Paul Van Eck

How large is your dataset and how long are you training for? Depending on the size of the dataset, it can take a long time for the model to start...

Is the image you posted showing the training run for the white background images? Based on that image, it looks like you've reached a decent accuracy, but it seems like...

For the first issue, I think something with API changed that altered the shape of the input image. I will push an update. But seems like all you need to...

Sorry, I was using the `optimized_hangul_tensorflow.pb` file already contained in the Android folder (https://github.com/IBM/tensorflow-hangul-recognition/blob/master/hangul-tensordroid/app/src/main/assets/optimized_hangul_tensorflow.pb) which does work since it was trained using an older version of TF. I did get...

Once https://github.com/kserve/kserve/pull/2296 merges in master, I'll cherry-pick to the KServe `release-0.8` branch, updating the [install manifests ](https://github.com/kserve/kserve/tree/release-0.8/install/v0.8.0)in the process (which is what is used in kubeflow/manifests).

release branch updated with fixes: https://github.com/kserve/kserve/tree/release-0.8

We have decided to go with Knative 1.4 as the desired version of Knative for KF 1.6. I can help with with the initial PR updating the manifests.

Thanks, @kimwnasptd. I think that sounds good, though I wonder if a newer Knative Version would be preferred (e.g. v0.26). @yuzisun Do you have thoughts on the Knative version, or...

> @pvaneck @yuzisun is there going to be an issue with using Knative 1.2 and KServe 0.8? @kimwnasptd I think Knative 1.2 should also be fine.

Ran into this issue when trying to use Kustomize with a Knative serving [YAML](https://github.com/knative/serving/releases/download/v0.17.4/serving-crds.yaml). Seems like the error stems from the usage of `!!merge