Elton Chen-Yu Ho
a5f26eeef7
Set GPU index based on local_rank instead of rank
2021-08-22 17:45:54 +08:00
Szymon Migacz
ef98b2cef9
[TXL/PyT] update: ( #989 )
...
* changed API calls to torch.einsum
* added export OMP_NUM_THREADS=1 to all launcher scripts
* additional runtime checks to ensure that launch configuration is valid
2021-08-20 08:39:12 -07:00
Mikolaj Blaz
706ef498c9
[Jasper/PyT] Clean up inference flags
2021-08-13 19:25:20 +00:00
Mikolaj Blaz
568e246408
[Jasper/PyT] Add missing package requirements
2021-08-13 19:21:56 +00:00
Jan Baczek
233287038c
[Transformer/PyT] Removing obsolete profiling code
2021-08-13 16:36:58 +00:00
gkarch
793027c9da
removed triton stuff from model dockerfile and moved it to ./triton folder
2021-08-13 16:33:09 +00:00
Michal
ee3f3db4c0
[nnUNet/PyT] update dlprof
2021-08-13 16:31:23 +00:00
Andrei Shumak
6a642837c4
Revert changes to FastPitch and BERT on PyTorch
2021-08-13 16:19:34 +02:00
Krzysztof Kudrynski
a860701665
[BART/PyT] Initial release
2021-08-11 17:08:16 +00:00
Lukasz Pierscieniewski
69a26f5c76
[Multiple models] Added protocol specification to requirements.txt
2021-08-11 16:54:02 +00:00
Andrei Shumak
9c5766102f
[PyT] BERT: new CI configurations
2021-08-06 15:18:29 +00:00
Andrei Shumak
9a6ccc9e23
Add MaskRCNN PyT new CI config support + tests
...
Update tools + regenerate all tests
Add one more configuration file
Update tools
Update configuration + update tools
check: remove `ntasks=1` for 8 gpu test
add --gpus-per-node=8
Upadte tools (task_per_gpu parameter added)
Update tools
2021-08-06 16:12:45 +02:00
Andrei Shumak
9ea704497c
[PyT] FastPitch: new CI configurations
2021-08-06 14:08:33 +00:00
Andrzej Sulecki
ab6267ad33
[internal/CI][PyT/ConvNets] Move to launcher
2021-07-26 21:18:30 +00:00
nv-kkudrynski
d788e8d496
Merge pull request #970 from nv-kkudrynski/kkudrynski/readme_notice
...
Adding links to performance benchmark page
2021-07-21 14:46:11 +02:00
Krzysztof Kudrynski
49e23b4597
Adding links to performance benchmark page
2021-07-21 14:39:48 +02:00
Yuanzhe Dong
3d8d878489
[ConvNets/PyT] update triton repo url in readmes
2021-07-21 12:53:10 +02:00
Joaquin Anton
41e4a07a26
[Jasper/PyT] Update DALI Jasper pipeline to functional API
...
Signed-off-by: Joaquin Anton <janton@nvidia.com>
2021-07-21 12:47:05 +02:00
Michal
ce7926fc80
[nnUNet/PyT] fix pyprof import
2021-07-19 20:02:36 +02:00
gkarch
5be8a3cae2
fix for github issue 965
2021-07-15 12:42:26 +00:00
Lukasz Pierscieniewski
33110132cc
[Convnets/TF] Basic CPU model support
2021-07-15 08:22:43 +00:00
Andrzej Sulecki
a1bbe6687e
[PyT/ConvNets] Fixing bug in model2onnx
2021-07-12 10:20:43 +00:00
Piotr Marcinkiewicz
96b70f5d16
[ResNet50/TF1][Triton] Bermuda 0.6.11 update
2021-07-12 10:19:44 +00:00
Andrzej Sulecki
c481324031
[PyT/EfficientNet] Update README
2021-07-01 15:11:18 +02:00
Mikolaj Blaz
0d4dd6b523
[Jasper/PyT] Minor update in metrics and CLI params
2021-06-30 17:38:22 +00:00
Lukasz Pierscieniewski
36c2d7e8b8
[ResNet50/MX] Fixing perf bug with latest DALI introduced with HW decoder on Ampere
2021-06-30 11:52:43 +00:00
Jan Baczek
7e8102d0ec
[Transformer/PyT] Update to 21.05
2021-06-30 11:50:08 +00:00
kkudrynski
81ee705868
[PyTorchHub] Adding licences to entrypoint scripts
2021-06-18 14:56:07 +00:00
byshiue
39c349f81a
Merge pull request #958 from jdemouth/patch-1
...
