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.
Updated 2026-09-18 · How the library is built
2,760 models, sorted by most downloaded.
This is the model card of NLLB-200's distilled 600M variant. Here are the metrics for that particular checkpoint. - Information about training algorithms, parameters, fairness constraints or other applied approaches, and features. The exact training algorithm, data and the strategies to handle data imbalances for high and low resource languages that were used to train NLLB-200 is described in the paper. - Paper or other resource for more information NLLB Team et al, No Language Left Behind: Scaling Human-Centered Machine Translation, Arxiv, 2022 - Where to send questions or comments about the model: https://github.com/facebookresearch/fairseq/issues • Model performance measures: NLLB-200…
ONNX port of sentence-transformers/all-MiniLM-L6-v2 for text classification and similarity searches. Here's an example of performing inference using the model with FastEmbed.
https://huggingface.co/openai/whisper-tiny 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).
We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures, which were thoroughly validated in DeepSeek-V2. Furthermore, DeepSeek-V3 pioneers an auxiliary-loss-free strategy for load balancing and sets a multi-token prediction training objective for stronger performance. We pre-train DeepSeek-V3 on 14.8 trillion diverse and high-quality tokens, followed by Supervised Fine-Tuning and Reinforcement Learning stages to fully harness its capabilities. Comprehensive evaluations…
The Pythia Scaling Suite is a collection of models developed to facilitate interpretability research (see paper). It contains two sets of eight models of sizes 70M, 160M, 410M, 1B, 1.4B, 2.8B, 6.9B, and 12B. For each size, there are two models: one trained on the Pile, and one trained on the Pile after the dataset has been globally deduplicated. All 8 model sizes are trained on the exact same data, in the exact same order. We also provide 154 intermediate checkpoints per model, hosted on Hugging Face as branches. The Pythia model suite was designed to promote scientific research on large language models, especially interpretability research. Despite not centering downstream performance as a…
This model card focuses on the LTX-2.3 model, which is a significant update to the LTX-2 model with improved audio and visual quality as well as enhanced prompt adherence. LTX-2 was presented in the paper LTX-2: Efficient Joint Audio-Visual Foundation Model. If you want to dive in right to the code - it is available here. LTX-2.3 is a DiT-based audio-video foundation model designed to generate synchronized video and audio within a single model. It brings together the core building blocks of modern video generation, with open weights and a focus on practical, local execution. LTX-2.3 is accessible right away via the API Playground. You can use the models - full, distilled, upscalers and any…
CLIPSeg model with reduce dimension 64, refined (using a more complex convolution). It was introduced in the paper Image Segmentation Using Text and Image Prompts by Lüddecke et al. and first released in this repository. This model is intended for zero-shot and one-shot image segmentation. Refer to the documentation.
Qwen3-Coder is available in multiple sizes. Today, we're excited to introduce Qwen3-Coder-30B-A3B-Instruct-FP8. This streamlined model maintains impressive performance and efficiency, featuring the following key enhancements: - Significant Performance among open models on Agentic Coding, Agentic Browser-Use, and other foundational coding tasks. - Long-context Capabilities with native support for 256K tokens, extendable up to 1M tokens using Yarn, optimized for repository-scale understanding. - Agentic Coding supporting for most platform such as Qwen Code, CLINE, featuring a specially designed function call format. Qwen3-Coder-30B-A3B-Instruct-FP8 has the following features: NOTE: This model…
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the OPENSLRSLR53 - bengali dataset. It achieves the following results on the evaluation set. With 5 gram language model trained on 30M sentences randomly chosen from AI4Bharat IndicCorp dataset: Note: 5% of a total 10935 samples have been used for evaluation. Evaluation set has 10935 examples which was not part of training training was done on first 95% and eval was done on last 5%. Training was stopped after 180k steps. Output predictions are available under files section. The following hyperparameters were used during training: - datasetname="openslr" - modelnameorpath="facebook/wav2vec2-xls-r-300m"…
Surya is a 650M param OCR model with these features: - Accuracy - scores 83.3% on olmOCR-bench (top under 3B params) - Multilingual - scores 87.2% on an internal benchmark set of 91 languages (more here) - Layout analysis (table, image, header, etc.) with reading order - Table recognition (rows + columns) It works on a range of documents (see usage and benchmarks). Our managed platform runs both Surya, and variants of our highest accuracy model, Chandra. Get started with $5 in free credits — sign up (takes under 30 seconds) or try our free public playground. Surya is named for the Hindu sun god, who has universal vision. The Surya code is licensed under Apache 2.0. The model weights use a…
SPECTER2 is the successor to SPECTER and is capable of generating task specific embeddings for scientific tasks when paired with adapters. This is the base model to be used along with the adapters. Given the combination of title and abstract of a scientific paper or a short texual query, the model can be used to generate effective embeddings to be used in downstream applications. Note:For general embedding purposes, please use allenai/specter2. To get the best performance on a downstream task type please load the associated adapter with the base model as in the example below. Model usage updated to be compatible with latest versions of transformers and adapters (newly released update to…
Following the February release of the Qwen3.5 series, we're pleased to share the first open-weight variant of Qwen3.6. Built on direct feedback from the community, Qwen3.6 prioritizes stability and real-world utility, offering developers a more intuitive, responsive, and genuinely productive coding experience. This release delivers substantial upgrades, particularly in For more details, please refer to our blog post Qwen3.6-27B. For streamlined integration, we recommend using Qwen3.6 via APIs. Below is a guide to use Qwen3.6 via OpenAI-compatible API. Qwen3.6 can be served via APIs with popular inference frameworks. In the following, we show example commands to launch OpenAI-Compatible API…
https://huggingface.co/cross-encoder/ms-marco-MiniLM-L-6-v2 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).
