Enable TensorFlowModelDataset to overwrite existing model, and add support for tf.device
#1915
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
merelcht
changed the title
deepyaman
requested changes
Oct 12, 2022
* Change a file extention to match the previous article Signed-off-by: dinotuku <kuan.tung@epfl.ch> * Add a missing import Signed-off-by: dinotuku <kuan.tung@epfl.ch> * Change both preprocessed datasets to parquet files Signed-off-by: dinotuku <kuan.tung@epfl.ch> * Change data type to ParquetDataSet for parquet files Signed-off-by: dinotuku <kuan.tung@epfl.ch> * Add a note for installing seaborn if it is not installed Signed-off-by: dinotuku <kuan.tung@epfl.ch> Signed-off-by: dinotuku <kuan.tung@epfl.ch> Signed-off-by: William Caicedo <williamc@movio.co>
Signed-off-by: William Caicedo <williamc@movio.co>
Signed-off-by: William Caicedo <williamc@movio.co>
Signed-off-by: William Caicedo <williamc@movio.co>
Signed-off-by: William Caicedo <williamc@movio.co>
Signed-off-by: William Caicedo <williamc@movio.co>
Co-authored-by: Deepyaman Datta <deepyaman.datta@utexas.edu> Signed-off-by: William Caicedo <williamc@movio.co>
williamcaicedo
force-pushed
the
fix-696-tfmodeldataset
branch
from
321a859 to
8813536
Compare
williamcaicedo
requested review from
deepyaman
and removed request for
idanov and
yetudada
deepyaman
approved these changes
Oct 19, 2022
There was a problem hiding this comment.
I haven't tested this myself yet, but LGTM from a code review perspective!
Edit: @williamcaicedo I added a note to PR title that added support for tf.device; can you actually include that in the release notes, too?
deepyaman
changed the title
TensorFlowModelDataset to override existing model, and add support for tf.device
deepyaman
changed the title
TensorFlowModelDataset to override existing model, and add support for tf.deviceTensorFlowModelDataset to overwrite existing model, and add support for tf.device
noklam
approved these changes
Oct 20, 2022
…icedo/kedro into fix-696-tfmodeldataset
@deepyaman done 👍 |
Amazing, thanks! |
nickolasrm
pushed a commit
to ProjetaAi/kedro
that referenced
this pull request
Oct 26, 2022
…support for `tf.device` (kedro-org#1915) * Fix issue with save operation. Add gpu option Signed-off-by: William Caicedo <williamc@movio.co> * Add tests Signed-off-by: William Caicedo <williamc@movio.co> * Update RELEASE.md Signed-off-by: William Caicedo <williamc@movio.co> * Update test description Signed-off-by: William Caicedo <williamc@movio.co> * Remove double slash and overwrite flag in fsspec.put method invocation Signed-off-by: William Caicedo <williamc@movio.co> * Allow to explicitly set device name Signed-off-by: William Caicedo <williamc@movio.co> * Update RELEASE.md Co-authored-by: Deepyaman Datta <deepyaman.datta@utexas.edu> Signed-off-by: William Caicedo <williamc@movio.co> * Update docs Signed-off-by: William Caicedo <williamc@movio.co> Co-authored-by: Deepyaman Datta <deepyaman.datta@utexas.edu> Signed-off-by: nickolasrm <nickolasrochamachado@gmail.com>
mle-els
pushed a commit
to mle-els/kedro
that referenced
this pull request
Nov 7, 2022
…support for `tf.device` (kedro-org#1915) * Fix issue with save operation. Add gpu option Signed-off-by: William Caicedo <williamc@movio.co> * Add tests Signed-off-by: William Caicedo <williamc@movio.co> * Update RELEASE.md Signed-off-by: William Caicedo <williamc@movio.co> * Update test description Signed-off-by: William Caicedo <williamc@movio.co> * Remove double slash and overwrite flag in fsspec.put method invocation Signed-off-by: William Caicedo <williamc@movio.co> * Allow to explicitly set device name Signed-off-by: William Caicedo <williamc@movio.co> * Update RELEASE.md Co-authored-by: Deepyaman Datta <deepyaman.datta@utexas.edu> Signed-off-by: William Caicedo <williamc@movio.co> * Update docs Signed-off-by: William Caicedo <williamc@movio.co> Co-authored-by: Deepyaman Datta <deepyaman.datta@utexas.edu> Signed-off-by: Minh Le <m.le@elsevier.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
prevents TF models from being saved correctly using an unversioned dataset.
Development notes
For models using the SavedModel format, a check was added to find out if the save path already exists. In case it does, path is deleted prior to saving the new model.
Removed the code that manually created all save path intermediate directories if they didn't exist, as
_fs_args.setdefault("auto_mkdir", True)is already being set.Model uses gpu by default unless
tf_device: cpuload arg is supplied.Checklist
RELEASE.mdfile