Metadata-Version: 2.1
Name: TorchFly
Version: 0.0.1
Summary: Pytorch Fast Development Kit
Home-page: https://github.com/qywu/TorchFly
Author: Qingyang Wu
Author-email: wilwu@ucdavis.edu
License: UNKNOWN
Description: # TorchFly
        
        TorchFly is a PyTorch Fast Development Kit. The purpose is to learn the pipelines of the SOTA algorithms in Deep Learning areas like CV, NLP and RL. The utilities provided in this kit will shorten the time needed to rebuild some basic functions. Now, the kit is mainly for personal use, but will be updated from time to time. 
        
        ## Installation
        
        ### Installing via pip:
        Installing is simple using `pip`.
        
           ```bash
           pip install torchfly
           ```
           
        ### Installing from source
        You can clone the repository.
            ```bash
            git clone https://github.com/qywu/TorchFly.git
            ```
        
        ### Docker
        It is recommended to run on Nvidia Docker for better performance.
        
            ```bash
            FROM nvcr.io/nvidia/pytorch:18.12.1-py3
            RUN apt-get update
            
            RUN pip install torchfly
            ```
         
         ## TODOS
         
        1. Custom Bucket Sampler
        
        2. Custom Beam Search
        
         
         ## Code References
         
         [FastAI](https://github.com/fastai)
         
         [AllenNLP](https://github.com/allenai/allennlp/)
         
         [Pytorch BERT](https://github.com/huggingface/pytorch-pretrained-BERT)
        
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Description-Content-Type: text/markdown
