-
Notifications
You must be signed in to change notification settings - Fork 4.2k
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.
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
Add README for image classification example #21758
Changes from 2 commits
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change | ||||
---|---|---|---|---|---|---|
@@ -0,0 +1,111 @@ | ||||||
<!-- | ||||||
Licensed to the Apache Software Foundation (ASF) under one | ||||||
or more contributor license agreements. See the NOTICE file | ||||||
distributed with this work for additional information | ||||||
regarding copyright ownership. The ASF licenses this file | ||||||
to you under the Apache License, Version 2.0 (the | ||||||
"License"); you may not use this file except in compliance | ||||||
with the License. You may obtain a copy of the License at | ||||||
|
||||||
http://www.apache.org/licenses/LICENSE-2.0 | ||||||
|
||||||
Unless required by applicable law or agreed to in writing, | ||||||
software distributed under the License is distributed on an | ||||||
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY | ||||||
KIND, either express or implied. See the License for the | ||||||
specific language governing permissions and limitations | ||||||
under the License. | ||||||
--> | ||||||
|
||||||
# Example RunInference API Pipelines | ||||||
|
||||||
This module contains example pipelines that use the Beam RunInference | ||||||
API. <!---TODO: Add link to full documentation on Beam website when it's published.--> | ||||||
|
||||||
## Pre-requisites | ||||||
|
||||||
You must have `apache-beam>=2.40.0` installed in order to run these pipelines, | ||||||
because the `apache_beam.examples.inference` module was added in that release. | ||||||
``` | ||||||
pip install apache-beam==2.40.0 | ||||||
``` | ||||||
|
||||||
### Pytorch dependencies | ||||||
The RunInference API has support for the Pytorch framework. To use Pytorch locally, first install `torch`. | ||||||
``` | ||||||
pip install torch==1.11.0 | ||||||
``` | ||||||
|
||||||
For installation of the `torch` dependency for Dataflow pipelines, refer to these | ||||||
[instructions](https://beam.apache.org/documentation/sdks/python-pipeline-dependencies/#pypi-dependencies). | ||||||
|
||||||
<!--- | ||||||
TODO: Add link to full documentation on Beam website when it's published. | ||||||
|
||||||
i.e. "See the | ||||||
[documentation](https://beam.apache.org/documentation/dsls/dataframes/overview/#pre-requisites) | ||||||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. is this a leftover? |
||||||
for details." | ||||||
--> | ||||||
|
||||||
### Datasets and Models for RunInference | ||||||
Data related to RunInference has been staged in | ||||||
`gs://apache-beam-ml/` for use with these example pipelines. You can see this by using the [gsutil tool](https://cloud.google.com/storage/docs/gsutil#gettingstarted). | ||||||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. (optional) maybe link to the cloud console here: https://pantheon.corp.google.com/storage/browser/apache-beam-ml There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Is that link accessible to only Google employees though? would There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. whoops, yes it would. I copied the wrong thing :) There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Done. |
||||||
``` | ||||||
gsutil ls gs://apache-beam-ml | ||||||
``` | ||||||
|
||||||
--- | ||||||
## Image Classification with ImageNet dataset | ||||||
|
||||||
[`pytorch_image_classification.py`](./pytorch_image_classification.py) contains | ||||||
an implementation for a RunInference pipeline thatpeforms image classification | ||||||
on [ImageNet dataset](https://www.image-net.org/) using the MobileNetV2 | ||||||
architecture. | ||||||
|
||||||
The pipeline reads the images, performs basic preprocessing, passes them to the | ||||||
PyTorch implementation of RunInference, and then writes the predictions | ||||||
to a text file in GCS. | ||||||
|
||||||
### Dataset and model for Image Classification | ||||||
|
||||||
<!--- | ||||||
TODO: Add once benchmark test is released | ||||||
- `gs://apache-beam-ml/testing/inputs/imagenet_validation_inputs.txt`: | ||||||
text file containing the GCS paths of the images of all 5000 imagenet validation data | ||||||
- gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00000001.JPEG | ||||||
- ... | ||||||
- gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00050000.JPEG | ||||||
--> | ||||||
- `gs://apache-beam-ml/testing/inputs/it_imagenet_validation_inputs.txt/`: | ||||||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more.
Suggested change
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Fixed. |
||||||
text file containing the GCS paths of the images of a subset of 15 imagenet | ||||||
validation data | ||||||
- gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00000001.JPEG | ||||||
- ... | ||||||
- gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00000015.JPEG | ||||||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. (optional) It might be nice to clarify that these sub-bullets are the file contents with something like: $ gsutil cat gs://apache-beam-ml/testing/inputs/it_imagenet_validation_inputs.txt
gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00000001.JPEG
...
gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00000015.JPEG There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Added. |
||||||
|
||||||
- `gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_*.JPEG`: | ||||||
JPEG images for the entire validation dataset. | ||||||
|
||||||
- `gs://apache-beam-ml/models/torchvision.models.mobilenet_v2.pth`: Path to | ||||||
the location of the saved state_dict of the pretrained mobilenet_v2 model | ||||||
from the `torchvision.models` subdirectory. | ||||||
|
||||||
### Running `pytorch_image_classification.py` | ||||||
|
||||||
To run the image classification pipeline locally, use the following command: | ||||||
```sh | ||||||
python -m apache_beam.examples.inference.pytorch_image_classification \ | ||||||
--input gs://apache-beam-ml/testing/inputs/it_imagenet_validation_inputs.txt \ | ||||||
--output predictions.csv \ | ||||||
--model_state_dict_path gs://apache-beam-ml/models/torchvision.models.mobilenet_v2.pth | ||||||
``` | ||||||
|
||||||
This will write the output to the `predictions.csv` with contents like: | ||||||
``` | ||||||
gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00005002.JPEG,333 | ||||||
gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00005003.JPEG,711 | ||||||
gs://apache-beam-ml/datasets/imagenet/raw-data/validation/ILSVRC2012_val_00005004.JPEG,286 | ||||||
... | ||||||
``` | ||||||
where the second item in each line is the integer representing the predicted class of the | ||||||
image. | ||||||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. it would be cool if one of the ptransforms in the example joined to integer prediction to the actual name of the image. for example: etc. But that is outside of the scope of this PR. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I have that for a different example #21766 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
(Doesn't need to be exactly that text, just in general Beam docs should mention Dataflow as a distributed runner and be clear that there are others)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, fixed.