models microsoft beit base patch16 224 pt22k ft22k

microsoft-beit-base-patch16-224-pt22k-ft22k

Overview

The BEiT is a vision transformer that is similar to the BERT model, but is also capable of image analysis. The model is pre-trained on a large collection of images, and uses patches to analyze images. It uses relative position embeddings and mean-pooling to classify images, and can be used to extract image features for downstream tasks by placing a linear layer on top of the pre-trained encoder. You can place a linear layer on top of the [CLS] token or mean-pool the final hidden states of the patch embeddings, depending on the specifics of your task.

The above summary was generated using ChatGPT. Review the original-model-card to understand the data used to train the model, evaluation metrics, license, intended uses, limitations and bias before using the model.

Inference samples

Inference type	Python sample (Notebook)	CLI with YAML
Real time	image-classification-online-endpoint.ipynb	image-classification-online-endpoint.sh
Batch	image-classification-batch-endpoint.ipynb	image-classification-batch-endpoint.sh

Finetuning samples

Task	Use case	Dataset	Python sample (Notebook)	CLI with YAML
Image Multi-class classification	Image Multi-class classification	fridgeObjects	fridgeobjects-multiclass-classification.ipynb	fridgeobjects-multiclass-classification.sh
Image Multi-label classification	Image Multi-label classification	multilabel fridgeObjects	fridgeobjects-multilabel-classification.ipynb	fridgeobjects-multilabel-classification.sh

Model Evaluation

Task	Use case	Dataset	Python sample (Notebook)
Image Multi-class classification	Image Multi-class classification	fridgeObjects	image-multiclass-classification.ipynb
Image Multi-label classification	Image Multi-label classification	multilabel fridgeObjects	image-multilabel-classification.ipynb

Sample inputs and outputs (for real-time inference)

Sample input

{
  "input_data": {
    "columns": [
      "image"
    ],
    "index": [0, 1],
    "data": ["image1", "image2"]
  }
}

Note: "image1" and "image2" string should be in base64 format or publicly accessible urls.

Sample output

[
    {
        "probs": [0.91, 0.09],
        "labels": ["can", "carton"]
    },
    {
        "probs": [0.1, 0.9],
        "labels": ["can", "carton"]
    }
]

Model inference - visualization for a sample image

Version: 10

Tags

Preview license : apache-2.0 model_specific_defaults : ordereddict([('apply_deepspeed', 'true'), ('apply_ort', 'true')]) task : image-classification

View in Studio: https://ml.azure.com/registries/azureml/models/microsoft-beit-base-patch16-224-pt22k-ft22k/version/10

License: apache-2.0

Properties

SHA: 9da301148150e37e533abef672062fa49f6bda4f

datasets: imagenet, imagenet-21k

evaluation-min-sku-spec: 4|1|28|176

evaluation-recommended-sku: Standard_NC6s_v3

finetune-min-sku-spec: 4|1|28|176

finetune-recommended-sku: Standard_NC6s_v3

finetuning-tasks: image-classification

inference-min-sku-spec: 4|0|14|28

inference-recommended-sku: Standard_DS3_v2, Standard_D4a_v4, Standard_D4as_v4, Standard_DS4_v2, Standard_D8a_v4, Standard_D8as_v4, Standard_DS5_v2, Standard_D16a_v4, Standard_D16as_v4, Standard_D32a_v4, Standard_D32as_v4, Standard_D48a_v4, Standard_D48as_v4, Standard_D64a_v4, Standard_D64as_v4, Standard_D96a_v4, Standard_D96as_v4, Standard_FX4mds, Standard_F8s_v2, Standard_FX12mds, Standard_F16s_v2, Standard_F32s_v2, Standard_F48s_v2, Standard_F64s_v2, Standard_F72s_v2, Standard_FX24mds, Standard_FX36mds, Standard_FX48mds, Standard_E4s_v3, Standard_E8s_v3, Standard_E16s_v3, Standard_E32s_v3, Standard_E48s_v3, Standard_E64s_v3, Standard_NC4as_T4_v3, Standard_NC6s_v3, Standard_NC8as_T4_v3, Standard_NC12s_v3, Standard_NC16as_T4_v3, Standard_NC24s_v3, Standard_NC64as_T4_v3, Standard_NC24ads_A100_v4, Standard_NC48ads_A100_v4, Standard_NC96ads_A100_v4, Standard_ND96asr_v4, Standard_ND96amsr_A100_v4, Standard_ND40rs_v2

model_id: microsoft/beit-base-patch16-224-pt22k-ft22k