reference/generated/tensordict.TensorDictBase.to_store

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

TensorDictBase.to_store

TensorDictBase.to_store(*, backend: str = 'redis', host: str = 'localhost', port: int = 6379, db: int = 0, unix_socket_path: str | None = None, prefix: str = 'tensordict', device=None, **kwargs) → Any

Uploads the tensordict to a key-value server and returns a tensordict backed by it.

The entries are written to a Redis or Dragonfly server, under the key namespace prefix. The result reads and writes the server. A lazy stack is uploaded in chunks, without building the dense stack in memory, and gives a LazyStackedTensorDictStore. Requires the redis Python package and a running server. See Large datasets and storage backends.

Keyword Arguments:

• backend (str, optional) – the server type: "redis" or "dragonfly". Defaults to "redis".

• host (str, optional) – the server hostname. Defaults to "localhost".

• port (int, optional) – the server port. Defaults to 6379.

• db (int, optional) – the database number. Defaults to 0.

• unix_socket_path (str, optional) – the path of a Unix domain socket to connect through instead of host and port. Defaults to None.

• prefix (str, optional) – the key namespace of the data on the server. Defaults to "tensordict".

• device (device, str or int, optional) – the device of the tensors that the result reads. Defaults to None, which uses the device of the tensordict.

• **kwargs – passed to the connection of the client.

Returns:

A TensorDictStore, or a LazyStackedTensorDictStore for a lazy stack.

Examples

The example needs a Redis server on localhost:6379, and the last line a Dragonfly server on the host dragonfly-host, so it is not run when the documentation is tested.

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict({"obs": torch.randn(10, 84)}, [10])
>>> store_td = td.to_store(host="localhost")  
>>> store_td["obs"].shape  
torch.Size([10, 84])
>>>
>>> # Using Dragonfly instead of Redis
>>> store_td = td.to_store(backend="dragonfly", host="dragonfly-host")  

Every tensordict class and every tensorclass has this method. For the other methods, see Methods by task.