erasure code.pdf: 项目的总体方向
FusedLayer_test.py: 它对应于pdf中给出的第一步操作, 给定output的range, 计算input的range
其他可能需要的代码:
-
Learning-Based_Coded_Computation 的仓库:
它对应于编码和解码的过程
-
EdgeLD 的仓库:
它对应于更后续的工作, 研究异质设备下, workload分配的部分
- MNIST
- CIFAR-10 / CIFAR-100
- ResNet: Deep Residual Learning for Image Recognition
- VGGNet: Very Deep Convolutional Networks for Large-Scale Image Recognition
- DenseNet: Densely Connected Convolutional Networks
- EfficientNet: EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
- Wide ResNet: Wide Residual Networks
input shape:
- MLP with two layers:
$k a a' \times k a a'$ ,$kaa' \times raa'$ - CNN with ? layers
input shape:
- MLP with three layers:
$nbb' \times kbb'$ ,$kbb' \times kbb'$ ,$kbb' \times kbb'$ - CNN with ? layers