4e43ec6995
new usage of amp training. ( #564 )
...
* new usage of amp training.
* change the usage of amp and pure fp16 training.
* modified code as reviews
2021-02-26 09:25:54 +00:00
ba17052a54
add eta info ( #613 )
...
* add eta info
* rm duplicate desc
2021-02-26 13:40:40 +08:00
c8a155635e
fix train and save ( #594 )
2021-02-01 22:24:25 +08:00
aa8e3c1183
Fix the mertirc_list when \'use_mix==True\' ( #529 )
2020-12-30 17:36:56 +08:00
8c82d89469
Fix the calculation method of batch_time and reader_time ( #528 )
2020-12-30 17:07:50 +08:00
c213c9fc5f
Fix the training log in static graph ( #525 )
...
* Adapt to PaddleHub2.0 to eliminate warning
* Fix the training log format
2020-12-30 14:43:00 +08:00
e7dbecd22e
fix predict ( #527 )
...
* fix predict
* fix export model
* fix doc
2020-12-30 14:28:06 +08:00
918e68a934
fix bug for kunlun ( #518 )
2020-12-28 13:43:05 +08:00
11ef05ca32
Update the hubserving ( #517 )
...
* Fix the timing of hubserving
2020-12-28 13:28:06 +08:00
aa65556433
Adapt to PaddleHub2.0 ( #512 )
2020-12-24 18:21:37 +08:00
62fd192784
optimizer fleet distributed strategy ( #500 )
2020-12-18 11:18:58 +08:00
eb945e5d69
Fix "tar" to "pdparams" to adapt to dygraph ( #496 )
...
* Fix "tar" to "pdparams" to adapt to dygraph
* Update the download link of Paddle Inference Library
2020-12-17 23:20:16 +08:00
6ddd8049e0
Add "cpu_num_threads" and "enable_profile" ( #494 )
2020-12-17 15:23:29 +08:00
8fd56a4503
fix static train ( #478 )
2020-12-15 16:43:15 +08:00
29b305d228
add profile for pred ( #476 )
2020-12-15 14:32:07 +08:00
dc3020ab4a
support fp16 training ( #435 )
...
* support fp16 training
* Use compiled training program
* Change timing ips.
* Use dali
* add pure fp16 training
* fix a bug, which will not use fuse pass using pure fp16 training.
* modify code as review
* modify loss, so that it will use different loss when using pure fp16 training.
* remove some fluid API
* add static optimizer.
2020-12-11 11:04:51 +08:00
9992415867
add mkldnn speed problm ( #465 )
2020-12-11 00:39:12 +08:00
b9d243f854
Update utils.py
2020-12-11 00:24:14 +08:00
15b18973f1
fix eval script ( #464 )
...
* fix eval script
* fix dali shell
2020-12-10 23:45:58 +08:00
e3801c5515
Fix the bug about calling create_predictor repeatedly in hubserving ( #462 )
2020-12-10 01:15:24 +08:00
066d53f8ec
support cpu/xpu/gpu in static graph ( #460 )
2020-12-08 20:59:23 +08:00
a76e404d9c
fix time sta ( #457 )
2020-12-08 17:05:00 +08:00
1ecf8334ff
fix pred ( #450 )
...
* fix pred
* fix resnest
* fix hrnet se
* fix se for export
2020-12-06 03:08:44 +08:00
0df5ba63df
Update test_hubserving.py
2020-12-04 10:43:56 +08:00
049cbb26d7
Update run_dali.sh
2020-12-03 15:58:07 +08:00
1e4925704d
fix dali train ( #446 )
2020-12-03 15:48:35 +08:00
e92cb0b93c
fix init model in static mode ( #444 )
2020-12-03 12:50:33 +08:00
fff2a92b25
add shell scripts to run training on single node and multiple nodes. ( #424 )
...
* add shells, test=develop
2020-12-03 11:11:39 +08:00
2b77c71459
Support DALI ( #442 )
2020-12-02 22:06:23 +08:00
832364e191
support static graph train for kunlun ( #441 )
2020-12-02 18:36:51 +08:00
8e5bde6c39
Update predict.py
2020-12-01 01:20:10 +08:00
6796bca110
fix logger ( #426 )
2020-11-25 19:45:43 +08:00
463980f331
add support for multi cards eval ( #413 )
...
* add support for multi cards eva
2020-11-20 18:10:06 +08:00
e83e3038e1
fix local rank get word size in dist ( #402 )
...
* fix local rank
* fix export model
2020-11-18 13:59:34 +08:00
6a5f4626d7
add static running in dygraph ( #399 )
...
* add static running in dygraph
2020-11-18 09:48:56 +08:00
10e73342ff
Add MKL-DNN ( #398 )
...
add mkldnn config
2020-11-16 19:44:17 +08:00
7d0f694c7b
unify model and model_name ( #386 )
...
* unify model and model_name
* rm unsed archs fix bug of se_resnext
* fix distillation model for distributed.launch
* fix export model for googlenet
2020-11-12 00:17:28 +08:00
18a4c14233
Update preprocess and HubServing ( #382 )
...
add hub serving code and doc
2020-11-10 17:19:50 +08:00
c933dcd8db
minor fix init ( #384 )
...
* add densent init
* fix export model
2020-11-10 17:09:39 +08:00
95a8990066
fix py infer ( #376 )
...
fix python and cpp inference
2020-11-07 22:09:42 +08:00
f921b9bd5b
fix feature map visualization ( #377 )
...
fix feature map visualization
2020-11-07 22:07:33 +08:00
ee10207811
fix log benchmark ( #369 )
...
fix log benchmark
2020-11-05 10:35:19 +08:00
d4b1f6aff3
Update export_model.py
2020-11-04 22:49:20 +08:00
c3922941de
fix dist training ( #363 )
...
