Issues
- 1
Feature Request: Instance Normalization
#604 opened by andevellicus - 1
- 6
Make CuArrays work with Knet.load/save.
#587 opened by denizyuret - 7
New GPUArrays indexing causes scalar indexing for some Knet operations (i.e., cat)
#618 opened by andevellicus - 9
libknet8 library not found
#611 opened by bioorg - 2
LeNet model for MNIST Example
#617 opened by nelsonmarcelino - 3
Error converting CuArray to KnetArray
#619 opened by ngphuoc - 0
`k[1:2:3,:] .= 0` broken for KnetArray
#620 opened by denizyuret - 0
AlphaZero benchmark
#616 opened by denizyuret - 4
Building Knet fails in 1.5.1
#615 opened by philtomson - 4
Knet v1.4 loading v1.3.9 model: JLD2- Knet.softmax not defined in workspace
#600 opened by radonnachie - 7
ERROR: LoadError: MethodError: no method matching unsafe_free!(::Nothing)
#610 opened by andevellicus - 3
Knet.Layers20
#607 opened by denizyuret - 5
Knet GPU OOM with recent CUDA/Knet updates
#602 opened by andevellicus - 0
- 5
Use CUDA.jl functions where possible
#573 opened by denizyuret - 0
Knet.CUDNN: a submodule for cudnn calls
#605 opened by denizyuret - 0
Knet.Data
#608 opened by denizyuret - 0
Create Knet.Layers21
#584 opened by denizyuret - 0
Create Knet.Models21
#585 opened by denizyuret - 18
Seg fault after upgrading to Julia 1.5
#589 opened by andevellicus - 0
Feature request: `DataParallel` like in PyTorch
#603 opened by ngphuoc - 9
CPU computation (Switching off GPU computation)/@diff of Array{} values returns 0-element Array
#594 opened by radonnachie - 2
Knet.randn! not defined in v1.4.0
#598 opened by ngphuoc - 1
- 0
- 2
pool() Discrepancy between KnetArray (pass) and Arrays (fails) for shape (X,1,D,N)
#591 opened by radonnachie - 3
LBFGS optimizer
#590 opened by qzhu2017 - 5
- 2
- 2
Knet GPU selection and auto device selection
#586 opened by ngphuoc - 0
- 1
Create a profiling suite
#588 opened by denizyuret - 7
- 2
CUDA.pool not defined
#595 opened by radonnachie - 1
Missing website
#592 opened by zlatanvasovic - 7
cudnn 8.0 compatibility
#569 opened by kadir-gunel - 0
Get rid of @cu calls, replace them with CUDA calls
#580 opened by denizyuret - 0
Get rid of libknet8 calls, replace with Julia kernels, CUDA calls, Torch.jl etc.
#581 opened by denizyuret - 2
- 8
- 2
CUDAapi doesn't seem to have find_host_compiler
#567 opened by ViralBShah - 1
Warning: Hardware is unsupported by NNPACK so falling back to default NNlib
#576 opened by denizyuret - 0
addtoindex! ambiguity
#578 opened by denizyuret - 1
- 4
Copy pasted example code doesn't work
#572 opened by Cvikli - 2
Feature request: Depthwise and groupwise convolution
#574 opened by ngphuoc - 1
travis does not push new docs
#571 opened by denizyuret - 1
CuArrays.jl: DEPRECATED, use CUDA.jl instead!
#570 opened by denizyuret - 4
GPU tests suggested in the documentation failing
#568 opened by ViralBShah