- 10 1月, 2021 1 次提交
-
-
由 wangchaochaohu 提交于
reduce the occupied size of memory for the fused pattern of elementwise_add Op and activation Op(relu Op for example) (#29885)
-
- 09 1月, 2021 2 次提交
-
-
由 zhang wenhui 提交于
-
由 Jacek Czaja 提交于
* - Added UT for testing elementwise_mul caching * lint fixes
-
- 08 1月, 2021 6 次提交
-
-
由 Zhen Wang 提交于
* add cast ops before and after unsupported fp16 ops. * Keep partial net in FP32 pattern. * Support check_finite_and_unscale and update_loss_scaling for FP16 calculation mode. * Add fp16 support for adam op. * add multi precision attr for adam. * Fix the bug of test_multi_precision_fp16_train UT. * Code format for CI. * Fix the redefine error about MPTypeTrait on windows. * fix bugs of the _create_accumulators func in Momentum. * fix bug when inserting post cast op. * Add the update_loss_scaling op in allow_set of UnusedVarCheck. * Update for ci coverage. * Add some doc for OptimizerWithMixedPrecision. * Fix the code style. * Imporve the doc of `amp_init`. * Change for fp16 testing if users have the infer program defined in separate way.
-
由 Leo Chen 提交于
-
由 Leo Chen 提交于
* fix dtype of ungenerated grad var * update ut * refine code * set default dtype * fix could_use_cudnn bug * remove debug code * re-implement * fix bug
-
由 Wilber 提交于
-
由 ruri 提交于
-
由 Chengmo 提交于
* add tensor table
-
- 07 1月, 2021 7 次提交
-
-
由 lijianshe02 提交于
* enhance error message of nll_loss op test=develop
-
由 Huihuang Zheng 提交于
Improve some error messages in parallel_executor.cc, conditional_block_op.cc, recurrent_op.cc
-
由 123malin 提交于
* test=develop, add index_select_cuda kernel
-
由 wangchaochaohu 提交于
-
由 Wilber 提交于
* enhance error info for py_func * update
-
由 liuyuhui 提交于
-
由 Jack Zhou 提交于
-
- 06 1月, 2021 6 次提交
-
-
由 ShenLiang 提交于
-
由 wangguanzhong 提交于
-
由 QingshuChen 提交于
* add aarch64 and sunway kunlun lib * minor * optimize elementwise_add for kunlun * update kunlun dependence * minor * minor
-
由 石晓伟 提交于
-
由 xiemoyuan 提交于
-
由 liym27 提交于
1. when slice_item is a slice: 1) the start of __getitem__ should be std::max(start, 0) if slice 2) the start of __getitem__ should be std::min(end, dim) 2. when slice_item is an integer, it should be in [-dim_len, dim_len) 3. Fix error message to use accurate data
-
- 05 1月, 2021 4 次提交
-
-
由 chentianyu03 提交于
-
由 chentianyu03 提交于
* add trace op_register_version and fix defaulf bug; test=op_version * add trace op_register_version; test=op_version * add trace op_register_version; test=op_version * add trace op_register_version; test=op_version * fix missing the template bug of vector; test=op_version
-
由 cc 提交于
* fix the formate of raising error in randperm op
-
由 wangchaochaohu 提交于
-
- 04 1月, 2021 7 次提交
-
-
由 WangXi 提交于
-
由 myq406450149 提交于
* p_norm fix op version info. test=develop
-
由 tangwei12 提交于
* for inference checkpoint Change-Id: I36c979240ffa55bf1ef0c9315402960762af6be4 * for inference checkpoint Change-Id: I82025365d5b792cbea1ead506df685aecc8ac198
-
由 whs 提交于
-
由 ceci3 提交于
* register instance norm, test=op_version
-
由 channings 提交于
Co-authored-by: Nzhoushunjie <zhoushunjie@baidu.com>
-
由 GaoWei8 提交于
-
- 31 12月, 2020 7 次提交
-
-
由 yinhaofeng 提交于
* Register op version for linspace,test=op_version * Register op version for linspace,test=op_version * Register op version for linspace,test=op_version * Register op version for linspace,test=op_version * Register op version for linspace,test=op_version
-
由 123malin 提交于
* test=develop, add op_register_version for roll_op
-
由 chentianyu03 提交于
* dot op support complex types * matmul support complex types * add test case * matmul broadcast gradient support complex * move conjFunctor to complex_functor.h
-
由 ShenLiang 提交于
* fix rank_attention, test=op_version
-
由 Adam Osewski 提交于
* Add operator checkpoints for new attributes. * Fix adding subsequent checkpoint to quantize op.
-
由 wangguanzhong 提交于
-
由 cc 提交于
* Add mkldnn nearest_interp and bilinear_interp op * don't run mkldnn interpolate in default * add interpolate_mkldnn_pass
-