Timm-converted ViT-Small weights
According to the model card, the checkpoint converts weights from the timm repository for ViT-Small availability on Hugging Face.
Open Source Model Profile · WinKawaks
vit-small-patch16-224
vit-small-patch16-224 is a 22.05M-parameter ViT-Small image classifier from WinKawaks. According to the model card, its weights were converted from timm for ViT-base-style use.
- Publisher
- WinKawaks
- Task
- image-classification
- Model type
- vit
- License
- apache-2.0
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
vit-small-patch16-224 is published by WinKawaks as an image-classification model. The captured configuration identifies ViTForImageClassification with model type vit, and Safetensors metadata reports 22,050,664 parameters. According to the model card, it converts ViT-Small weights from timm and follows ViT-base usage.
Recorded capabilities
ViT-base-compatible use
According to the model card, this model is used in the same way as ViT-base.
Documented torch 2.0 safetensors note
According to the model card, the safetensors model requires a torch 2.0 environment.
22.05M-parameter ViT record
Captured configuration identifies ViTForImageClassification with model type vit, and Safetensors metadata reports 22,050,664 parameters; tags record imagenet association.
Use cases in the source record
- Lightweight ViT-Small image classification using the same workflow as ViT-base.
- Small-model ViT experiments in a torch 2.0 safetensors environment, following the publisher's conversion and compatibility notes.
Limitations and unknowns
- No evaluation results were extracted from this record.
- No training details beyond the publisher-described timm conversion were extracted.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: WinKawaks/vit-small-patch16-224
Captured: Unknown. Processed: 2026-09-07T19:34:38.524576+00:00.
Google didn't publish vit-tiny and vit-small model checkpoints in Hugging Face. I converted the weights from the timm repository . This model is used in the same way as ViT-base . Note that [safetensors] model requires torch 2.0 environment.
F001F002F003F004F005F006F007F009F010F011F012