- 30 6月, 2020 13 次提交
-
-
由 WuHaobo 提交于
-
由 Kaipeng Deng 提交于
-
由 Yiqun Liu 提交于
-
由 Huihuang Zheng 提交于
Support Various-Length Return Grammar in Dy2stat. This PR is a follow-up of https://github.com/PaddlePaddle/Paddle/pull/25176 . The basic idea is putting no-value placeholder variables at `return` statement to make all `return` statement have same length, after that the static graph can have fixed fetch output (code at return_transformer.py). Then remove those no-value placeholder when we finally return dygraph result (code at partial_program.py). However, various length return in Bert model is still not supported. The dy2stat can change the code as I wish but some ops which check shape at compile time (e.g. Reshape, MatMul) will throw error because of the no-value-placeholder may not have the required shape. Is this a matter? To me, those no-value placeholder will be replaced as really values meeting shape requirements at run time, so I think the solution should be some way to do the compile-time checking. By the way, every time when we have dynamic shape, it often causes problem in dy2stat. We should find a way to handle it in the future. Fixing various return in Bert is my TODO thing and I will also find some other existing models for verification.
-
由 Aurelius84 提交于
* fix GPU diff test=develop * refine code test=develop
-
由 Wojciech Uss 提交于
test=develop
-
由 123malin 提交于
-
由 Yi Liu 提交于
test=develop
-
由 hong 提交于
-
由 tangwei12 提交于
-
由 Tao Luo 提交于
* add test_parallel_executor_profiler.py test=develop * set exec_strategy.num_threads=1 * skip check_profile_result * add TODO for test_profiler test=develop
-
由 FDInSky 提交于
-
由 Sylwester Fraczek 提交于
-
- 29 6月, 2020 6 次提交
-
-
由 zhangchunle 提交于
-
由 Wojciech Uss 提交于
test=develop
-
由 Aurelius84 提交于
* support to save varBase created in __init__ test=develop * polish code test=develop * refine to_static_var test=develop * refine warning test=develop * add unitteset for to_static_var test=develop * fix logger test=develop * polish buffers en doc test=develop * fix param_guard test=develop * refine en doc test=develop
-
由 Wilber 提交于
-
由 Wilber 提交于
-
由 Huihuang Zheng 提交于
This PR added basic support for 'return' grammar in dy2stat. It supports the control flow of 'return'. The basics idea is using a return value variable to store the early return statements and boolean state variables with if-else to skip the statements after the return statements. **This PR is very basic support. There are some corner cases I didn't develop/test**. For example, 'return None', 'return different length of variables', 'return non-tensor and tensor together', 'no return statement'. **These corner cases will be done in my next PRs**. Target date is this week. **Note**: 1. for the unit test, I changed test_program_translator.py because the StaticCode of `dyfunc_with_if_else` will change. To guarantee the correctness of `dyfunc_with_if_else`, I also run it in `TestRecursiveReturn` in test_return.py. 2. I commented the early return code in bert_dygraph_model.py because 'return different length of variables' is unsupported now. I also know that there are some other models used early return and we didn't enable it in the unit test. I will add support for it in next PRs and then re-enable those tests.
-
- 28 6月, 2020 2 次提交
-
-
由 liym27 提交于
-
由 Aurelius84 提交于
-
- 26 6月, 2020 1 次提交
-
-
由 liu zhengxi 提交于
* fix beam_search infershape, test=develop * fix beam search op unittest, test=develop
-
- 25 6月, 2020 1 次提交
-
-
由 chalsliu 提交于
-
- 24 6月, 2020 4 次提交
-
-
由 Chen Weihang 提交于
* add default cudnn lib path, test=develop * change default path in func, test=develop * move to linux branch, test=develop * fix var error in other plat, test=develop
-
由 Aurelius84 提交于
* add cycle_gan_model * align train test=develop * modify image_size into 64 to avoid TimeOut test=develop * TODO in GPU test=develop
-
由 Leo Chen 提交于
* refine PADDLE_THROW, test=develop * refine error msg, test=develop * refine cuda error, test=develop * follow comments, test=develop * fix compile problem, test=develop * fix bug, test=develop
-
由 FDInSky 提交于
* test=develop fix conv3d_transpose_test error
-
- 23 6月, 2020 5 次提交
-
-
由 Adam 提交于
* Refactor of conv fp32 oneDNN operator test=develop * Formatting fix test=develop * Return Enforces test=develop * GetWeights improvements test=develop
-
由 Pei Yang 提交于
* Paddle-TensorRT support slim QAT. test=develop * add comments. test=develop * use RenameInput instead of ResetInputs. test=develop
-
由 Tao Luo 提交于
test=develop
-
由 LielinJiang 提交于
-
由 Shibo Tao 提交于
* generate dummy file using cmake configure_file function to avoid re-generating it. test=develop * add cmake/dummy.c.in. test=develop
-
- 22 6月, 2020 3 次提交
- 21 6月, 2020 1 次提交
-
-
由 Shibo Tao 提交于
* don't re-generate header file if content doesn't change. test=develop * add copy_if_different function. test=develop
-
- 19 6月, 2020 4 次提交
-
-
由 iducn 提交于
-
由 T8T9 提交于
-
由 Aurelius84 提交于
* Refine caches of converting func test=develop * fix if statement test=develop * refine cache code test=develop * rm unuse import statement test=develop * Polish code comment test=develop
-
由 tianshuo78520a 提交于
-