Philippe Tillet
|
7bfbb89612
|
[python] now packaging include and libtriton in triton._C submodule
|
2019-09-05 15:37:00 -04:00 |
|
Philippe Tillet
|
9ab2880fba
|
[python][examples] cleaned up dot example
|
2019-09-05 12:54:35 -04:00 |
|
Philippe Tillet
|
2d6c8311e8
|
[python] upgraded pybind11 ; forcing torch tensors to be contiguous()
|
2019-09-05 12:30:51 -04:00 |
|
Philippe Tillet
|
58544d0523
|
[python] renamed src/tensorflow.cc -> src/bindings.cc
|
2019-09-05 09:39:58 -04:00 |
|
Philippe Tillet
|
b2629da1fe
|
[python] more cleaning of frameworks logic
|
2019-09-05 02:21:07 -04:00 |
|
Philippe Tillet
|
44896ee777
|
[pytorch] clean-up of dynamic framework load
|
2019-09-05 02:16:27 -04:00 |
|
Philippe Tillet
|
65133cdf33
|
[python] basic support for pytorch seems to be working
|
2019-09-05 01:32:21 -04:00 |
|
Philippe Tillet
|
ed0f706005
|
[python] fixed various issues in pytorch supoport
|
2019-09-05 00:19:42 -04:00 |
|
Philippe Tillet
|
945b5d0de9
|
[python] modularized triton package
|
2019-09-04 21:55:47 -04:00 |
|
Philippe Tillet
|
f6e9c24fe8
|
[python] more progress towards tensorflow/pytorch unification
|
2019-09-04 19:45:50 -04:00 |
|
Philippe Tillet
|
cdbc9d4ecd
|
[python] more generic gradient registration
|
2019-09-04 03:12:23 -04:00 |
|
Philippe Tillet
|
b747959a57
|
trying to work around tensorflow limitations
|
2019-09-04 01:54:43 -04:00 |
|
Philippe Tillet
|
2ccc915011
|
[python][examples] added template for blocksparse
|
2019-09-03 20:44:27 -04:00 |
|
Philippe Tillet
|
5e03f0a065
|
[codegen][align] reverted some changes
|
2019-09-03 15:28:07 -04:00 |
|
Philippe Tillet
|
97fdb5b6be
|
[tests] added missing files
|
2019-09-03 12:44:35 -04:00 |
|
Philippe Tillet
|
a842d337c5
|
[general] various cleaning and bugfix:
* added copy1d and copy2d benchmark
* fixed issue in reassociation pass
|
2019-09-02 23:00:49 -04:00 |
|
Philippe Tillet
|
90d80c3b2e
|
[codegen][selection] bugfix in scanline dot lowering
|
2019-09-01 16:30:53 -04:00 |
|
Philippe Tillet
|
2d4ddab4d0
|
[ir][print] improved pretty-printing of constants and instructions
|
2019-08-30 18:02:33 -07:00 |
|
Philippe Tillet
|
5db3a7adfe
|
[python][examples] some more cleaning of dot product example
|
2019-08-30 17:05:03 -07:00 |
|
Philippe Tillet
|
7e0af2118c
|
[codegen] worked around bug seemingly from nvptx/ptxas by simplifying multiplications by 1:
- Generated LLVM-IR looked correct
- Illegal addressing disappeared when running cuda-memcheck
- Illegal addressing disappeared when using nvptx-short-pointer
|
2019-08-30 16:45:14 -07:00 |
|
Philippe Tillet
|
141a823799
|
[python] refactoring in anticipation of pytorch support
|
2019-08-29 18:08:51 -07:00 |
|
Philippe Tillet
|
e3c953e79f
|
[test] added more re-usable code in common/util.h
|
2019-08-28 18:06:36 -07:00 |
|
Philippe Tillet
|
d457482539
|
