- 11 3月, 2020 7 次提交
-
-
由 Wilber 提交于
* add skip_layernorm pass. test=develop
-
由 wawltor 提交于
In the op of gemm, we use the gemm to replace batch gemm, speed up the matmul op
-
由 Huihuang Zheng 提交于
As the title
-
由 Adam 提交于
-
由 lidanqing 提交于
-
由 Zhaolong Xing 提交于
* 1. add embedding eltwise layernorm fuse 2. add embedding eltwise layernorm op 3. refine inplace_add_relu 4. refine fc_eltwise_layernorm test=develop * 1. refine fc test=develop * fix comments test=develop * fix comments test=develop
-
由 Aurelius84 提交于
+ Refine code structure and move code related with control flow `if` into `ifelse_transformer.py` + Merge code of `ast_utils.py` into `utils.py`
-
- 10 3月, 2020 6 次提交
-
-
由 guofei 提交于
As the title.
-
由 Chen Weihang 提交于
-
由 yaoxuefeng 提交于
* fix add grad bug test=develop * update style test=develop
-
由 zhangchunle 提交于
-
由 WangXi 提交于
-
由 Aurelius84 提交于
-
- 09 3月, 2020 9 次提交
-
-
由 Zeng Jinle 提交于
* refine grad maker, test=develop * refactor tracer stage 1, test=develop * merge develop to solve conflict third times, test=develop
-
由 mapingshuo 提交于
* fix API param bug of recompute.backward(), test=develop
-
由 liu zhengxi 提交于
* fix fc padding during fusion, test=develop * fix optim model inference after SaveOptimModel, test=develop
-
由 gongweibao 提交于
-
由 tangwei12 提交于
* fix communicator when breaking under PyReader mode, test=develop * revert some vlog level to 0, test=develop
-
由 mapingshuo 提交于
add lookup_table_dequant_op
-
由 zhaoyuchen2018 提交于
As model fails when enable int8 quant, so disable allocate memory in cpu for small variable.
-
由 Zhaolong Xing 提交于
* change the ci trt from version 5. to 6.0 * paddle-trt dynamic shape support init * conv+bias or conv+bn dynamic shape support test=develop * modity trt engine opconvert test=develop * fix ci error test=develop
-
由 WangXi 提交于
-
- 08 3月, 2020 1 次提交
-
-
由 tangwei12 提交于
-
- 07 3月, 2020 3 次提交
-
-
由 liym27 提交于
-
由 Zhang Ting 提交于
-
由 wangchaochaohu 提交于
* refine the profiler print test=develop
-
- 06 3月, 2020 6 次提交
-
-
由 Michał Gallus 提交于
-
由 liym27 提交于
* support basic tensor.shape. * Support tensor.shape with dependencies.
-
由 Chen Weihang 提交于
* polish detail implement of data loader, test=develop * solve coverage ci problem, test=develop
-
由 wawltor 提交于
In elementwise_min api doc has the error, all the api just use the elementwise_max op, we just correct the doc.
-
由 zhangchunle 提交于
-
由 Huihuang Zheng 提交于
1. copy.deepcopy in NameVisitor should be changed to copy.copy to make hash or set work 2. read_context should be type of gast.Load()/gast.AugLoad(), not gast.Load/gast.AugLoad
-
- 05 3月, 2020 7 次提交
-
-
由 Wilber 提交于
fix concat_mkldnn op when encounter extreame conditions.
-
由 Aurelius84 提交于
-
由 hong 提交于
* reduce default attrs for dynamic graph, test=develop * add some explanations for explicit attr, test=develop * tweak explicit attr comments, test=develop
-
由 Huihuang Zheng 提交于
Add basic support for while in translating dygraph to static 1. Analysis the variable liveness in class NameVisitor 2. Replace while key word using while_loop API
-
由 cc 提交于
* Added an option to use external FP32 model in QAT comparison test
-
由 Chen Weihang 提交于
-
由 Zhaolong Xing 提交于
test=develop
-
- 04 3月, 2020 1 次提交
-
-
由 Aurelius84 提交于
* Modify the way of inserting newly defined func_nodes test=develop
-