Adeel Hassan

Make Raster Vision easier to use as a library (and not just as a framework)

9

## Problem Raster Vision (RV) is a [framework](https://en.wikipedia.org/wiki/Software_framework). To use it, users must configure the RV pipeline by defining a `get_config()` function. This has the following disadvantages: - Even the...

## 🚀 Feature Support multi-GPU training and prediction. ## Motivation Distributing the workload across multiple GPUs can considerably speed up the training and prediction process.. ## Pitch PyTorch provides [at...

## 🚀 Feature See the following discussion on Gitter: https://gitter.im/azavea/raster-vision?at=6298bff6ef00bd1dc60fc7ee `ObjectDetectionGeoDataWindowConfig` should allow `neg_ratio` to be `None` since `ObjectDetectionRandomWindowGeoDataset` allows it. ## Motivation It's a feature that already exists. This...

## 🐛 Bug See the following discussion on Gitter: https://gitter.im/azavea/raster-vision?at=6298bff6ef00bd1dc60fc7ee Whenever a `RandomWindowGeoDataset` for a scene fails to find an eligible chip in `max_sample_attempts` attempts, it throws a `StopIteration` error...

## 🚀 Feature RV uses https://github.com/azavea/mask-to-polygons for vectorization, but the library is no longer being maintained. We should move the vectorization code into RV itself and remove this dependency. ##...

pytorch_learner examples need to be updated

1

[pytorch_learner/learner_pipeline.py#L16](https://github.com/azavea/raster-vision/blob/master/rastervision_pytorch_learner/rastervision/pytorch_learner/learner_pipeline.py#L16) calls `ClassificationLearnerConfig.build()` ``` learner = learner_cfg.build(learner_cfg, self.tmp_dir) ``` which has the signature ``` def build(self, tmp_dir, model_path=None): ... ``` So the above call should, in effect, map `learner_cfg` to...

Add example notebooks

9

As a follow-up to #1395, we should add Python notebooks that demonstrate how to use Raster Vision as a library. If feasible, it would also be nice to automatically add...

### for #1395 - `core` - [x] Add a `Labels.from_predictions()` method to easily convert model predictions to `Labels` (#1434) - [x] Refactor `ClassConfig` to use pydantic validators (#1436) - [x]...

[Migration guide](https://docs.pydantic.dev/2.5/migration/).