uoguelph-mlrg/Theano-MPI
MPI Parallel framework for training deep learning models built in Theano
PythonNOASSERTION
Issues
- 1
- 0
how and where to initial the learing rate
#27 opened by Lx1010 - 32
- 0
Experience from "Accurate, Large Minibatch SGD"
#22 opened by hma02 - 2
- 5
Cifar10, with BSP vs EASGD
#25 opened by dlunga - 3
Cifar10_model not found
#24 opened by benliebersohn - 6
Segmentation fault for all examples with python3
#21 opened by Nqabz - 1
mpi run not found
#20 opened by dlunga - 4
ImportError: No module named 'rules'
#19 opened by dlunga - 0
- 1
layer definition inconsistency
#9 opened by hma02 - 1
need a CudaNdArray.as_buffer() function
#11 opened by hma02 - 5
Weird 16 GPU training time
#7 opened by FredericMao - 1
Error with adjust_lr() fuction
#17 opened by FredericMao - 5
is it suitable for cpu cluster?
#16 opened by fancyerii - 7
ZMQ error
#13 opened by FredericMao - 1
MKL_CBWR setting
#12 opened by FredericMao - 0
16GPU segfault bug on copper
#10 opened by hma02 - 1
cnmem wrong result
#8 opened by FredericMao - 2
Remove cudaconvnet (pylearn2) support
#1 opened by FredericMao - 1
Couldn't run on one GPU
#5 opened by FredericMao - 3
Googlenet not working
#6 opened by FredericMao - 3
Make central crop as a default
#2 opened by FredericMao - 1
hickle file loading in read-only mode
#4 opened by hma02 - 2
fp16 error
#3 opened by FredericMao