-
Thanks a lot for your work and your research-contribution to neural architecture search with bayesian optimization. I am implementing BONAS for genomic sequence data and was wondering, wether the EA-s…
-
Things we care about in the upstream:
- Kubeflow
- Training
- Elastic Training
- https://github.com/kleveross/ftlib Fault-tolerant for DL frameworks
- https://github.com/kube…
-
# One-Shot
- 将搜索空间定义为超级网络 `SuperNet` - 包含全部网络
- `One-Shot` 共享权重
- `One-Shot` vs `One-Stage`
- `One-Stage` - 搜索完之后模型可以直接使用,而不再需要retrain等操作
- `One-Shot` - 构造 `SuperNet`
- 大部分 `One-Stag…
-
如题,和现有的NAS方面有啥区别,有木有文档参考一下~
wwjjy updated
4 years ago
-
We've gotten some good responses on our survey (https://github.com/deepchem/deepchem/issues/1722) about the future of DeepChem, and are starting to think about synthesizing the responses together. As …
-
## 🚀 Feature
I am interested in contributing an [exponentiated gradient descent optimizer](https://ttic.uchicago.edu/~tewari/lectures/lecture4.pdf) to pytorch.
## Motivation
Optimizing variabl…
-
Is anybody aware of any hardware that can efficiently train/infer depthwise convolutions? (Other than full datacentre TPUs) Is it in the pipeline for Nvidia GPUs? It really seems like the way forward …
-
**What would you like to be added**:
More case studies on the byproducts of Neural Architecture Search:
* MobileNetV3
* MobileDet
* Swish
These case studies do not need to fully imple…
-
Some ideas for figures to add to the PPT
- [ ] Linear regression, single-layer neural network
- [ ] Multilayer Perceptron with hidden layer
- [ ] Backpropagation
- [ ] Batch Normalization and al…
-
thank you for this nice program !
i compiled it on Linux and it seems to work well.
but i wonder about the interpretation .. here's one of my recent own games, see https://lichess.org/iFSd9qaS/bla…