0d7d29fa57
Enable caching allocator for CUDA pinned memory ( #275 )
...
Also add binding for CUDA "sleep" kernel
2016-12-02 01:33:56 -05:00
1f5951693a
Change torch.randperm to return Long tensors
2016-12-01 23:14:41 +01:00
3928f7740a
Implement functional interface for Variables (torch.*)
2016-11-08 16:13:25 -05:00
ebc70f7919
Look for libcudart in default CUDA installation paths ( #195 )
2016-11-02 19:36:10 -04:00
f2d7e94948
Use torch.Size for Tensor sizes and tuple for strides
...
See issue #20
The torch.Size class is a tuple subclass which distinguishes sizes from
other tuples so that torch.Tensor(size) is interpreted as size instead
of data.
2016-10-28 19:37:09 +02:00
ad2d413c0b
Add C++ bindings for cuDNN ( #167 )
...
The Python ctypes bindings overhead was high enough that it slowed down
multi-gpu training when using 4+ Maxwell GPUs.
2016-10-26 19:51:48 -04:00
9000f40e61
Add torch.from_numpy
2016-10-24 22:30:11 +02:00
f137c0c05a
Improve error messages of stateless functions
2016-10-24 22:29:43 +02:00
79ead42ade
Add CUDA Stream and Event API ( #133 )
2016-10-18 12:15:57 -04:00
3931beee81
Use THSetNumThreads instead of omp_set_num_threads
...
Set OMP num threads to one in the data loader.
Fixes #81
Fixes #82
2016-10-17 15:15:00 -04:00
ee14cf9438
Add support for pinned memory: ( #127 )
...
torch.Storage/Tensor.pin_memory()
torch.Storage/Tensor.is_pinned()
2016-10-15 18:38:26 -04:00
3d6ebde756
qr and ormqr tests and bugfix
2016-10-14 03:10:16 -04:00
0325e2f646
Major autograd refactor
...
Improves autograd performance by more than 2x and fixes a couple
of bugs. All core functions have been moved to C.
2016-10-13 17:17:49 -07:00
2acee24332
Add keyword argument support to most tensor functions
2016-10-13 12:32:04 -04:00
96f61bff30
Add LAPACK functions
2016-10-08 20:37:37 -07:00
dbe540e49f
Use the custom TH error handler in all threads by default
2016-09-30 14:59:50 -07:00
3f7ab95890
Finish implementation of prng related functions
2016-09-29 11:33:25 -07:00
941cf4e63d
Add ffi utils for user C extensions
2016-09-29 09:35:56 -07:00
1828e7c42f
Add async CUDA copy
2016-09-27 15:12:48 -07:00
ddf1598ef8
Add a method for catching exceptions thrown in ctypes
2016-09-25 12:25:54 -07:00
e71204b52f
Improve error messages in storage and tensor C functions
2016-09-23 17:17:35 -07:00
06ab3f962f
Refactor _C extension to export some utilities
2016-09-21 08:36:54 -07:00
8fdec15a55
Codemod to remove camel case method naming
2016-09-20 08:40:28 -07:00
1f2695e875
adding cuda driver check functions for runtime checking
2016-09-13 10:34:13 -07:00
58f507f9e3
Add file descriptor sharing mode to multiprocessing
2016-09-08 11:23:33 -07:00
f9d186d33a
Add initial version of multiprocessing module
2016-08-31 19:46:08 -07:00
1902bc0bfb
Interface with numpy
2016-08-13 20:19:17 -07:00
12bed8dc0d
Add CUDA device selection
2016-08-12 07:46:46 -07:00
e9f9fd3727
Major refactor
2016-08-10 09:24:53 -07:00
554a1d8336
Add optim
2016-07-21 16:42:06 -04:00
bc7bd7a8b3
Add unit tests and fix detected bugs
2016-07-21 13:46:59 -04:00
c574295012
Various fixes
2016-07-19 10:45:59 -04:00
3a44259b32
Add support for CUDA
2016-07-19 10:45:59 -04:00
93ed433de3
Add rand and randn
2016-07-18 23:59:27 -04:00
3cec305524
Restructure python code
2016-06-23 22:55:05 +02:00
486ea76b98
Add more Tensor methods
2016-06-19 00:24:18 +02:00
4f66ea42af
Add random-related Tensor methods
2016-06-18 21:36:10 +02:00
857c32bc21
Add all mm methods
2016-06-16 23:40:35 +02:00
0eb2b9e756
Add more Tensor and Storage methods
2016-06-15 23:03:47 +02:00
fdfe9d836e
Add index* Tensor methods
2016-06-13 13:58:09 +02:00
a9282edf79
Add THPPointer and more Tensor methods
2016-06-13 13:26:00 +02:00
5ee3358a92
python 2 support
2016-06-08 19:14:57 -04:00
0b61c3f233
Add more Tensor methods
2016-05-13 22:38:51 +02:00
56c98f7897
Add more Tensor methods
2016-05-13 00:01:54 +02:00
c3f7aac4f9
Add logical functions
2016-05-12 01:22:51 +02:00
449ac4ca2a
Add torch.* functions
2016-05-09 19:14:40 +02:00
842e1b6358
Add exception handling
2016-05-05 20:58:13 +02:00