The following model is a Pytorch pre-trained model obtained from converting Tensorflow checkpoint found in the official Google BERT repository. This is one of the smaller pre-trained BERT variants, together with bert-mini bert-small and bert-medium. They were introduced in the study Well-Read Students Learn Better: On the Importance of Pre-training Compact Models (arxiv), and ported to HF for the study Generalization in NLI: Ways (Not) To Go Beyond Simple Heuristics (arXiv). These models are supposed to be trained on a downstream task. If you use the model, please consider citing both the papers: - prajjwal1/bert-tiny (L=2, H=128) Model Link - prajjwal1/bert-mini (L=4, H=256) Model Link…
Model on Google Cloud Model Garden: MedGemma GitHub repository (supporting code, Colab notebooks, discussions, and Foundations terms of use](https://developers.google.com/health-ai-developer-foundations/terms). This section describes the MedGemma model and how to use it. MedGemma is a collection of Gemma 3 variants that are trained for performance on medical text and image comprehension. Developers can use MedGemma to accelerate building healthcare-based AI applications. MedGemma currently comes in three variants: a 4B multimodal version and 27B text-only and multimodal versions. Both MedGemma multimodal versions utilize a SigLIP image encoder that has been specifically pre-trained on a…
Optical character recognition for Japanese text, with the main focus being Japanese manga. It uses Vision Encoder Decoder framework. Manga OCR can be used as a general purpose printed Japanese OCR, but its main goal was to provide a high quality text recognition, robust against various scenarios specific to manga: - both vertical and horizontal text - text with furigana - text overlaid on images - wide variety of fonts and font styles - low quality images Code is available here.
Model trained with TASS 2020 corpus (around ~5k tweets) of several dialects of Spanish. Base model is RoBERTuito, a RoBERTa model trained in Spanish tweets. Uses POS, NEG, NEU labels. Use it directly with pysentimiento Results for the four tasks evaluated in pysentimiento. Results are expressed as Macro F1 scores Note that for Hate Speech, these are the results for Semeval 2019, Task 5 Subtask B If you use this model in your research, please cite pysentimiento, RoBERTuito and TASS papers
Anima is a 2 billion parameter text-to-image model created via a collaboration between CircleStone Labs and Comfy Org. It is focused mainly on anime concepts, characters, and styles, but is also capable of generating a wide variety of other non-photorealistic content. The model is designed for making illustrations and artistic images, and will not work well at realism. It is trained on several million anime images and about 800k non-anime artistic images. No synthetic data was used for training. The knowledge cut-off date for the anime training data is September 2025. - The pretrained, unrefined base model. Maximum flexibility, diversity, and style adherence. - LoRAs should be trained using…
Table Transformer (DETR) model trained on PubTables1M. It was introduced in the paper PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting the structure (like rows, columns) in tables.…
WhereIsAI/UAE-Large-V1 is licensed under MIT. Feel free to use it in any scenario. If you use it for academic papers, you could cite us via citation info. Welcome to using AnglE to train and infer powerful sentence embeddings. Achievements - May 16, 2024 | AnglE's paper is accepted by ACL 2024 Main Conference - Dec 4, 2023 | Our universal English sentence embedding WhereIsAI/UAE-Large-V1 achieves SOTA on the MTEB Leaderboard with an average score of 64.64! - WhereIsAI/UAE-Code-Large-V1: This model can be used for code or GitHub issue similarity measurement. There is no need to specify any prompts. For retrieval purposes, please use the prompt Prompts.C for query (not for document). Infinity…
Kronos is the first open-source foundation model for financial candlesticks (K-lines), trained on data from over 45 global exchanges. It is designed to handle the unique, high-noise characteristics of financial data. Kronos is a family of decoder-only foundation models, pre-trained specifically for the "language" of financial markets—K-line sequences. It leverages a novel two-stage framework: 1. A specialized tokenizer first quantizes continuous, multi-dimensional K-line data (OHLCV) into hierarchical discrete tokens. 2. A large, autoregressive Transformer is then pre-trained on these tokens, enabling it to serve as a unified model for diverse quantitative tasks. The success of large-scale…
Qwen3.6-35B-A3B uncensored by HauhauCS. 0/465 Refusals. No changes to datasets or capabilities. Fully functional, 100% of what the original authors intended - just without the refusals. These are meant to be the best lossless uncensored models out there. Stronger uncensoring — model is fully unlocked and won't refuse prompts. May occasionally append short disclaimers (baked into base model training, not refusals) but full content is always generated. For a more conservative uncensor that keeps some safety guardrails, check the Balanced variant when it's available. All quants generated with importance matrix (imatrix) for optimal quality preservation on abliterated weights. KP ("Perfect")…
DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data. In DeBERTa V3, we further improved the efficiency of DeBERTa using ELECTRA-Style pre-training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our paper. Please check the official repository for more implementation details and updates. The DeBERTa V3 large model comes with 24 layers and a hidden size of 1024. It has 304M…
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.