Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 29, 2020

• feat: print epoch during testing

  weiaicunzai

  committedDec 29, 2020
  2149cb5

• fix: google large gpu memory usage

  weiaicunzai

  committedDec 29, 2020
  49ddb26

• feat: change argument gpu to store_true

  weiaicunzai

  committedDec 29, 2020
  b9cca00

• feat: print gpu info only once when testing

  weiaicunzai

  committedDec 29, 2020
  0df7238

Commits on Dec 28, 2020

• fix: large gpu memory required by googlenet, add more downsampling

  weiaicunzai

  committedDec 28, 2020
  a9c8225

• refactor: change output to x in googlenet forward function

  weiaicunzai

  committedDec 28, 2020
  5bd0fa8

• refactor: put lr.step() after optimizer.step()

  weiaicunzai

  committedDec 28, 2020
  ca235c2

• refactor: add runs and checkpoints folders to .gitignore file

  weiaicunzai

  committedDec 28, 2020
  ffdb814

• dcos: remove gpu memory util, memory util is effected by input image

  weiaicunzai

  committedDec 28, 2020
  ddf1db4

• docs: add sotchastic depth 101 result

  weiaicunzai

  committedDec 28, 2020
  1c7ab20

• docs: add stochasticdepth50 result

  weiaicunzai

  committedDec 28, 2020
  01781f9

• docs: add stochasticdepth34 result

  weiaicunzai

  committedDec 28, 2020
  1d96935

• feat: put tensor p inside a function instead of class member

  weiaicunzai

  committedDec 28, 2020
  bf2fa9d

• fix: always return false due to different tensor shape in survial method

  weiaicunzai

  committedDec 28, 2020
  ca25ee9

• fix: set torch.ones(1) on the same device as the output of torch.bernoulli()

  weiaicunzai

  committedDec 28, 2020
  5974f22

• fix: remove redundant newline

  weiaicunzai

  committedDec 28, 2020
  e471d6a

• fix: torch.jit.tracke error: change block type to torch.jit.ScriptMedule

  weiaicunzai

  committedDec 28, 2020
  60b7236

• refactor: use pytorch buildin bernoulli distribution sampler

  weiaicunzai

  committedDec 28, 2020
  4cc6a35

• feat: add stochastic depth network

  weiaicunzai

  committedDec 28, 2020
  ba50283

• feat: push for test

  weiaicunzai

  committedDec 28, 2020
  aecf3af

• feat: put tensor p inside a function instead of class member

  weiaicunzai

  committedDec 28, 2020
  eaa4090

• feat: print gpu usage

  weiaicunzai

  committedDec 28, 2020
  979c8ab

• fix: always return false due to different tensor shape in survial method

  weiaicunzai

  committedDec 28, 2020
  09bc55e

• fix: set torch.ones(1) on the same device as the output of torch.bernoulli()

  weiaicunzai

  committedDec 28, 2020
  f9c63d2

• fix: remove redundant newline

  weiaicunzai

  committedDec 28, 2020
  8c38d9d

• fix: torch.jit.tracke error: change block type to torch.jit.ScriptMedule

  weiaicunzai

  committedDec 28, 2020
  186db4e

• refactor: use pytorch buildin bernoulli distribution sampler

  weiaicunzai

  committedDec 28, 2020
  72cd1b3

• fix: remove redundant print function

  weiaicunzai

  committedDec 28, 2020
  e8d7ce4

• feat: add stochastic depth network

  weiaicunzai

  committedDec 28, 2020
  8c6a1d7

• feat: output weights file path when saving

  weiaicunzai

  committedDec 28, 2020
  5c14898

• fix: fewer epoch number than desired

  weiaicunzai

  committedDec 28, 2020
  ce88ba9

Commits on Dec 23, 2020

• fix: integer division error in shufflenetv2

  weiaicunzai

  committedDec 23, 2020
  225510f

Commits on Oct 28, 2020

• fix: input tensor type and network weights tensor type should be the

  weiaicunzai

  committedOct 28, 2020
  22c4fae

Commits on Oct 19, 2020

• dcos: add wide resnet results

  weiaicunzai

  committedOct 19, 2020
  75cd6e6

Commits on Oct 18, 2020

• fix: wrong output message format

  weiaicunzai

  committedOct 18, 2020
  94694af

Pagination

PreviousNext