ImageNet-1k pre-training
According to the model card, the model is pre-trained on ImageNet-1k at 224x224 resolution for 1,000-class image classification.
Open Source Model Profile · microsoft
resnet-34
microsoft/resnet-34 is a 21.8M-parameter ResNet image-classification model. According to the model card, it is a ResNet-34 v1.5 variant pre-trained on ImageNet-1k at 224x224 resolution.
- Publisher
- microsoft
- Task
- image-classification
- Model type
- resnet
- License
- apache-2.0
- Library
- transformers
- Publication status
- Approved for indexing
Model overview
resnet-34 is published by Microsoft as an image-classification model. The captured configuration identifies ResNetForImageClassification with model type resnet. According to the model card, it is a ResNet-34 v1.5 model pre-trained on ImageNet-1k at 224x224 resolution, introduced via the Deep Residual Learning for Image Recognition paper.
Recorded capabilities
Residual learning design
According to the model card, the ResNet design uses residual learning and skip connections so much deeper models can be trained.
Documented v1.5 stride change
According to the model card, v1.5 moves stride 2 into the 3x3 convolution of downsampling blocks, with a small reported accuracy gain cited to Nvidia.
Transformers usage recipe
According to the model card, images can be classified with AutoFeatureExtractor and ResNetForImageClassification, illustrated on a COCO 2017 example.
Use cases in the source record
- Classifying images into one of the 1,000 ImageNet classes using the documented Transformers recipe.
- PyTorch or TensorFlow vision workflows that load the model through the Transformers ResNet stack.
Limitations and unknowns
- No context-window value applies to this image-classification record and none was extracted.
- No evaluation results were extracted from this record.
- Provider state is historical snapshot data, not independently refreshed current availability.
- The card states it was written by the Hugging Face team rather than the team releasing ResNet.
Source and provenance
Source: microsoft/resnet-34
Captured: Unknown. Processed: 2026-09-07T19:34:51.789403+00:00.
ResNet-34 v1.5 ResNet model pre-trained on ImageNet-1k at resolution 224x224. It was introduced in the paper Deep Residual Learning for Image Recognition by He et al. Disclaimer: The team releasing ResNet did not write a model card for this model so this model card has been written by the Hugging Face team. Model description ResNet (Residual Network) is a convolutional neural network that democratized the concepts of residual learning and skip connections. This enables to train much deeper models. This is ResNet v1.5, which differs from the original model: in the bottleneck blocks which require downsampling, v1 has stride = 2 in the…
F001F002F003F004F005F006F007F008F009F010F011F012F013F014F015F016