Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
I Have A Pipeline With Tasks A->B->C. I Want To Be Able To Trigger It Manually, And Skip A Regardless Of It’S Cache Status. I Want To Pass B Value That Represents A’S Output If Needed. What’S A Good Way To Achieve This (Can Be Ui-Based, Or Pipeline-Gymnas

I have a pipeline with tasks A->B->C.
I want to be able to trigger it manually, and skip A regardless of it’s cache status.
I want to pass B value that represents A’s output if needed.
What’s a good way to achieve this (can be UI-based, or pipeline-gymnastics based solution)

Posted one year ago
Votes Newest

Answers 7

Hi RoughTiger69
Is the pipeline in question based on decorators or is it based on existing Tasks?

Posted one year ago

Sort of.
I would go with something which is more like:
` execution_plan = {'step_b':'b_result', step_c: None, ...}

def pipeline(execution_plan):
step_results = {}
for step in pipeline.get_dag():
if step.name in execution_plan.keys():
step_results[step.name] = execution_plan[step.name] or step(**step_results)

`The ‘execution plan’ specifies list of steps to run (keys) and for each, whether we should use a user-specified value.

Then we iterate on the steps in the dag.
If the step is not at all in the execution plan, skip it entirely.
If the step is in the execution plan and has a user-provided value - use that value.
If it doesn’t have a user provided value - execute the step with all **kwargs it exects.
if decided to execute - fall back on clearml behaviour (including cache configuration)

Posted one year ago

I think the part that is missing for me is the context, in other words how would one configure the execution_plan and why would they configure it in a specific way?
My intuition, without fully understanding it, is that for some reason the internal DAG/decision is exposed to the user, and it feels like too much information. Basically I have a hunch that the users should not need to have such deep understanding to control the flow, and they should end up with an abstraction on top of it. Does that make sense ?

Posted one year ago

Decorators are good 🙂
Something along the lines of
` @PipelineDecorator.pipeline(...)
def pipeline(skip_a=False):
if not skip_a:
a = step_a()
# somehow get a previous A?
# let's call it cached A
a = "replace with real'

... `Is this the gist?
If it is, this looks like, "how can I control whether A is cached or not", is that correct?

Posted one year ago

It’s more like this:
I have a pipeline, ran on all data.
Now I change/add a sub-dag to the pipeline
I want to run only that sub-dag on all historical data in ad-hoc manner
And then next runs will run the full dag (e.g. only on new data)

Posted one year ago

I want to run only that sub-dag on all historical data in ad-hoc manner

But wouldn't that be covered by the caching mechanism ?

Posted one year ago

AgitatedDove14 decorators. but I would consider to convert it to whatever in order to achieve the above

Posted one year ago
7 Answers
one year ago
one year ago