Reputation
Badges 1
100 × Eureka!i had a misconception that the conf comes from the machine triggering the pipeline
My use case is developing the code, i don’t want to spam the UI
BTW the code above is from clearml github so it’s the latest
I am currently on vacation, I'll ask my team mates. But if not I'll get to it next week
when i did this with a normal task it worked wonderfully, with pipeline it didn’t
And i am logging some explicitly
BTW, i would expect this to happen automtically when running “local” and “debug”
TimelyMouse69
Thanks for the reply, this is only regarding automatic logging, where i want to disable logging all together (avoiding the task being added to the UI)
also, i don’t need to change it during execution, i want it for a specific run
and the agent is outputting sdk.development.default_output_uri =
although it’s different in both the original config, and the agent extra config
@<1523701435869433856:profile|SmugDolphin23> @<1523701087100473344:profile|SuccessfulKoala55> Yes, the second issue still consists, currently breaking our pipeline
It’s a lot of manual work that you need to remember to undo
i’m following this guide
https://docs.fast.ai/distributed.html#Learner.distrib_ctx
so you run it like thispython -m fastai.launch <script>
tnx, i just can’t use 1.7.1 because of the pipeline problem from before
don’t have one ATM
i didn’t, prefer not to add temporary workarounds
looks like it’s working 🙂 tnx
not sure about this, we really like being in control of reproducibility and not depend on the invoking machine… maybe that’s not what you intend
` from clearml.automation import PipelineDecorator
from clearml import TaskTypes
@PipelineDecorator.component(task_type=TaskTypes.data_processing, cache=True)
def run_demo():
from transformers import AutoTokenizer, DataCollatorForTokenClassification, AutoModelForTokenClassification, TrainingArguments, Trainer
from datasets import load_dataset
dataset = load_dataset("conllpp")
model_checkpoint = 'bert-base-cased'
lr = 2e-5
num_train_epochs = 5
weight_decay =...
i’ll try to work on something that works on 1.7.2
SmugDolphin23 SuccessfulKoala55 ^
This is the next step not being able to find the output of the last step
ValueError: Could not retrieve a local copy of artifact return_object, failed downloading
We also wanted this, we preferred to create a docker image with all we need, and let the pipeline steps use that docker image
That way you don’t rely on clearml capturing the local env, and you can control what exists in the env
@<1523701205467926528:profile|AgitatedDove14>
Only got some time to work on it now, i created a small reproducible example.
I also tried to use your suggestion with import accelerate, it also had issues.
overall, when using debug_pipeline
it works ok, but both methods don't work without it, i think it has something to do with wrapping accelerate.
Problem with launching through python module (your suggestion), the argparse breaks.
Problem with launching using a new process - rank0 proce...