Reputation
Badges 1
45 × Eureka!Hello , I passed following variables to the agent followingly: CLEARML_AGENT_SKIP_PIP_VENV_INSTALL=1 CLEARML_AGENT_SKIP_PYTHON_ENV_INSTALL=1 clearml-agent daemon --queue "default" --foreground
and it works perfectly ! Thanks for advise.
I just set up my server from following url : https://clear.ml/docs/latest/docs/deploying_clearml/clearml_server_linux_mac/
command i run:learml-agent daemon --queue default --foreground
response i get:clearml_agent: ERROR: create.<locals>.Validator.__init__() got an unexpected keyword argument 'types'
isnt brew for mac ? i am on ubuntu
for the requirements how do you mean it please? To add requirements.txt into root directory ith the description of packages is enough ? or do you have to put somewhere you want to use this file? Thanks
Still not solved, idk if these dependencies are cached somewhere but when i change requirements.txt or i add it manually into code it still have problems with the torch and is looking for 'torch==1.10.0+cu113'
Insight 2 <- repository is cloned correctly but the uncomitted changes are not.
OPENBLAS="$(brew --prefix openblas)" pip install pandas
Probably i figured out everything. For me it is better to work with https and token (for git).
Thanks for help ! I will try and let you know.
when i run it locally it was python script.py and for the remote you are right
Hi @<1523701070390366208:profile|CostlyOstrich36> , the worker cloned the repo correctly, however in the nested scipt if you use task.init
it wont work / wont overwrite anything.
I understand what you mean. I am just describing different case. Lets assume i have my docker image already (all dependencies , data solved). Right now I run my task and it automatically looks for requirements.txt file in the repository. My question is -> is there any way to avoid this (simplest solution for now will be to -> rename requirements.txt to different filename)? I tried the things you sent already. The thing is that in requirements.txt in this repos can not be installed that easily...
To describe the use-case. Lets say we have someapp which can export specific training script. I would like to create this as a specific "draft" task and later execute it.
RE: When people do not install the server on the same machine, how is it possible for them then ? I cant reach apiserver / clearml-apiserver.
After new installation of clearml-agent and clearml I still do have the same problem.
Example: I have a simple python script and defined default_docker in clearml.config. When i clone this experiment and run it from clearml dashboard my clearml-agent running in docker mode should execute this task in docker. However, it is stucked after dependencies in...
The agent simply try to install requirements from requirements.txt , however i dont want to do that because i have my docker image ready.
I went into app.clear.ml dashboard -> project -> archived experiments -> and i wanted to delete experiment from there -> that is the output i got
but for the local execution virtual env (conda) named detectron2 was used
Server: 3.13.0-832. I just waited for some time and suddenly it was possible to delete it again. Idk what was that issue. No matter what , if it will appear again I can announce you.
you can edit the requirements section directly <- where ? if i create requirements.txt it seems to be ignored
I'm at the point where it looks like the clearml-agent is stuck (How i execute the agent: clearml-agent daemon --queue "default" --gpus 0 --foreground --docker. After the last message: " Successfully installed:<dependencies>" nothing really happens. I do attach logs from experiment. And I also do provide the config:
So, I just just define all the requirements for the docker in clearml.conf
in default_docker
part?
So probably you are right - > nc -vz localhost 8080
Output when run locally not in docker: Connection to localhost (127.0.0.1) 8080 port [tcp/http-alt] succeeded!
Output when inside docker bash: localhost [127.0.0.1] 8080 (http-alt) : Connection refused
I can see the docker in docker ps
but it seems like it never gets to code execution. I do not have an idea where it got from. Seems like somewhere it gets "pip" + "pip".