search

macournoyer / neuralconvo

Neural conversational model in Torch
776 stars 347 forks source link

Crash on 2nd epoch: invalid arguments: DoubleTensor number FloatTensor - adam.lua #75

Open TTN- opened 7 years ago

TTN- commented 7 years ago 
user@pc:~/Scripts/neuralconvo$ th train.lua --dataset 1000 --hiddenSize 500 --maxEpoch 10
-- Loading dataset  
Loading vocabulary from data/vocab.t7 ...   

Dataset stats:  
  Vocabulary size: 2536 
         Examples: 1569 

-- Epoch 1 / 10  (LR= 0.001)    

 [================ 1569/1569 ==========>]  Tot: 3m17s | Step: 128ms     

Finished in 3m18s 7.9049796615389 examples/sec. 

Epoch stats:    
  Errors: min= 4.3083517415234  
          max= 19.967736960488  
       median= 9.4001431829178  
         mean= 10.031499396177  
          std= 3.1296645971186  
          ppl= 22731.329274047  

(Saving model ...)  

-- Epoch 2 / 10  (LR= 0.0009505)    

/home/user/Scripts/torch/install/bin/luajit: ...user/Scripts/torch/install/share/lua/5.1/optim/adam.lua:59: invalid arguments: DoubleTensor number FloatTensor 
expected arguments: *DoubleTensor* [DoubleTensor] double | *DoubleTensor* [DoubleTensor] [double] DoubleTensor
stack traceback:
    [C]: in function 'add'
    ...user/Scripts/torch/install/share/lua/5.1/optim/adam.lua:59: in function 'adam'
    train.lua:131: in main chunk
    [C]: in function 'dofile'
    ...ipts/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00406670

Anyone know how to fix this or a work around?

TTN- commented 7 years ago

@macournoyer if you could accept the pull request once you're happy with it that would be great :-)