Need Optional parameter for Keypoint class as we have one for BoundingBox

In case, we want to track the Keypoint transformations and pickout the labels which we want to use, It would be useful if we add label parameter.

As of now, we can do something like the below code in BB's :
`ia.BoundingBox(
x

To get started using this repo quickly using a Google Cloud Platform (GCP) Deep Learning Virtual Machine (VM) follow the instructions below. New GCP users are eligible for a $300 free credit offer. Other quickstart options for this repo include our [Google Colab Notebook](https://colab.research.google.com/github/ultralytics/yolov3/blob

Python : 3.7.4
OS : Mac OSX 10.14.3
Bug : 'OSError' object has no attribute 'message' when using sample method
Solution : message is deprecated in latest Python version. One must use msg while catching such exceptions

Description
In some rare cases, for example, when you need to finetune a large model on a small dataset the majoring part of training loop is waiting for saving model checkpoints to a hard drive.

Proposal
Would be logically to add a CheckpointCallback with parameter save_n_best=0 to a configuration and do not store best checkpoints and instead use the latest state of the model.

A very simple and naive augmenter, which just randomly crop part of the original text. Can work on char, word or sentence level.

I myself found it useful using with tf-idf, especially when you have only a very small dataset. I can provide an implementation if you'd like.

Hi PBA Team,

Spotted this warning in my IDE when looking through your code:

https://github.com/arcelien/pba/blob/a9cc308d936dee488aab7e3582ab7a47167b587a/pba/augmentation_transforms.py#L224

getargspec() has been deprecated since Python 3.0 in favour of getfullargspec()

It may be worthwhile adding a try/except block for this import because getfullargspec() isn't available in Pyt

The bundle command is really handy for installation, but can be a bit "magical." We should put a note in the output (and the configuration file) about what it's actually doing and make sure it is obvious and easy to change.

cc @edisgreat

Using the same concept used in keras segmentation example add pytorch segmentation example

Hello okankop!

This is a really impressive repo and it definitely saved me from writing those data augmentation functions. Thanks greatly for the great work!

But I do think it will be even better if you can provide a document, however easy it is. Maybe just add something to explain the input parameters at the bottom of the form in README.md.

Thanks again!

Below is the traceback and the libraries installed in my conda env.
My suspect is Python 3.5 let me know how to resolve this.
I Believe I need to update the Python 3.5 to 3.7 without disturbing the environment( idk how to do it ) suggestions are welcome

> Traceback (most recent call last):

  File "C:\Users\admin\Miniconda3\envs\machine_learning\lib\site-packages\IPython\core\interactive