Update README.md
2021-06-17 08:27:45 +08:00
Julien Demouth
a079758fb9
Update README.md
2021-06-16 22:34:05 +02:00
Piotr Marcinkiewicz
5c33a8289b
[ResNet50/PyT] Triton perf fix
...
* ResNet50/PyT Triton ONNXruntime fix with env flag
Scripts were modified to fix missing ORT_TENSORRT_FP16_ENABLE flag for
Triton Inference Server with ONNXRuntime and TensorRT execution provider.
* ResNet50/PyT TensorRT FP16 support fixed
ONNX to TensorRT converter was fixed to force FP16 precision for
TensorRT networks.
2021-06-16 16:04:22 +02:00
nv-kkudrynski
e22cfdd617
Merge pull request #957 from alancucki/jasper/fix_pyprof
...
[Jasper/PyT] Fix pyprof import
2021-06-14 16:56:17 +02:00
Adrian Lancucki
f41d86db2a
[Jasper/PyT] Fix pyprof import
2021-06-14 16:53:47 +02:00
Alex Olmucci Poddubnyy
15af494a8e
[DLRM/PyT] Container update to 21.04
2021-06-10 03:56:12 -07:00
Lukasz Pierscieniewski
70e0d55e80
Add missing QAT training file
2021-06-10 10:20:15 +00:00
Jan Golda
778583481b
[TorchHub] restructured hubconf and updated SSD and Tacotron2/WaveGlow entrypoints
2021-06-07 12:20:51 +00:00
Szymon Migacz
ff6f7c6532
[GNMT/PyT] Update for PyT GNMT ( #944 )
2021-05-27 09:48:37 -07:00
Andrzej Sulecki
555b84b3b1
[ConvNets/PyT] Adding checkpoints for EfficientNet/PyT, Squeeze&Excitation can use Conv or Linear layer depending on --trt switch.
2021-05-26 11:03:13 +00:00
Adrian Lancucki
8d8c524df6
[FastPitch/PyT] Fix updated regulate_len
2021-05-21 11:35:47 +02:00
Grzegorz Karch
5a8521ee05
[Tacotron2/PyT] rename common->tacotron2_common
2021-05-18 16:28:11 +00:00
Jan Golda
5afd63dde0
[MaskRCNN/TF2] Update
2021-05-17 09:22:33 +02:00
Michal Futrega
905e9e507e
[nnUnet/PyT] Add support for Triton
2021-05-14 15:41:09 +02:00
nv-kkudrynski
2a2735fed1
Merge pull request #935 from alancucki/readme-update
...
[FastPitch/PyT] Update model description
2021-05-14 14:09:58 +02:00
Adrian Lancucki
0c04f3caaf
Update model description
2021-05-14 14:05:32 +02:00
nv-kkudrynski
97942685fb
Merge pull request #934 from PiotrBigajNV/master
...
[Wide&Deep/TF2] Add Dockerfile symlink to W&D
2021-05-10 12:01:17 +02:00
Piotr Bigaj
ceaa6c63bd
Add Dockerfile symlink to W&D
2021-05-10 11:39:14 +02:00
hXl3s
2d555548b6
[ResNet/TF] Fix gradient calculation for sync variable
2021-04-26 15:00:10 +02:00
nv-kkudrynski
01201316f8
Merge pull request #924 from NVIDIA/gh/release
...
[MaskRCNN/TF2] Readme fix
2021-04-24 17:55:58 +02:00
kkudrynski
e438e68236
[MaskRCNN/TF2] Readme fix
2021-04-24 17:54:33 +02:00
Subhankar Ghosh
ec2bef8444
[EfficientNet/TF2] Fix inference numbers and calculations
...
Co-authored-by: Subhankar Ghosh <subhankarg@nvidia.com>
2021-04-22 13:52:13 +02:00
