Multi-Scale Efficient Global Context Vision Transformer (MS-EffGCViT) is a hybrid CNN-ViT architecture for deepfake detection. It fuses CNN-driven spatial inductive bias with hierarchical global-context attention to catch both local artifacts (textures, blending seams) and global artifacts (lighting, structural inconsistency). A single architecture ships in two sizes and three domain-tuned checkpoints, working on both static images and video at the frame level. - Frame-level — one model handles both images and videos (frame-level inference + aggregation). - Cross-domain — robust on both East-Asian (KoDF) and Western (Celeb-DF-v2, FaceForensics++) faces. - Two variants — Fast (b0) for…
SAVRN Model Hub
Open-Weight Models
An open-weight model is an AI model whose trained weights are published for anyone to download. The weights are what the model learned in training. With a copy of them you can run the model on hardware you control and train it further on your own data.
Open weights are not the same as open source. Many publishers release the weights without the training data or code, and the license sets what you may do with the model. This library puts each model's full card, architecture, files, license and published evaluations on one page.
2,760Models
859Datasets
254Papers
1,692Publishers
5,040Sourced relationships
Updated 2026-09-18 · How the library is built
2,760 models, sorted by most downloaded.
Multi-Scale Efficient Global Context Vision Transformer (MS-EffGCViT) is a hybrid CNN-ViT architecture for deepfake detection. It fuses CNN-driven spatial inductive bias with hierarchical global-context attention to catch both local artifacts (textures, blending seams) and global artifacts (lighting, structural inconsistency). A single architecture ships in two sizes and three domain-tuned checkpoints, working on both static images and video at the frame level. - Frame-level — one model handles both images and videos (frame-level inference + aggregation). - Cross-domain — robust on both East-Asian (KoDF) and Western (Celeb-DF-v2, FaceForensics++) faces. - Two variants — Fast (b0) for…
Mask2Former model trained on COCO panoptic segmentation (base-sized version, Swin backbone). It was introduced in the paper Masked-attention Mask Transformer for Universal Image Segmentation and first released in this repository. Disclaimer: The team releasing Mask2Former did not write a model card for this model so this model card has been written by the Hugging Face team. Mask2Former addresses instance, semantic and panoptic segmentation with the same paradigm: by predicting a set of masks and corresponding labels. Hence, all 3 tasks are treated as if they were instance segmentation. Mask2Former outperforms the previous SOTA, MaskFormer both in terms of performance an efficiency by (i)…
1. pure torch and huggingface-based implementation of the original microsoft/BiomedCLIP-PubMedBERT256-vitbasepatch16224 2. rename the checkpoint state key names.
MaskFormer model trained on COCO panoptic segmentation (tiny-sized version, Swin backbone). It was introduced in the paper Per-Pixel Classification is Not All You Need for Semantic Segmentation and first released in this repository. Disclaimer: The team releasing MaskFormer did not write a model card for this model so this model card has been written by the Hugging Face team. MaskFormer addresses instance, semantic and panoptic segmentation with the same paradigm: by predicting a set of masks and corresponding labels. Hence, all 3 tasks are treated as if they were instance segmentation. You can use this particular checkpoint for semantic segmentation. See the model hub to look for other…
Model · Audio classification
speech-emotion-recognition-with-openai-whisper-large-v3
This project leverages the Whisper model to recognize emotions in speech. The goal is to classify audio recordings into different emotional categories, such as Happy, Sad, Surprised, and etc. The dataset used for training and evaluation is sourced from multiple datasets, including: The dataset contains recordings labeled with various emotions. Below is the distribution of the emotions in the dataset: This distribution reflects the balance of emotions in the dataset, with some emotions having more samples than others. Excluded the "calm" emotion during training due to its underrepresentation. The model used is the Whisper Large V3 model, fine-tuned for audio classification tasks: I map the…
