A Swin Transformer image classification model. Pretrained on ImageNet-22k and fine-tuned on ImageNet-1k by paper authors. Explore the dataset and runtime metrics of this model in timm model results.
Open-weight model · Image classification
siglip2-x256-explicit-content
by Prithiv Sakthi prithivMLmods/siglip2-x256-explicit-content
SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features https://arxiv.org/pdf/2502.14786 The model classifies each image into one of the following content categories: This model is intended for…
Parameters93M
Context64
Weights3.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads112.3k
Runs On
What it takes to serve siglip2-x256-explicit-content (93M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 0.2 GB | 0.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.0 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.
Model Card
By Prithiv Sakthi, published under apache-2.0, revision 143a4508c50d.
siglip2-x256-explicit-content is a vision-language encoder model fine-tuned from siglip2-base-patch16-256 for multi-class image classification. Built on the SiglipForImageClassification architecture, the model is trained to identify and categorize content types in images, especially for explicit, suggestive, or safe media filtering.
[!note] SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features https://arxiv.org/pdf/2502.14786
Classification Report:
precision recall f1-score support
Anime Picture 0.8940 0.8718 0.8827 5600
Hentai 0.8961 0.8935 0.8948 4180
Normal 0.9100 0.8895 0.8997 5503
Pornography 0.9496 0.9654 0.9574 5600
Enticing or Sensual 0.9132 0.9429 0.9278 5600
accuracy 0.9137 26483
macro avg 0.9126 0.9126 0.9125 26483
weighted avg 0.9135 0.9137 0.9135 26483
Label Space: 5 Classes
The model classifies each image into one of the following content categories:
Class 0: "Anime Picture"
Class 1: "Hentai"
Class 2: "Normal"
Class 3: "Pornography"
Class 4: "Enticing or Sensual"
Install Dependencies
pip install -q transformers torch pillow gradio
Inference Code
Configuration
- Architecture
- SiglipForImageClassification
- Context length (tokens)
- 64
- Layers
- 12
- Hidden size
- 768
- Feed-forward size
- 3,072
- Attention heads
- 12
- Vocabulary size
- 256,000
- Stored precision
- float32
- Model type
- siglip
Identity and Version
- Repository
- prithivMLmods/siglip2-x256-explicit-content
- Publisher
- Prithiv Sakthi
- Task
- Image classification
- Modality
- Image
- Library
- transformers
- Parameters
- 93M parameters
- Languages
- en
- Revision
- 143a4508c50d75299159a410d4649afe5b7e38fb
- First published
- 2025-04-29
- Last updated
- 2025-04-30
Files and Weights
30 files, 3.5 GB in total. The weights are 17 files totalling 3.5 GB in bin, pt, pth, safetensors.
Weights17 files · 3.5 GB
Configuration11 files · 12.3 KB
Documentation1 file · 3.8 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| checkpoint-1656/model.safetensors | Weights | 371.8 MB | 886acfda2690 |
| checkpoint-1656/optimizer.pt | Weights | 686.9 MB | 9a80622a0edf |
| checkpoint-1656/rng_state.pth | Weights | 14.2 KB | 66b395f57c7d |
| checkpoint-1656/scheduler.pt | Weights | 1.1 KB | d36413dbf261 |
| checkpoint-1656/training_args.bin | Weights | 5.3 KB | 77f0ca87b399 |
| checkpoint-2484/model.safetensors | Weights | 371.8 MB | 846e9b76948a |
| checkpoint-2484/optimizer.pt | Weights | 686.9 MB | 754fddb5628d |
| checkpoint-2484/rng_state.pth | Weights | 14.2 KB | bb29b7c6a866 |
| checkpoint-2484/scheduler.pt | Weights | 1.1 KB | a31cd561974c |
| checkpoint-2484/training_args.bin | Weights | 5.3 KB | 77f0ca87b399 |
| checkpoint-828/model.safetensors | Weights | 371.8 MB | c82837ce983c |
| checkpoint-828/optimizer.pt | Weights | 686.9 MB | 04edb7400da1 |
| checkpoint-828/rng_state.pth | Weights | 14.2 KB | 8ef17702a072 |
| checkpoint-828/scheduler.pt | Weights | 1.1 KB | 317a7aa06c6e |
| checkpoint-828/training_args.bin | Weights | 5.3 KB | 77f0ca87b399 |
| model.safetensors | Weights | 371.8 MB | 846e9b76948a |
| training_args.bin | Weights | 5.3 KB | 77f0ca87b399 |
| checkpoint-1656/config.json | Configuration | 1.3 KB | — |
