- 18 11月, 2021 9 次提交
-
-
由 zmx 提交于
* fix pslib. test=develop * add device to train_from_dataset. test=develop * refine fleet.stop_worker. test=develop * fix ut. test=develop * fix ut. test=develop * fix executor & ut. test=develop * fix executor & ut. test=develop * fix executor & ut. test=develop
-
由 xiayanming 提交于
* fleet support elastic train * fleet support elastic train * support elastic * add unittest * fix unitest bug * fix unittest bug * fix unittest bug * fix unittest coverage * fix unittest coverage * fix unittest coverage * fix unittest coverage * fix unittest coverage * fix elastic bug * fix ci fail * fix ci fail * fix elastic bug * fix elastic bug * fix joint debugging bug * fix joint debugging bug * fix windows ci failed * fix windows ci failed * Optimize fleet elastic scale in/out * elastic support pre hook * add prehook unittest
-
由 zhangbo9674 提交于
-
由 Shang Zhizhou 提交于
-
由 zmx 提交于
-
由 Yuang Liu 提交于
-
由 LielinJiang 提交于
* fix cache * Fix unittest
-
由 LielinJiang 提交于
* Fix the slow running speed of kl_div when option reduction is set * fix unittest coverage
-
由 LielinJiang 提交于
-
- 17 11月, 2021 9 次提交
-
-
由 Leo Chen 提交于
* init * add feed ops in python side * import LRScheduler * update_feed * refine code format
-
由 Sing_chan 提交于
-
由 YUNSHEN XIE 提交于
* remove test_hapi_hub from mac * fix format error
-
由 zhaocaibei123 提交于
-
由 zmx 提交于
* fix. test=develop * fix. test=develop * fix. test=develop * fix. test=develop * fix. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * refactor heter trainer. test=develop * fix. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop * fix. test=develop * fix. test=develop * fix. test=develop * fix. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop
-
由 LiYuRio 提交于
-
由 WangXi 提交于
-
由 Tongxin Bai 提交于
* [Einsum] correct output dimension errors due to single element tensors. * [Einsum] format polish.
-
由 xiongkun 提交于
* add * add BuildOperatorDependences * fix bug * add unittest for write after write * fix merge bug * fix
-
- 16 11月, 2021 6 次提交
-
-
由 arlesniak 提交于
* Added BF16 Pool2d grad * upstream pulled * fix for CI * fixes after review
-
由 Weilong Wu 提交于
-
由 Zeng Jinle 提交于
-
由 Zeng Jinle 提交于
* make pass ut timeout smaller * increate ut timeout
-
由 jakpiase 提交于
-
由 Li Min 提交于
fused_attention_op的实现中,使用了bias_add,且其实现是通过使用kernel primitive来实现的,之后kernel primitive的WriteData api接口及函数内部实现发生了更改,将判断越界的逻辑移到了template的参数中,使得调用的分支有错误,产生了越界赋值操作,污染了别的显存空间的内容。具体表现为:test_fused_attention_op_api.py 单次执行基本上不会报错,多次循环执行不同shape的输入,结果计算不对,具有偶发性,bug不易察觉。
-
- 15 11月, 2021 10 次提交
-
-
由 Chen Weihang 提交于
* move extension into pten [no-verify] * append tensor methods by ext_tensor [no-verify] * append other tensor methods [no-verify] * ext related files tidy [no-verify] * include relation tidy [no-verify] * add pten tensor test [no-verify] * replace tensor in custom op & compile success * refine tensor constructor for unittest * custom relu jit run success * fix all custom op unittests * add inference cmake adapt [no-verify] * fix failed unittests * fix windows failed unittests * try to fix kunlun and inference failed * fix test_elementwise_api error * try to fix win compile failed * fix kunlun fp16 type error * remove useless haddle error macro * add custom linear op test * fix compile failed & add win symbols * fix non pten kernel cast failed * add dll decl for api * polish several deetails * polish details by review comment * add dll_decl for register
-
由 baoachun 提交于
* remove input dim check of activation in op_teller * remove input dim check of concat in op_teller * remove input dim check of clip in op_teller * remove input dim check of scale in op_teller * remove input dim check in op_teller * update attr check of slice in op_teller
-
由 wanghuancoder 提交于
* fix 3 bug, test=develop * refine, test=develop
-
由 arlesniak 提交于
* Added BF16 to mean op * fix for CI * fix for CI * fix for CI
-
由 Weilong Wu 提交于
* Add elementwise_mul triple grad kernel * Removed InplaceInferer and polished code
-
由 Zeng Jinle 提交于
* add split_program * make ut faster * increase ut timeout * make result deterministic * add fuse_all_reduce pass * add ut framework, update * fix ut framework * remove useless code * add coverage support * update * fix CI * fix some bugs and fix ci coverage * fix conflict
-
由 zyfncg 提交于
-
由 jiangcheng 提交于
-
由 Liu-xiandong 提交于
* modify sparse_attention docs, test=develop * add warning * add warning ,test=document_fix
-
由 zmx 提交于
* fix. test=develop * fix. test=develop * fix. test=develop * fix. test=develop * fix. test=develop * fix ut. test=develop * fix ut. test=develop * fix ut. test=develop
-
- 12 11月, 2021 6 次提交
-
-
由 zhangkaihuo 提交于
* fix bug: 1. atten: set the default value of attn_dropout_rate to None 2. ffn: add activation parameter
-
由 Chen Weihang 提交于
-
由 Yuang Liu 提交于
-
由 Leo Chen 提交于
* split declaration and implementation * remove initdevices * refine VariableMetaInfo * add ut * fix compile
-
由 Fan Zhang 提交于
[CPU-PSLIB] Fix bug for consistency insepection of op's embedding name and sparse table name in config_fleet.py (#36753) * [CPU-PSLIB] Fix bug for consistency insepection of op's embedding name and sparse table name in config_fleet.py * [CPU-PSLIB] Fix bug for consistency insepection of op's embedding name and sparse table name in config_fleet.py
-
由 Aganlengzi 提交于
-