Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · facebook

convnextv2-large-22k-224

convnextv2-large-22k-224 is a 198M-parameter ConvNeXt V2 image-classification model from facebook. According to the model card, it was pretrained with FCMAE and fine-tuned on ImageNet-22K at 224x224.

Publisher
facebook
Task
image-classification
Model type
convnextv2
License
apache-2.0
Library
transformers
Publication status
Accepted · not indexed

Model overview

convnextv2-large-22k-224 is published by facebook as an image-classification model. The captured configuration identifies ConvNextV2ForImageClassification with model type convnextv2, and Safetensors metadata reports 197,956,840 parameters. According to the model card, it is a large-sized ConvNeXt V2 model pretrained with FCMAE and fine-tuned on ImageNet-22K at 224x224.

Recorded capabilities

FCMAE pretraining and ImageNet-22K fine-tune

According to the model card, the model was pretrained using the FCMAE framework and fine-tuned on the ImageNet-22K dataset at 224x224 resolution.

Pure ConvNet design with GRN

According to the model card, ConvNeXt V2 is a pure convolutional model introducing a fully convolutional masked autoencoder framework and a Global Response Normalization layer.

Documented Transformers classification workflow

According to the model card, the model classifies images into 1,000 ImageNet classes with AutoImageProcessor and ConvNextV2ForImageClassification, illustrated on a COCO 2017 example.

ImageNet-22K and paper association

Hub tags record dataset:imagenet-22k and arxiv:2301.00808, matching the card's description of ImageNet-22K tuning and the Woo et al. ConvNeXt V2 paper.

Use cases in the source record

  • Raw image classification into 1,000 ImageNet classes using the documented AutoImageProcessor and ConvNextV2ForImageClassification workflow.
  • Starting point for locating task-specific fine-tuned versions through the model hub, as suggested by the model card.

Limitations and unknowns

  • No evaluation results were extracted from this record.
  • The card states it was written by the Hugging Face team because the releasing team did not provide one, so its descriptions are secondary documentation rather than publisher-authored claims.
  • Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: facebook/convnextv2-large-22k-224

Captured: Unknown. Processed: 2026-09-07T19:34:44.256640+00:00.

ConvNeXt V2 (large-sized model) ConvNeXt V2 model pretrained using the FCMAE framework and fine-tuned on the ImageNet-22K dataset at resolution 224x224. 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. Model description ConvNeXt V2 is a pure convolutional model (ConvNet) that introduces a fully convolutional masked autoencoder framework (FCMAE) and a new Global Response Normalizatio…

F001F002F003F004F005F006F007F009F010F011F012F013F014F015F016