oh ahahah you meant the sdk right? yea i noticed some new pipeline functionalities...was gonna wait for an official release but yea sure i will try it. Thanks mate!

Same error without the VENV_DIR variable. oops sorry typo it was already double spaced

task.set_base_docker("<image> -e CLEARML_AGENT_SKIP_PIP_VENV_INSTALL=/opt/conda/envs/rapids/bin/python -e CLEARML_AGENT__AGENT__PACKAGE_MANAGER__TYPE=conda")

Yes actually, i'm trying to access cudf/cuml libraries on rapids and the official guide insist that these libraries within the image has to be used with conda

it was pre-built by rapidsai themselves

ahhh okay the logs are in a closed environment but i will try to extract what i can 🙏

CostlyOstrich36 hmmm i doubt so, i'm the only one using the machine for this particular experiment at the moment.

It intermittently reads the requirements.txt between my repo and the cache. Im wondering if there is anyway to circumvent the cache?

mmm are there any methods to approach this (toggling between pip and conda mode) at the code level? i'm actually not allowed to reconfigure the agents as a developer-user.

yup yup the code runs good locally

Hmm that's strange... AgitatedDove14 i'm using the 1.0.5 pypi package as well as the most recent server from this command - curl https://raw.githubusercontent.com/allegroai/trains-server/master/docker/docker-compose.yml -o docker-compose.yml (iirc it should be 1.1.1)

Ahh okay this was the specific replication of the environment
task.set_base_docker("rapidsai/rapidsai-dev:21.10-cuda11.0-devel-ubuntu18.04-py3.8 -e CLEARML_AGENT_SKIP_PIP_VENV_INSTALL=/opt/conda/envs/rapids/bin/python -e CLEARML_AGENT__AGENT__PACKAGE_MANAGER__TYPE=conda")

ideally the code should be able to import cuml

This is the parent task...shown in the pic of the experiment list above

Oops just closing the loop here, it turned out it was a permissions error on the .clearml cache that was blocking. All's well now hahah thanks!

ahhhh i see...so if i were to logically split a single agent into multiple queues would that allow me to run more tasks?

yea the agent appears idle based on utilisation. It's actually my laptop beside me...

Erm i'm not sure if the pics show what you mean. but essentially the hpo is running on the hpopt queue and my worker "fixitfelix" is assigned to the queue and is supposedly running the experiment already

within the docker image and the conda environment

Hey SuccessfulKoala55 , i figured a workaround to the problem and just wanted to close the loop. Rapids requires c++ code to be integrated into their package and also auxiliary packages inside their prebuilt image and the pip ecosystem currently doesn't support their requirements https://medium.com/rapids-ai/rapids-0-7-release-drops-pip-packages-47fc966e9472 (hence the need to use conda). Instead of trying to run conda with clearml-agent i figured it might be possible to pass the ` PYTHON...

SuccessfulKoala55 Yes, i believe if it's within the .set_base_docker(...) method i should be able to? Is there a specific env variable i can set?

okay will try this. Thanks so much!

AgitatedDove14 yup it's with target project, the code was really just...

# Creating the pipeline
pipe = PipelineController(target_project= "pipeline-demo", default_execution_queue='default', add_pipeline_tags=False)
pipe.add_step(name='model-predict', base_task_id='17cc79f0dae0426d9354ds08d979980g')
pipe.start()
# Wait until pipeline terminates
pipe.wait()
# cleanup everything
pipe.stop()
print('pipeline completed')

AgitatedDove14 i'm still getting this error in 1.0.6rc2 tho

Exception in thread Thread-1:
Traceback (most recent call last):
File "/home/aaron/.pyenv/versions/3.7.8/lib/python3.7/threading.py", line 926, in _bootstrap_inner
self.run()
File "/home/aaron/.pyenv/versions/3.7.8/lib/python3.7/threading.py", line 870, in run
self._target(*self._args, **self._kwargs)
` File "/home/aaron/.pyenv/versions/dev_env/lib/python3.7/site-packages/clearml/a...

oops sorry i found the repo in the .clearml/venv-builds/ folder but but im not sure why the remainder of the code isn't executed still

Hey AgitatedDove14 , so i have gotten the latest server version (as shown in the image from the bottom right of the user profile page) and still no luck with a simple test example like this using clearml-1.0.6.rc2

from clearml import Task, StorageManager, Dataset, PipelineController

# Creating the pipeline
pipe = PipelineController(target_project= "pipeline-demo", default_execution_queue='128RAMv100', add_pipeline_tags=False)

` pipe.add_step(name='predict', base_task_id...

hmmm is this the latest?

oh okay i assumed the new docker-compose would pull the latest image. erm i pulled the new image from docker and restarted the server. But it still seems to be sending me the same error tho

CostlyOstrich36 I'm using clearml_agent v1.1.2 on multiple agents in the same machine

Hey Martin, thanks for the reply. I did set up an agent and the reference template task runs perfectly fine on the worker. I only encounter this problem with the hyper parameter tuning experiments

yes it is running the parent hpo task.

CostlyOstrich36 ahhh i suspect the error might be coming from using a cached repository? e.g. Using cached repository in "/root/.clearml/vcs-cache/<my repository> it seems like it is trying to install the requirements.txt that was cached but isnt available anymore and there are occasions where the installed packages do not reflect a complete list of what was specified in the repository's requirements.txt. Could this be a possibility for the error (either not detecting the complete list...