| checkpoint-1656/preprocessor_config.json | Configuration | 394 B | — |
| checkpoint-1656/trainer_state.json | Configuration | 1.9 KB | — |
| checkpoint-2484/config.json | Configuration | 1.3 KB | — |
| checkpoint-2484/preprocessor_config.json | Configuration | 394 B | — |
| checkpoint-2484/trainer_state.json | Configuration | 2.4 KB | — |
| checkpoint-828/config.json | Configuration | 1.3 KB | — |
| checkpoint-828/preprocessor_config.json | Configuration | 394 B | — |
| checkpoint-828/trainer_state.json | Configuration | 1.3 KB | — |
| config.json | Configuration | 1.3 KB | — |
| preprocessor_config.json | Configuration | 394 B | — |
| README.md | Documentation | 3.8 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 3.5 GB
Download from Prithiv Sakthi
Released by Prithiv Sakthi through its official repository on Hugging Face. Read the license.
Built From
- Derived from google/siglip2-base-patch16-256
- Described by arXiv:2502.14786
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 3.5 GB |
| 16-bit | 0.2 GB |
| 8-bit | 0.1 GB |
| 4-bit | 0.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About siglip2-x256-explicit-content
How much GPU memory does siglip2-x256-explicit-content need?
About 0.2 GB at 16-bit and 0.1 GB at 4-bit: the weights (93M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run siglip2-x256-explicit-content on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use siglip2-x256-explicit-content commercially?
Yes. siglip2-x256-explicit-content is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
What is siglip2-x256-explicit-content's context length?
64 tokens, from the maximum position embeddings in its published configuration.
Similar Models
ConvNeXt V2 model pretrained using the FCMAE framework and fine-tuned on the ImageNet-22K dataset at resolution 384x384. It was introduced in the paper ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders by Woo et al. and first released in this repository. Disclaimer: The team releasing ConvNeXT V2 did not write a model card for this model so this model card has been written by the Hugging Face team. ConvNeXt V2 is a pure convolutional model (ConvNet) that introduces a fully convolutional masked autoencoder framework (FCMAE) and a new Global Response Normalization (GRN) layer to ConvNeXt. ConvNeXt V2 significantly improves the performance of pure ConvNets on various…
A ConvNeXt-V2 image classification model. Pretrained with a fully convolutional masked autoencoder framework (FCMAE) and fine-tuned on ImageNet-22k and then ImageNet-1k. Explore the dataset and runtime metrics of this model in timm model results. All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
A ConvNeXt image classification model. Pretrained on ImageNet-22k and fine-tuned on ImageNet-1k by paper authors. Explore the dataset and runtime metrics of this model in timm model results. All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
A ConvNeXt image classification model. CLIP image tower weights pretrained in OpenCLIP on LAION and fine-tuned on ImageNet-12k followed by ImageNet-1k in timm bby Ross Wightman. Please see related OpenCLIP model cards for more details on pretrain: https://huggingface.co/laion/CLIP-convnextxxlarge-laion2B-s34B-b82K-augreg-soup https://huggingface.co/laion/CLIP-convnextlarged.laion2B-s26B-b102K-augreg https://huggingface.co/laion/CLIP-convnextbasew-laion2B-s13B-b82K-augreg https://huggingface.co/laion/CLIP-convnextbasew320-laionaesthetic-s13B-b82K - Learning Transferable Visual Models From Natural Language Supervision: https://arxiv.org/abs/2103.00020 Explore the dataset and runtime metrics…
A Vision Transformer (ViT) image classification model. Trained on ImageNet-21k and fine-tuned on ImageNet-1k (with additional augmentation and regularization) in JAX by paper authors, ported to PyTorch by Ross Wightman. - How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers: https://arxiv.org/abs/2106.10270 - An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale: https://arxiv.org/abs/2010.11929v2 Explore the dataset and runtime metrics of this model in timm model results.