialized…

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

Qwen3-0.6B-Distilled-30B-A3B-Thinking-SFT

Convergent Intelligence

A 0.6B parameter model built in two stages: knowledge distillation from a 30B Thinking teacher to establish a structured reasoning backbone, then supervised fine-tuning on legal instruction data. 50x compression. Under 500MB quantized. Runs on a phone. The training order is the thesis: teach the model how to reason first (distillation from Thinking teacher), then teach it what to reason about (legal SFT). The Thinking teacher's extended deliberation traces transfer deeper reasoning structure than an Instruct teacher — critical when the student has only 0.6B parameters to work with. Qwen3-0.6B distilled from Qwen3-30B-A3B-Thinking-2507 — a Mixture-of-Experts model with 30B total parameters…

Model · Zero-shot classification

ELISARCyberAIEdge7B-LoRA-GGUF

Dr Sabri ALLANI

Offline-ready, quantized LLaMA edge model for cybersecurity use cases ELISAR: An Adaptive Framework for Cybersecurity Risk Assessment Powered by GenAI - Sabri ALLANI, PhD – AI & Cybersecurity Expert - Karam BOU-CHAAYA, PhD – AI & Cybersecurity Expert - Helmi RAIS – Global Practice Lead, Expleo France May 31, 2025 This work will be published by Springer in the following book: ELISAR is a fine-tuned LoRA model based on Mistral-7B, designed for contextualized cybersecurity risk assessment using Retrieval-Augmented Generation and Agentic AI capabilities. The model targets real-world use cases including: ELISARCyberAIEdge7B-LoRA-GGUF is a LoRA-finetuned, GGUF-quantized version of the Mistral-7B…

Model · Text generation

Qwen3-0.6B-Distilled-30B-A3B

Convergent Intelligence

A 0.6B parameter model distilled from Qwen3-30B-A3B-Thinking on 6,122 STEM chain-of-thought samples. 50x parameter compression. The Thinking variant teacher produces richer extended reasoning traces than the Instruct variant, transferring deeper deliberation structure into the smallest possible student. The result: a model under 500MB quantized that produces structured STEM derivations because a 30B thinking model showed it how to reason. Two key differences from standard small-model distillation: 1. Thinking teacher, not Instruct teacher. The Qwen3-30B-A3B-Thinking variant generates extended internal reasoning before committing to an answer. Its softmax distributions are higher-entropy…

Model · Robotics

xvla-base

LeRobot

X-VLA is a Vision-Language-Action foundation model that uses soft prompts to handle cross-embodiment and cross-domain robot control within a unified Transformer architecture. Original paper: X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Model For full installation details (including optional video dependencies such as ffmpeg for torchcodec), see the official documentation: https://huggingface.co/docs/lerobot/installation If you’re training / fine-tuning, you typically call forward(...) to get a loss and then: - -policy.chunksize=... - -policy.nactionsteps=... - -policy.maxactiontokens=... - -policy.gradientcheckpointing=true You can use the record…

Model · Robotics

Giga-World-Policy-0.5

GigaAI

The MoT design keeps a visual expert stream (reference + future latents) and an action expert stream (state + action), with multi-modal self-attention across both. Weights are sharded at ~10GB per file. This repo contains the transformer only; runtime also needs the Wan2.2 VAE / scheduler from the base Diffusers checkpoint. For usage, training, and inference details, see our open source code page.

Model · Summarization

bert-small2bert-small-finetuned-cnn_daily_mail-summarization

Manuel Romero

This model is a warm-started BERT2BERT (small) model fine-tuned on the CNN/Dailymail summarization dataset. The model achieves a 17.37 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 · Image and text to text

RynnBrain-2B

DAMO Academy

[2026.02.02] Release RynnBrain family weights and inference code. - [2026.02.02] Add cookbooks for cognition, localization, reasoning, and planning. RynnBrain aims to serve as a physics-aware embodied brain: it observes egocentric scenes, grounds language to physical space and time, and supports downstream robotic systems with reliable localization and planning outputs. - Comprehensive egocentric understanding Strong spatial comprehension and egocentric cognition across embodied QA, counting, OCR, and fine-grained video understanding. - Diverse spatiotemporal localization Locates objects, target areas, and predicts trajectories across long episodic context, enabling global spatial…

