提交 · 91a0b7c68e5cfb2821a59f2768287bc15f2a9880 · PaddlePaddle / models

09 4月, 2018 1 次提交
- G
  
  Refine the pad_idx in Transformer · 91a0b7c6
  由 guosheng 提交于 4月 09, 2018
  
  91a0b7c6
08 4月, 2018 7 次提交
- G
  
  Remove pad_idx in Transformer · d2d973d2
  由 guosheng 提交于 4月 08, 2018
  
  d2d973d2
- G
  
  Avoid predicting <pad> by restricting the size of the final fc_layer in Transformer. · 10de2bf3
  由 guosheng 提交于 4月 08, 2018
  
  10de2bf3
- G
  Merge pull request #783 from guoshengCS/fix-transformer-batchsize · f14db82d
  由 Guo Sheng 提交于 4月 08, 2018
```
Decouple the program desc with batch_size in Transformer.
```
  f14db82d
- G
  
  Refine the validation in Transformer. · baa01f6f
  由 guosheng 提交于 4月 08, 2018
  
  baa01f6f
- X
  Ssd coco reader (#767) · 76526e51
  由 Xingyuan Bu 提交于 4月 08, 2018
```
* ready to coco_reader

* complete coco_reader.py & coco_train.py

* complete coco reader

* rename file

* use argparse instead of explicit assignment

* fix

* fix reader bug for some gray image in coco data

* ready to train coco

* fix bug in test()

* fix bug in test()

* change coco dataset to coco2017 dataset

* change dataset from coco to coco2017

* change learning rate

* fix bug in gt label (category id 2 label)

* fix bug in background label

* save model when train finished
```
  76526e51
- G
  
  Append the <pad> token to the dict when converting ids to words · afe55c9e
  由 guosheng 提交于 4月 08, 2018
  
  afe55c9e
- G
  
  Increase the batch size in inference config of Transformer · f82f47ca
  由 guosheng 提交于 4月 08, 2018
  
  f82f47ca
04 4月, 2018 13 次提交
- G
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/models into fix-transformer-batchsize-dev · 86fe83f2
  由 guosheng 提交于 4月 04, 2018
  
  86fe83f2
- G
  Merge pull request #810 from guoshengCS/add-transformer-ppl · a4d00a37
  由 Guo Sheng 提交于 4月 04, 2018
```
Add avg_cost and ppl for Transformer
```
  a4d00a37
- G
  
  Refine the training output in Transformer by following comments · 40c68724
  由 guosheng 提交于 4月 04, 2018
  
  40c68724
- G
  
  Fix the merged conflicts in Transformer inference · 8ca89039
  由 guosheng 提交于 4月 04, 2018
  
  8ca89039
- G
  
  Refine Transformer by following comments · 70f9601b
  由 guosheng 提交于 4月 04, 2018
  
  70f9601b
- G
  
  Refine the training output in Transformer · ef88e63f
  由 guosheng 提交于 4月 04, 2018
  
  ef88e63f
- G
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/models into fix-transformer-batchsize-dev · 2e7494d9
  由 guosheng 提交于 4月 04, 2018
  
  2e7494d9
- G
  
  Delete the test code in Transformer · ad1c3857
  由 guosheng 提交于 4月 04, 2018
  
  ad1c3857
- G
  
  Format the ppl output for Transformer · dc6e3748
  由 guosheng 提交于 4月 04, 2018
  
  dc6e3748
- G
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/models into add-transformer-ppl · ab1b1de9
  由 guosheng 提交于 4月 04, 2018
  
  ab1b1de9
- G
  
  Add avg_cost and ppl for Transformer · a7196c88
  由 guosheng 提交于 4月 04, 2018
  
  a7196c88
- G
  Merge pull request #809 from guoshengCS/refine-transformer-token · c2de925f
  由 Guo Sheng 提交于 4月 04, 2018
```
Refine the inference to output special tokens optionally in Transformer
```
  c2de925f
- W
  Merge pull request #799 from wanghaoshuang/fix_policy_gradient · 288664c1
  由 whs 提交于 4月 04, 2018
```
Adapt usage of reduce_mean to the latest fluid API.
```
  288664c1
03 4月, 2018 5 次提交
- Y
  Merge pull request #808 from kuke/fix_test · fccdf3c9
  由 Yibing Liu 提交于 4月 03, 2018
```
Fix data feed in test & profile
```
  fccdf3c9
- G
  
