-
## 🚀 Feature
Improved image batch plots during training
## Motivation
It's nice to see labeling performance as training progresses, but there are a few issues:
* Boxes are not coloured per-c…
-
As a proof of concept, I rendered an object in blender, and re-used the same object position and camera parameters to render an image in Blender. I tried to regress a texture in both Lambertian and Ve…
-
## Summary
I used Mindtpy to solve syn40m04m from the minlplib with gurobi as the mip solver and baron as the nlp solver and received a "this should never happen" error.
### Steps to reprodu…
-
We need to add a gradient penalty in the GAN discriminator similar to StyleGAN for stability.
Katherine Crowson suggested spectral norm as an alternative and maybe EMA.
-
* Orion version: 0.3.1.dev0
* Python version: 3.7.11
* Operating System: macOS
### Description
There is a significant increase in training time per signal between TadGAN implemented in TensorFlo…
-
Heya,
this might be nothing, but when starting from a previous checkpoint/epoch, I get the following warning:
```
Using GPU with mixed precision enabled...
Calculating total number of sample…
-
**Describe the bug**
The same issue as https://github.com/microsoft/DeepSpeed/issues/3228, except for stage3 with zero init
**To Reproduce**
Steps to reproduce the behavior:
1. Install `accelera…
-
hello, I have a question about the implementations of **CASE U3**:
why the formula of calculating bitwidth is different between weight and activation?
https://github.com/sony/ai-research-code/blob/m…
-
In this line, a "ridge penalty" is applied, but actually it seems to be a lasso penalty (and the weight parameter says "L1", indicating as such).
https://github.com/matsengrp/multidms/blob/bea2c93dbc…
-
Sample a set of positive / negative graph paths from entities to answers.
Make it fairly balances something between 1-1 to 1-5 fraction of positives/negatives.
Add the dataset sample here to d…