Code

Pruning, save on GPU, test on CPU.#152

Code

Open

bryanbocao wants to merge 55 commits intokuangliu:masterkuangliu/pytorch-cifar:masterfrom bryanbocao:wipbryanbocao/pytorch-cifar:wipCopy head branch name to clipboard

+1,795-45

Commits

Commits on Jun 22, 2022

• Add args input and pruning

  bryanbocao

  committed
  318e504

  Copy full SHA for 318e504

• Update args and pruning reference

  bryanbocao

  committed
  38912e4

  Copy full SHA for 38912e4

• Add commands with args

  bryanbocao

  authored
  713fa09

  Copy full SHA for 713fa09

• Add nohup version

  bryanbocao

  authored
  2937365

  Copy full SHA for 2937365

Commits on Jun 24, 2022

• --train --test with action | Save on GPU, test on CPU

  bryanbocao

  authored
  b782bba

  Copy full SHA for b782bba

• --train --test with action

  bryanbocao

  authored
  aea8bde

  Copy full SHA for aea8bde

• Test on CPU

  bryanbocao

  authored
  b9f693a

  Copy full SHA for b9f693a

• Clean main.py

  bryanbocao

  authored
  3102ecf

  Copy full SHA for 3102ecf

• Print summary of net

  bryanbocao

  authored
  c85353c

  Copy full SHA for c85353c

• Print layer params

  bryanbocao

  authored
  18edadd

  Copy full SHA for 18edadd

Commits on Jun 30, 2022

• Add Trained Weights

  bryanbocao

  authored
  ba1df42

  Copy full SHA for ba1df42

Commits on Jul 2, 2022

• test_sim.py

  bryanbocao

  committed
  b4cbcd2

  Copy full SHA for b4cbcd2

Commits on Jul 7, 2022

• main_n_cls.py

  bryanbocao

  committed
  10d8828

  Copy full SHA for 10d8828

• Start training for n_cls experiments

  bryanbocao

  authored
  16cb8ab

  Copy full SHA for 16cb8ab

• Save n_cls checkpoints

  bryanbocao

  committed
  f8a84cf

  Copy full SHA for f8a84cf

• Merge branch 'wip' of https://github.com/bryanbo-cao/pytorch-cifar into wip

  bryanbocao

  committed
  40399b7

  Copy full SHA for 40399b7

• main_n_cls_nohup.py

  bryanbocao

  committed
  383d959

  Copy full SHA for 383d959

• Check if inputs are empty before feeding into the network

  bryanbocao

  committed
  c2c7714

  Copy full SHA for c2c7714

• Check if inputs are empty before feeding into the network

  bryanbocao

  committed
  f978249

  Copy full SHA for f978249

Commits on Jul 8, 2022

• Correct num_classes layer for MobileNetV2

  bryanbocao

  committed
  610e452

  Copy full SHA for 610e452

• Correct num_class layer for MobileNetV2

  bryanbocao

  committed
  dad10af

  Copy full SHA for dad10af

• Move inputs,targets to GPU

  bryanbocao

  committed
  7ea9c17

  Copy full SHA for 7ea9c17

• Save models at epoch=0

  bryanbocao

  committed
  62a444a

  Copy full SHA for 62a444a

Commits on Jul 9, 2022

• Move net.to(device) after loading the weights

  bryanbocao

  committed
  a27d19c

  Copy full SHA for a27d19c

• Save models

  bryanbocao

  committed
  2e8d7ae

  Copy full SHA for 2e8d7ae

• Save models

  bryanbocao

  committed
  9ac704a

  Copy full SHA for 9ac704a

• Load epoch

  bryanbocao

  committed
  4ae4620

  Copy full SHA for 4ae4620

Commits on Jul 10, 2022

• Add num_classes args for models

  bryanbocao

  committed
  e007fe5

  Copy full SHA for e007fe5

• rm __pycache__

  bryanbocao

  committed
  4aded52

  Copy full SHA for 4aded52

• Add resnet num_classes

  bryanbocao

  committed
  5217d90

  Copy full SHA for 5217d90

• Create sweep_n_cls_0.sh

  bryanbocao

  authored
  861f042

  Copy full SHA for 861f042

• Create sweep_n_cls_1.sh

  bryanbocao

  authored
  1113952

  Copy full SHA for 1113952

• Add BasicBlock arg for simpleDLA

  bryanbocao

  committed
  5ba5748

  Copy full SHA for 5ba5748

• Split works

  bryanbocao

  committed
  148b18b

  Copy full SHA for 148b18b

• training and testing devices

  bryanbocao

  committed
  1644d32

  Copy full SHA for 1644d32

Commits on Jul 11, 2022

• Add more bash scrpits for training

  bryanbocao

  committed
  bb0c72b

  Copy full SHA for bb0c72b

• Add CUDA_3

  bryanbocao

  committed
  fa96e89

  Copy full SHA for fa96e89

• Add CUDA_3

  bryanbocao

  committed
  f048194

  Copy full SHA for f048194

• Add arg num_classes in EfficientNetB0

  bryanbocao

  committed
  d6aca36

  Copy full SHA for d6aca36

• Correct bash scripts

  bryanbocao

  committed
  e349cf0

  Copy full SHA for e349cf0

• GoogLeNet CUDA_0

  bryanbocao

  committed
  6bc1098

  Copy full SHA for 6bc1098

Commits on Jul 12, 2022

• Add CUDA

  bryanbocao

  committed
  9adf370

  Copy full SHA for 9adf370

Commits on Jul 13, 2022

• Add num_class

  bryanbocao

  committed
  aa20d8e

  Copy full SHA for aa20d8e

• Save 10 classes by default

  bryanbocao

  committed
  ecac0bf

  Copy full SHA for ecac0bf

• Correct args

  bryanbocao

  committed
  c3a1e13

  Copy full SHA for c3a1e13

Commits on Jul 18, 2022

• D and S groups

  bryanbocao

  committed
  848a806

  Copy full SHA for 848a806

• Add num_class groups

  bryanbocao

  committed
  0f86d7f

  Copy full SHA for 0f86d7f

• Add D2G1

  bryanbocao

  committed
  c792b84

  Copy full SHA for c792b84

• Correct class group

  bryanbocao

  committed
  2a98e5d

  Copy full SHA for 2a98e5d

• Convert target into [0..n] range

  bryanbocao

  committed
  1c8bc3d

  Copy full SHA for 1c8bc3d

• Add tensor

  bryanbocao

  committed
  f55c39d

  Copy full SHA for f55c39d

• Correct the bug of index

  bryanbocao

  committed
  00fd6fb

  Copy full SHA for 00fd6fb

• Correct the bug of index

  bryanbocao

  committed
  4fd9728

  Copy full SHA for 4fd9728

• Correct the bug of index

  bryanbocao

  committed
  5535497

  Copy full SHA for 5535497

Commits on Jul 19, 2022

• Add class group in checkpoint path

  bryanbocao

  committed
  39b2f62

  Copy full SHA for 39b2f62