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

Update for sample repo restructuring #1828

Merged
merged 1 commit into from
Aug 15, 2019
Merged

Update for sample repo restructuring #1828

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 8 additions & 10 deletions samples/contrib/image-captioning-gcp/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,15 +24,13 @@ In order to run this pipeline, make sure to upload the notebook to your notebook
### Create base image
In order to run this pipeline, you need to first build the docker base image and upload it to a container registry. This can be done locally with the following commands:

`git clone https://github.com/kubeflow/pipelines.git`

`cd pipelines/samples/notebooks/image-captioning-gcp/src`

`docker build -t img-cap .`

`docker tag img-cap gcr.io/[PROJECT-ID]/img-cap:latest`

`docker push gcr.io/[PROJECT ID]/img-cap:latest`
```
git clone https://github.com/kubeflow/pipelines.git
cd pipelines/samples/contrib/image-captioning-gcp/src
docker build -t img-cap .
docker tag img-cap gcr.io/[PROJECT-ID]/img-cap:latest
docker push gcr.io/[PROJECT ID]/img-cap:latest
```

### Download dataset
To download the dataset, run the first few cells in the notebook.
Expand All @@ -50,4 +48,4 @@ Below are some screenshots of the final pipeline and the model outputs.
3. Use `file_io.FileIO()` instead of `open()` when loading files from GCS.
4. To pass multiple outputs downstream, pass them as a tuple of strings. Kubeflow converts this tuple to a string when you pass it downstream. So, you have to convert it from a string back to a tuple in the downstream component to get the multiple outputs.
5. To pass many numpy arrays to downstream components, first save them on GCS. Put the paths to the saved numpy files in a new array, and then save that array on GCS as well. Pass the path to this array to the downstream components.
6. Use `tf.io.read_file` and then `tf.image.decode_jpeg` instead of `PIL.Image` to be compatible with GCS
6. Use `tf.io.read_file` and then `tf.image.decode_jpeg` instead of `PIL.Image` to be compatible with GCS.