Components - Added the "Keras - Train classifier" component #809
Conversation
Ark-kun
force-pushed
the
Added-sample-component
branch
5 times, most recently
from
3cd1f3c
to
4344806
Compare
Ark-kun
force-pushed
the
Added-sample-component
branch
from
4344806
to
8b1c7d3
Compare
Some people are confused.
| metrics=['accuracy']) | ||
|
|
||
| x_train = x_train.astype('float32') | ||
| x_train /= 255 |
There was a problem hiding this comment.
Just curious - why do we need this?
There was a problem hiding this comment.
I should remove this.
This is a bit of a leftover. When I was creating the sample I started from CIFAR dataset classifier training. It uses integer values for pixel colors, so the values are 0-255.
Usually people try to keep the activations around -1 .. 1, so they scale the inputs to that scale.
But this should probably be done on the network side.
|
/lgtm |
|
/approve |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: paveldournov The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
1 similar comment
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: paveldournov The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
Ark-kun
changed the title
Ark-kun
changed the title
Ark-kun
changed the title
|
/test kubeflow-pipeline-e2e-test |
Ark-kun
changed the title
* Added sample component * Replaced human-readable names with pythonic names Some people are confused. * Removed the tag from image name.
* We should be using the issue-label-bot-user GCP SA as that's the cluster where its running.
This change is