This model is a fine-tuned version of mo-thecreator/wav2vec2-base-finetuned on the None dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 3e-05 - trainbatchsize: 8 - evalbatchsize: 8 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 32 - lrschedulertype: linear - lrschedulerwarmupratio: 0.1 - numepochs: 5 - Transformers 4.39.3 - Pytorch 2.1.2 - Datasets 2.18.0 - Tokenizers 0.15.2 - mo-thecreator
A fast and efficient 32B model optimized for CPU inference. The model was refactored with BitNet features and an updated tokenizer that includes new Routing, Media, Vision, Sound, Tool call, and Robotics tags. Built on a DeepSeek R1-32B architecture with native ternary (BitNet-style) support and ready-to-run GGUF quantizations. - JiRack is a cloud-ready model that helps save money on cloud infrastructure. It can be used as an expert model in RAG deployments, with the ONNX JiRack Java server as an alternative. - Benefits high quality CPU inference TQ2 on Llama.cpp and Ollama via QAT - Robotcs, Routing, Coding, Multimedia, Advanced tool calling via CMSManhattan/JiRackPrecisionTokenizer…
Multi-Scale Efficient Global Context Vision Transformer (MS-EffGCViT) is a hybrid CNN-ViT architecture for deepfake detection. It fuses CNN-driven spatial inductive bias with hierarchical global-context attention to catch both local artifacts (textures, blending seams) and global artifacts (lighting, structural inconsistency). A single architecture ships in two sizes and three domain-tuned checkpoints, working on both static images and video at the frame level. - Frame-level — one model handles both images and videos (frame-level inference + aggregation). - Cross-domain — robust on both East-Asian (KoDF) and Western (Celeb-DF-v2, FaceForensics++) faces. - Two variants — Fast (b0) for…
This is new updated version of Moirai-1.0-R (https://huggingface.co/Salesforce/moirai-1.0-R-base). The new Moirai model achieved significant improvements (~20%) for low-frequency cases like Yearly and Quarterly data in Normalised Mean Absolute Error (NMAE) for 40 datasets on the Monash repository. This release is for research purposes only in support of an academic paper. Our models, datasets, and code are not specifically designed or evaluated for all downstream purposes. We strongly recommend users evaluate and address potential concerns related to accuracy, safety, and fairness before deploying this model. We encourage users to consider the common limitations of AI, comply with…
This model is fine-tuned version of microsoft/conditional-detr-resnet-50. You can find details of model in this github repo -> fashion-visual-search And you can find fashion image feature extractor model -> yainage90/fashion-image-feature-extractor This model was trained using a combination of two datasets: modanet and fashionpedia The labels are ['bag', 'bottom', 'dress', 'hat', 'shoes', 'outer', 'top'] In the 96th epoch out of total of 100 epochs, the best score was achieved with mAP 0.7542. Therefore, it is believed that there is a little room for performance improvement.
Fine-tuned Wav2Vec2 model for detecting AI-generated speech. Determines if audio was spoken by a human or created by AI text-to-speech/voice cloning software. Fine-tuned Wav2Vec2 transformer for binary audio classification (real vs AI-generated speech). Trained to distinguish authentic human speech from synthetic audio generated by AI text-to-speech and voice cloning services including: Note: This model uses transfer learning from a base model already trained for deepfake detection. Fast convergence is expected due to task similarity and TTS engine overlap with the base model's training data. The model outputs logits (raw, unnormalized scores) for two classes: Apply softmax to convert raw…
This model was produced by fine-tuning the OpenVLA 7B model via LoRA (r=32) on the LIBERO-Spatial dataset from the LIBERO simulation benchmark. We made a few modifications to the training dataset to improve final performance (see the OpenVLA paper for details). Below are the hyperparameters we used for all LIBERO experiments: - No gradient accumulation (i.e. gradaccumulationsteps == 1) - shufflebuffersize == 100000 See the OpenVLA GitHub README for instructions on how to run and evaluate this model in the LIBERO simulator.
