AMBER

AMBER copied to clipboard

https://github.com/zj-zhang/AMBER/blob/master/amber/architect/optim/controller/pytorch/generalController.py#L120C18-L120C45 default = True add benchmarking between `rescale_advantage_by_reward=False` vs `rescale_advantage_by_reward=True`

zj-zhang

`GeneralController.train(self)` vs `pmbga.train(self, episode, working_dir)`

zj-zhang

https://github.com/zj-zhang/AMBER/blob/945d4e8d560ccb48efa2af2a03c7624f100b7398/amber/architect/buffer.py#L77 `max_size` is the number of iterations, not the number of total sampled arcs. cc- @tinghui333

zj-zhang

Need better handling for multi-label tasks and take infer: https://github.com/zj-zhang/AMBER/blob/master/amber/backend/pytorch/model.py#L43

zj-zhang

how to call custom reward_fn

2

Hi, How can I parse any other (custom) reward function other than the default auc/aupr? I am following [this example](https://colab.research.google.com/gist/zj-zhang/48689d8bdc8adf3375719911f7e41989/amber-epigenetics-tutorial-v2.ipynb#scrollTo=KMX9a_0d72Lq) on colab how to use AMBER. Say i want to...

dvespasiani

What caused this error？ KeyError: 'loss'

6

I believe there is no problem with my data set construction. The error message is as follows: ` ['traindata', 'trainxdata'] ['validdata', 'validxdata'] BUILDING ---------- controller = GeneralController modeler = EnasCnnModelBuilder...

panzhengsen