Thank you for the great effort, you are putting into this project :) There is, however, a feature I miss; rotated bounding boxes. Especially when objects are thin and diagonal, an ordinary bounding box fits poorly. Examples of such cases are shown here: rotated bounding boxes

A way annotation could be

I have set up Postgres in Kubernetes and also setup Doccano in Kubernetes however it's working well but wants to know the mount point for Kubernetes to attach Persistence volume.

My deployment.yaml

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  labels:
    app: doccano
  name: doccano
  namespace: default
spec:
  progressDeadlineSeconds: 600
  replicas: 1
  revisi

Request type

• Please close this issue, I accidentally submitted it without adding any details
• New documentation
• Correction or update

Details

Burried in the Formal Syntax is <media-or> that allows a list of media rules to use the or keyword. As I understand this change was added in [CSS Conditional Rules Module Level 3](https

📚 Documentation

Description

It is not clear how (and when) to use SubwordField from the documentation. And it is hard to find usage examples. It would be great if someone who used it would add at least a few lines to its doc.

For example, if I am using github.com/VKCOM/YouTokenToMe tokenizer - should I create SubwordField or Field. And what is the difference between them?

Expected Behavior

I want to convert torch.nn.Linear modules to weight drop linear modules in my model (possibly big), and I want to train my model with multi-GPUs. However, I have RuntimeError in my sample code. First, I have _weight_drop() which drops some part of weights in torch.nn.Linear (see the code below).

Actual Behavior

RuntimeError: arguments are located on different GPUs at /

It would be nice to have some general developer documentation for potential contributors to help in cases such as #510, etc.

What are the best steps to take towards accomplishing this? Maybe something similar (albeit not all details needed) to the Pandas developer docs?

I've begun an implementation of this on my fork, basicall

I noticed that you used image height param format as the font size.
https://github.com/Belval/TextRecognitionDataGenerator/blob/33d8985521645280e102987e773bf1e424a045df/TextRecognitionDataGenerator/computer_text_generator.py#L14
In my test, image_font = ImageFont.truetype(font=font_size=500), no error was reported, but it was time consuming.
So I am confused, why set format, font_size

I would be useful to implement support for various photoreceptor models so that it is possible to generate custom cone fundamentals down the road. I have started with Caroll et al. (2000), Stockman and Sharpe (2000), Lamb (1995) photoreceptor models in that notebook: https://colab.research.google.com/drive/1snhtUdUxUrTnw_B0kagvfz015Co9p-xv

We will obviously need support for various pre-receptor

The documentation file appears to have been generated with no space between the hashes and the header text. This is causing the headers to not display correctly, and is difficult to read. See below for an example of with and without the space:

##

Mobius API Documentation