vis4d.common.util

Utility functions for common usage.

Functions

create_did_you_mean_msg(keys, query)	Create a did you mean message.
init_random_seed()	Initialize random seed for the experiment.
set_random_seed(seed[, deterministic])	Set random seed.
set_tf32(use_tf32, precision)	Set torch TF32.

create_did_you_mean_msg(keys, query)

Create a did you mean message.

Parameters:

• keys (list[str]) – List of available keys.

• query (str) – Query.

Returns:

Did you mean message.

Return type:

str

Examples

>>> keys = ["foo", "bar", "baz"]
>>> query = "fo"
>>> print(create_did_you_mean_msg(keys, query))
Did you mean:
    foo

init_random_seed()

Initialize random seed for the experiment.

Return type:

int

set_random_seed(seed, deterministic=False)

Set random seed.

Parameters:

• seed (int) – Seed to be used.

• deterministic (bool) – Whether to set the deterministic option for CUDNN backend, i.e., set torch.backends.cudnn.deterministic to True and torch.backends.cudnn.benchmark to False. Default: False.

Return type:

None

set_tf32(use_tf32, precision)

Set torch TF32.

Parameters:

• use_tf32 (bool) – Whether to use torch TF32. Details: https://pytorch.org/docs/stable/notes/cuda.html#tf32-on-ampere

• precision (str) – Internal precision of float32 matrix multiplications. Details: https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision # pylint: disable=line-too-long

Return type:

None