Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Refactoring of function RNNForwardTrainingPackedTensors #2750

Draft
wants to merge 28 commits into
base: develop
Choose a base branch
from
Draft

Refactoring of function RNNForwardTrainingPackedTensors #2750

wants to merge 28 commits into from
+1,309 −71

Conversation

kikimych
Copy link
Contributor

Removed gru related code to separate function

kikimych added 25 commits June 14, 2023 18:01
@kikimych
Added RNN_RELU and RNN_TANH forward training refactored functions
13da580
@kikimych
Added RNNForwardTrainingGRU method
c3d8ab4
@kikimych
Merge branch 'develop' of https://github.com/ROCmSoftwarePlatform/MIOpen
263c278 
 into rnn_training_forward_refactor
@kikimych
Added RNN Backward Data refactor
f247d7d
@kikimych
Added bidirection support for RNNBackwardData
7bb3be5
@kikimych
Unified relu and gru offset interfaces
3c889de
@kikimych
Lstm minor refactoring
cf96713
@kikimych
Lstm minor refactoring
c2bfee2
@kikimych
Reverted back RNNForwardTraining_MS method
cdce2f5
@kikimych
Merge branch 'rnn_training_forward_refactor' of https://github.com/ki…
b114b9b 
…kimych/MIOpen into rnn_training_forward_refactor
@kikimych
Rnn relu bidirectional support draft
12e4a43
@kikimych
Refactored Relu forward
d2a166b
@kikimych
RNNBackwardDataPackedTensorsRelu refactor
f0d1d0a 
 Пожалуйста, введите сообщение коммита для ваших изменений. Строки,
@kikimych
Simplified RNNBackwardDataPackedTensorsRelu
e8de935
@kikimych
Relu forward minor refactor
ee524b6
@kikimych
Removed hidden_size from offset helpers
5cc2e09
@kikimych
Enabled RNNForwardTrainingTanhRelu by default
7ca8c32
@kikimych
Merge branch 'develop' into rnn_training_forward_refactor
25dc932
@kikimych
Extracted GRU code from RNNForwardTrainingPackedTensors to RNNForward…
26d7373 
…TrainingGRU
@kikimych
temporary working commit
96883eb
@kikimych
Merge branch 'develop' into gru_refactor_bidirectional
dec2542
@kikimych
working draft
aca2181
@kikimych
Small fixes
6b687de
@kikimych
Added RnnSkip mode support
e39d149
@kikimych
Merge branch 'develop' into gru_refactor_bidirectional
17aa218
@kikimych kikimych marked this pull request as draft February 12, 2024 21:37
@junliume
Copy link
Contributor

@kikimych we are running into CI issues:

