Issues
- 1
- 0
same sentence different padding length result different embedding because softmax precision
#1326 opened by JJplane - 1
How to resume training from a certain epoch or how to update an already trained model in torch ?
#1305 opened by maryam089 - 0
Compatibility issues
#1324 opened by kysunami - 0
install/share/lua/5.1/nn/LookupTable.lua:73: bad argument #3 to 'index' (Tensor | LongTensor expected, got torch.CudaLongTensor)
#1323 opened by m0dulo - 3
SpatialDepthWiseConvolution swaps height and width
#1307 opened by diovisgood - 0
Backprop procedure for nn.Narrow
#1322 opened by pskrunner14 - 0
- 1
getParamsByDevice Branch problem
#1293 opened by MikeChenfu - 0
TripletCriterion (aka triplet loss) missing
#1320 opened by jatinsha - 0
parallelcriterion has no length operator
#1317 opened by 980380446 - 1
Ask a question of nn.SplitTable?
#1312 opened by xtanitfy - 1
- 0
FastLSTM out of memory error
#1316 opened by tastyminerals - 0
Why gradInput of nn.Mul is computed by gradOutput add by weight rather than multiply by weight?
#1315 opened by ChunyanZhao - 0
ClassNLLCriterion.cu t < n_classes failed.
#1314 opened by skynbe - 0
Simple real time object detection using Torch
#1313 opened by VictorQueiroz - 0
- 0
- 0
Using nn.Padding throws luajit: C++ exception
#1309 opened by tastyminerals - 0
Strange behavior of nn.Padding during training
#1308 opened by tastyminerals - 11
Simple neural net has different output for inputs in the same batch that are identical
#1306 opened by happypepper - 4
The getParameters() function cannot work with the module designed in the criterion part of document
#1304 opened by hfxunlp - 4
ParallelCriterion input form?
#1302 opened by jmfu95 - 0
/torch/install/share/lua/5.1/nn/LookupTable.lua:59: attempt to call method 'Contiguous' (a nil value)
#1301 opened by jmfu95 - 0
- 4
Bad Argument
#1299 opened by zaktab - 1
Uses the deprecated package ffi
#1298 opened by yurivict - 1
Issue with type conversion for BatchNorm inference in THNN
#1296 opened by libfun - 0
How to separate a network into submodules
#1295 opened by SeunghyunMoon - 2
- 0
MSECriterion_updateGradInput error: cannot convert 'struct THCudaHalfTensor *' to 'struct THCudaTensor *
#1290 opened by michaelhuang74 - 1
nn.Recurrent() function is missing!
#1289 opened by Elyorcv - 0
Inconsistence in SpatialMaxPooling and SpatialMaxUnpooling operations for torch.CudaTensor that leads to an error
#1287 opened by borismu - 1
How to use nn.ParallelTable() inside nn.Sequential()
#1286 opened by mbcel - 0
- 0
- 0
Bug: Inconsistence of the channel of `input` and that of `self.gradInput` can also work!
#1279 opened by Naruto-Sasuke - 0
multi target not supported issue in ClassNLLCriterion
#1278 opened by dephaniehe - 4
- 0
- 0
- 3
torch.nn in context of a cuda stream
#1273 opened by uriv - 0
VolumetricConvolution + ReLU
#1268 opened by ebsrn - 0
VolumetricUpSamplingNearest module
#1272 opened by mks0601 - 1
Inconsistency in nn.HardShrink() activation function?
#1270 opened by NataliaDiaz - 2
[Linear Layer] Possibly wrong gradient?
#1269 opened by cdluminate - 2
Why is not nn.MSDCriterion() listed in https://github.com/torch/nn/blob/master/doc/criterion.md ?
#1266 opened by NataliaDiaz - 7
Volumetric Upsampling Modules Release Date
#1262 opened - 1
nn.SpatialUpSamplingBilinear (a nil value)
#1263 opened by abdullahjamal