https://github.com/rom1504/deepfashion_to_tfrecords

Convert deepfashion to tfrecords to learn multimodal models
https://github.com/rom1504/deepfashion_to_tfrecords

Last synced: 14 days ago
JSON representation

Convert deepfashion to tfrecords to learn multimodal models

• Host: GitHub
• URL: https://github.com/rom1504/deepfashion_to_tfrecords
• Owner: rom1504
• Created: 2021-05-08T13:06:21.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2021-05-08T13:41:11.000Z (over 3 years ago)
• Last Synced: 2024-10-17T07:24:18.398Z (25 days ago)
• Language: Jupyter Notebook
• Size: 972 KB
• Stars: 2
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# deepfashion_to_tfrecords

Convert deepfashion to tfrecords to learn multimodal models

![image](https://user-images.githubusercontent.com/2346494/117541312-996f2f80-b013-11eb-973c-7d60ee68fb5d.png)

## Usage

First you need to get https://github.com/switchablenorms/DeepFashion2 then:

`pip install tensorflow fire`

`python deepfashion_to_tfrecords --dataset_path=deepfashion/train --num_shards=100 --output_folder=outputs`

It runs in 5min when using 20 cores

Check the results at [deepfashion to tfrecords](read_with_tfdata.ipynb)