Model · Zero-shot classification
roberta-base-zeroshot-v2.0-c
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…
Moondream 3.1 is a vision language model with a mixture-of-experts architecture (9B total parameters, 2B active). It delivers state-of-the-art visual reasoning and detection while staying fast and cheap to deploy. Skills include query, detect, point, and caption, all native and all returning structured output. For the full story on what's new — including how we trained it and how it holds up on your own tasks — see the release notes. Photon is Moondream's high-performance inference engine. It runs the model locally on NVIDIA GPUs (Ampere or newer) and Apple Silicon Macs, with the same API as Moondream Cloud. No API key is required to run the base model locally. (You'll only need one to run…
This model is a fine-tuned version of Google's T5-Base model tailored for summarizing financial 10K report sections. T5-Base-10K-Summarization is optimized to condense lengthy 10K reports into manageable summaries, enabling quick insights into financial data and trends. Ideal for use by financial analysts and regulatory agencies needing rapid insights from 10K reports. It may not be suited for summarizing non-financial documents or informal texts. Trained on a diverse collection of 10K reports from various industries, annotated for summarization to ensure broad applicability and accuracy. The following hyperparameters were used during training: - learningrate: 5e-05 - trainbatchsize: 8…
Alpamayo 2 Super is a 34B-parameter foundation model designed to tackle multiple autonomous vehicle (AV) development tasks. It combines a 32B VLM backbone with a 2B diffusion expert. Alpamayo 2 Super was developed by NVIDIA as a part of the broader Alpamayo Open Platform. Model weights: The model weights are released under the OpenMDW-1.1 license. Source code: Apache License 2.0, as provided in the Alpamayo 2 Super source repository. Global Developers and researchers working on autonomous vehicle systems who need a foundation model for perception, planning, and decision-making tasks. Alpamayo 2 Super supports multiple AV development tasks such as trajectory prediction, visual question…
However, we observe that the speed and accuracy of YOLOs are negatively affected by the NMS. Recently, end-to-end Transformer-based detectors (DETRs) have provided an alternative to eliminating NMS. Nevertheless, the high computational cost limits their practicality and hinders them from fully exploiting the advantage of excluding NMS. In this paper, we propose the Real-Time DEtection TRansformer (RT-DETR), the first real-time end-to-end object detector to our best knowledge that addresses the above dilemma. We build RT-DETR in two steps, drawing on the advanced DETR: first we focus on maintaining accuracy while improving speed, followed by maintaining speed while improving accuracy.…
Note: Following the release of NVIDIA Alpamayo at CES 2026, Alpamayo-R1 has been renamed to Alpamayo 1. Alpamayo 1 integrates Chain-of-Causation reasoning with trajectory planning to enhance decision-making in complex autonomous-driving scenarios. Alpamayo 1 (v1.0) was developed by NVIDIA as a vision-language-action (VLA) model that bridges interpretable reasoning with precise vehicle control for autonomous-driving applications. This model is ready for non-commercial use. Commercial licensing available upon request. Model weights: The model weights are released under the OpenMDW-1.1 license. Source code: Apache License 2.0, as provided in the Alpamayo 1 source repository. Global Researchers…
Finetuned T5 base summarization model for Indonesian. t5-base-indonesian-summarization-cased model is based on t5-base-bahasa-summarization-cased by huseinzol05, finetuned using idliputan6 dataset.