Wan2.2-TI2V-5B video DiT + 48-joint action head, trainingmode=joint. The base is suhyeok's finalized B-method recipe: a teacher-forced (sigma=0.25) self-EMA teacher plus an iBOT prototype loss at L18 L18, gamma=0.01, two-view. On top of it the 3 PAST cond latent frames are pooled into one motion frame before a chosen block. These are NOT the surrogate ctxpool runs. The surrogate line (older base, pd8 x GA1) lives in hmkang/wamctxpoolxattn and hmkang/wamctxpoolavg. Do not compare across the two sets. Geometry: 4-latin (numframesin=25, numframesout=41, fdf 2) = 4 cond + 2 future latent slots, 96 tokens per latent frame, 576 tokens per row. Effective batch 16 clips x GA 2 x 2 views = 64 rows…

Model · Summarization

led-large-book-summary

Peter Szemraj

This model is a fine-tuned version of allenai/led-large-16384 on the BookSum dataset (kmfoda/booksum). It aims to generalize well and be useful in summarizing lengthy text for both academic and everyday purposes. - See the Colab demo linked above or try the demo on Spaces To improve summary quality, use encodernorepeatngramsize=3 when calling the pipeline object. This setting encourages the model to utilize new vocabulary and construct an abstractive summary. Load the model into a pipeline object: Feed the text into the pipeline object: Important: For optimal summary quality, use the global attention mask when decoding, as demonstrated in this community notebook, see the definition of…

Model · Text classification

opir-multitask-large-v1.0

Knowledgator Engineering

Opir-multitask-large is the English, highest-accuracy multi-task checkpoint in the Opir family: an encoder-based GLiClass guardrail model for real-time LLM safety filtering. It supports binary safe/unsafe classification, toxicity detection, jailbreak and prompt-injection detection, and zero-shot harmful-content categorization over a hierarchical safety taxonomy. This card is for knowledgator/opir-multitask-large. The model is used through GLiClass zero-shot classification: pass text plus the candidate labels you want scored. Use single-label mode for binary safe/unsafe decisions and multi-label mode for taxonomy, toxicity, jailbreak, or custom policy labels. Use multi-label mode when you…

Model · Text generation

MoA-155M

Convergent Intelligence

This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019). 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 should produce and what it actually produces. DISC treats training…

Model · Text generation

LFM2.5-8B-A1B-Opus-Distil

Convergent Intelligence

This model is a fine-tuned version of LiquidAI/LFM2.5-8B-A1B, adapted on the angrygiraffe/claude-opus-4.6-4.7-reasoning-8.7k dataset for English text-generation and reasoning-style responses. The fine-tuning run used a custom Convergent Intelligence optimizer stack, CIxOpt, designed for heterogeneous routing across parameter types. The goal of this checkpoint is to test whether a Liquid Foundation Model backbone can be adapted efficiently through targeted sparse participation rather than broad full-model modification. This is an experimental research checkpoint intended for continued evaluation, domain adaptation, and architecture/optimizer testing.…

Model · Zero-shot classification

deberta-v3-large-tasksource-nli

Damien Sileo

DeBERTa-v3-large fine-tuned with multi-task learning on 600 tasks of the tasksource collection You can further fine-tune this model to use it for any classification or multiple-choice task. This checkpoint has strong zero-shot validation performance on many tasks (e.g. 77% on WNLI). The untuned model CLS embedding also has strong linear probing performance (90% on MNLI), due to the multitask training. This is the shared model with the MNLI classifier on top. Its encoder was trained on many datasets including bigbench, Anthropic rlhf, anli... alongside many NLI and classification tasks with a SequenceClassification heads while using only one shared encoder. Each task had a specific CLS…