twitter-archive/torch-autograd

Issues

• Failed to parse generated code when optimize=true

  #178 opened 6 years ago by farleylai
  0

• FloatTensor.cdata not currently supported by autograd when implementing SSIM loss

  #177 opened 6 years ago by farleylai
  1

• "Failed to parse generated code"

  #157 opened 8 years ago by abdullahjamal
  6

• gradfuns.lua dimension problem on line 130

  #176 opened 7 years ago by HeylenJonas
  0

• Feature: Autograd for tensor:split

  #174 opened 7 years ago by zzhang1987
  0

• Help with using autograd in training with wrapped NN modules

  #169 opened 8 years ago by synchro--
  4

• using `optimfn` with wrapped `nn` modules

  #173 opened 8 years ago by sebastiangonsal
  0

• mailing list for autograd

  #172 opened 8 years ago by biggerlambda
  0

• Autogradient of function cannot handle batched forward-prop

  #171 opened 8 years ago by NaimKabir
  1

• ABS still not implemented error but ABS code is actually there?

  #168 opened 8 years ago by synchro--
  0

• autograd.optim wrapper does not update state value

  #121 opened 8 years ago by hbqjzj
  5

• Wrong implemention for logsoftmax?

  #166 opened 8 years ago by callowbird
  1

• About tying the weights

  #165 opened 8 years ago by johnny5550822
  0

• "torch.DoubleTensor.t not currently supported". Workarounds?

  #159 opened 8 years ago by vadimkantorov
  5

• torch-cl?

  #163 opened 8 years ago by cathalgarvey
  0

• Autograd UPDATES weight and bias that DOESN'T CONTRIBUTE to the output of the module!

  #151 opened 8 years ago by dlmacedo
  1

• Question about implementing a pairwise L2 distances function

  #161 opened 8 years ago by vadimkantorov
  1

• Reusable AutoModule's

  #160 opened 8 years ago by vadimkantorov
  1

• Adding a torch-signal function

  #158 opened 8 years ago by ssakhavi
  0

• Support to torch.FloatTensor.resize

  #156 opened 8 years ago by abdullahjamal
  0

• torch.add(tensor1, scalar, tensor2) not supported.

  #155 opened 8 years ago by rtqichen
  1

• discrepency between licenses

  #153 opened 8 years ago by korymath
  1

• Make autograd ignore computing gradients for certain functions

  #152 opened 8 years ago by MLpatzer
  2

• I can't pass tensors to AutoModule

  #150 opened 8 years ago by dlmacedo
  0

• Different results due to optimize(true)

  #122 opened 8 years ago by mys007
  3

• Nil arguments prevent parsing later non-nil arguments

  #148 opened 8 years ago by allanzelener
  5

• autograd initialization error

  #149 opened 8 years ago by eladrich
  2

• Input is not a package name or nn object

  #131 opened 8 years ago by aleSuglia
  2

• Stop gradients

  #144 opened 8 years ago by ikostrikov
  3

• torch.bmm

  #111 opened 8 years ago by szagoruyko
  6

• Confusion with using optim with autograd

  #130 opened 8 years ago by zenna
  5

• GPU support

  #134 opened 8 years ago by Zealcui
  2

• Is it possible to segment the definition of the function to be taken the gradient?

  #118 opened 8 years ago by dlmacedo
  7

• {optimize = true} doesn't work with closures (even without updating variables)?

  #129 opened 8 years ago by oal17
  5

• Having issue autograd RecurrentLSTMNetwork

  #128 opened 8 years ago by rnunziata
  1

• Better documentation about optimization

  #125 opened 8 years ago by mys007
  2

• Gradient direct assignment is missing

  #126 opened 8 years ago by bartvm
  5

• util.makeContiguous

  #123 opened 8 years ago by mys007
  1

• Torch.gt autograd fails

  #112 opened 8 years ago by DmitryUlyanov
  3

• How to create a module to behave like a simple activation function?

  #120 opened 8 years ago by dlmacedo
  4

• Is it possible to use autograd.nn.AutoModule but not autograd.nn.AutoCriterion?

  #119 opened 8 years ago by dlmacedo
  2

• Wrong error message due to operation overloading

  #113 opened 8 years ago by albanD
  1

• Allow AutoModule to have no bias

  #117 opened 9 years ago by yangky11
  0

• Didn't catch an unsupported torch:copy / wrong gradient computation

  #106 opened 9 years ago by as1986
  3

• sigmoid operating on tensors only

  #115 opened 9 years ago by tuananhle7
  3

• "failed to parse generated code."

  #114 opened 9 years ago by tuananhle7
  1

• dot product in autograd

  #110 opened 9 years ago by pdakwal
  2

• attempt to call field 'nDimension' (a nil value)

  #109 opened 9 years ago by tuananhle7
  1

• "How it works" documentation

  #108 opened 9 years ago by nicholas-leonard
  1

• "Failed to parse generated code"

  #104 opened 9 years ago by xq111
  3