[2024-02-13T06:31:06.674Z]   27/2049 Test   #26: test_gru ................................................................................................................................................................................................................................................................................................................................................................................................................***Failed  Error regular expression found in output. Regex=[FAILED]  4.54 sec
[2024-02-13T06:31:06.674Z] /home/jenkins/workspace/MLLIBS_MIOpen_PR-2750/build/bin/test_gru --float --disable-verification-cache 
[2024-02-13T06:31:06.674Z] PRNG seed: 12345678
[2024-02-13T06:31:06.674Z] Empty batch sequence. Filling uniformly with batch size: 17
[2024-02-13T06:31:06.674Z] /home/jenkins/workspace/MLLIBS_MIOpen_PR-2750/build/bin/test_gru --float --batch-size 17 --seq-len 2 --vector-len 13 --hidden-size 67 --num-layers 1 --use-dropout 0 --in-mode 0 --bias-mode 0 --dir-mode 0 --batch-seq 17 15 
[2024-02-13T06:31:06.674Z] FAILED: 0.778508
[2024-02-13T06:31:06.674Z] Iteration: 0
[2024-02-13T06:31:06.674Z] ./bin/MIOpenDriver rnn -n 17,15 -m gru -k 2 -H 67 -W 13 -l 1 -F 0 -r 0 -b 0 -p 0
[2024-02-13T06:31:06.675Z] inputMode: 0 biasMode: 0 dirMode: 0
[2024-02-13T06:31:06.675Z] hz: 67 batch_n: 32 seqLength: 2 inputLen: 13 numLayers: 1 useDropout: 0
[2024-02-13T06:31:06.675Z] Forward Train GRU: 
[2024-02-13T06:31:06.675Z] Output tensor output failed verification.
[2024-02-13T06:31:06.675Z] Max diff: 0.367524
[2024-02-13T06:31:06.675Z] Mismatch at 0: 0.0417352 != -0.173738
[2024-02-13T06:31:06.675Z] /home/jenkins/workspace/MLLIBS_MIOpen_PR-2750/build/bin/test_gru --float --batch-size 17 --seq-len 2 --vector-len 13 --hidden-size 67 --num-layers 1 --use-dropout 0 --in-mode 0 --bias-mode 0 --dir-mode 0 --batch-seq 17 15 
[2024-02-13T06:31:06.675Z] FAILED: 0.880309
[2024-02-13T06:31:06.675Z] Iteration: 0
[2024-02-13T06:31:06.675Z] ./bin/MIOpenDriver rnn -n 17,15 -m gru -k 2 -H 67 -W 13 -l 1 -F 0 -r 0 -b 0 -p 0
[2024-02-13T06:31:06.675Z] inputMode: 0 biasMode: 0 dirMode: 0
[2024-02-13T06:31:06.675Z] hz: 67 batch_n: 32 seqLength: 2 inputLen: 13 numLayers: 1 useDropout: 0
[2024-02-13T06:31:06.675Z] Forward Train GRU: 
[2024-02-13T06:31:06.675Z] Hidden state tensor failed verification.
[2024-02-13T06:31:06.675Z] Max diff: 0.367524
[2024-02-13T06:31:06.675Z] Mismatch at 0: 0.0465669 != -0.283446
[2024-02-13T06:31:06.675Z] /home/jenkins/workspace/MLLIBS_MIOpen_PR-2750/build/bin/test_gru --float --batch-size 17 --seq-len 2 --vector-len 13 --hidden-size 67 --num-layers 1 --use-dropout 0 --in-mode 0 --bias-mode 0 --dir-mode 0 --batch-seq 17 15 
[2024-02-13T06:31:06.675Z] FAILED: 0.44586
[2024-02-13T06:31:06.675Z] Iteration: 0
[2024-02-13T06:31:06.675Z] ./bin/MIOpenDriver rnn -n 17,15 -m gru -k 2 -H 67 -W 13 -l 1 -F 0 -r 0 -b 0 -p 0
[2024-02-13T06:31:06.675Z] inputMode: 0 biasMode: 0 dirMode: 0
[2024-02-13T06:31:06.675Z] hz: 67 batch_n: 32 seqLength: 2 inputLen: 13 numLayers: 1 useDropout: 0
[2024-02-13T06:31:06.675Z] Forward Train GRU: 
[2024-02-13T06:31:06.675Z] Weight tensor failed verification.
[2024-02-13T06:31:06.675Z] Max diff: 0.63235
[2024-02-13T06:31:06.675Z] Mismatch at 0: 0.009554 != -0.362995

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@JehandadKhan JehandadKhan Awaiting requested review from JehandadKhan

@junliume junliume Awaiting requested review from junliume

@BrianHarrisonAMD BrianHarrisonAMD Awaiting requested review from BrianHarrisonAMD BrianHarrisonAMD will be requested when the pull request is marked ready for review BrianHarrisonAMD is a code owner

@BradPepersAMD BradPepersAMD Awaiting requested review from BradPepersAMD BradPepersAMD will be requested when the pull request is marked ready for review BradPepersAMD is a code owner

At least 1 approving review is required to merge this pull request.

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@kikimych @junliume