# Keep PyTorch DataLoader when using Ray Data

**URL:** <https://discuss.ray.io/t/keep-pytorch-dataloader-when-using-ray-data/12730>\
**Category:** Ray Data\
**Created:** [November 7, 2023, 2:00am UTC](https://discuss.ray.io/t/keep-pytorch-dataloader-when-using-ray-data/12730 "2023-11-07T02:00:54Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![xuyifangreeneyes](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/xuyifangreeneyes/32/5300_2.png) [@xuyifangreeneyes](https://discuss.ray.io/u/xuyifangreeneyes)\
**Post date:** [November 7, 2023, 2:00am UTC](https://discuss.ray.io/t/keep-pytorch-dataloader-when-using-ray-data/12730/1 "2023-11-07T02:00:54Z")

</div>

**How severe does this issue affect your experience of using Ray?**

- High: It blocks me to complete my task.

I wonder whether there is a solution to use Ray Data with my current PyTorch DataLoader.

Here is my current code.

```auto
for inputs in next(dataloader):
    outputs = model(inputs) # inference
    postprocess_on_cpu(outputs)

```

Now I hope to decouple inference and post process to fully utilize CPU/GPU resources. I guess my code can be rewritten using Ray Data as follows.

```auto
ds = ray.data.read_parquet(data_uri)
ds = ds.map(preprocess_on_cpu)
ds = ds.map_batches(model, compute=ray.data.ActorPoolStrategy(size=8), num_gpus=1, batch_size=1024)
ds = ds.map(postprocess_on_cpu)
# then pull data from ds

```

The problem is to migrate data loading logic from PyTorch DataLoader to Ray Data. [Ray Data Doc](https://docs.ray.io/en/latest/data/working-with-pytorch.html#migrating-from-pytorch-datasets-and-dataloaders) says:

> Any logic for reading data from cloud storage and disk can be replaced by one of the Ray Data `read_*` APIs, and any transformation logic can be applied as a [`map`](https://docs.ray.io/en/latest/data/api/doc/ray.data.Dataset.map.html#ray.data.Dataset.map) call on the Dataset.

However, it may take some effort when data loading logic is complicated. Besides, I don’t want to risk data loading performance degradation.

It seems that `ray.data.Dataset` is always constructed from some data source(for example, `ray.data.read_parquet`). Is there any way to use Ray Data with my current PyTorch DataLoader, or I may need to use more primitive Ray APIs(for example, manipulate tasks and actors directly) to decouple inference and post process.