Hi, thanks for the great code!

I wonder do you have plans to support resuming from checkpoints for classification? As we all know, in terms of training ImageNet, the training process is really long and it can be interrupted somehow, but I haven't notice any code related to "resume" in scripts/classification/train_imagenet.py.

Maybe @hetong007 ? Thanks in advance.

I tried building the docs, but was met with a graphviz error. Typically this means I can spend a few hours pecking away at the dependencies until I get stable build... or someone that has it working can export their environment, and publish an environment.yml that we can use with the build instructions.
I was going off of the d2l book since that's a dep here, but their [environment.yml](https://g

Hi there and thanks for this product. Currently, I see zero build instructions overall. How to install just Gluon on my computer, so I can work with it? Cooperating with Rust is a non-goal by now and it seems the current Readme assumes I am a Rust developer and familiar with its package system.

I think we can remove the check of mutable_args in the get_model API for BERT, GPT, Language model and transformer. We can introduce a separate flag "allow_override" if users want to override any configuration. Otherwise overriding any configuration is forbidden.

Great work!

I discovered in https://github.com/arogozhnikov/einops/blob/master/einops/einops.py#L199 that you also support ellipsis. Its an important feature so you may want to add it to the documentation.

Bug report

I am doing first experiments with multi-domain support following this. After setting up the domain, I used these instructions to configure my node:

uci set gluon.core.domain="test"
gluon-reconfigure
gluon-reload

What is the problem?
After doing that, the node still sends RAs for the old networ

mixins sets a default fetchPolicy on a per-component basis. This may conflict with the user's

In Colab notebook :

https://github.com/arita37/mlmodels/blob/dev/mlmodels/example/README_usage.md

你好，我按照你的方法制作了多目标检测的.rec和.idx文件，但是读取的时候出现 Encounter sample with no valid label 错误，能帮我看一下吗？多谢了
这是我生成.lst文件的程序
with open('dataset/dataset.lst', 'w+') as f:
----f.truncate()
----for i in range(len(labels)):
--------lst = str(i) + '\t' + str(4) + '\t' + str(5) + '\t' + str(1360) + '\t' + str(800) + '\t'
--------for label in labels[i]:
------------lst += str(label[4]) + '\t' + str(round(label[0]/