[codegen] fixed issue in double buffering pointer update
|
2019-08-28 17:50:45 -07:00 |
|
Philippe Tillet
|
59281f5794
|
[structure] better directory structure for tests
|
2019-08-27 20:33:38 -07:00 |
|
Philippe Tillet
|
37cbcfabd0
|
[examples] back to 96 TFLOPS on V100
|
2019-08-26 22:49:14 -07:00 |
|
Philippe Tillet
|
b4ae06a714
|
tracking down performance regression
|
2019-08-26 20:38:39 -07:00 |
|
Philippe Tillet
|
7cb73f66e2
|
testing some register gradient
|
2019-08-26 19:25:58 -07:00 |
|
Philippe Tillet
|
9ece3eccc6
|
some cleaning
|
2019-08-26 17:28:24 -07:00 |
|
Philippe Tillet
|
4075949f80
|
[python] basic tensorflow wrapper working
|
2019-08-26 16:53:49 -07:00 |
|
Philippe Tillet
|
0e0399f866
|
more tests
|
2019-08-26 11:00:00 -07:00 |
|
Philippe Tillet
|
321d268a4a
|
more progress
|
2019-08-25 21:26:09 -07:00 |
|
Philippe Tillet
|
96b4d5e411
|
[examples] multiple transposition schemes now supported
|
2019-08-24 13:08:38 -07:00 |
|
Philippe Tillet
|
0b1c389894
|
[lang] changed array declarations from [{}] to []
|
2019-08-23 20:34:24 -07:00 |
|
Philippe Tillet
|
44eb3891ae
|
[lang] added support for restrict; added macros for attributes
|
2019-08-23 20:29:12 -07:00 |
|
Philippe Tillet
|
8c6bac49d1
|
[lang][codegen] added basic attribute support
|
2019-08-23 19:49:06 -07:00 |
|
Philippe Tillet
|
cb04ec0b3b
|
some more cleaning
|
2019-08-23 19:22:38 -07:00 |
|
Philippe Tillet
|
732156b942
|
[general] rename *.cpp -> *.cc
|
2019-08-23 19:06:39 -07:00 |
|
Philippe Tillet
|
6158d96ff7
|
[general] cleaned include guards and added #pragma once
|
2019-08-23 18:08:05 -07:00 |
|
Philippe Tillet
|
606e799948
|
[LICENSING] updated license to incorporate credit for wgtcc
|
2019-08-23 17:56:30 -07:00 |
|
Philippe Tillet
|
a110a7e8cf
|
[ir] changed type of tile shapes from constant_int* to int
|
2019-08-23 17:49:21 -07:00 |
|
Philippe Tillet
|
c9371c7234
|
[general] error messages no longer depend on a program name
|
2019-08-23 17:32:05 -07:00 |
|
Philippe Tillet
|
f98b0b8e2a
|
[general] deleted the old compiler frontend
|
2019-08-23 17:28:02 -07:00 |
|
Philippe Tillet
|
8798d240dc
|
matmul test passes
|
2019-08-23 17:13:30 -07:00 |
|
Philippe Tillet
|
64a6910644
|
[lang][parser] better support for attributes
|
2019-08-22 21:02:38 -07:00 |
|
Philippe Tillet
|
845c0e5b93
|
adding tunable parameters
|
2019-08-22 19:21:01 -07:00 |
|
Philippe Tillet
|
87072203c1
|
[codegen] triton-ir code generation does not crash
|
2019-08-22 17:27:10 -07:00 |
|
Philippe Tillet
|
a6ec807223
|
more debugging
|
2019-08-21 21:53:41 -07:00 |
|
Philippe Tillet
|
a23225ad37
|
more progress
|
2019-08-21 18:27:02 -07:00 |
|
Philippe Tillet
|
5224bbbe06
|
preparing codegen
|
2019-08-20 18:06:30 -07:00 |
|
Philippe Tillet
|
61f25f90eb
|
basic parsing doesn't throw error
|
2019-08-20 16:22:43 -07:00 |
|