This is a fine-tunned object detection model for fashion. For more details of the implementation you can check the source code here the dataset used for its training is available here this model supports the following categories: CATS = ['shirt, blouse', 'top, t-shirt, sweatshirt', 'sweater', 'cardigan', 'jacket', 'vest', 'pants', 'shorts', 'skirt', 'coat', 'dress', 'jumpsuit', 'cape', 'glasses', 'hat', 'headband, head covering, hair accessory', 'tie', 'glove', 'watch', 'belt', 'leg warmer', 'tights, stockings', 'sock', 'shoe', 'bag, wallet', 'scarf', 'umbrella', 'hood', 'collar', 'lapel', 'epaulette', 'sleeve', 'pocket', 'neckline', 'buckle', 'zipper', 'applique', 'bead', 'bow', 'flower'…
Segments ink directly in 3D in micro-CT of PHerc. Paris 4, where ink is volumetrically visible. This model provides the independent volumetric validation of surface-conditioned ink recovery reported in "Complete virtual unwrapping and reading of a rolled Herculaneum papyrus" (Angelotti et al., arXiv:2606.29085, 2026). (PHerc. Paris 4, 0139, 0500P2, 0814, 0841, 1667, MAN5, 9B); (2) a DINO-guided student using dense ink-likeness from the 3D DINOv2 representation model (3) + background masking; (4) self-distillation. DINO guidance compares each 864-D patch token to a reference ink embedding (avgrefembedding.npy, the L2-normalised mean of 256 expert-clicked tokens stored in…
Mask2Former model trained on ADE20k semantic segmentation (tiny-sized version, Swin backbone). It was introduced in the paper Masked-attention Mask Transformer for Universal Image Segmentation and first released in this repository. Disclaimer: The team releasing Mask2Former did not write a model card for this model so this model card has been written by the Hugging Face team. Mask2Former addresses instance, semantic and panoptic segmentation with the same paradigm: by predicting a set of masks and corresponding labels. Hence, all 3 tasks are treated as if they were instance segmentation. Mask2Former outperforms the previous SOTA, MaskFormer both in terms of performance an efficiency by (i)…
Paper: Zheng et al., 2025, “X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Model” (arXiv:2510.10274) Successful generalist Vision-Language-Action (VLA) models rely on effective training across diverse robotic platforms with large-scale, cross-embodiment, heterogeneous datasets. To facilitate and leverage the heterogeneity in rich robotic data sources, X-VLA introduces a Soft Prompt approach with minimally added parameters: we infuse prompt-learning concepts into cross-embodiment robot learning, introducing separate sets of learnable embeddings for each distinct embodiment. These embodiment-specific prompts empower VLA models to exploit cross-embodiment…
Model Collections
Hand-picked starting points, each with the reason it exists.
Collection · 4 entries
Embedding models for retrieval
Sentence and document embedding models used to build retrieval systems. Dimension and sequence length matter more than size here, and both come from the publisher.
Collection · 4 entries
Models that fit on one accelerator
Models whose publisher-reported parameter count puts them within reach of a single accelerator at common precisions. Memory needed depends on precision and serving configuration, so treat the parameter count as the starting point, not the answer.
Collection · 6 entries
Open-weight text models worth knowing
Widely used open-weight language models, chosen because each one is a distinct family rather than a variant of the one above it. Selection, not a ranking.
Collection · 3 entries
Speech and audio models
Recognition and synthesis models, grouped so the two directions are easy to compare.
Open-Weight Models Explained
What is an open-weight model?
An AI model whose trained weights are published for anyone to download, so it can be run, tested and fine-tuned on hardware the user controls.
Is an open-weight model the same as open source?
Not always. Open weights means the trained model can be downloaded. Open source usually also means the training code and data are available and the license allows broad reuse. Many open-weight models release the weights only.
Can I use an open-weight model commercially?
It depends on the license. Apache 2.0 and MIT allow commercial use. Other licenses limit it, for example to non-commercial use or below a set number of users. Every model page here shows its license.
How much memory does an open-weight model need?
About two bytes per parameter at 16-bit precision, so a 7-billion-parameter model needs roughly 14 GB for its weights, plus memory for the context it processes. Each model page lists its parameter count and the size of its files.
Related SAVRN Research
The hub sits beside SAVRN's market data and infrastructure research: what models cost to run, and what it takes to run them.
SAVRN Index
What open models cost to run
The same open-weight model priced by every host that serves it, per million tokens.
Research Hub
Data center trackers and maps
Moratoriums, permits, power, water and capital behind the facilities that run these models.
Method
How the Model Hub is built
Sources, evidence labels, refresh behaviour, and the limits of every comparison here.