A vision transformer finetuned to classify the age of a given person's face.
Open-weight model · Image classification
fairface_age_image_detection
by Dmytro Iakubovskyi dima806/fairface_age_image_detection
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.
Parameters86M
Context—
Weights4.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads1.2M
Runs On
What it takes to serve fairface_age_image_detection (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.
SAVRN's Notes on fairface_age_image_detection
The publisher's own figure, about 59 percent accuracy at assigning an age group from a photo, frames every decision here. Hardware is not the constraint: 86 million parameters need 0.2 gigabytes of memory at 16-bit, so the cheapest listed setup, one MI300X at $1.85 an hour, is a card you would rent only if this classifier were one of dozens sharing it. It belongs on whatever GPU already runs your image pipeline. The repository weighs 4.5 gigabytes across 38 files, stored in float32.
Apache 2.0 allows commercial use, modification and redistribution with the notices kept, but the real diligence is on lineage. The weights derive from google/vit-base-patch16-224-in21k and were trained on nateraw/fairface, so results depend on how close your images sit to that set. With no evaluations reported beyond the publisher's own 59 percent, it wants a test on your own labeled images before any production use.
Model Card
By Dmytro Iakubovskyi, published under apache-2.0, revision 4e02ab8057ea.
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.
Read Dmytro Iakubovskyi's full model card
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.
Classification report:
precision recall f1-score support
0-2 0.7803 0.7500 0.7649 180
3-9 0.7998 0.7998 0.7998 1249
10-19 0.5361 0.4236 0.4733 1086
20-29 0.6402 0.7221 0.6787 3026
30-39 0.4935 0.5083 0.5008 2099
40-49 0.4848 0.4386 0.4606 1238
50-59 0.5000 0.4814 0.4905 725
60-69 0.4497 0.4685 0.4589 286
more than 70 0.6897 0.1802 0.2857 111
accuracy 0.5892 10000
macro avg 0.5971 0.5303 0.5459 10000
weighted avg 0.5863 0.5892 0.5844 10000
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
- dima806/fairface_age_image_detection
- Publisher
- Dmytro Iakubovskyi
- Task
- Image classification
- Modality
- Image
- Library
- transformers
- Parameters
- 86M parameters
- Languages
- vit
- Revision
- 4e02ab8057ea7fd74b1670940995c5dfda3e6ec0
- First published
- 2024-12-06
- Last updated
- 2024-12-15
Files and Weights
38 files, 4.5 GB in total. The weights are 22 files totalling 4.5 GB in bin, pt, pth, safetensors.
Weights22 files · 4.5 GB
Configuration14 files · 20.7 KB
Documentation1 file · 1.2 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| checkpoint-32/model.safetensors | Weights | 343.2 MB | 8c2e34c076a3 |
| checkpoint-32/optimizer.pt | Weights | 686.6 MB | ecf20baba72b |
| checkpoint-32/rng_state.pth | Weights | 14.2 KB | 6b3ee827a7a0 |
| checkpoint-32/scheduler.pt | Weights | 1.1 KB | 56eec20d0c97 |
| checkpoint-32/training_args.bin | Weights | 5.3 KB | 28ebf2b6dbb0 |
| checkpoint-4688/model.safetensors | Weights | 343.2 MB | 1f6f3b28c3de |
| checkpoint-4688/optimizer.pt | Weights | 686.6 MB | 8791320f4d58 |
| checkpoint-4688/rng_state.pth | Weights | 14.2 KB | 6b3ee827a7a0 |
| checkpoint-4688/scheduler.pt | Weights | 1.1 KB | b93e187257ca |
| checkpoint-4688/training_args.bin | Weights | 5.3 KB | 28ebf2b6dbb0 |
| checkpoint-8752/model.safetensors | Weights | 343.2 MB | cd3a725b1119 |
| checkpoint-8752/optimizer.pt | Weights | 686.6 MB | 087e51816b13 |
| checkpoint-8752/rng_state.pth | Weights | 14.2 KB | 48ee9b73399c |
| checkpoint-8752/scheduler.pt | Weights | 1.1 KB | 197fc8dcfe2b |
| checkpoint-8752/training_args.bin | Weights | 5.3 KB | e97ef976af56 |
| checkpoint-9376/model.safetensors | Weights | 343.2 MB | 1265d8e5a8a3 |
| checkpoint-9376/optimizer.pt | Weights | 686.6 MB | db14e1baf0fc |
| checkpoint-9376/rng_state.pth | Weights | 14.2 KB | 48ee9b73399c |
| checkpoint-9376/scheduler.pt | Weights | 1.1 KB | 162830a25651 |
| checkpoint-9376/training_args.bin | Weights | 5.3 KB | 5ba8df84fa27 |
| model.safetensors | Weights | 343.2 MB | 1265d8e5a8a3 |
| training_args.bin | Weights | 5.3 KB | 5ba8df84fa27 |
| checkpoint-32/config.json | Configuration | 967 B | — |
| checkpoint-32/preprocessor_config.json | Configuration | 351 B | — |
| checkpoint-32/trainer_state.json | Configuration | 1.3 KB | — |
| checkpoint-4688/config.json | Configuration | 967 B | — |
| checkpoint-4688/preprocessor_config.json | Configuration | 351 B | — |
| checkpoint-4688/trainer_state.json | Configuration | 2.9 KB | — |
| checkpoint-8752/config.json | Configuration | 967 B | — |
| checkpoint-8752/preprocessor_config.json | Configuration | 351 B | — |
| checkpoint-8752/trainer_state.json | Configuration | 4.9 KB | — |
| checkpoint-9376/config.json | Configuration | 967 B | — |
| checkpoint-9376/preprocessor_config.json | Configuration | 351 B | — |
| checkpoint-9376/trainer_state.json | Configuration | 5.0 KB | — |
| config.json | Configuration | 967 B | — |
| preprocessor_config.json | Configuration | 351 B | — |
| README.md | Documentation | 1.2 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 4.5 GB
Download from Dmytro Iakubovskyi
Released by Dmytro Iakubovskyi through its official repository on Hugging Face. Read the license.
Built From
- Derived from google/vit-base-patch16-224-in21k
- Trained on (disclosed) nateraw/fairface
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 4.5 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 fairface_age_image_detection
How much GPU memory does fairface_age_image_detection 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 fairface_age_image_detection 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 fairface_age_image_detection commercially?
Yes. fairface_age_image_detection 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
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
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.