The timm/convnext_base.fb_in22k_ft_in1k is a ConvNeXt Base model for image classification. It was pre-trained on the ImageNet-22k dataset and fine-tuned on ImageNet-1k, as indicated by its name and associated tags. The model forms part of the timm PyTorch Image Models collection hosted on Hugging Face.
It can be loaded and used through two primary libraries. With timm, the model is instantiated via a single create_model call that references the Hugging Face hub identifier and enables pretrained weights. The Transformers library supports it through a high-level pipeline configured for image-classification tasks as well as direct loading of the underlying model with AutoModel, including options for automatic device mapping. Example code demonstrates inference on an image URL such as a photograph of parrots.
The model card lists it under the Apache-2.0 license. It includes PyTorch and Safetensors file formats along with references to the arXiv paper 2201.03545. The entry appears in the context of image classification tasks and carries tags for both imagenet-1k and imagenet-22k datasets.
This implementation is delivered as a downloadable model repository on the Hugging Face platform rather than a standalone application. It targets developers and researchers who integrate it into Python-based computer vision workflows for inference, further fine-tuning, or as a feature extractor.
Convnext Base.fb In22k Ft In1k sits in PulseGate's Other AI category. It focuses on performing accurate image classification using a modern convolutional neural network without training from scratch. Convnext Base.fb In22k Ft In1k is an open-source project aimed at developers. The project is open source (MIT). It ships for the web and API.
PyTorch Image Models builds and maintains Convnext Base.fb In22k Ft In1k, and it first shipped in 2022. The GitHub repository has been archived. PulseGate's similarity index places it among 6 comparable projects. Among its 3 catalogued features are image classification, convNeXt architecture, and imageNet pretrained.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do