Skip to content
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.

Sign up for GitHub

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

Lift restrictions on cache overrides #907

Merged
merged 4 commits into from
Oct 13, 2022
Merged

Lift restrictions on cache overrides #907

Changes from 3 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
84cfc63
Cache key overrides are now open to all types
eapolinario Oct 4, 2022
46cdf6a
Remove other mentions to FlyteSchema and StructuredDataset
eapolinario Oct 4, 2022
c743d7b
Merge branch 'master' into lift-restrictions-on-cache-overrides
eapolinario Oct 6, 2022
52b5d27
Merge branch 'master' into lift-restrictions-on-cache-overrides
eapolinario Oct 12, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 2 additions & 7 deletions cookbook/core/flyte_basics/task_cache.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -127,7 +127,7 @@ def wf(a: int, b: str):

# %%
# If run twice with the same inputs, one would expect that ``bar`` would trigger a cache hit, but it turns out that's not the case because of how dataframes are represented in Flyte.
# However, with release 0.19.3, Flyte provides a new way to control memoization behavior of `FlyteSchema` and `StructuredDataset` literals. This is done via a ``typing.Annotated`` call on the task signature.
# However, with release 1.2.0, Flyte provides a new way to control memoization behavior of literals. This is done via a ``typing.Annotated`` call on the task signature.
# For example, in order to cache the result of calls to ``bar``, you can rewrite the code above like this:


Expand Down Expand Up @@ -169,11 +169,6 @@ def wf_1(a: int, b: str): # noqa: F811
#
# This feature also works in local execution.
#
# How to Enable Caching of Offloaded Types for a Different Type Transformer?
# ***************************************************************************
#
# The ability to cache offloaded objects is controlled by the `hash_overridable <https://github.com/flyteorg/flytekit/blob/9754a0f462f495dbd77df0572719885230bd909b/flytekit/core/type_engine.py#L56-L62>`_ property of the Type Transformer. For example, see how this is done in the case of `StructuredDataSetTransformer <https://github.com/flyteorg/flytekit/blob/9754a0f462f495dbd77df0572719885230bd909b/flytekit/types/structured/structured_dataset.py#L382-L383>`_.

# %%
# Here's a complete example of the feature:

Expand Down Expand Up @@ -204,7 +199,7 @@ def compare_dataframes(df1: pandas.DataFrame, df2: pandas.DataFrame):
def cached_dataframe_wf():
raw_data = uncached_data_reading_task()

# Execute `cached_data_processing_task` twice, but we force those
# Execute `cached_data_processing_task` twice, but force those
# two executions to happen serially to demonstrate how the second run
# hits the cache.
t1_node = create_node(cached_data_processing_task, df=raw_data)
Expand Down