finetuned-clothes

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

finetuned-clothes

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the clothes_simplifiedv2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2225
• Accuracy: 0.9417

Model description

This model classifies clothes category based on the given image.

Intended uses

You can use it in a jupyter notebook:

from PIL import Image
import requests

url = 'insert image url here'
image = Image.open(requests.get(url, stream=True).raw)

from transformers import AutoModelForImageClassification, AutoImageProcessor

repo_name = "samokosik/finetuned-clothes"

image_processor = AutoImageProcessor.from_pretrained(repo_name)
model = AutoModelForImageClassification.from_pretrained(repo_name)

encoding = image_processor(image.convert("RGB"), return_tensors="pt")
print(encoding.pixel_values.shape)

import torch
with torch.no_grad():
  outputs = model(**encoding)
  logits = outputs.logits

predicted_class_idx = logits.argmax(-1).item()
print("Predicted class:", model.config.id2label[predicted_class_idx])

Limitations

Due to lack of available data, we support only these categories: hat, longsleeve, outswear, pants, shoes, shorts, shortsleve.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1

Training dataset

This model was trained on the following dataset: https://huggingface.co/datasets/samokosik/clothes_simplifiedv2