Detects age group with about 59% accuracy based on an image. See https://www.kaggle.com/code/dima806/age-group-image-classification-vit for details.
Open-weight model · Image classification
vit-base-oxford-iiit-pets
by Ilias Strub ISxOdin/vit-base-oxford-iiit-pets
This model is a fine-tuned version of google/vit-base-patch16-224 on the pcuenq/oxford-pets dataset.
Parameters86M
Context—
Weights343.3 MB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads175.4k
Runs On
What it takes to serve vit-base-oxford-iiit-pets (86M 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 Ilias Strub, published under apache-2.0, revision 6fcb4ce433d4.
This model is a fine-tuned version of google/vit-base-patch16-224 on the pcuenq/oxford-pets dataset. It achieves the following results on the evaluation set: This model is a fine-tuned version of a pre-trained Vision Transformer (google/vit-base-patch16-224) for image classification on the Oxford-IIIT Pet Dataset. It uses transfer learning to adapt a generic vision model to identify 37 different cat and dog breeds. The model head is adjusted to output the number of classes in the dataset, and it is trained end-to-end using standard classification loss. - Educational demos on transfer learning and fine-tuning vision models. - Pet breed classification in structured datasets similar to Oxford…
Read Ilias Strub's full model card
This model is a fine-tuned version of google/vit-base-patch16-224 on the pcuenq/oxford-pets dataset. It achieves the following results on the evaluation set: - Loss: 0.1924 - Accuracy: 0.9445
Model description
This model is a fine-tuned version of a pre-trained Vision Transformer (google/vit-base-patch16-224) for image classification on the Oxford-IIIT Pet Dataset.
It uses transfer learning to adapt a generic vision model to identify 37 different cat and dog breeds.
The model head is adjusted to output the number of classes in the dataset, and it is trained end-to-end using standard classification loss.
Intended uses & limitations
Intended Uses: - Educational demos on transfer learning and fine-tuning vision models. - Pet breed classification in structured datasets similar to Oxford Pets. - Comparative analysis with zero-shot models like CLIP.
Limitations: - May not generalize well to breeds outside of the Oxford-IIIT dataset. - Not suitable for real-world medical or safety-critical applications. - Input images should be clear, centered, and close in style to the training data (cropped pet portraits).
Training and evaluation data
The model is trained and evaluated on the Oxford-IIIT Pet Dataset, which contains 7,349 images of cats and dogs spanning 37 different breeds. The dataset includes equal representation of pets and was split into training, validation, and test sets. Evaluation metrics used include accuracy, precision, and recall.
Training procedure
Training hyperparameters
The following hyperparameters were used during training: - learning_rate: 0.0003 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments - lr_scheduler_type: linear - num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.3716 | 1.0 | 370 | 0.3013 | 0.9242 |
| 0.2048 | 2.0 | 740 | 0.2342 | 0.9310 |
| 0.1764 | 3.0 | 1110 | 0.2124 | 0.9350 |
| 0.1617 | 4.0 | 1480 | 0.2050 | 0.9350 |
| 0.1235 | 5.0 | 1850 | 0.2032 | 0.9350 |
Zero-Shot Classification Evaluation (CLIP)
Evaluated the Oxford-IIIT Pet dataset using a zero-shot image classification model: openai/clip-vit-base-patch32.
Instead of training, the CLIP model was evaluated using a list of breed names (e.g., "Siamese", "Persian", "Chihuahua") as candidate labels for zero-shot classification.
Evaluation Results:
- Accuracy: 0.8800
- Precision: 0.8768
- Recall: 0.8800
Framework versions
- Transformers 4.50.0
- Pytorch 2.6.0+cu124
- Datasets 3.4.1
- Tokenizers 0.21.1
Configuration
- Architecture
- ViTForImageClassification
- Layers
- 12
- Hidden size
- 768
- Feed-forward size
- 3,072
- Attention heads
- 12
- Stored precision
- float32
- Model type
- vit
Identity and Version
- Repository
- ISxOdin/vit-base-oxford-iiit-pets
- Publisher
- Ilias Strub
- Task
- Image classification
- Modality
- Image
- Library
- transformers
- Parameters
- 86M parameters
- Languages
- vit
- Revision
- 6fcb4ce433d4b43a19bd9b4d2d906b2646858d45
- First published
- 2025-04-01
- Last updated
- 2025-04-11
Files and Weights
8 files, 343.4 MB in total. The weights are 2 files totalling 343.3 MB in bin, safetensors.
Weights2 files · 343.3 MB
Configuration2 files · 2.8 KB
Documentation1 file · 3.4 KB
Other2 files · 13.1 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 343.3 MB | 534df4dd01c2 |
| training_args.bin | Weights | 5.4 KB | d33510da3843 |
| config.json | Configuration | 2.5 KB | — |
| preprocessor_config.json | Configuration | 351 B | — |
| README.md | Documentation | 3.4 KB | — |
| runs/Apr01_15-21-10_ip-10-192-10-121/events.out.tfevents.1743520871.ip-10-192-10-121.15731.0 | Other | 12.7 KB | 1053f1d82f57 |
| runs/Apr01_15-21-10_ip-10-192-10-121/events.out.tfevents.1743521355.ip-10-192-10-121.15731.1 | Other | 411 B | 425f0f5ca67a |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 343.3 MB
Download from Ilias Strub
Released by Ilias Strub through its official repository on Hugging Face. Read the license.
Built From
- Derived from google/vit-base-patch16-224
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 343.3 MB |
| 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 vit-base-oxford-iiit-pets
How much GPU memory does vit-base-oxford-iiit-pets need?
About 0.2 GB at 16-bit and 0.1 GB at 4-bit: the weights (86M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run vit-base-oxford-iiit-pets 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 vit-base-oxford-iiit-pets commercially?
Yes. vit-base-oxford-iiit-pets 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.
Similar Models
A vision transformer finetuned to classify the age of a given person's face.
In today's digital world, user-generated content is a double-edged sword. While it fosters creativity and engagement, it also opens the door to inappropriate or illegal content being shared. Our NSFW Image Classifier is specifically designed to identify and filter out explicit images, including pornography, hentai, and sexually suggestive content, ensuring your platform remains safe, secure, and legally compliant. With more than 2M downloads, our NSFW Image Classifier has become the go-to solution for platforms looking to maintain a clean and safe environment for their users. Many developers and companies have already chosen our solution to protect their communities—will you be next? 1.…
The Fine-Tuned Vision Transformer (ViT) is a variant of the transformer encoder architecture, similar to BERT, that has been adapted for image classification tasks. This specific model, named "google/vit-base-patch16-224-in21k," is pre-trained on a substantial collection of images in a supervised manner, leveraging the ImageNet-21k dataset. The images in the pre-training dataset are resized to a resolution of 224x224 pixels, making it suitable for a wide range of image recognition tasks. During the training phase, meticulous attention was given to hyperparameter settings to ensure optimal model performance. The model was fine-tuned with a judiciously chosen batch size of 16. This choice not…
Autogenerated by HuggingPics Create your own image classifier for anything by running the demo on Google Colab. Report any issues with the demo at the github repo.
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 2e-05 - trainbatchsize: 8 - evalbatchsize: 8 - lrschedulertype: linear - numepochs: 5.0 - Transformers 4.36.0.dev0 - Pytorch 2.1.1+cu118 - Datasets 2.15.0 - Tokenizers 0.15.0