  Refine the inference to output special tokens optionally in Transformer · 27fd97ce
  由 guosheng 提交于 4月 03, 2018
  
  27fd97ce
- Y
  
  Fix data feed in test & profile · a9a79458
  由 Yibing Liu 提交于 4月 03, 2018
  
  a9a79458
- D
  Merge pull request #805 from daming-lu/develop · 9264e8ce
  由 daminglu 提交于 4月 02, 2018
```
update IP to new node
```
  9264e8ce
- D
  
  update IP to new node · 04225919
  由 daming-lu 提交于 4月 02, 2018
  
  04225919
02 4月, 2018 5 次提交
- G
  Merge pull request #796 from buaawht/my-dev · edee4bf6
  由 gx-wind 提交于 4月 02, 2018
```
Modify gradient method, add README.md and tutorials in advbox
```
  edee4bf6
- G
  Merge pull request #803 from guoshengCS/fix-transformer-dataset · 35a407f5
  由 Guo Sheng 提交于 4月 02, 2018
```
Replace v2.dataset with dataset in Transformer
```
  35a407f5
- G
  
  Replace v2.dataset with dataset in Transformer · 40580c44
  由 guosheng 提交于 4月 02, 2018
  
  40580c44
- G
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/models into fix-transformer-batchsize-dev · b6ab2f4a
  由 guosheng 提交于 4月 02, 2018
  
  b6ab2f4a
- W
  
  Adapt usage of reduce_sum to the latest fluid API. · ab30b9ba
  由 wanghaoshuang 提交于 4月 02, 2018
  
  ab30b9ba
01 4月, 2018 1 次提交
- W
  Merge pull request #789 from wanghaoshuang/refine_ctc · 9b0d8621
  由 whs 提交于 4月 01, 2018
```
Refine OCR CTC model.
```
  9b0d8621
31 3月, 2018 3 次提交
- B
  
  modify README.md · b65b4000
  由 buaawht 提交于 3月 31, 2018
  
  b65b4000
- B
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/models into my-dev · 3da4acc3
  由 buaawht 提交于 3月 31, 2018
  
  3da4acc3
- B
  
  add tutorials and README.md · 090cf436
  由 buaawht 提交于 3月 31, 2018
  
  090cf436
30 3月, 2018 5 次提交
- G
  Merge pull request #795 from guoshengCS/refine-transformer-loss · 958812f7
  由 Guo Sheng 提交于 3月 30, 2018
```
Replace cross_entropy with softmax_with_cross_entropy in Transformer
```
  958812f7
- G
  
  Replace cross_entropy with softmax_with_cross_entropy in Transformer · f853e70c
  由 guosheng 提交于 3月 30, 2018
  
  f853e70c
- Y
  Merge pull request #788 from zhxfl/fix-661 · 9e1459d9
  由 Yibing Liu 提交于 3月 30, 2018
```
data load for inference
```
  9e1459d9
- G
  Merge pull request #794 from guoshengCS/fix-transformer-softmax · 218d199d
  由 Guo Sheng 提交于 3月 30, 2018
```
Fix the softmax in Transformer.
```
  218d199d
- G
  
  Refine Transformer by following comments. · cbd63162
  由 guosheng 提交于 3月 30, 2018
  
  cbd63162

PaddlePaddle / models 大约 1 年 前同步成功

PaddlePaddle / models
大约 1 年前同步成功