1. fix dist training
2. fix cpp infer to support dir inference
2020-11-04 22:48:19 +08:00
14063829db
add random seed in train ( #358 )
2020-11-03 10:08:30 +08:00
5704a255dc
fix eval ( #355 )
2020-10-30 21:02:49 +08:00
7ef474169d
polish api to Paddle2.0-rc ( #346 )
...
polish codes and docs to adapt Paddle2.0-rc
2020-10-30 00:20:48 +08:00
260dcb5079
add pre-label function ( #341 )
...
add pre-label function
2020-10-24 01:14:10 +08:00
b5cd360b0d
fix getting started
2020-10-22 18:18:37 +00:00
0e0967545a
fix load ckp
2020-10-22 06:12:03 +00:00
d3ab5bd3da
fix benchamark list
2020-10-17 17:51:54 +00:00
ebadacd7f8
add acc benchmark
2020-10-17 17:27:36 +00:00
2cb86a9bbb
fix export model
2020-10-16 06:25:55 +00:00
2d86d1ac0c
fix msg
2020-10-10 05:35:48 +00:00
a0b125d99a
imporve msg info
2020-10-10 05:33:44 +00:00
e70b3351bf
fix eval
2020-10-09 01:56:05 +00:00
7cf6e9ae93
add dataloader inferface
2020-09-28 16:47:22 +00:00
53c5850df7
Merge pull request #291 from TingquanGao/ppcls-dev
...
fix export_model to support dygraph
2020-09-25 13:15:27 +08:00
5ac48b7764
fix export_model to support dygraph
2020-09-24 13:03:42 +08:00
fa471435a7
fix export_model to support dygraph
2020-09-23 22:51:52 +08:00
c5cf3c154a
fix export_model to support dygraph
2020-09-22 16:49:12 +08:00
be9b45d9e1
add reader time
2020-09-22 07:17:28 +00:00
922f33fef1
fix train
2020-09-16 08:35:37 +00:00
630cb60ae3
fix dropout
2020-09-16 07:26:21 +00:00
5ce653d68c
fix eval
2020-09-16 05:28:09 +00:00
b8a7d186d7
fix optimizer and regularizer
2020-09-15 09:43:19 +00:00
d29700944b
fix eval
2020-09-13 09:57:23 +00:00
b7967b4109
fix infer
2020-09-13 09:26:39 +00:00
edd7382dd3
fix googlenet
2020-09-13 09:21:20 +00:00
7ed80c5232
fix program
2020-09-13 07:09:43 +00:00
d609a81df9
fix train
2020-09-13 06:49:47 +00:00
a52efec34e
fix train.py
2020-09-13 06:18:20 +00:00
670eaf3105
fix py infer
2020-09-07 02:44:02 +00:00
a958e356de
fix benchmark
2020-09-07 02:33:04 +00:00
29e0bd0115
fix infer for dir
2020-09-07 02:12:39 +00:00
8a80a7da75
remove python path config and support cpu train/val/infer
2020-09-04 09:02:50 +00:00
a43aac3250
fix optimizer
2020-09-03 08:39:14 +00:00
7a4b2b1f85
fix layer helper
2020-09-02 16:15:13 +00:00
32dc1c1c0d
improve dygraph model
2020-09-02 15:15:49 +00:00
994d198eb6
add reader
2020-09-01 06:57:24 +00:00
48afe86f68
fix dygpaph training speed
2020-08-29 09:44:30 +00:00
5a15c16581
add dygrapgh load and static load
2020-08-28 09:43:27 +00:00
c384773f00
fix single card dygraph train process
2020-07-23 14:23:28 +00:00
b06bd21e94
modify eval and vgg
2020-07-20 06:34:51 +00:00
a68d90c53b
fix interpolation
2020-07-20 04:16:02 +00:00
5c20b55ead
modify position
2020-07-20 03:28:10 +00:00
0644c970a6
add feature maps visualization
2020-07-04 05:19:09 +00:00
f9f8d03297
add feature maps visualization
2020-07-04 05:08:48 +00:00
ac321194a5
fix fetchs
2020-06-29 07:19:59 +00:00
7713736e69
fix bs
2020-06-29 06:53:54 +00:00
4338730eb4
add ls training for dygraph
2020-06-29 06:23:13 +00:00
6a0dce8aff
support label smooth for dyg
2020-06-29 06:09:09 +00:00
4691e9e4e3
add correct infer
2020-06-15 11:22:56 +08:00
5588734626
Merge branch 'dygraph' into dynamic
2020-06-09 08:19:20 +08:00
b1a6f2d193
fix effnet
2020-06-08 10:22:32 +00:00
0cae84a111
fix effnet
2020-06-08 07:25:53 +00:00
798fe1aa61
add dynamic train
2020-06-08 12:26:35 +08:00
bb622b24ed
fix bugs
2020-06-04 04:07:38 +00:00
04e9720b6d
Merge branch 'master' of https://github.com/PaddlePaddle/PaddleClas into add_efficientnet
2020-06-03 13:56:56 +00:00
dea8eabb45
fix
2020-06-03 13:27:12 +00:00
huangxu96
littletomatodonkey
Tingquan Gao
QingshuChen
WangXi
QingshuChen
lilong12
littletomatodonkey
shippingwang
wqz960
dyning
WuHaobo
WuHaobo
root